Climatic Geomorphology PDF

ELSEVIE R 8.\'. ELSEVIER Inc.
ELSEVIER Ltd ELSEVIER ltd

Rada r"''l: 29 525 B Street. Suite 1900 The Boulc'''trd. L.tngfo rd l ane 84 Theobald> Ro.td
P.O. nu, 211. 1000 AE. Amsterdam S.1n Diego. CA 92101-4495 K1dlington, O.<ford OX~ tG B London WCI X 8RR
The Neth• rlands USA K UK
~ 200 I Origindl Spamsh La nguage Ed nion. Ed1cio ncs Omega. S.A., Barcclouu. Spaiu.
0 2005 Engli'h transla tion. Elsevier B.V.
l 'his work i~ prot~'Ctcd under COp}nght by Ekvicr B.V.• " "d the following tern» aud condition, apply 10 its u'"':
l'hotocopymg
Single photocopies of stngle ~:hap tcr> m.1y lx: made for persona l use as allo,,ed by na tional cop~ right la,.s . l'crmi.sio n of
the Publi her and pa}menl of a fee is required for all other pho tocopying. including multiple or systen1.11ic copying.
co p}1ng fo r odvcnising nr promotional purposes. rcs:tk·. and all fomt~ of document deh,·ery. Specull nues are a\ailable
for educa tional in>tituLions that ,,;.h to ma ke pho tocopies fo r non·prolit edu ational cl.tssroom ~.
Pcnn1ssions n><l) be >ought direct I~ from El.e,'icr's Rights Deparcmrnt in Oxford, UK : phone (- 441 1865 843S30.
faA ( • .I-IJ 1865 85J3J J, c·mail: pcrmi.sion. 'a cl~•'irr.com. Requests mly ulso ~com pleted on-line \ ia th~ Else vier
horne page (http : ""' w.else\'ier.com loc:uelpcrmissions)
In the USA. users may clear pcnnis.ions and maLe payments through the Copyright C lc:or:mcc Center. Inc.. 222
Rose\\ood Dm e. Danvers. MA 01923. US1\ : phone: (+I) (978) 7508400. fa.\: (-I) (978) 7504744. ~nd in the K
through the CoJ>yngiH L1cen5mg Ag~nC) Ra pid Ckarnncc Service (CLARCS). 90 Tottcnham Court Ro;1d. Lo ndon
\ I P OLP. K : phone: ( ~ 44) 20 763 1 5555: f:tx : ( - .1-1) 20 763 1 5500. O ther countr ies may have a local r~prographic
right> .lgcncy for p:t) mcnts.
Deri><llhe Work'
T nbles of content> ma) be reproduced for internal c trculatto n, but permission o f the Publi,hc r;, required fo r external
rc.s.~le or di>tribution of such material. Permission o f the Publi\her is required for all o ther deriv:lli\'C work>. including
compilAtion' a nd translatiOns.
Electro nic Storage o r Usage
Pcrmiss1on of the Publisher 1 required to ' tore or u;e tl<:ct ro nically any material conta ined in this work. including a ny
chapter or pan of a c hapter.
E.1c.:pt (15 o uthncd abo•·c. no p:trt of this wo rk rna) lx: reproduced, •lorcd in a rctrie1al sys1em or t r:tthlniued in any
form <•r b) any rnc.ons. ekctronic. mechanical. photocopying. record ing or o thcm isc. \lltho ut prior wriuen permt.sion
of the Publi'h~r.
Addrc>s pcrmis-iuns roqursts to: Elsc1icr's Righ ts Drpartmcnl. attbc fax and e·mail ndd re""" noted ab<>\ c.
Notice
No responsibility IS as•umed by rhe Publisher for a ny injury a nd <o r damage to tx•rso ns or property as a mancr o f products
lr~b1lit) , n~ghgentt o r o therwiSe. or from an) use or o pcrntion of any methods. products. 1nstruct10ilS or ide.ts containod in
Lhc material herein. &'C'ausc ofrnpid ad ~anccs in the medical sciences. in particular. independent verificat io n of diagnose>
and drug dosage> >hould be made.
Onginal Tille: G eomorfologia C lim~ti(.,t. published 2001 by Edicionrs Omega. S.A .. Barcelona. Spain. first cdilion
Eng.lish tr.mslauo n publrshcd 1005 by Elsc\'lcr O.V.
Lihrar)' of Congre s \atalogmg in Publication Data
/\ catalog re-cord is av:oila ble from the Libra ry of Congress.
Sriti'h Lrbra ry Ctta loguing in Public:otion D a ta
A c.1talogue r~-cord is a' ailablc from the Bri t i~h Library.
ISBN: 041~ - 1794·4 (H ardback)
ISO ': ~4·51 128-3 (Paperback)
IS BN: 0928-2025 (Scric')
@ Titc paper u<ed in tht> pubhcatio n me.:ts the rcquirrotcnts of A 'S I NISO Z39.4~· 1 992 ( Permanence of Paper).
Prime,( 10 h.tly
Working together to grow

libraries in developing countries
www.clscvier.com I Y,'WW.bookaid.org I www.sabre.org
Contents
Preface ix
Preface to the English language edition xiii
FIRST PART
Introduction and concepts
I Climatic geomorphology 3
SECO D PART
Geomorphology of glacial zones 33
2 Glaciers 35
3 Glacial erosion 61
4 Glacial transport and sedimentation 91
5 Fluvioglacial ero ion and sedimentation 105
6 Geomorphology applied to glacial regions 125
THlRD PART
Geomorphology of periglacial zones 143
7 The periglacial environment 145
8 Peri glacial landfom1s 17 1
9 Some aspects on applied geomorphology in perigl acial regions 209
FOURTH PART
Geomorphology of arid regions 227
l 0 T he arid zone 229
II Weathering proce. ses and resulting fonn · 241

viii Contents
12 Desert surfaces: pave ments, panerned ground, vamishcs and crusts 259
13 T he action of water in arid regions 285
14 Slopes in arid zones 3 17
15 The arid region piedmonts: glacis and alluvial fans 343
I6 Desert Jakes: playa Jakes and sebkhas 383
FIFTH PART
Aeolian geomorphology 407
17 Aeolian processes and erosive landforms 409
18 Aeolian accumulations 429
SIXTH PART
Applied geomorphology and arid regions 459
19 Applied geomorpho logy and arid regions 46 1
SEVENTH PART
Geomorphology of tropical zones 499
20 The humid tropics: weathering and laterites 50 1
21 Tropical land forms 533
22 Applied geomorphology in tropical regions 583
EIGHTH PART
Geomorphology and climate change 599
23 Environmental change 601
24 Climate change in g lacial and periglacial regions 6 15
25 Climate change in arid and tropical humid regions 643
References 697
Subject Index 753
c
Preface
Climatic Geomorphology constitutes a new look at a fairly old subject that has been
largely ignored in other more recent trends in the field, particularly in much process
geomorphology. But now Mateo Guti6rrez, well-known Professor of Geomorphology at
the University of Zaragoza in northeast Spain, has presented us with this new volume,
translated into English by ten of his colleagues. We are fortunate indeed to have his
authoritative writing and artful photography of so many parts of the world to grace our
understanding of landform character and evolution. Originally published in Spanish in
2001, this volume attracted the notice of many other geoscientists and was reviewed quite
favourably in more than five of our main journals of geomorphology, physical geography,
and environmental geology. The uniform praise led to suggestions that a translation into
English might be warranted, with the results that you now see before you. The relative lack
of such modem and well-illustrated texts devoted to climatic geomorphology in the
English language indicates that this new book is a deserving addition to the literature.
In recent years the great interest in the climate change so obviously underway in the
world, as well as the consequences of even greater climate change in the geological past,
have led to considerable interest in the topic. This book provides a global overview of the
different climatic zones of the Earth, and it reviews the advances in climatic
geomorphology that have been achieved in the past decades. Certain landform assemblages
are well known to be associated with different regional climate-process systems, a
statement that underscores the obvious, despite the fact that the details of some climate-
controlled processes are not yet that well understood. Thus we all recognize certain glacial,
periglacial, or humid tropical landform types, even though processes such as the exact
efficacy of deep glacial erosion, or long term rock-glacier formation and mechanics, or
bornhardt generation may remain somewhat elusive. Furthermore, in the temperate
latitudes where so many geomorphologists live and work, with the plethora of
morphoclimatic process overprints, such as, for example, the Pleistocene glacial signatures
impressed upon so much of the landscape of Europe and North America, it may surprise
some that aspects of this now vanished glacial process still remain somewhat enigmatic.
Nevertheless, the fact that climate change has occurred so many times in the past is not
in dispute, although the lability or amount of time that a given landform may take to come
into equilibrium existence in a particular climate type may not be at all well understood.
Thus a renewal of interest in the topic of climatic geomorphology seems entirely justified
as we seek to better understand the many ways that climate forcings produce landforms.
Nonetheless, one point that Professor Guti6rrez makes in this book that must always be
kept in mind is expressed by the principle of convergence or equifinality, in which different
processes produce the same landform so that oversimplified climatic explanations do not
x Preface
work. For example, granites and some other massive crystalline and sedimentary rocks
weather and erode to produce tors in a variety of climatic regimes from tropical humid to
periglacial. Long-term chemical weathering of certain minerals in these rocks, with later
stripping of the weathered mantles, occurs in a variety of climate regimes. Similarly, frost
shattering and gelifluction processes can also produce these bedrock tots. In the same
equifinality fashion, horizontally bedded sedimentary and volcanic rocks can produce
similar cliffed topography in various climates wherever the free faces are maintained by
undercutting of the resistant rocks, even through dramatic climate changes. Excellent
examples include the buttes and mesas of the deserts and pluvial basins of the
southwestern USA, or the tepui mesas over which Angel Falls flows in the Amazonian
rainforest of Venezuela. Similarly, bog-burst mass movement landforms can occur
wherever rainfall and edaphic conditions allow saturated conditions to generate unstable
peat bogs, be it tropical humid highlands, humid temperate maritime regions, or
periglacial tundras. But in spite of such landform convergence, or the palimpsest landform
inheritance from past climatic regimes, understandings of climatic geomorphology still
have much to offer those who utilize the appropriate concepts.
This book is structured into eight main sections that are set up first as an introduction to
some of the history of geomorphology, then separate groupings of multiple chapters on
glacial, periglacial, arid, aeolian, tropical, and climate-change geomorphology. The 25
different chapters in all of the sections discuss the dominant weathering, erosion, and
deposition associated with the processes characteristic of each climate zone, along with
the resulting erosional and depositional landforms. Strikingly, four chapters on applied
aspects of glacial, periglacial, arid, and tropical landform processes and products provide
an exceptionally important primer for many parts of the less developed world. This aspect
will recommend the book to those in government offices or in the various development
agencies world-wide who may need to better understand hazards or opportunities for
assisting their citizenry to live more effectively in their particular climatic environments.
The dominant fluvial regimes of the temperate latitudes receive only limited attention
in this book, which may somewhat reduce interest for those geomorphologists who
specialize in such aspects in their regions. I suspect, however, that most geoscientists will
still recognize the importance of the arid and tropical zones to so many people of the less
developed world, and therefore will understand the reasons for the particular emphases in
this book. In addition, the lack of a quantitative approach in the book, or the overall dearth
of formulas may not meet the preferences of some, but still the many graphs and diagrams
taken from the primary references do cover adequate quantitative materials.
Although the author of this book is well versed in the whole of climatic
geomorphology, he has devoted about one third of the book to arid-zone topics, perhaps
reflecting some of his main interests in interior, semi-arid Spain. Out of a total of
25 chapters, after the introductory chapter, 10 are on aeolian, arid zone, and applied arid
topics. Next comes the glacial zone, with five chapters, and the periglacial and tropical
zone sections with three chapters each. The last section on geomorphology and climate
change also includes three chapters, which may add a measure of considerable interest to
those younger geomorphologists who are going to live through what promise to be the
most interesting climate changes to come in the 21st century. This book provides some
mental preparation for recognizing and perhaps dealing with the most likely changes that
can be expected.
Preface xi
The great majority of references in this book are in fact from the English-language
literature, and thus these concepts originally had to be converted into Spanish for the initial
volume. But because this book was then translated from the Spanish back into English by
some ten different translators, some of our phraseology may be a little less artful than
intended by the original authors of the primary references. If this is detected by anyone I
must plead guilty myself to the difficulty presented in editing some of the literal
translations back into the rather more idiomatic presentations that exactly reflect the
original intentions, as well as being also quite accurate scientifically. We of the editorial
staff did our best to convey the key ideas and information without error, but some few
problems may have crept in inadvertently in this fashion anyway, inasmuch as we did not
elect to rewrite the text from the original primary references. By and large the original text
was also left fairly unmodified in the editorial process, with the exception of a few added
observations and more up-to-date references that were felt to enhance the text.
The disconnect between the dominant studies of process geomorphology of the last half
of the 20th century, or what has been referred to as timeless geomorphology, and an adequate
exposition of time-bound landform evolution has become obvious to many geomor-
phologists at the beginning of the 21 st century. This book, Climatic Geomorphology, is put
forward as one approach to better understanding landform evolution from a climatic control
of process such that the time-bound nature of the control is seen to exert both subtle and
obvious process-directed changes to the landscape. An alternative and quantitatively robust
approach to achieving this same result can be observed at many of the more advanced
geoscientific meetings of the world today wherein the equations of climate-controlled
process mechanics are run iteratively with computers to synthesize landscape evolution.
Decades of new work in this vein are likely to be required to make adequate and realistic
progress with this new technology, but this robust methodology is already producing
dramatic results. It is likely that this book can serve as one of the background sources of
concepts and references that will enable better understanding of these process mechanics
under certain climate forcings, and the inevitable fluctuations that must result from climate
change. In this fashion we may eventually come to better understand the rich variety of
climatically controlled landscapes that dominate planet Earth.
The publication of this book, Climatic Geomorphology, represents the first of the now
rejuvenated series on Developments in Earth Surface Processes published by Elsevier.
Geomorphology in the 21st century seems to be undergoing a renaissance of theory and
methodologies as we recover from the paucity of robust theoretical underpinnings following
the partial collapse and/or dramatic revisions of the Davisian-Penckian-Kingian
paradigms in the middle of the prior century. After research excursions deep into the
reductionistic thickets of process geomorphology, we now seem poised to move on smartly
with the integration of short-term, timeless, landform process studies into long-term, time-
bound, landform evolution. This book is offered as a rejuvenated first in a series of what is
hoped will be an ever-increasing list of titles of new developments in geomorphology.
John Shroder
Preface to the English language edition
During the past few decades, knowledge in climatic geomorphology has been substantially
increased thanks to numerous detailed investigations, the application of a large number of
techniques, and the acquisition of abundant absolute dates. The challenge of predicting the
effects of the prophesied future global warming on morphogenetic processes and
landforms has encouraged geomorphologists to study the Late Pleistocene and Holocene
climatic changes from the geomorphological and geological record. The advances
achieved in the field of climatic geomorphology during the past years are reflected in the
publication of several specific monographs about the different morphoclimatic zones. The
aim of this book is to provide an up-to-date general view of this branch of geomorphology.
It includes a chapter on applied geomorphology for each morphoclimatic zone providing
an approximation of the main environmental problems.
I hope that the book will be useful to geomorphology students and to those interested in
the role played by climatic conditions on the generation of landforms and landscapes.
I would be fully satisfied if it helps some researchers working in disciplines close to
climatic geomorphology. I would like to express my warmest gratitude to the scientists
involved in the review of the book: G. Benito, A. Cendrero, J.L. Goy, J. Gracia, F.
Guti~rrez, J. Ldpez-Martfnez, A. G6mez-Ortiz, A. Martfn-Serrano, J.L. Pefia, A. P~rez-
Gonz~ilez, J. Rodrfguez-Vidal, J.M. Garcfa-Ruiz, C. Sancho and C. Zazo. Their wise
suggestions and the unpublished data supplied have contributed to the improvement of the
content of the book. Most of them have provided excellent photographs from different
parts of the Planet. I would also like to thank Dr A. G6mez-Sal, Dr B. Lerfinoz, C.
Maldonado, E. Pueyo and M. Urigtien for additional photographs providing.
I would especially like to thank G. Benito, G. Desir, J.M. Garcfa-Ruiz, J. Gracia,
F. Guti~rrez, J. Ldpez-Martfnez, C. Martf, J. Remondo, P. Silva, and B. Valero for their
help with translation of the text into English so that Editor J. Shroder could proceed with
completing the new edition.
We are very grateful to the other editors, societies, institutions and researchers for the
authorizations conceded to reproduce several figures in this publication. We have not re-
ceived responses from some editors and researchers to our request for a reproduction permit
for which we apologize. We would be thankful for any information that would facilitate this
task. Some of the figures came from journals that are now out of print or difficult to obtain.
My sincere gratitude goes to Ana Carmen Aguirre, who has typed all of the manuscripts
and has rigorously prepared the figures with great enthusiasm.
Finally, thanks to Nieves for her continuous help, tolerance and infinite patience.
Mateo Gutidrrez
Chapter 1
Climatic geomorphology
1. Introduction to geomorphology
The term "geomorphology" comes from the ancient Greek words geo (Earth), morphos
(shape) and logos (thesis). This is, therefore, the science of the form of the land surface.
Several authors limit the scope of geomorphology to the study of subaerial landscapes,
whereas others also include submarine topography. Even the study of extraterrestrial
landforms has recently been included in the so-called planetary or extraterrestrial
geomorphology (Baker, 1993).
The topography of the Earth's surface is the result of the balance between endogenic
and exogenic forces. Endogenic forces trigger vertical movements that generate large
mountain belts and depressions, whereas exogenic forces work progressively to denude
the resulting relief. This permanent interplay of forces has been performed at different
scales on the Earth's surface during the whole of geological history. The external
processes occur in a classical sequence of weathering, erosion, transport and sedimenta-
tion. The result is the generation of different erosive and depositional landscapes, with
different specific features depending on the dominant process working in the different
morphogenetic environments. The energy that powers these processes comes from
different sources. Solar radiation received by the Earth heats the land surface and the
atmosphere, as it constitutes the main source of energy for meteorological processes.
Such processes control rock weathering, soil formation, relief development and also
the biological activity of plants and animals. In addition, gravitational energy supports
the processes of sediment transport and mass wasting. Finally, the internal energy of the
planet is the ultimate source of energy for virtually all tectonic processes and the
associated crustal movements that generate relief (Biidel, 1968, 1977).
Traditionally, the study of geomorphology has focused on the analysis of the sequence
and nature of geomorphic events involved in the present configuration of the Earth's
surface through geological time. This approach is denoted by some authors as historical
geomorphology (Chorley, 1978). Recently, the study of the working processes at smaller
scales, together with the analysis of the correlative landform variability, has resulted in the
so-called quantitative (Chorley, 1978) or processes geomorphology (Hart, 1986; Thorn,
1988a). As pointed out by Chorley et al. (1984) such historical studies are based on
retrodiction, whereas processes analysis is mainly focused on forecasting. Both trends are
the dominant ones in most current geomorphological studies. In some countries, processes
analysis has been stimulated in opposition to traditional studies of relief development over
long time spans. From the study of micro-landform development, however, we can obtain
valuable information for a better understanding of the long sequence of events that
has resulted in present-day landforms. We could compare this situation to that experienced
4 Climatic Geomorphology
by the discipline of geotectonics after micro-structural analysis was developed, which

assisted far deeper knowledge of the structure of mountain belts (Gutirrrez, 1990).
These new trends in geomorphology have been instigated, as in other branches of the
Earth sciences, by the evolution of ideas, knowledge and scientific methodology over the
last centuries. The birth of geomorphology as a science took place when the explanatory
focus replaced its initial descriptive motive; only a century ago descriptive geomorphol-
ogy was known as "orography" or "hydrography" (Tricart, 1965). The Americans McGee
and Powell were probably the first authors to use the term "geomorphology" during the
decade of the 1880s (Hart, 1986).
2. History of climatic geomorphology
2.1. Geomorphology before the second half of the 19th century
The origins of geomorphology, as with many other scientific disciplines, are unclear and
the first observations and interpretations came from the philosophers of ancient Greece.
An extensive and detailed analysis of that primary epoch can be consulted in the excellent
treatise of Chorley et al. (1964). Although the term "geomorphology" was not coined then,
the first ideas and fundamentals were developed in concert with other intrinsic
observations of the natural world.
As Holmes (1965) pointed out, in spite of the important advances in mathematics and
the astronomy made by Greek philosophers, they considered natural phenomena to be
powerful manifestations of their ancient gods or myths. In fact, many of the natural
phenomena were personifications of the diverse aspects of nature (Poseidon, Zeus, Pluto,
etc.). These beliefs had a considerable bearing on the development of geomorphologic
theories until the 18th century (the teleological period of Chorley (1978)).
Nevertheless, in spite of the common conviction regarding the divinity of natural
phenomena, some of the philosophers set aside this framework. Thales of Miletus (624-
565 BC) considered natural processes as usual facts that could be understood by
observation and logical approach. Likewise, Anaxagoras (500-428 BC) pointed out that
the Earth revolutions were so slow, when compared with human life, that they go
unnoticed by humans (Cailleux, 1961). This same philosopher indicated that water was
elevated by the Sun and came back to the Earth's surface as rain. Plato (429-347 BC)
imagined that all the rivers came from, and returned to, a huge cavern of water inside the
Earth (Adams, 1938). This theory was possibly based upon field observations of the karstic
landscape of Greece. Aristotle (348-322 BC) also developed excellent ideas on
geomorphology, such as the hydrological cycle and the theory of relief development
due to fiver incision.
As indicated by Cailleux (1961) and Ellenberger (1988), almost all of the deductions
by the Romans about the natural world were based upon Greek theories. This fact can be
inferred from the Natural History of Pliny (23-79 BC), which constituted a summary
of Greek theories, superstitions and conjectures mixed together (Cailleux, 1961).
Consequently, scientific advances were scarce during the Roman period.
This generalized hiatus in science lasted until the 16th century, and was also manifest in
the development of geomorphology. During this millennium and a half, the ancient Greek
Climatic geomorphology 5
works were transmitted to Spain from the Middle East by means of the Arabs. The Arab
manuscripts were translated into Hebrew and Latin in the ancient Translators School of
Toledo (Cailleux, 1961). Therefore, during this time the diffusion of knowledge was
mainly carried out in the monasteries, but also in some cases by means of the individual
patronage of a few kings. In this fashion, Avicena (Ibn Sina, 980-1037), an expert on
Aristotle's work, developed hypotheses on the desert landscapes of Arabia. He thought
that mountains could be generated by elevation of the ground or by the work of water and
wind (Thornbury, 1954).
This period is also characterized by catastrophist or cataclysmic ideas partly based upon
Christian thinking, with the pervasive use of Noah's Flood for the explanation of many
natural processes or observations. This dogmatism produced absurd proclamations, such
as the one manifested in 1654 AD by Archbishop Ussher of Ireland, which stated that the
Sky and the Earth, centre and circumference, were created together by the Holy Trinity on
October 26 of 4004 BC, at 9 o'clock in the morning (King, 1976a,b).
Catastrophist ideas dominated until the 19th century. But meanwhile the brilliant
Leonardo da Vinci (1452-1519) exposed in his works significant considerations on the
origin of the mountains (pioneering ideas about isostasy). He also indicated that rock salt
was generated from the evaporation of seawater, as well as considering the importance of
fluvial erosion and the generally slow work of geomorphic processes. Unfortunately his
ideas were not published until the end of the 18th century, and his scientific influence
during his epoch was practically nil (Ellenberger, 1988). Catastrophism consequently was
the prevailing doctrine during this period.
During the 17th, 18th and part of the 19th centuries, the most significant advances in
geomorphology came mainly from the studies on applied hydrology, carried out by several
engineers. These were mainly French engineers, such as Perrault, De Chezy, Surrell,
Guetthard, Desmarest and others who discarded in their works many academic concepts
that were mostly inoperative.
Catastrophist theories had their opposite counterpart in the uniformitarian ideas
developed at the end of the 18th century, with the postulation of the principle of actualism.
Although previously noted by others in obscure works, this principle was first developed
by Hutton (in 1726-1797) in his work Theory of the Earth (in Hutton, 1788) and later
popularized by Lyell (in 1797-1875) in the first edition of Principles of Geology
(Lyell, 1830), which was extensively reedited and reprinted. These Scottish geologists
established the basis of modem geology, as well as geomorphologic methods. Unifor-
mitarian theory can be synthesized in the simple sentence: "the present is the key to the
past." This means that different landforms have been created by processes similar to those
of the present day, but operating over long periods of time. As we can see, it is totally
contrary to the ideas of catastrophism. For example, Lyell described mountains that were
eroded by weathering and fluvial transport, and later the transported materials were
deposited in sedimentary basins which would ultimately give rise to new mountains.
During this period Playfair (in 1802) published his work Illustrations of the Huttonian
Theory of the Earth, establishing that in areas of homogeneous lithology and structure that
are subject to fluvial erosion, the valleys are proportional to the size of the tributaries
contained in them. It is today is known as "Playfair's Law."
But the catastrophists, as with the cleric geologists Buckland and Sedgwick, argued that
the erratics of glaciers and the alpine hanging valleys could only be explained by
a universal flood (Beckinsale and Chorley, 1968). The gradual acceptance of the ideas of
Lyell resulted in the end of the diluvialists. Also, the publication in 1840 of the book by the
well-known Agassiz (1807-1873) on his studies of glaciers, by using the ideas of
Scheuchzer, Martel, Saussure, and especially Charpentier, further limited the ideas of
catastrophism. In this book the idea was launched of the occurrence of an ice age, and its
impact on the erosion and sedimentation of valleys. The development of these ideas by
Ramsay (in 1862), President of the Geologic Society of London, gave uniformitarism its
decisive scientific support (Price, 1973).
2.2. The dawning of climatic geomorphology
The second half of the 19th century and the beginnings of the 20th were marked by
pioneering scientific exploration of unknown or not very well-known regions of the
Earth's globe. The aims of these expeditions were diverse, but mainly focused on the
exploration of new territories for the evaluation of their mining and agricultural resources,
but also their colonization and Christianization. These objectives came from those of the
previous centuries that were specific to the different continents. Although there are earlier
scientific documents, these scientific studies were mainly focused upon mining and
geology, but detailed observations about landforms were scarce compared with the
excellent works on botany and zoology.
These diverse expeditions resulted in the publication of different descriptive works on
the Earth's relief, and barely contributed to the scientific framework of the early scientific
research. Some German naturalists, however, started to consider Earth surface processes
as illustrated in the works of Von Richtofen (in 1886) and Albrecht Penck (in 1894).
During this period Martonne (in 1913) introduced the term "climatic geomorphology."
The Swiss geologist Agassiz had proposed his glacial theory in his pioneering work
"Etudes sur les Glaciers" (Agassiz, 1840). Although many ideas in this work came first from
others whom he beat to publication (Charpentier, Venetz, Perraudin), it focused upon the
analysis of the set of materials deposited by glaciers, and represented the first well-known
step in glacial geomorphology. Later, the English geologist Geikie, in his work The Great
Ice Age (in Geikie, 1873), discussed the causes of glaciations, ice dynamics and the origin of
glacial and post-glacial deposits. Afterwards, the early 20th century was a fruitful time for
the recognition of existing glacial landforms and deposits. The work of Russell (in 1893) on
the Malaspina Glacier, the paper on Spitsbergen glaciers by Garwood (in 1899) and
Kendall's PhD (in Kendall, 1902) on glacier lakes illustrate this prolific period. During the
second half of this century studies were progressively focused on the origin of erosive
and depositional landforms of ancient glaciated zones (Price, 1973). The book of the
American geologist R.F. Flint (in 1947) on Glacial Geology and the Pleistocene, and
the two volumes of the treatise on the Quaternary Age by Charlesworth (1957), synthesized
and reviewed the state of the art on relict glacier landforms and deposits.
The analysis of surface processes related to frost action goes back to the early 19th
century, but the origin ofperiglacial geomorphology dates from the dawning of the 20th
century. The term "periglacial" was introduced by the Polish scientist Lozinski (1909) to
describe the processes and landforms resulting from freeze-thaw cycles occurring
around the margins of the ancient Pleistocene ice sheets. These periglacial processes,
however, are not only constrained to the periglacial climatic belt, but they can actually
extend into other climatic zones as well. In consequence, at present this term holds a
broader application than in its initial use (Thorn, 1992). At the same time the concept of
solifluction was introduced by Andersson (1906), who also considered the climatic
characteristics and morphologic features of these cold zones. Numerous investigations
were also carried out in the expanses of Siberia by Russian scientists, but these went
largely unnoticed by many European and American scientists due to language problems.
During the beginning of the 20th century, the colonization of Alaska produced the first
well-recognized advances in periglacial research (Cairnes, Capps, Eakin). In Europe, due
to the inaccessibility of the main northern territories, periglacial geomorphology
developed 30 years later (French and Karte, 1988). The first advances were mainly
related to paleogeographic and environmental reconstructions of the late Pleistocene ice
sheets in central and western Europe (i.e. Biidel, Cailleux, Dylik, Edelman, Poser, Tricart,
Troll).
The last part of the 19th century was the most active period in the early development of
desert geomorphology, mainly in the arid zones of the western United States of America
(King, 1976a,b). During different geological expeditions, Powell (1834-1918), Dutton
(1841-1912) and especially Gilbert (1843-1918) noticed a new kind of arid landscape in
which the work of water was manifest. Powell (in 1875), in his main work (Exploration of
the western Colorado River, 1875), introduced the key concept of base-level, as a
forethought to the broader ideas of peneplanation. This same author established the genetic
classification of fluvial networks as consequent, antecedent and superimposed drainages.
Dutton largely contributed to the knowledge of isostasy, but also gave us detailed
descriptions of alluvial fans, and established the earliest model of parallel retreat in the
evolution of slopes in desert zones. Gilbert was certainly the main American
geomorphologist of this epoch. His excellent reports on the Geology of the Henry
Mountains (Gilbert, 1877) and on Debris transport by overland flows (in Gilbert, 1914)
constitute relevant advances in the mechanics of fluvial processes, sediment transport,
lateral fiver erosion and pediment formation, slope development and so forth. Gilbert can
be considered to be a pioneer in surface process research. His report on the History of the
Bonneville Lake (Gilbert, 1890) is one of the classics of geomorphology. In this work
Gilbert underlined the use of ancient lake-levels and the resulting lake terraces for
understanding the origin of the present-day Great Salt Lake of the State of Utah, and as a
method that could be applied to other littoral zones. Also, he pondered the isostatic
rebound produced after the evaporation of most of the large Pleistocene pluvial lakes.
Scientific contributions from other arid zones of the world were comparatively less
important (Graf, 1988). Numerous expeditions were carried out by French scientists across
the Sahara desert. They resulted in merely descriptive works, but some of them clearly
introduced the climatic change problem (i.e. Chudeau, Hammond, Gautier, Urvoy). The
Kalahari desert was extensively described by Passarge (in 1904). The Namib desert was
analysed by Kaiser (in 1921) and Little studied part of the Arab desert in 1925. The Iranian
and Thar (India) deserts were explored in the decade of the 1870s (Blanford, Oldham).
Research on the Australian desert was focused on the age of the large erosional surfaces
and their relation to the resulting products of weathering (in Juston, 1934). After a first
descriptive period, desert research in South America began with the studies on the salt lakes
of the Atacama by Frenguelli (in 1928). Also notable was the research on La Puna carded
out by Walther Penck (in 1920). At the beginning of the 20th century the prevailing idea
was that the wind was the main geomorphologic agent in the development of planation
surfaces in deserts (i.e. Hedin, Passarge, Walther). Bryan (1922), however, highlighted the
idea that wind action only introduces small modification to these planation surfaces.
Research on the work of wind carried out by Bagnold in the Egyptian desert was compiled
in his classic dissertation entitled The physics of aeolian sands and desert dunes. This work
constitutes a basis for the understanding of aeolian processes and the fundamentals of
aeolian geomorphology.
The work of the North American scientist Dana (in 1849) on the landforms developed
in the volcanic islands of the Central Pacific was the key for the initiation of tropical
geomorphology (Chorley et al., 1964). Previously, Thomson (in 1822) had pointed out
the occurrence of numerous isolated hills upstanding from the plains located in the
region of the Mozambique - Tanzania border (Douglas and Spencer, 1985a,b,c). Similar
observations were made by Bornhardt (in 1900) in eastern Africa, who introduced the term
inselberg to denote this kind of isolated hill. On the other hand, Darwin (in 1890) and
Branner (in 1896) highlighted the occurrence of thick weathering horizons, but previously
Buchanan (in 1807) had made the first scientific descriptions of laterite weathering
horizons developed in India. During the first half of the 20th century, concurring with the
exploration of new territories all around the world's climatic zones, different works on
tropical geomorphology were published (i.e. Cushing, Falconer, Grund, Hayes, Hubert,
Passarge, Sapper) (Douglas, 1978). In 1926 Thorbecke presided over the first meeting on
climatic geomorphology (Diisseldorf, Germany) in which the bases of morphoclimatic
classifications were established. The more relevant works of this period may be those
undertaken by the German scientist Freise during the decade of the 1930s in the State of
Rio de Janeiro (Brazil), and that of Sapper (in 1935) on the Geomorphology of the Humid
Tropics, which constituted the most relevant review of the time.
2.3. The impact of landscape evolution models on climatic geomorphology
The three existing models of landscape evolution, developed by Davis, Penck and King,
were proposed during different periods of the first half of the 20th century. The models of
Davis and King were particularly constrained to a specific climate, whereas the Penckian
model emphasized the role of tectonic and climatic factors together. Knowledge of all
the scientific works of these authors can lead to a better understanding of the content,
progressive development and criticisms represented by their models.
The life and work of William Morris Davis (1850-1934), emblematic geomorphol-
ogist and Professor of Geology at Harvard University, have been the subject of copious
analytic studies, some of them made with considerable extension and depth (Chorley
et al., 1973; King and Schumm, 1980). His more essential works, carried out until 1906,
were compiled in a Special Volume (Davis, 1954), but the scientific production of this
prolific scientist continued up to 1938, the date in which his last paper was published,
long after the first one published in 1880.
Davis graduated from Harvard University. After a 3-year long stay in Crrdoba,
Argentina and a journey around the world, he began his lectures at Harvard University in
1875. His research work was copious, always written in a clear and brilliant language,
and accompanied by wonderful illustrations that improved his papers (Higgins, 1975;
King, 1976a,b) and constituted an excellent tool for the teaching of geomorphology.
His best quality was the capacity for assembling most of the previous, and unconnected,
knowledge of geomorphology. Some authors of the 20th century considered him to be the
father of geomorphology (Tricart, 1965).
The major contribution of Davis was the geographical cycle, also known variously as
the erosion cycle, the normal cycle, the humid cycle and so forth (Higgins, 1975), and
it was considered to be one of the first paradigms in geomorphology (Hart, 1986). It was
the first theory on landscape development with a general agreement from the scientific
community (Thorn, 1988a,b). He published his basic model in the paper The Rivers and
Valleys of Pennsylvania, but the definitive version was published with the title The
Geographical Cycle 10 years later.
According to the ideas of Davis, a sector of a planated land surface may undergo rapid
uplift leading to a subsequent period of valley incision under conditions of tectonic
stability. The initial landscape underwent sequential stages of different relief features
denoted as youth, maturity and old age. The final result was the elaboration of apeneplain
with some isolated residual relief (Fig. 1.1) that could correspond to the initial planated
land surface of a new cycle with renewed uplift. In the explanation of the cycle, Davis
analysed the development of hillslopes and fluvial valleys, as well as the probable
interruptions of the ideal cycle, in which erosion was reactivated in relation to a new or
different base level.
Criticism of this model mainly came from the fact that the model only represented
landscape development for humid temperate regions, considered to be the "normal
climate" by Davis. He considered glacial and arid landscapes as "accidents" to the
geographical cycle. Consequently he developed "arid" (Davis, 1905) and "glacial" (Davis,
1906) versions of the cycle, the latter divided into pre-glacial, optimum and post-glacial
stages (Fig. 3.21). In spite of numerous criticisms, some of them addressed in his papers,
the concept of the Davisian cycle of erosion had a general acceptance. Higgins (1975)
pointed out 12 reasons for this general acceptance, with simple arguments such as
simplicity, applicability, rationality and so forth.
On the other hand, the work of Davis also included different fields of geomorphology,
such as glacial, coastal, arid and volcanic landscapes. His relationships with European
scientists were important, especially those maintained with his German colleagues (Alfred
and Walter Penck), where he carried out long and successive visits.
The concept of landscape cyclicity gained a wide acceptance among the different
authors of that period, and was applied to a wide variety of other geomorphologic
environments, as was the development of specific cycles for karst (Fig. 1.2), littoral,
glacial, savannah and periglacial landscape cycles of erosion (Higgins, 1975; Hart, 1986).
The ideas of Davis were well received among many German and Anglo-American
scientists, but received little acceptance in the former USSR or countries of eastern and
northem Europe, where his work went virtually unnoticed. He had enthusiastic supporters
among French investigators, including Lapparent, Baulig and Martonne. The latter
published the Trait~ de G~ographie Physique in three volumes, in 1925-1927. This work
exerted a great deal of influence on many French specialists. As mentioned previously,
Martonne was also the first author who had introduced the term "climatic geomorphology"
in 1913, (Stoddart, 1969a; Beckinsale and Chorley, 1991).
(a) (d)
(b) (e)
(c) (f)
(g)
Figure 1.1. Erosion cycle of Davis. (a) Initial stage. (b and c) Youth stage. (d and e) Maturity stage.
(f) Old age stage with development of peneplains and monadnocks. (g) Uplift and development of renewed
erosion (drawn by Raisz in Strahler (1965)).
But Davis also had a large number of detractors. Even today, some geomorphologists
consider Davis to be responsible for the more than 50 years of backwardness of
geomorphology with respect other sciences (Hart, 1986). Tricart (1965) criticized the lack
of methodology and the deductive approach contrary to the scientific inductive method.
Other mainstream disapproval was voiced in relation to ideas of processes working in
distinct climatic zones, generating different landscapes. In Germany the birthplace of
climatic geomorphology (von Richtofen, Machatschek, Passarge and Troll, as represen-
tative authors), the Davisian model was largely critiqued on account of its marked
overlooking of the climatic approach. In this way, Passarge (1931) pointed out: "The
forces involved in landscape development are of climatic nature... For instance, vegetation
depends only on the climate; but, in turn, the vegetal cover partly influences, but in a
decisive way, the landscape demolition." Later, during the decade of the 1950s, most
of the French investigators abandoned Davisian concepts, supporting the new principles of
climatic geomorphology (Cholley, Dresch, Birot, Tricart and Cailleux).
(upper left) 'Young' karst with extensive flat surfaces between dolines, much of the original surface remaining.
(upper right) 'Adolescent' karst, with larger dolines and many coalescing.
(lower left) 'Mature' karst (Cockin'tlandschaft) : the original surface has disappeared; large dolines and uvalas lie between
the Cockpithiigeln.
(lower right) 'Old' karst. Only isolated hills remain upon the residual plain.
Figure 1.2. Karsticcycle (Grund, 1914).
As we can see, the response to the Davisian model was marked from the earliest
moments among some geomorphologists, but in other schools, after an initial, and fairly
generalized acceptance, Davisian ideas were abandoned. The scientific opponents of
Davis, however, never really offered alternative models to address the defects of the
normal erosion cycle.
Only Walther Penek (1888-1923) elaborated another alternative model for landscape
development in his work Morphological Analysis of Landforms. This important
geomorphologist, son of the famous researcher on glaciers Albrecht Penck, after his PhD
dissertation in 1910 at Heidelberg, worked for 2 years in the Andean Cordillera for the
Geologic Survey of Argentina. His expertise on the geology of Central Europe, together
with that acquired in the arid active tectonic zone of the Andean region, greatly influenced
the development of his theory. In addition, his primarily geological education was
complemented by the geomorphologic knowledge of his father. His book Morphological
Analysis of Landforms was written in his convalescence during a fatal illness that ended
with his death at the age of only 35 years (Bremer, 1983). His father eventually published
the book in 1924. Their thesis was an approach to geomorphology from the point of view of
a geologist who wants to find in geology an element for the interpretation of diastrophic
history (Thornbury, 1954). In this way, we can consider Penck's thinking as one of the
precursors of the geomorphologic method in neotectonic research.
Penck based his model on the study of the sedimentary sequences flanking the uplifted
blocks of the Andes and the Alps (Chorley et al., 1973) and he deduced that the relief was
generated during an initial period of slowly increasing uplift rates followed by a period
of declining uplift that eventually terminated in tectonic stability. In contradistinction,
Davis had started his theory with a planated land surface undergoing abrupt uplift followed
by a long period of tectonic stability. The model of Penck began with the gradual uplift
of a primary surface (Primiirrumpf) promoting the progressive development of a set of
benched erosional surfaces around the margins of the primary one. This process was
thought to give place to a staircase piedmont (Piedmonttreppen) (Fig. 1.3) in which each
one of the developed benches worked as a local base-level. The relief generation was thus
controlled by the balance between crustal uplift (endogenic forces) and river incision rates
(exogenic forces). When crustal uplift waned, fiver erosion dominated and landscape
progressively evolved towards a terminal planated land surface (Endrumpf). Parallel with
the development of this model, Penck also proposed his theory of slope evolution linked to
vertical crustal movements.
Penck's ideas were never popular in Germany (Bremer, 1983). His theory was really
only known 8 years later by means of a critical and inaccurate interpretation of Davis
(Chorley, et al., 1984). This misinterpretation was probably due to the rather obscure
terminology and writing style used by Penck. The English translation of his work, in Penck
(1953) resulted in an important revival of the Penckian model.
A third model for landscape development was developed by the London-born,
South African geologist, Lester Charles King. He was instructed by Cotton, a Davisian
geomorphologist and author of three important books (1942-1944) of the mid 20th
century, and then he tried to apply the Davisian theory to South Africa. King's ideas were
influenced, like the Davisian and Penckian ones, by the particular geomorphology of his
home landscapes, in this case in South Africa characterized by the occurrence of extensive
erosion surfaces separated by large scarps.
King's ideas were first introduced in the paper Canons of landscape evolution
(King, 1953) and later in his work The Morphology of the Earth (King, 1962) where they
were extensively explained. Resembling the model of Davis, King's own theory is based
on cyclicity. Landscape evolution was supposed to begin with rapid uplift, followed by a
long period of tectonic quiescence, during which extensive pediments developed, giving
rise eventually to a pediplain. Some resistant isolated relief of diverse morphology, such
as buttes in landscapes of flat-lying sedimentary rocks, or inselbergs and bornhardts in
regions of crystalline rock can dot these extensive planated surfaces. King considered
the arid and/or savannah regions with contrasted humid-dry seasons as the "normal
morphoclimatic systems." Renewed uplift was to give rise to the beginning of a new cycle
Figure 1.3. Davis's interpretation of the Penck's piedmonttreppen on a dome (Davis, 1932).
of erosion, in which previously developed pediplains continue evolving, but the associated
scarp retreat rates diminish at distal locations in response to the development of new
pediments (Twidale, 1992). Successive erosion cycles gave place to the staircase
development of repeated pediplains at continental scale in a similar assemblage to that
illustrated by the Piedmonttreppen of Penck.
King's model was based on the process of parallel retreat of slopes and, as
aforementioned, elaborates upon the ideas of Davis and Penck (Higgins, 1975; Thorn,
1988). His ideas did not generate the relevant discussion of the predecessor models,
mainly because his theory was published at a time during which emergent methods of
process analysis and quantification were convulsing geomorphology with other new
thinking (Higgins, 1975).
Another much discussed contribution of King's was his introduction in 1962 of the
concept of cymatogeny. This would be added to the concepts of orogeny and epirogeny.
This process, apparently based upon the "Hebung" of Cloos (1939), involved the
upwarping and flexure of continental margins. Nonetheless, in opposition to the theory of
plate tectonics because of his favoured position on the idea of an expanding Earth, King
(1983) considered cymatogeny to be the dominant crustal movement.
The models of Davis, Penck and King today are the three main historical approaches to
large scale landscape development. The Davis and King proposals are based upon
cyclicity and have a clear climatic background, whereas the Penck approach represents a
continuous contest between the rates of fluvial incision and crustal uplift. All these models
are illustrated in Figure 1.4.
2.4. Process geomorphology
As another branch of science, geomorphology passed from its initial descriptive character
into a more quantitative approach, for which it was necessary to use the methods of
mathematics and physics. Although in the decades of the 1940s and 1950s some authors
analysed the dynamics of glaciers, beaches, dunes, fluvial bars and other landforms, it
was primarily during the second half of the 20th century when the revolution occurred in
the process - geomorphology paradigm. During that time geomorphologists were more
concerned with the workings of geomorphic processes than with the shape of the
landforms (Barry, 1997). Some authors considered processes analysis to be the key topic
of modem geomorphology (Hart, 1986). It has been estimated that in the year 1980, 75 per
cent of the geomorphologic research in the United Kingdom was focused on process
analyses at detailed scales (Gardner, 1983).
For many authors the revolution started with the paper by Strahler (1952), which can be
considered as the first essay on dynamic geomorphology (Higgins, 1975). Others think that
the works of Leopold and collaborators, partly synthesized in the volume Fluvial
Processes in Geomorphology (Leopold et al., 1964) are the genuine pioneers and
instigators of the implementation of quantitative process analysis.
Process analysis is included in the framework of process-response systems, in which
the process is the working geomorphic agent and the response is the resultant landform.
Process analysis may include numerous and different approaches, according to the variety of
the different existing geomorphic environments (Goudie, 1981 a,b). Some of the techniques
DAVIS
~ t TIME
3I- OF II YOUTH I
I
MATURITY I
I
OLD AGE Many times as long as youth and maturity
~/UPLIFT
_. I I. I
</ _iI_. . . . . . ~No further uplift9
tv~u II /Convexo-concave valley side slopes
_/
/
.~
III ' V/DES
....... ALTITUDE of HI
/FlOOdplaindevelOpment I . peneplain
i~AS~ L E V ~ A i LE~yFLO0~. ~ / ,~,~"~~"~-~ ,
(a) TIME
PENCK ACCORDING TO VON ENGELN

T Waxing I Uniform I Waxing development
development I development I
I I
I I
.,J--- . . . . No furthe! uplift GRAVITY SLOPE 'B(Eschung'
/ SHSLOP
: '.a,o
'Haldenhang'
I endrumpf
~/~7 1 .iT/ ...... 7 e '~ ~ ~.,_~lberg
ASE LEVEL , I
I ! ...
(b) TIME
PENCK
~T Valley deepening to obtain

maximum relief Concave slopes produced by parallel
/" I retreat of 'inclined flats'
.,- ..
endrumpf .
BASE LEVEL
.
(c) TIME
KING
A
~::)/
1 !I YOUTH Incision I ceases MATURITY I. OLD AGE9"
_~/ I max relief / _. .
~J J attained 'constant' convex ,i .. . pealplaln . . ~.
i ,,.%.~<~,,o~t~. No furtheJ stabe s ope crest ~cran I initial surface may relief and slope decline
/ o,~"~'~r "--uDlift ! forms produced / / " ~. I remainto Katestage at last stage only
/r,,~ ,,~"~,~" ]_ ALTITUDE of IJIIGHI~ST DIVIDES///gddebris slopel
~-- ~ i~ I -- r a ~ / / l l ~e fl ~i n
/ "~'~,~.40Y/f/'~ \ / \ [ / retrea,,~t/r ~ ( " ' / . !lateral. par
~ A ~ ~ 'V~/pedlmentl corraslon r e t r e a ~ j / ' . ~ . , , . ' ~ ' al form -9
biconcave
LTI~UDE of MAIN V,~.LLEYs . ~ . "' / _

[ ....... i-- , .] ~ . >
(d) TIME
Figure 1.4. Cyclic models of landscape evolution showing the relationships between elevation and time
for a fixed base-level. Also, schematic slope profile development is shown (Thornes and Brunsden, 1977).
used are very new, and in many cases not only require handmade but also complex
equipment. As pointed out by Hart (1986), in many cases "necessity is the mother of
invention." Process research is often difficult not only because of the complexity of the
instrumentation to be used, but also because of the variety of geomorphic agents working
together, as well as to our incomplete understanding of the working mechanisms of
some process (Baker and Twidale, 199 l). The study of processes also promoted a drastic
reduction in the spatial and temporal scales being analysed. For instance, research can be
focused on the analysis of micro-drainage basins and the individual slope segments. During
the last few decades efforts in this analytic field have been important. It constitutes part
of the basic subject matter on geomorphology, and the expected advances in it will give a
better understanding to landform evolution. Today, a variety of erosion rates have already
been reported for different lithologies and climates. These sets of data were compiled by
Saunders and Young (1983) and Goudie (1995). For example, on abrupt glaciated slopes
erosion rates are --- 1-4 mm/yr, slopes in temperate zones erode about 0.01-0.1 mm/yr,
in semiarid regions, --~0.1-1 mm/yr, in badlands, --~ 1-10 mm/yr, and in the abrupt slopes
of tropical forests, --- 0.1 - 1 mm/yr.
Field and laboratory research on processes requires use of some of the tools of statistics
for the adequate management of the obtained data. This fact gave rise to quantitative
geomorphology, with pioneering work by the hydrologist R. Horton on the morphometry
of drainage basins (Horton, 1945). The further development of quantitative analysis of
landforms has been controlled by the increasing capacity of computers in data management.
As a consequence of the complexity of geomorphological processes, the use of models
has become common as another important aspect of modem geomorphology. Scaled
models partly reproduce natural processes, such as the use of wind tunnels, fluvial erosion
tanks and other laboratory tests. Other models are the analogue ones, such as the ice
deformation simulations by kaolin analogues, or the mathematical models in which Ahnert
(1987a) and Kirby (in 1994) are among the more outstanding present-day geomorphol-
ogists. These two kinds of models may achieve all the complexity of process interaction
(Baker and Twidale, 1991). At the beginning of the 21st century it now appears possible
that mathematical process models iterated successively in supercomputers to simulate
deep geologic time may at last make possible the integration of timeless process models
with timebound landform evolution; something that has not been possible heretofore.
2.5. Structure of climatic geomorphology
On 22-23rd September 1926, in Dtisseldorf, Germany, researchers held a meeting on

landform development in different climatic zones, as well as on morphoclimatic classifi-
cation (Thorbecke, 1927). The results of that conference constituted the first straight-
forward and comprehensive approach to climatic geomorphology. For some of the
participants, however, their approach to problem formulation for particular climatic
regions lacked a true morphoclimatic method (Beckinsale and Chorley, 1991).
In about the middle of the 20th century, Btidel (1944, 1948) published works that he
afterwards expanded on and analysed in depth in his book on Climatic Geomorphology
(Biidel, 1977), as well as in some subsequent papers (i.e. Btidel, 1980). In these works,
morphoclimatic zones were defined as an assemblage of landscapes resulting from the
activity of the processes working in the different types of climates. In his paper of 1948, Btidel
differentiated eight different morphoclimatic zones using morphologic and climatic criteria.
These were expanded to 10 in his treatise of 1977, in which mountainous environments were
excluded. This author also proposed the term "climatogenic geomorphology" (Btidel, 1963),
defined as the science of the reconstruction of past climates from the analysis of
ancient working processes and landforms. As pointed out by Bremer (1996),
however, ideas of climatic geomorphology lead directly to climatogenic geomorphology,

because almost all the world regions are constituted by relict landforms from past climates.
Tricart and Callieux (1965) developed the concept of zonation in geomorphology and
elaborated a map of morphoclimatic regions, giving to the vegetation a potentially
dominant role in its designations. They differentiated 13 morphoclimatic regions around
the world, including an azonal mountain zone. Birot (1949a) analysed the climate
influence on slope development in bedrock of different lithologies. During the decade of
the 1950s the climatic differentiations of karst (Lehmann et al., 1954) and granitic
(Wilhelmy, 1958) landscapes were established. Peltier (1950) related a variety of
exogenic processes to mean annual temperatures and total annual precipitation and
proposed the differentiation of nine morphoclimatic regions; which representation was
later modified by Wilson (1968, 1969). This classification was criticized for its inherent
limitation from the bi-variable approach (Stoddart, 1969a). The book of Birot (1960) on
Le Cycle d'Erosion sous les Diffdrents Climats constituted a relevant contribution to the
geomorphology of arid and humid tropical zones.
2.6. The development of climatic geomorphology
Until the second half of the 20th century, geomorphology was mainly focused on the age
determination and reconstruction of the different sequential stages over which the erosive
and tectonic processes built up the present-day landscapes. This was denoted by Chorley
(1978) as the historic stage of geomorphology, in which its development was based upon
the establishment of erosion models, the analysis of denudation chronology, and studies on
structural landscapes. Later the increasing interest in process research and quantitative
analysis over different climatic zones resulted in the further development of climatic
geomorphology. This branch of geomorphology has made much progress during the last
decades, as indicated by the profusion of the published thematic volumes on different
morphoclimatic zones.
During the second half of the 20th century glacial geomorphology received relevant
advances in the mechanics of ice-flow (i.e. Hambrey, Lliboutry, Nye, Paterson,
Weertman). In the same way, research on glacial processes and landforms (i.e. Boulton,
Dreimanis, Drewry, Iverson, Linton, Menzies, Sugden), but also on fluvio-glacial
environments and glaciotectonics (i.e. Price, Menzies, Van der Wateren) is abundant.
Aside from the papers published in specialized or general scientific journals, numerous
thematic volumes have been published on tills, moraines, drumlins, glaciotectonics,
fluvio-glacial and glacio-lacustrine environments. General volumes on glacial geomor-
phology were also produced, such as those of Embleton and King (1975), Sugden and John
(1976), Menzies (1995a,b,c- 1996) and Benn and Evans (1998).
During this same period, research on periglacial geomorphology continued, but
focused on field and laboratory analysis of processes such as frost action (rock and soil
cracking, weathering, heaving, thrusting and classification), mass movement, fluvial and
wind action. There was a broad group of authors dealing with these different topics (i.e.
Clark, Corte, French, Lachenbruch, Lautridou, Mackay, Pissart, Washburn). Landform
recognition from this variety of working processes was diverse, but well known today. It
may develop in specific micro and/or meso environments that indirectly generate great
gaps in the understanding of periglacial environments (Barsch and Caine, 1984). There
are many relevant geomorphologists in the study of periglacial landforms and it is
difficult to highlight the outstanding contributors (i.e. Barsch, Biidel, Czudek, Demek,
French, Mackay, Pissart, Rapp, Tricart, Washburn). Some treatises on periglacial
geomorphology have been also published, stressing the work of Washburn (1979), which
was considered by Thorn (1992) to be the "bible" of this discipline. The thematic volume
edited by Clark (1988) is a collection of papers covering the whole periglacial geomor-
phology. Finally, the treatise of French (1996) updated the present knowledge on
periglacial research.
Extreme conditions in most of the present deserts of the world made it necessary to
carry out numerous expeditions and group research there during the last 50 years. In this
way, landscapes and landforms of these inhospitable regions have been progressively
studied. Satellite imagery, however, become a much appreciated tool for research on
arid zone geomorphology since the decade of the 1970s. It made possible large-scale
desert mapping and analysis, especially of the great sand deserts that were otherwise
intractable. The increasing interest in the knowledge of arid environments is manifest by
the recent creation of several "desert research centres" around the world. Research on
desert weathering and resultant micro-landforms has been undertaken by Cooke, Evans,
Goudie, Mabbutt and Twidale. Dorn and Oberlander have mainly done rock ("desert")
varnish studies, and Machette, Marion, Reeves, Summerfield and Watson developed
analyses of calcrete and silcrete crusts. There is also abundant and recent literature on
experimental analysis of complex process response on slope and drainage basin activities
(i.e. Abrahams, Bryan, De Ploey, Parsons, Schumm, Thornes). Fluvial geomorphology
on deserts has, as its main authors, Leopold, Wolman, Miller, Schumm and Shick.
Badland research has been focused on slope evolution and the quantification of slope
retreat (i.e. Grossman, Howard, King, Mabbutt, Oberlander, Selby, Twidale). The always
challenging topic of desert glacis and pediments has been undertaken by many different
authors (i.e. Birot, Dresch, King, Mabbutt, Mensching, Tricart, Twidale). The same can
be said of research on alluvial fans on which Blair, Bull, Denny, Harvey, Hooke and
Mcpherson, among others, can be considered as the key authors. There is also abundant
research on playa-lakes and sebkhas (i.e. Eugster, Glennie, Hardie, Krinsley, Thomas).
Aeolian erosion and sedimentation have been the subject of numerous research papers
in different scientific journals by Kocurek, Lancaster, Livingstone, Mainguet, McCauley,
Prw~, Pye, Thomas, Tsoar, Warren among others, but also thematic volumes on aeolian
geomorphology have been recently published (Pye, 1987; Pye and Tsoar, 1990;
Lancaster, 1995; Livingstone and Warren, 1996). On the other hand, the recurring
publication of thematic volumes and special issues on arid zone geomorphology is
relevant, such as those of Tricart (1969), Mabbutt (1977), Cooke et al. (1993), Abrahams
and Parsons (1994) and Thomas (1997a). These last two works are excellent special
volumes that comprise the contribution of many specialists on drylands.
Tropical geomorphology has not been a systematic research topic until very recent
times (Gupta, 1993). Some authors (Stoddart, 1969a,b; Twidale and Lageat, 1994) do not
even consider it a true morphoclimatic zone. Studies on weathering processes and laterite
development have been carried out by Berry, Birot, Goudie, Loughnan, McFarlane, Oilier,
Ruxton and Thomas, as the most relevant authors. Research on water erosion in tropical
forests has increased during the three last decades, and research was mainly focused
on experimental basins. Studies on fluvial transport were undertaken by Douglas, Spencer

and Walling. Mass movements are common in deep-weathered slopes of tropical zones as
can be inferred from the works of Brand, Lal, So and Tricart. Btidel (1957) carried out the
analysis of erosion surfaces and inselbergs in the tropics. This author developed his theory
of double planation differentiating an upper denudation surface and a lower weathering
front, in which the constant removal of the weathered horizon results in the exhumation
of the so-called etchsurfaces. The development of these ideas was mainly achieved by
Bremer, Ollier, Thomas, and Twidale. Carbonate outcrops are subject to intense
dissolution in tropical climates generating typical cone karst and tower karst landscapes
(i.e. Balazs, Lehmann, Sweeting). Books on tropical geomorphology are scarce in relation
to those dealing with other morphoclimatic zones. The works of Birot (1973) and Tricart
(1974b) offer important contributions, but it is the thematic volume edited by Thomas
(1994) that is the most relevant treatise on tropical processes and landforms, subsequent
to his earlier contribution in 1974. Another important paper is that of Faniran and Jeje
(1983) which offers a state of the art on this subject in the early 1980s. Finally, the
thematic volumes of Twidale (1982b), Gerrad (1988) and Vidal and Twidale (1998) also
consider problems on granite geomorphology closely related to tropical geomorphology.
2.7. Applications and future trends in climatic geomorphology
Despite the fact that climatic geomorphology has always played a relevant role in applied
studies, it was the decade of the 1960s when a relevant applied nature was begun as a
consequence of the increasing development of process research. One of the more
outstanding fields is the research on natural hazards derived from the dynamics of surface
processes, such as salt weathering, fluvial and aeolian erosion, flooding, mass movement
and karstic subsidence, among others. Another field with relevant development during the
last few decades is environmental geomorphology. In response to the constant population
increase and exploitation of natural resources, the physical environment is undergoing
related changes. Humans are now an important geomorphic agent, and some authors even
contemplate the term anthropic or anthropogenic geomorphology. It is necessary to live
according to nature, and it is also essential to gain a better understanding of the processes
operating in the global ecosystem. It is in this subject matter where climatic
geomorphology should play an outstanding role.
Today, geomorphology has become an encyclopaedic body of information and research
(Dury, 1978), as indicated in the Encyclopedia of Geomorphology edited by Fairbridge
(1968b). This same author opined that at the end of the 20th century, scientific production
in geomorphology would be close to 200,000 papers per year, which clearly indicates the
amazing, but expectable, development of this science.
It may be expected that research on climatic geomorphology will be focused in the
following years on: (1) constant progress in geomorphic process research; (2) increasing
interdisciplinary problem approaches; (3) more accurate advances in remote sensing data
management; (4) increasing research on paleogeomorphology in relation to global climate
prediction and (5) applied studies in environmental geomorphology (Guti6rrez, 1990).
As previously mentioned, the panorama of geomorphology has moved from the initial
genetic studies of denudational chronology to the current emphasis on surface processes
research. This fact has triggered a metamorphosis and breaking up of the discipline
(Chorley, 1978; Thorn, 1988a). As indicated by the first author, geomorphology today has
been submerged in a processes syndrome. The subjacent problem is related to translating
the detailed-scale studies of processes research to broad-scale landscape evolution
(Barsch, 1990). This is the motive as to why geomorphology should carry on research at a
wide variety of temporal and spatial scales, in an attempt to better understand the different
workings of landscape development at different scales (Chorley, 1978). In addition there is
hope that modem computer simulations will also be able to merge surface process research
with landscape evolution studies.
3. Climatic geomorphology: processes and morphoclimatic zonation
Climatic geomorphology can be defined as the discipline that identifies climatic factors
such as the intensity, frequency and duration of precipitation, frost intensity, direction
and power of wind, and it explains the development of landscapes under different climatic
conditions (Ahnert, 1996). It is noteworthy that in this recent definition the magnitude and
frequency of surface processes are taken into account. Initially, in the pioneering work of
Peltier (1950) only two climatic parameters (mean annual temperature and total annual
precipitation) were considered in establishing their relation with five key geomorphic
processes: chemical weathering, frost action, pluvial erosion, mass movement and wind
action (Fig. 1.5). For Peltier, "this set of graphics simply represents the schematic
illustration of a concept." As a whole, he distinguished two different morphogenetic
elements, which are weathering processes and the transport agents of the resultant
materials. From this perspective this author postulated nine different morphogenetic
regions (Fig. 1.6), which may be distinguished by a characteristic assemblage of geo-
morphic processes. Distinctions made by Peltier were mainly qualitative and very
subjective (Derbyshire, 1973b). A similar approach was that developed by Leopold et al.
(1964). Wilson (1968, 1969) also developed a comparable approach, but he changed the
frost action graphic for a mechanical weathering one, modified the fields for relative
intensities of geomorphic processes defined by Peltier, and distinguished six climatic
regimes denominated as climate-process systems. Wilson underlined the monthly
variation of temperature and precipitation (seasonality), as well as its influence on the
activity of geomorphic processes (Fig. 1.7).
Tanner (1961) used potential evapotranspiration instead of temperature as a climatic
variable, because this factor gives us a more realistic idea of moisture availability and, in
addition, is linked to the proportion of vegetal cover. Likewise, Tanner also assumed the
four main morphogenetic regions to be glacial, temperate, arid and selva, but also added
tundra, savannah, and semiarid regimes.
An interesting approach is that proposed by Common (1966) in his work about
landslides and morphoclimatic regions. He considered that mass movement is a part of the
denudational process and may be used to recognize different landscapes. Instead of using a
large number of involved variables, Common elaborated three global maps based upon
particular characteristics of precipitation and temperature of geomorphic interest
(i.e. precipitation regimes, intervals of total precipitation and temperature).
Chemical weathering Frost action

80 70 60 50 40 30 20 10 nches 80 70 60 50 40 30 20 10 inches
IO~ -12~ IOOF -12oC

k _
20 -7 20 -7
30 -1 30 -1
40 4 40 4
50 10 50 10
60 16 60 16
70 21 70 21
27
80 i 1 l, v v' v__v ~__" 27 80
(a) 203178152127102 76 50 25 "m (b) 20317815212710276 50 25 cm
Weathering regions Pluvial erosion

80 70 60 50 40 30 20 10 inches 80 70 60 50 40 30 20 10 inches
- i " I t '1 I/i I 1
lOOF -12~ lOOF - 1 Minimum _ -12~
20 -7 20 - -- -7
30 -1 30
40
50
60
4
10
16
40
50
60
i
- ~ /pMaximum
I
i
aa i rUs i 10
16
4
70 21 70
80 27 80 ".Minimur~! \w I I [ i - 27
(c) 203178152127102 76 50 25 cm (d) 20317815212710276 50 25 cm
Mass movement Wind action

80 70 60 50 40 30 20 10 inches 80 70 60 50 40 30 20 10 inches
~ I-'J' | " , ~l~axir~u'm I "'
IOOF -12~ lOOF -12oC
20 -7 2O -7
-1 30 _ / ~ Moderate _. -1
30
40 4 40 4
50 I-.. Maximu~n I \movement -4 10
/mas.s" l \ i~0 SO - / l" jr Maximu~
60 LI movement
/
/
/Moderate
!i
~
"1
_j 16 60 / f ~ wind ,
16
70 [ - / [ mass I 7 21 70 {w?~d' am~mn~ , ~ction ~ 21
80 ~!. - . i_ I! movement / _ _-4
I I l # i i i 27 80 "- i .... i - i , t \ v i II 1 - 27
(e) 20317815212710276 50 25 cm (f) 203178152127102 76 50 25 cm
Figure 1.5. "Peltier diagrams" illustrating the relative importance of some geomorphic processes as a
function of mean annual rainfall and temperature (Peltier, 1950).
Worldwide scale approaches to climate and vegetation controls on weathering intensity

and products have been also undertaken (Strakhov, 1967). This author recognized a clear
climatic zonation of weathering, and identified lixiviation or solution maxima in the
tropical forest and taiga-podsol zones, where precipitation is also higher. Weathering is
more intense and deeper in the tropics in response to the acceleration of the chemical
80 70 60 50 40 30 20 10 inches
I I-' I" I I ~/'I ' I l J
IOOF _ -12o0
20 - -7
30 -- -1
40 ~ S a v a n ~ 4
50 10
60 16
70 21
80 27
203 178 152 127 102 76 50 25 cm
Figure 1.6. Morphogenetic regions: mean annual rainfall and temperature (Peltier, 1950).
0-1 1 3 5 ,
8 ,
10 _ t i. _
15 =
20 ,inches
_
lOOOF-~, ~ " / SA l ~JS 9 i- I I
90 -
/ \ 35~
Trooical'~wet-dry -9 30
80 , - D e s e r ~ . . / / j / Z ' ~f- s. , . ~ ..... ..__~.
Z'=" ~ -t2s
Equatorial
70 ,,L~/"-/-~" ..... .. . ,. . .
i
Medit
M editerranea~7~.
"'--. ?-----
k , - - 20
- 15
$,r ~.
50 , ~ a , ~ Latitude
Contined~ %..~'~;,~'------ ~ "
"-------------"10
40 ~ -
. ~.-.Z. Climatic r6gimes and

5
climate-process systems
32 = - - - - ! --------- 0
"" - - 5
20 I 1 ,.~
- -10
10 tG"
J
0 ""--*'Polar - - -20
o
-20 - ~ - -40
-4o - I My ----"--4o
-60 - ~ ' -60
-i '
-8O - i i _~-~,i 8 0
t t ,;t _ t | .i v v ... , - ,'I
0 0.5 1 3 5 10 20 30 40 50 60 cm
Figure 1.7. Climatic regimes and climate-processes. Graphs show monthly rainfall and temperature
data reported by Strahler (1965). Scales are modified to emphasize arid regions. System boundaries are
given by dash dot-lines. Systems indicated by letter abbreviation: A - arid; SA - semiarid; S - selva;
HT - h u m i d - t e m p e r a t e ; P - periglacial and G - glacial (Wilson, 1968).
reactions by the higher temperatures. On the contrary, in the tundra and desert zones
weathering is small and shallow due to scarce precipitations. The north-south global
cross-section presented by Strakhov (1967) led to differentiation of a set of weathering
profiles linked to the various climatic zones (Fig. 1.8). This same author also elaborated a
global map showing the geographical distribution of weathering profiles.
Another attempt at morphoclimatic zonation is based on the regional differences in
fluvial erosion, but this parameter is greatly influenced by extra-environmental factors.
Several regression lines relating the mean annual precipitation (Langbein and Schumm,
1958; Fournier, 1960; Wilson, 1973) or mean annual runoff (Douglas, 1967; Dendy and
Bolton, 1976) with sediment production in metric ton/kmZ/yr have been proposed. Most of
the authors assume that maximum fluvial erosion occurs around 300 mm of precipitation,
whereas others highlight the occurrence of another denudational peak up to 1500 mm.
Corbel (1964) synthesized the set of erosion data for the different climatic zones, and
considered that erosion is smaller in the tropics. Fournier (1960) obtained very different
conclusions from the analysis of 78 drainage basins between 2460 and 1,060,000 km 2.
For this he correlated the climatic parameter p2/p, where p is the maximum monthly
precipitation, and P the total annual precipitation, with the suspended load transported
by the rivers. Opposite to Corbel, Fournier (1960) found that the increment of erosion
is directly linked to the increase of precipitation. This author elaborated the map of
worldwide erosion distribution presented in Figure 1.9 and illustrated the occurrence of
highest values in the humid-seasonal tropics that decrease towards equatorial and arid
regions. The main conclusions of Fournier (1960) are retained in the above-mentioned
work of Strakhov (1967).
300~. Tundra Taiga-podsol zone Steppes Semidesert and desert Savannas Tropic forest zone > 3000 Savannas
27001
2400 I
E2100 I- t ~
~ 15001-25 oration
"~ 12002- 0 E.v. a. .9. . . . . . . . . . . . . . . . . . . . . . . . . . . ~ ~ .....
,oof::
~" 600900F .... i o n 600-700 ............. ... t ~ C . . . . .
+++++
Figure 1.8. Weatheringzones in relation to latitude. 1 - Fresh rock; 2 - Rock debris with small chemical
weathering; 3 - Zone of dominant hydrolysis; 4 - Kaolinite zone; 5 - Zone of iron and alumina oxides;
6 - Soil armour or ferricrete (Strakhov, 1967).
/I>2000 [~IIII1M \ ;~YJ e ~ "

looo-2000 ~ ~ ~
600- looo ~llt~r" "~ ~ %
6o- 6oo ~llY '~ ~"
'o~ ~ ,,,
Figure 1.9. World distribution of fluvial erosion (Fournier, 1960).
Chorley et al. (1984) pointed out that a simple morphoclimatic classification may be
developed considering parameters such as mean annual temperature, mean annual
precipitation and seasonality, expressed by the temperature of the warmest month or by the
number of months with precipitation up to 50 mm (Fig. 1.10). The diagram of this figure
is composed of eight different regions that may be divided in two main groups:
'900 O
,',-~
9 /'AFID/'.//" ~ "q'~
zCe,T
/-,.#"
-'y'~ ~ / / SEMI-
\ ,o
-,,, -oO --i<,,~
._O ~ / ARID .~~ ~e'O.
..x'bv" Oo , A ' ~ ' D R Y ~ / ,,~",2 ~ z~,,.,
~'~" )/x~\CONTIN',IX ,,4/'/ ",,4"O %2".
.~'~'~\ ,/ -~,'k~NTAL~ ~. ~ / / / " TROPICAL ~ "V~Q
,k..~ ZOo / ~'~" .... ~_,,~%.~z/~ WET-DRY /
\ @a"o o ,~, LATiTUD E'~"X,~/~ TROPICAL

IUMID~ %
. . , e o
TEMPERATURE WARMEST MONTH (~ I NUMBER OF MONTHS
COLD - WARM > 50 mm RAINFALL
DRY - WET
SEASONALITY
Figure 1.10. Main present morphoclimatic zones classified according to mean annual temperature (~
mean annual precipitation (mm), mean number of wet months (> 50 mm) and mean temperature of
warmest month (~ (Chorley et al., 1984).
(a) First-order morphogenetic regions, considered as non-seasonal by Beckinsale and

Chorley (1991), comprising the glacial, arid and humid tropical regions. These are
characterized by the occurrence of non-seasonal processes with low erosion rates,
except those related to sporadic events such as surging glaciers, desert storms and
mass movements. The central zones of these non-seasonal regions persist latitudinally
during climatic changes.
(b) Second-order morphogenetic regions, considered as seasonal ones by the same
authors, comprising tropical wet-dry, semiarid, dry continental, humid mid-latitude
and periglacial regions. These have seasonal processes, occasionally of high intensity
accompanying important changes. In these regions it is possible to differentiate
warmer climates (tropical wet-dry and semiarid) where geomorphic processes differ
significantly in terms of length of the wet season; and cooler climates (dry continental,
humid mid-latitude and periglacial) whose geomorphic processes differ mainly in
respect of summer temperatures and also partly with the precipitation amounts. These
authors detail a map modifying that elaborated by Tricart and Cailleux (1965) of the
current distribution of these eight morphogenetic regional types.
Data usually managed in geomorphology are those directly provided by climatology,
such as those used by Peltier (1950), Tanner (1961), Leopold et al. (1964), Wilson (1968,
1969), Chorley et al. (1984), and others. These data can be reasonably adequate to develop
broad overviews about geomorphic processes. Whatever the case, the utilized mean
numeric values are inadequate because most of the quoted surface processes have a
discontinuous nature with relevant variations in frequency and magnitude (Ahnert, 1987b,
1996; De Ploey et al., 1991). Therefore, it seems to be necessary to use other kinds of
numerical parameters such as rainfall intensity, wind velocity, frequency and duration
of freeze and so forth. Once these numerical data have been selected, a frequency-
magnitude analysis can be performed to characterize the morphoclimatic elements. Ahnert
(1987b) proposed a magnitude-frequency index for precipitation from which can be
recognized the distribution of meteorological events of geomorphologic meaning.
De Ploey et al. (1991), using daily precipitation values, obtained a sort of quantification
of the accumulative potential erosion, which may serve to evaluate slope erosion by
laminar overland flows. This kind of approach is not only useful for the characterization of
morphoclimates but also can be applied to other branches of the science (Ahnert, 1987b).
This method cannot identify non-episodic climatic features such as seasonality; however,
it is a first step towards a more comprehensive morphoclimatology.
There is a general agreement about the concept and focus of climatic geomorphology,
but also some criticisms and objections have been made by some authors. Stoddart
(1969a) considered that is not realistic to take into account only a particular set of
factors, because clear misrepresentations may arise. Climatic factors are, of course,
important, but not the dominant ones; landscape builds from complex interactions among
climate, bedrock lithology and vegetation, and it is not possible to separate climatic
geomorphology from other geomorphologic approaches. Baker and Twidale (1991)
indicated that only the glacial and arid morphogenetic regions can be easily identified,
whereas the other climatic landscapes can hardly be recognized and defined. Twidale and
Lageat (1994) pointed out that the glacial, periglacial and arid morphoclimatic regions
represent 50 per cent of the Earth surface, and the other half is occupied by humid zones
(selva, savannah, maritime and boreal regions of Peltier). In these latter zones most of
the landscapes may be considered of doubtful designation (Stoddart, 1969b). Another
fact analysed by Twidale and Lageat (1994) is related to the climatically triggered
processes and to the shared mechanisms operating in the different climatic regions. In
this sense, landforms linked to fluvial erosion and sedimentation can be observed all over
the climatic regions (gulling, gorges, meanders, alluvial fans, braided channels, etc.).
Tafonis, gnammas and weathering pits can be recognized in the wide variety of climatic
zones, and patterned soils (grounds) in periglacial and arid zones represent a problem of
landform convergence or "equifinality" throughout different climatic environments
(Birot, 1955). In these cases, the activity of different processes, and with different
intensity, eventually give rise to similar morphologies. Consequently, these authors
admitted that the impact of climate has been largely overvalued, but did not dispute that
it is a key influence. Also they assessed critically the establishment of a humid tropical
region on the basis of the abundance of mass movement and etched surfaces, because
these morphologic features also develop in other climatic zones.
4. The zonal concept in climatic geomorphology
From a planetary perspective, the variations occurring at the lithosphere-atmosphere

contact may be considered as zonal, and they form wide bands from the poles to the
equator derived from the different amounts of both precipitation and insolation
(Derbyshire, 1976a,b). The zonal concept has been previously used, in other branches
of the physical geography, such as climatology, pedology and biogeography. The sub-
division of the continental zones into climatically featured edaphic regions was first
proposed by Dokuchayev (1883), who distinguished zonal, azonal and intrazonal soils.
The first ones are well developed soils that differ according to the ecological-climatic area
in which they were formed. The azonal soils are poorly developed, with many features
close to those present in the host rock. Finally, the intrazonal soils display good
development, but they are not well drained or are affected by salinity, and so forth.
The work of Troll (1944) on periglacial micro-landforms is considered to be an
introduction of the zonation concept in climatic geomorphology. The term zonal applies
to those processes or events in which geographical distribution occurs in latitudinal
bands parallel to the Equator, such as the equatorial forests, coral reefs and inland ice
caps. Each one of these environments constitutes a morphoclimatic area (Tricart and
Cailleux, 1965). The term azonal refers to those processes with a world-wide geographic
distribution such as fluvial, aeolian and coastal environments. Finally, the term extra-
zonal is used to denote some processes that, even though characteristic of a certain
environment, can also locally occur in other areas, as in the case of the periglacial and
glacial processes and landforms occurring in low latitude mountain systems.
The mountain areas display a characteristic climatic zonation or altitudinal gradient.
In large mountains located at low latitudes we can find, in a transect from the bottom to
the summit, the following climatic sequence: temperate, periglacial and, lastly, glacial
environments. In each altitudinal step the landscape is controlled by a specific morpho-
climatic zone (Biidel, 1948). This climatic altitudinal gradient is due to the variations
of temperature and precipitation with elevation. Temperature diminishes gradually with
altitude, the air becomes more rarified and solar radiation increases. In these conditions,
the rocks are subjected to more intense thermal effects (Tricart and Cailleux, 1965).
A height increment of 1000 m in the Alps is equivalent to a decrease of temperature
experienced during a 1000 km journey toward the North Pole. This thermal drop is
normally accompanied by altitudinal and latitudinal variations in vegetation (Bfidel, 1977,
1982). Also with increasing altitude, a characteristic succession in the morphology
and size of patterned soils commonly occurs, as reported for the Swiss Alps and
Karakoram by Furrer (1972). Similar cases have been described in other mountainous
areas, constituting a reliable proof of the altitudinal gradient. Precipitation varies with
height, displaying an initial increasing trend, which after a maximum, diminishes upwards.
All the big mountains are characterized by a more rainy and cloudy intermediate step.
Another important effect is that displayed in the north-facing areas at mid-latitudes where
snow is preserved during long periods, thereby impressing a periglacial landscape within
the altitudinal variation. In the intertropical zone, the windward areas are more humid
than those located to the leeward, which are generally dry (Tricart and Cailleux, 1965).
On the other hand, it can be pointed out that at scales of 106-107 k m 2 the variety of
landscapes is better explained by climatic differences; at scales of 104-105 km 2, landscape
variability is dominated by structural and bedrock lithology variations; and at scales of
102 k m 2 the landforms are the result of the activity of different erosive processes (Chorley
et al., 1984).
5. The main morphoclimatic zones
The differentiation of the different morphoclimatic areas must take into account the
dominant working processes in each of these areas, but it is also essential to consider
the possible occurrence of inherited Quaternary and Tertiary landforms. This interest in
climatic change was already manifest for Penck and Bruckner (in 1901-1909) in their
dissertation on the ages of the alpine glaciations. Similarly, Passarge (1904), in his work
on the Kalahari desert, found clear evidence of environments of alternating arid and
savannah conditions. Bfidel (1948) insisted on the importance of paleo-landforms, and in
1963 proposed the term climatogenic geomorphology to define the discipline of the study
of relic landforms developed under different climatic conditions, and for the deduction
of the succession of climatic environments over time. Obviously climatic change is
particularly relevant in mid-latitude areas. Also, Birot (1960) pointed out that the
influence of climatic change during Tertiary and Quaternary times extended from the
Arctic to the Equator. The distinction between modem and ancient inherited landforms
is essential to understanding landscape evolution in climatic terms, but in fact, numerous
misinterpretations of landscape development may actually arise.
The division of the Earth surface into different morphoclimatic areas, for different
reasons, has always been a complex task. There are different parts of the world for which
our knowledge of the main morphoclimatic controls is very limited (central Asia, South
America, etc.) and therefore it is difficult to trace precise boundaries between them (Tricart
and Cailleux, 1965). The increasing use of satellite imagery, however, has partially solved
this problem. On the other hand, it is necessary to keep in mind that a morphoclimatic map
does not exactly come from the total or partial overlapping of existing climatic, vegetation
and soil charts, but evidently there are clear relationships between them. A map of
morphoclimatic areas should be an original map in which the morphoclimate constitutes a
peculiar category of the natural phenomena (Tricart and Cailleux, 1965; Bfidel, 1977,
1982). It is also necessary to take into account that, except for the case of coastal
environments, the morphoclimatic boundaries are hardly sharp. There are, however,
physiographical areas, like the Black Forest, the Po plain and the Appalachian Mountains
that can be clearly differentiated (Btidel, 1977, 1982). In contrast, some transitional areas
(i.e. those between the savannas and semiarid areas on the African continent) can reach
huge dimensions. Only those areas corresponding to the cores of the different
morphoclimatic zones have no problems for their identification.
The main reviews on climatic geomorphology are those written by Btidel (1948, 1963,
1977) and Tricart and Cailleux (1965), which elaborated different maps of the postulated
morphoclimatic areas. Biidel (1948), using climatic and morphologic approaches,
divided the Earth into seven climatomorphological zones, illustrating this work with a
map of the distribution of these morphoclimatic zones in the "Old World" (Europe, Asia
and Africa). Tricart and Cailleux (1965) criticized this global zonation due to: (1) its
lack of homogeneity; (2) the segregation of three different permafrost regions; (3) the
consideration of the Mediterranean area as a single zone and (4) the use of a blurred
terminology. Bi.idel (1963) considered that the aclimatic factors (i.e. petrovariance or
change of bedrock resistance, epeirovariance or epeirogenesis, distance from base level,
topography), and human influences would play a relevant role in landscape evolution if
the Earth climate was uniform. These influences occur because the existing differences
between climatic zones are superior to these five aclimatic factors. For Btidel, on each
point on the Earth, climate determined the elemental combination of working
morphogenetic processes. From the effects produced in each climatic region, in which
these active complex morphogenetic processes operate, the Earth's surface can be
divided into five morphoclimatic zones (Fig. 1.11) not including the altitudinal climatic
gradient effects. This differentiation uses the morphologic approach as its main criteria,
which in turn is correlated to the climate. These morphoclimatic zones are denoted by
the modem climate, but for mid-latitude zones the influence of the past climates may be
relevant, because these zones undergo a slower geomorphic evolution.
The classification proposed by Tricart and Cailleux (1965) is based on two main
criteria: (1) The large climatic and biogeographic natural areas that provide the main
zonal boundaries; and (2) some subdivisions, assisted by the aforementioned set of criteria,
combined with palaeoclimatic differences. They differentiated the following areas, but
were only concerned with low-elevation zones where the altitudinal climatic gradient is
not relevant:
(1) Cold Zone.

(a) Glacier domain.
(b) Periglacial domain.
(2) Mid-latitude Forest Zone.
(a) Maritime domain with Quaternary glacial and periglacial landform heritage.
(b) Continental domain with the great influence of Quaternary and present-day ice
caps, with the possible occurrence of permafrost at depth.
(c) Mediterranean domain with a smaller influence of relict Quaternary landforms.
Equator
Subtropical zone of pediment

1:,',1 Glacier zone l~\\\\'i
and valley formation
-
11111111111
Zone of pronounced valley
formation
[L~~] Tropical zone of planation
surface formation
~ Extratropical zone of valley
formation
Figure 1.11. Present-dayclimatomorphological zones of the Earth (Btidel, 1963).
(3) Arid and subarid zones. They are divided:

(a) In relation to the moisture deficit in xerophitic bush-land steppes and deserts.
(b) In relation to the temperature of cold and warm areas.
(4) Humid intertropical zone.
(a) Savannah domain, with less dense vegetal cover, characterized by an intense
chemical weathering, but discontinuous over time, and for the occurrence of
significant overland flow events.
(b) Forest domain, with important vegetal cover and intense chemical weathering.
From these primary climatic zones Tricart and Cailleux (1965) subdivided the
Earth's surface into 12 different morphoclimatic regions (Fig. 1.12). But in this division
the morphoclimatic influences are not clearly distinguished from the climatogenic ones
(Beckinsale and Chorley, 1991). Perhaps it is also excessive to subdivide the periglacial
domain into five different morphoclimatic regions characterized by the occurrence of
modern or relict landforms (2-7). According to Stoddart (1969a) perhaps the map of these
authors gave greater importance to the landform controlling factors than to the landscape
itself. Derbyshire (1973b) indicated that vegetation is considered as a dominant geomor-
phic factor and that the terminology utilized is more pedologic and biogeographic than
climatic or geomorphologic.
The map of morphoclimatic zones presented by Btidel (1977) (Fig. 1.13) indicated that
the morphoclimatic boundaries enclose areas in which the set of active geomorphic
processes depend on a particular and uniform climate. In climatic zonation it is essential
o'O o
(M C~ @ e E
f ~m .E
~
E
o
I
i
I o
I ~"
k
I
I
I
E I
I
m m
I
~ c
~~'~
5
0 ~ ~
I
(-
"~.--
-6 x:
~ ~~
t'--
! . c
1 .. 9- - r" ~
9 I ,,_ E :~ '-'
I
4. .
..-
t. . j~,
,.
I
.//
1i~ , ~
I
I
1
6
- 6 ,o,oo"/~ ' ~ .' ~ "

-..i-.-.-i i . i . . , i . . . .. .. . . . . . . "~. . ~.~~-~ 6o~
.'->91',%
9 " .. .. " f i i i i i i~_~_..._._a~"
--4~ . . . . '":..:.:...:-:.....- . . . . . . . . . . . . . k~-,,, ,~
..~.m
Z,
9 . . . . ' ,
- 20~ ~ / ~ ~ I ~'" ~
. . . . . . . - --
;2. . . . . . ---------- __ . . . . . . - - - _ - -
- 60 ~ 60~
.- 80 ~ ' ~3000km80~
120~ 80 ~ 40 ~ 0o 40 ~ 80 ~ 120 ~ 160OE Entwurf 9J.B0del 1974
Glacial zone (and immediately Subtropic zone of mixed relief

adjacent area) development, monsoonal region
Subpolar zone of excessive Winter cold arid zone of surface transformation,
valley-cutting largely through pediments and glacis
Taiga valley-cutting zone, Warm arid zone of surface preservation
in the permafrost region and traditionally continued development,
largely through fluvio-eolian sandplains
Ectropic zone of retarded Peritropical zone of excessive planation
valley-cutting
[ ~ ubtropic zone of mixed relief Innter tropical zone of partial planation
development, etesian region
Figure 1.13. Morphoclimatic zones of the present, excluding the high mountains (Btidel, 1977).
to consider the central cores of the suspected zones, because their limits are usually not
well defined. On the other hand, only the low and middle elevation zones are represented,
whereas the high mountains have been misplaced. The establishment of different
morphoclimatic regions is very complex because it is necessary to distinguish a large and
complex set of processes implicated in its formation. It is also sometimes very difficult
to differentiate the modern landforms from the relict ones (from the upper Cretaceous to
the Holocene). If we understand the present active landscapes in each morphoclimatic zone,
we can recognize the relict landforms in other zones. The present active processes operating
in mid-latitude areas are exceptionally weak in their workings, and in these circumstances,
95 per cent of the relief of these regions is constituted by relict landforms (Bidel, 1977).
The most significant features of the main morphoclimatic areas have been collected
from the work of Bidel and from the great review on this author's ideas, carried out by
Kiewietdejonge (1984).
(1) Glacier zone, to which Bidel did not dedicate a special interest. In this zone the
mechanisms operating in landscape development are more easily studied in post-
glacial regions than in the presently active glacier regions.
(2, 3) Periglacial zone of pronounced valley-cutting. This corresponds to the subpolar and
also solifluction regions of Biidel (1948), due to the importance of mass movements
on periglacial slopes. Its northern area is limited by glaciers and the southern one by
the solifluction boundary of the tree-cover or permafrost termination. The forest
limits usually coincide with the solifluction one, whereas the permafrost boundary
can expand over lower latitudes. The upper horizon of the ground is structured,
thus generating patterned ground. One of the main problems in these zones is rapid
fluvial incision. Slopes may develop by parallel retreat, generating tripartite
hillslopes.
(4) Ectropic zone of retarded valley-cutting. The most diverse and possibly the most
complex region, with the occurrence of a large number of different landforms.
Some 95 per cent of the landscape in the European sector of this area is constituted
by inherited landforms. The stepped planation surfaces (Piedmonttrepen of
W. Penck) are very common. Glacis surfaces and not very deep, wide valleys
may be characteristic. In those areas subjected to ancient glacial erosion, extensive
landscapes of hills and lakes are developed. Finally, an important sheet-cover of
loess derived from materials deposited during previous cold periods commonly
coats the ground. This, together with the following region, are the most important
ones for the study of climatogenic geomorphology.
(5, 6) Subtropical zone ofmixed relief development. This comprises the Mediterranean and
monsoonal regions which hold different morphological and climatic characteristics.
Bi.idel did not develop descriptions of the monsoonal regions, and instead just
adjusted his observations for the Mediterranean areas. This zone is a transitional area
for which its evolution has been affected by tropical, periglacial, glacial, temperate
humid and arid climates. In addition in this climatic type it is necessary to consider
human influences, which have an extraordinary impact on the occurrence of fluvial
erosion. The morphologic result is the assemblage of relict and modern landforms,
commonly arranged in an altitudinal zonation. A considerable development of
Tertiary planation surfaces is recognized in associated calcareous and in karstic
landforms. In the fluvial valleys and intramontane basins the development of glacis is
common in relation to the occurrence of fluvial terrace systems.
(7, 8) Arid zone of surface preservation and transformation through pediments and glacis.
This area includes that of winter-cold deserts and tropical arid zones. In this
latter zone a relevant intensity of wetting and drying and salt weathering processes
occurs. Within this zone erosion plains with inselbergs, extensive hamadas and
stone deserts are the usual features. In areas of limited relief the alluvial materials
of glacis and terraces usually develop calcrete crusts. These alluvial materials
constitute the main source area for the adjacent sand deserts. The winter-cold arid
zones constitute a transformation area rather than a region for later development of
planation surfaces. Frost cracking is the dominant weathering process. The most
representative landforms are the glacis systems.
(9) Peritropical zone (humid-dry) of excessive planation. In this region thick horizons
of chemical weathering commonly develop, which can reach several hundred
meters thick, with the eventual development of laterites. The formation of extensive
plains can be explained by the double planation mechanism of Bfidel (1957) for
tectonically stable areas, in which etchplains and inselbergs are characteristic.
(10) Equatorial zone ofpartial planation. This is not a very well studied zone due to the
relative difficulties of research because of its dense tree cover. Chemical weathering
is more intense than in the other areas. Planation surfaces and inselbergs that
originated in prior humid-dry climates of the Tertiary occur in those regions that
have limited development of preceding climate zonation. In the central areas of these
zones planation surfaces undergo little reworking because the formative mechanisms
have greatly decreased in space and time.
These classifications of analysed morphoclimatic areas reveal an important variation
within the different main climatic zones: Btidel (1948) 7; Peltier (1950) 9; Biidel (1963) 5;
Tricart and Cailleux (1965) 13; Wilson (1968) 6; Bfidel (1977) 10 and Chorley et al.
(1984) 8. These modifications follow the different approaches used by the different authors
to carry out the subdivision of the Earth's surface in different morphoclimatic areas.
The monographs published on different morphoclimatic zones have been focused in
the following four categories: glacial, periglacial, arid and tropical humid regions. The
temperate regions have not been the object of a joint treatment due to the large number of
presently existent relict landforms. Evidently, the development of the published reviews
on these four climatic regions takes into account more detailed differentiations within the
master climatic areas.
Chapter 2
Glaciers
1. Introduction
Since Precambrian times, several ice ages have occurred in Earth history and have
been printed in the geologic registry by erosive and sedimentary features (Eyles, 1993)
(Fig. 2.1). Within this general perspective, the Quaternary encompasses a glacial period
that began to develop fundamentally in upper Tertiary. The remaining ice accumulations
at the present time are left occupying only --~ 10 per cent of the terrestrial surface above sea
level, whereas in the last glacial maximum about 18,000 years ago, they covered almost a
third of the continental areas (Flint, 1971; Sugden and John, 1976). The activity of the
glacier masses is, therefore, reflected by different erosive and depositional landform
modelling as an important activity of the planet.
Investigations on the activity of past and present glaciers occur in different fashions.
Works on glacial geology are carried out by Quaternary researchers, who fundamentally
worry about glacial stratigraphy and chronology; whereas geomorphologists analyse
glacier processes and model the results of ice mass performance, as well as its
evolution. Glaciologists are interested in the study of the nature, physical behaviour, and
work done by glaciers (Lliboutry, 1965). In most cases, glaciologic investigations and
those carried out by glacial geomorphologists show a clear disconnection, although
there seems to be some overlap (Sugden and John, 1976). This union of methods is
necessary because glaciers constitute a means of rapid activity, with modifications in
days or months, and are therefore suitable for studying present-day processes and the
resulting forms.
In the past few decades a new field of work related to glacial geomorphology
has been opened within planetary geology. Thus on Mars, snow and ice sheets covering
30 per cent of its surface are recognized as occurring during winter and being reduced
to residual ice caps in summer, occupying a single 1 per cent. Also, the interpretation
of some Martian channels as originating by valley glacier activity has been expressed.
Some investigators establish for this planet a climatic change similar to that of Earth's
during upper Cenozoic (Cutts et al., 1979; Baker, 1981a; Carr, 1981). These
investigations of planetary geomorphology can provide important data for a greater
understanding of our glacial ages.
2. Past and present extension of glaciers
The present-day glacial ice occupies an approximate surface of 15.8 million km 2

(Table 2.1), or about 10 per cent of the total emergent terrestrial surface above sea level.
m 9 m
~-.~, Neogene m"
:~. . Palaeogene
100 o Cretaceous 9 9
"ON 9 , ,
o Jurassic ? ,2
200
~: Triassic 9 ,9
~. Permian
300
5N iferous
I m
m
m 9 ,,
o , .Devonian m
400
~_ ~. Silurian m , m
m 9
500 ~
III I Ordovician
, ,
m 9
z
u.i Cambrian
(f)
i.iJ
n- 600
o.
iJJ
n--
Vendian m
m
,,
mmm.
O
II
700
" I
iaJ
o3
Sturtian
m
I I!
rr 800
<
I.iJ
>..
u_ 900
0 .O
m
oo O
N
z O
o ~ooo Riphean
..J
O
_J
~E n~
". m
O
1500
Animikean
2000
m m
Huronian mm
m m
2500
3000
< <
C--- -<. o o
4600 oc ~ w w
I- u ~
co n- < ,~
LL
< Z
< G) G)
< w :~ o5
Figure 2.1. Ice ages in Earth history (Eyles, 1993).
Only 3 per cent of glaciers are outside both the great Antarctic and G r e e n l a n d ice sheets
that constitute the largest fresh water reserve of the h y d r o s p h e r e (Lvovitch, 1967). The
large polar caps cover an area of 13.6 million k m 2, with an average thickness of 2.2 k m
and a m a x i m u m close to 5 km. The water e q u i v a l e n t v o l u m e of ice of this ice sheet is
Glaciers 37
Table 2.1. Present area occupied by ice (World Glacier Monitoring Service, 1989) and maximum
expansion during the Pleistocene (after data elaborated by Flint, 1971).
Region Present area Maximum expansion

(km 2) during Pleistocene (km 2)
Antarctica 13,593,310 13,800,000

Greenland 1,726,400 2,295,300
Mexico, United States, 276,100 16,217,091
Canada and Alaska
Fennoscandian 51,046 7,169,145
Africa 10 1900
Asia (Russia included) 185,211 3,951,000
Alps and Pyrenees 2921 39,000
South America 25,908 870,000
Australasian 860 30,000
Total 15,861,766 44,373,436
considered to be --~30.1 million km 3. If it melts completely it would produce a rise of sea

level of 59 m. The ice sheet of Greenland occupies a surface of 1.7 million km 2, with an
average thickness of 1.6 km, a maximum of 3.4 km, and a volume of 2.4 million km 3
(Donn et al., 1962; Flint, 1971; Sugden and John, 1976).
The remaining ice accumulations do not surpass a half million km 2 and are scattered all
over the planet, although they are fundamentally concentrated in the northern hemisphere
and in mountains of middle and low latitudes (Alps, Karakoram, Himalaya). The extension
of glaciers in Africa is practically negligible in contrast to the ice accumulations in the
Andes, which stands out by its extension. The area in the world occupied by glaciers has
recently been calculated by the World Glacier Monitoring Service in 1989 (see Hambrey,
1994). The total surface corresponds to a rather overly precise estimation of
--~ 15,861,766 k m 2.
In their maximal extension during the Pleistocene, ice sheets covered an area of more
than 44 million k m 2 (Flint, 1971), representing some 30 per cent of the emerged lands. Its
volume on the terrestrial surface is considered to be about 100 million km 3 (Donn et al.,
1962). In Table 2.1 the great difference between the area occupied by present glaciers,
corresponding to an interglacial age, and the surface occupied during the last Quaternary
glacial maximum can be observed.
The Antarctica and Greenland ice sheets have undergone a relatively small fluctuation,
whereas, for example, the Fennoscandian and Laurentian complexes have experienced
enormous advances and retreats. The last advance reached dimensions close to that of
present-day Antarctica and extended below latitudes of 40~ (Fig. 2.2).
3. Mass balances of glaciers: accumulation and ablation
The mass balance in a glacier deals with the gain and loss of snow and ice and is normally
measured by its equivalent in water (amount of resulting melt water). The term
Figure 2.2. Maximal extension of Pleistocene ice sheets in the Northern Hemisphere (after Antevs and
Flint; in Holmes (1944)).
accumulation refers to all those processes by which material is added to the glacier.
Precipitation is the most important one, although the feeding by avalanches to valleys can
sometimes be significant. Wind action plays a secondary role with its effects on snow-mass
redistribution.
Ablation alludes to the set of processes by which part of the glacier mass is lost. The
main one, melting, is influenced by solar radiation, rain, air temperature, cloud cover,
detritus amount, and other factors (Sharp, 1988). Fine rocky particles, with a greater
capacity for solar radiation absorption, penetrate when the snow around the edges melts
generating small cylindrical conduits. On the contrary, thick detrital layers isolate the
underlying ice from melting, which produces gravel knolls containing ice. If the detritus
mantle has an irregular superficial disposition, a topography of knolls and depressions
can be produced by differential melting. In the case of the presence of great blocks, the
resulting morphology will be of pedestals or glacial tables (Fig. 2.3) in which the block is
based on underlying ice. These tables frequently appear tilted until they finally fall, due to
a greater melting of the ice oriented to the sun. Sometimes differential melting produces
numerous and deep depressions, denoted as karstic glacial topography by its similarity
to the landscape of limestone areas (Clayton, 1964). On the other hand, other important
ablation processes in glaciers that lose their mass to lakes or seas is the generation of great
Glaciers 39
Figure 2.3. Glacial table on Godwin Austen Glacier (Karakoram, Pakistan). K2 (8611 m), the second
highest summit on Earth, is visible in the background. Photo: J. L6pez-Martfnez.
ice blocks or icebergs floating in the water. This process is influenced by the degree of
fracturing of the ice mass.
Any glacier is formed by two parts: the superior or upper one, where feeding or
alimentation is greater than loss by melting, thus constituting the accumulation zone, and
the inferior or lower one, with an opposite balance, denoted as the ablation zone. In
between the two zones is the equilibrium line where the balance is zero (Mtiller, 1962).
The difference between accumulation and ablation in a year for the whole glacier
constitutes the net balance (Andrews, 1975; Paterson, 1994) (Fig. 2.4). If the balance is
positive the glacier has experienced a gain of snow and ice (annual net accumulation)
and the opposite if it is negative (annual net ablation). A balance equal to zero indicates
an equilibrium between accumulation and ablation. The winter balance is usually
positive and the summer balance negative. The difference among them is the net
balance. Balances are calculated by means of field measures or by aerial photography
and hydrological methods (Benn and Evans, 1998). Few mass balances for glaciers
exist and a most complete one is the Storglaci~iren in Sweden. In a 28 years period, 5
years resulted in positive balance, one with balance zero and in the rest ablation
dominated (Ostrem et al., 1973; in Sugden and John (1976)) (Fig. 2.5).
"•'Iv
~)
1 ,--
1
,_ ~
9~
Eo~
Accumulation (c) I "
9
~.~-
.~F "~ ..... s -'~
~8
= [ = [ o~ v~ / , ./"
/ - Balance(b) (~ ~, l E ,.-
o o
j/ 2 ~ X / ~s
1 .- i
o - ~ , " , , s
i ='',, . . . . . . . . . . ......
,~Dlat~on [a) ,,'Summer
I ..... n "~
r Winter season / ~ -~
I - "d ~',. -Q
-~1I E_~
E o ~ Balance year X
~ z
Figure 2.4. Graphical explanation of the mass balances in glaciers (in Andrews, 1975).
!
I1970-71
1965-66
1960-61
1955-56
1950-51 ! : ! i I
1945-46 "~,I1 !
300cm 200 100 0 0 100 200cm

ablation (-) accumulation (+)
Figure 2.5. Mass balance for the period 1945-1973 in Storglaci~iren (Sweden) in which the annual
accumulation, ablation and net balance are expressed (Ostrem et al., 1973).
Glaciers 41
4. S n o w - i c e transformation
Glaciers are developed in situations in which snow accumulation exceeds melting.

Suitable climatic conditions are required, as well as topographic form providing ideal
places for their storage (Price, 1973).
Snow fallen in the accumulation zone undergoes deep changes until it finally
transforms into glacier ice. The grains constituting this ice are made up of crystals that vary
from millimetre to decimetre sizes and have been formed by numerous snowflakes.
The newly fallen snow has a very low density of sometimes 0.05, a great porosity
(95 per cent) and sometimes can be considered as a wind-deposited sediment (Fig. 2.6).
This fresh snow changes in a short time, through pressure induced melting and refreezing,
to granular or old snow with a density of 0.3 formed by loose spherical grains of snow with
great permeability. Transformation through minor melting and refreezing continues with
an increase of density by the modification of grains, loss of porous spaces and tightness of
the packing. When density is 0.4, the firn state is reached with 50 per cent porosity and a
grain size of one to several millimetres. As compaction increases, pores between grains get
sealed andfirn is transformed into glacier ice when density equals 0.80-0.85. This mass is
impervious, the existing air is contained as bubbles and any increase of density is due to its
compression. This snow-ice transformation is accompanied by displacements between
crystals, changes in size, form, and internal deformation (Paterson, 1994).
The speed of the snow-ice transformation varies from one place to another, depending
on the sensitivity of the process to the temperature and the speed of snow accumulation.
Figure 2.6. Wind-bornesnow, Heritage Range, Antarctica, about 80~ Photo: J. L6pez-Mart/nez.
_ '~".-.x Upper Seward Glacier
x~,x
"Y~'Y''"K
S i t e 2, G r e e n l a n d
40-
E
.E
80-
I . I . I I I ..
120 400 500 600 700 800 900
Density, kg/m3
Figure 2.7. Variation of firn density with depth in the temperate Upper Seward Glacier (Yukon) and in
the Greenland ice sheet (Paterson, 1994).
The presence of melting water in summer accelerates the transformation. The curves
of Figure 2.7 (Paterson, 1994) show the variation of firn density as a function of depth for a
dry snow glacier in Greenland and for a wet snow one in the Yukon. In the first one, firn
turns to ice at 66 m depth, requiting a time of more than 100 years, whereas in the second
one it happens at 13 m in a 3 - 5 years interval.
5. Classification of glaciers
Numerous classifications for ice accumulations have been proposed, but the most used,
because they allow easy differentiation, are those based on the internal regime of the
temperature of glaciers and on their geometry.
5.1. Thermal classification
Pure water is transformed into ice at 0~ under a pressure of one atmosphere, but a lower
temperature is required when pressures are important. Thus, for example, at the base of a
glacier of 1500 m of thickness the melting point is - 1~ It is possible thus to differentiate
cold ice in which temperature is below the melting point and warm ice in which it is so
close to fusion that it can contain water (Ahlmann, 1935).
Cold ice is found when surface temperatures of the glacier are very low in winter
and with little or no melting in summer, as in Antarctica, Greenland and the elevated zones
Glaciers 43
of great mountains. In Antarctica air temperatures of - 8 9 . 2 ~ have been recorded;

warmer temperatures of - 30~ are reached as well, but temperatures increase with depth
as a result of the geothermal heat and of the internal deformation of the ice mass in the
glacier flow (Sugden, 1982).
Warm ice is formed when sufficient heat to produce melting is available. During
summer, melting of surface ice is common. This water percolates into the ice mass and can
freeze again, releasing its latent heat and by each gram of refrozen water elevating 160 g of
ice by 1~ This can be one of the most important heat sources. In these circumstances the
whole warm-ice glacier can be formed by ice at its pressure-melting point. At the ice base
geothermal heat is usually sufficient to elevate the temperature of the ice to its melting
point. This warm ice can also be at the bottom of glaciers whose superficial layers are of
cold ice (Budd et al., 1970).
These ice types allow differentiation of polar, or cold-based, and temperate or warm-
based glaciers. These classifications are very simplistic, because the thermal conditions
of glaciers can vary in space and time (Sugden, 1977; Paterson, 1994), so that polythermal
ice is becoming increasingly recognized. Antarctica is dominantly a polar or cold-based
glacier, but some parts have warm ice at their base. In temperate glaciers the transfor-
mation of firn into ice is faster than in polar glaciers. These differences in ice temperature
are of extraordinary importance in glacial geomorphology and mainly related to the
presence of a thin water film in temperate glaciers lubricating the bed of the glacier and
facilitating its movement. At the same time this subglacial water facilitates quite important
erosive and sedimentation work.
5.2. Morphologic classification
Morphologic classifications are based upon geometry, position and size of the ice masses.
Some ice accumulations are restricted by the topography, whereas others do not present
any type of confinement. Continental ice sheets are represented by those of Antarctica
(Fig. 2.8) and Greenland. They show a domal form with a convex surface produced by
the ice flow. In the centre, ice accumulation thickness is larger and towards the margins
the surface slopes smoothly, increasing its gradient progressively and with diminishing
thickness. These ice sheets preserve or fossilize the underlying topography, although in
some cases the relief of the bedrock substrate exceeds that of the ice mass, and the rock
outcrops protrude through the ice as nunataks (Fig. 2.9).
Ice shelves (Fig. 2.10) are great masses that extend out into the sea > 500,000 km 2,
like the Ross and Filchner-Ronne ice shelves in Antarctica that occupy great bays and
have extensions corresponding to 7 per cent of the ice sheet. The ice thickness can reach
up to 200 m and the cliffs up to 30 m with the loosening of icebergs (Fig. 2.11) and
melting of its base. These floating ice masses contain particles of different sizes that fall
to the bottom of the sea when the ice melts (dropstones) (Fig. 2.12). On the ocean
surface remains a floating mass of split and rather thinner ice, denoted as pack ice
(Fig. 2.13). If the size of ice domes is smaller than 50,000 km 2 (Sugden and John, 1976)
they are denoted as mountain, plateau or plain ice caps, according to their topographic
position. The first is located on a plateau zone (Fig. 2.14), being the ice cap of
Figure 2.9. Nunatak in the Sentinel Range, Ellsworth Mountains, Antarctica. Photo: J. L6pez-Martfnez.
Glaciers 45
Figure 2.11. Iceberg in Bransfield Strait, Western Antarctica. Photo: J. L6pez-Martfnez.

Figure 2.13. Fractured pack ice in the proximity of the Antarctic Peninsula. Some seals are visible on the
ice blocks. Photo: J. L6pez-Martfnez.
Glaciers 47
Figure 2.14. Plateau ice cap of Myrdalsj6kull (Iceland). Photo: F. Guti6rrez.
Vatnaj6kull, in Iceland, the most characteristic example. The plain ice caps are located
in areas of little relief, such as the Barnes ice cap on Baffin Island in the Arctic.
Cirque glaciers occur in the valley heads of mountainous areas (Fig. 2.15). They are
small ice masses located in amphitheatres, generally of steep walls, hectometric dimen-
sions and thicknesses of tens of metres. During a glaciation period they are the first to be
developed and last ones to disappear.
Valley glaciers are laterally confined by rocky walls. If their feeding comes from
cirques we are speaking of valley glaciers of the alpine type (Embleton and King, 1975a)
(Fig. 2.16). They flow radially from the great massifs and can be fed by secondary valley
glaciers. A clear hierarchy is given by the relation between valley width and number of
tributaries. The glacier mass is long, narrow, and generally finishes in a single ice tongue
or terminus. The longitudinal slopes are variable. Some glaciers exceed 100 km in
length. At the present time they are common in high mountains. In the peripheral zone of
the icecaps the ice masses can escape from the margins giving origin to valley glaciers or
outlet glaciers (Figs 2.17 and 2.18), which are like appendices arising from the edge of
ice caps. When the front of valley glaciers is constituted by an extensive plain the ice
spreads out considerably over it, increasing in width and producing a piedmont glacier.
The most classical example is the Malaspina Glacier in Alaska extending as a wide lobe
in the coastal plains of the Pacific. It constitutes the ablation zone of the lower Seward
Glacier, being 600 m thick and occupying a great depression that reaches 250 m below
sea level (Benn and Evans, 1998). Its compressive flow produces an intense deformation
of the debris bands and medial moraines (Sharp, 1958) (Fig. 2.19).
Figure 2.16. Valley glacier of the alpine type. Spegazzini Glacier, Patagonia, Argentina. Photo:
C. Sancho.
Glaciers 49
Figure 2.18. Outlet glacier and pack ice in Whisky Bay, James Ross Island, to the East of Antarctic
Peninsula. Photo: A. Martfn-Serrano.
~:o..
MALASPINA
GLACIER
'ii ~F
___.___.-..---
.'~ 9
f~ * o Q"
:SZ--..-.-.'" ...<
;? ....;..:.-
Bay
f
Forest on debris-covered Ice Point Manby
Figure 2.19. Map showing similar folds developed in the ice and moraines of the Malaspina Glacier in
Alaska (Sharp, 1958).
6. Movement of glaciers
Ice is a crystalline solid that flows easily under the pull of gravity. This movement
implies a continuous transference of material from the accumulation to the ablation
zone. Two types of processes can be distinguished in this flow: internal deformation and
basal slipping (Sharp, 1988) (Fig. 2.20). The first is a creeping flow resulting from the
application of force over a long period. Intergranular sliding, re-crystallization and
sliding along crystal network planes in the ice crystals are the mechanisms of this flow
(Weertman, 1983).
The basal slipping indicates the displacement of the ice mass on its bed (Weertman,
1957). This movement on the bottom also implies a process of basal plastic flow in
which the ice flows over and around greater obstacles. These advances can take place
whatever the temperature of the ice is (Fig. 2.21). The other mechanism is denoted
sliding by refreezing. The bottom of the glacier presents a rough surface with small
thresholds and depressions of centimetre to metric scale. Melting by increase of pressure
in the basal ice can take place if the glacier is of the temperate type, near its pressure-
melting point, and a refreezing or regelation, when the water pressure is reduced. If the
protuberances are very small, latent heat transference can take place and facilitate new
melting. Refreezing layers of clear ice, laminated and of a few centimetres thickness, can
occur at the base of the glacier (Kamb and LaChapelle, 1964).
Glaciers 51
Total Surface M o v e m e n t
o * ~ 9
9. . . . . . / .
. 9 9 Basal Slip " 9 9 " 9
9 .
9 9 . . 9 9 9 . InternalFlow . ". . 9" 9
9 9 9 . 9 9 ., ,, 9 9 . 9 . . 9 9
9 . 9 9 9 . . 9 " 9 ,, * . 9
9 9 9 . 9 . m
9 . . 9 . 9 9 o " 9 ,, ., . 9 9 9 ,, "
9 " 9 " " 9 9 9 9 9 9 9 " 9 s

9 . 9 9 . 9 . ,, . .
9 9 9 . . . " . . " " ". . . " . Ice 9 .
9 9 9 .
9 . 9 9 9 . 9 9 . 9 9 9 * o
9 9 9 9 9 9 9 " . 9 ~ 9
.~" 9 .
9 Basal Slip . . 9 .
'. ' 9 .
- . . .
" 9 .
/ I,,i ~/ I ~ . ? ~ / i /~
,<, " - - - ' " -, " . / . . ' , , / \ , > , x v
.
. .
9
. .
.
9
* ,
9
9 .
"=+
9
~'-"~-'-:;.q-S'
.
"';' """';'"'"<:~''-
+'/ 9
" -
.
"
. *
"
'..,.
- .,, :-, ,- -,,, , : , -, ,. . . ,.';.
, . ' ,.~ . . ,c,.'~,',.-,.
-..
t .<S/.,'f /~,,/ ', "~,",,-r,,J :-~ ~ ;; ~ " ~/3",,~" { "7,,"~"~~~1/>~, .
,~,' i"-':/r ' "~ Rock '-, ~ , ' , , - , . / ~ , , , , \ , > ; ~ ~ ~ ' '
;,'i ,:-'- ,~.'-,'- i' ''t "%#1 " / ~ " \
"f,
.~','
x~.t.l~/|--.t x ,I.~.1. ~,,\..~, \ ,/~,--/.1~..I + .,Ix . / . , , t X . , X ~ , x ./,~l.l,.I _~ , i x . I . / l \ . , ~ . X I~-.,", I
Figure 2.20. Longitudinal section of a glacier that indicates how a vertical sounding gets deformed over
time. Total movement AA t measured on the surface is the sum of internal flow and basal slipping (Sharp,
1988).
Polar glaciers show little basal slipping because they are frozen to the glacier bed,
although a relatively fast sliding next to the i c e - r o c k interface can exist. In temperate
glaciers the basal slipping is favoured by the presence of a water film at the bottom, which
reduces the friction between ice and rock. If the substrate of a temperate glacier is
composed of unconsolidated sediments, such as basal tills, these materials can be saturated
with water to become deformable tills and, therefore, the speed of basal slipping increases
considerably.
The surface speeds of a glacier vary in time and space. The speeds throughout a glacier
fluctuate between 3 and 300 m/yr, being able to reach values of 1 - 2 km/yr in steep-walled
areas. The cross-sectional speeds are maximal in the centre and undergo a fast reduction
towards the margins as a result of the wall friction. The edge values fluctuate between 10
and 65 per cent of the central maximal dimension. These calculations of the superficial
il [
(b)
~'~'~" ~ ~ REGELATIONICE
~"1~1~ ,
Figure2.21. Influenceof bed irregularities on the ice flow. (a) Process of basal plastic flow, sight in plan
view (Weertman, 1957). (b) Sliding mechanism by refreezing, in section (Kamb and LaChapelle, 1964).
speeds of a glacier are obtained by installing stakes in the ice and locating a theodolite on
fixed points of the rocky margin. In order to obtain the speed at depth, borings to the base
are carried out (Fig. 2.20) and later a flexible tube is introduced into the conduit. By means
of a clinometer the movement can be followed. The speed is higher in the upper part than
in the lower.
Some valley glaciers experience fast and abnormal flows in which the ice mass
moves as a wave at speeds of up to one hundred times the average value and with
displacements up to 11 km. They are known as glaciers with spasmodic flows or surging
glaciers (Fig. 2.22). Vibrations and intense noises accompany their sudden movements
(Meir and Post, 1969; Sharp, 1988; Clarke, 1991; Dowdeswell et al., 1991; Evans,
2004a). These swellings can have a regularity of a few years to several centuries. In the
glacier they cause the presence of deformed chaotic surfaces, intensely sheared margins,
great changes in its thickness and strongly folded medial moraines. The intense shear at
the margins of surging glaciers can produce foliation in the ice (Pfeffer, 1992). In
Svalbard 504 surge-type glaciers have been described. These glaciers result from a
polythermal regime and fine-grained deformable beds (Jiskoot et al., 2000). High
pressures of the subglacial water motivate these fast flows and can produce catastrophic
advances. The surge outbreak can also be associated with seismic activity and important
modifications in precipitation.
Stress regimes (Fig. 2.23) vary throughout the length of the glacier and compressive
flow occurs as the speed of the glacier reduces (Nye, 1952), in such places as below
steep sections of the glacier. The sliding planes are curved in an ascending direction and
Glaciers 53
0 , , , , 5km
1963
I
~p'~d~o~
I
i
1964
~.,
,966 iikes.
Figure 2.22. Evolution of the spasmodic surge flow of glacier Tikke (British Columbia) between 1963
and 1966, in which deformation of the moraine surface is observed (Meir and Post, 1969).
the detritus can be transported with them up to the glacier surface. In contrast, extending
flow is located in zones, commonly with steeper gradients as at icefalls in which the ice
speed increases and the shear planes are curved downwards until becoming tangent to the
bed. This flow predominates through the equilibrium line.
7. Structures of glaciers
The deformation suffered by the ice masses as a result of its movement gives rise to
different structure types, similar to those observed on deformed rocks. The analysis
of glacial deformation is easier than that of rocks, however, because they are affected
solely by the action of gravity (Menzies, 1995a) and rocks are affected by many more
processes.
~,~ Glacier ~~~

> movement .... ~"
(a)
t~~..-)...-~..~ Equilibriumline
I
si ', ; :: T7"~%~-~,--_ Glaciersurface i
I ~ ' ' li t! 'l "1 li !

I I I | e., I , I | I 9 " t
__ ~ llllllllllllllII!!llllllllllll
. . . .
(b) ~ Extending ~ Compressive

Figure 2.23. (a) C o m p r e s s i v e and extensive flow and sliding planes. (b) Distribution of compressive and
extensive flows in a glacier (Nye, 1952).
In the accumulation areas stratification predominates, marked by an alternation of

winter and summer ice layers containing sediment formed in the summer melting.
During glacial flow, planar foliation structure originates and can be recognized by the
grain of the ice with alternation of clear blue ice bands and white ice with air bubbles,
with these last being the most abundant ones (Sharp, 1988). These bands vary in their
dimensions from millimetres to several metres. Foliation has a greater development next
to the walls of valleys, where it runs parallel to the rocky outcrops. It is arranged in
cross-section with the direction of the ice flow far from the margins and near the glacier
terminus (Rutter, 1965). Sometimes it is difficult to distinguish the stratification of the
foliation where it is parallel to sedimentary layers.
Commonly glaciers appear folded and faulted (Hambrey, 1977), where it is possible to
observe these structures on the glacier surface and in marginal cliffs (Fig. 2.24). Folds
have different sizes and directions and are very evident in glaciers with surging flows
(Fig. 2.22) as well as in some piedmont glaciers, such as Malaspina Glacier where great
folds similar to those affecting moraines and ice bands are observed (Fig. 2.19).
In steep sloped zones of a glacier, gravity faults and rotational slidings can develop.
At the glacier terminus and where the advance of the ice is interrupted by moraines they
produce thrust planes.
Glaciers 55
Figure 2.24. Fold in the ice, shown by the alternation of clear and dark bands. Livingston Island, South
Shetland Islands, Antarctica. Photo: J. L6pez-Martfnez.
Ogives or Forbes bands are alternating layers that extend over the surface of temperate
valley glaciers (Paterson, 1994). They display a curved disposition with their convexity
pointed in the flow direction. This curvature is due to the higher ice speed in its central
parts (Fig. 2.25). They are spaced between 50 and 200 m apart. The bands of the ogives are
constituted of clear and dark ice. Dark bands are formed by ice and sediment with an
intense foliation, originated by broken ice mixed with mud and snow and later
compressed. Clear bands have less foliation and contain rich white ice with air bubbles.
The model that best explains the formation of band ogives at Bas Glacier d'Arolla
(Switzerland) constitutes a slight variation to the "reverse faulting" hypothesis proposed
by Posamentier (1978). According to this model, multiple shear zones are formed and
basal ice is uplifted to the glacier surface to give rise to the dark ogive bands (Goodsell
et al., 2002). Some authors consider that each pair of bands has an annual origin. In steep
gradient zones of the glacier, ice advances during the winter from the great slope zones
of longitudinal extension, to the lower areas of intense compression, where cracks were
filled with snow originating the clear band. In summer, the cracks were opened and filled
up with ice fragments and detritus formed the dark band when being recompressed (Nye,
1958; Sharp, 1988).
Crevasses are the most abundant structures of glaciers and form because of tensional
stresses (Menzies, 1995a). Generally, they are straight or weakly bent and subvertical.
Their length is tens to several thousands of metres and their width fluctuates from
millimetres to several metres. The maximal depth is of the order of 35 m, because the ice
Figure 2.25. In the foreground in shadow, ogives in the Mer de Glace Glacier, Mont Blanc Massif,
French Alps. Photo: J. L6pez-Martfnez.
underneath shows a plastic behaviour. They constitute excellent routes for the penetration
of melt water. When the system of crevasses intersects, the surface of the glacier becomes
a fragile mass of dentated pinnacles known as sdracs (Fig. 2.26).
Crevasses are grouped in systems (Fig. 2.27) (Sharp, 1960), differentiated into the
marginal or in chevron crevasses formed by extensive stress generated by a higher speed
of the glacier flow in the centre rather than at the margins as a result of the friction with
the rocky walls. They intersect the edge of the glacier at angles of --~45 ~ Transverse
crevasses (Fig. 2.28), located in zones of higher speed, are perpendicular to the flow.
Extension crevasses occur where the valley glacier gets wider, or in bed threshold zones.
These crevasses form smaller angles of --~45 ~ with the edges. At the end of the glacier
terminus, crevasses o f radial extension are developed. Once formed, the systems of
crevasses are modified by glacier flow. Marginal crevasses can undergo rotation and the
transverse and extension crevasses are narrowed until they finally close, being then
recognized by a blue ice vein.
Glaciers 57
Figure 2.26. S6racs in Perito Moreno Glacier at its outlet in the Argentine Lake, Patagonia, Argentina.
Photo: C. Sancho.
.'//I/////77,,,{
I
J, / / / / .
I
. . .
(a) (b)
(c) (d)
Figure 2.27. Types of crevasses in a valley glacier. (a) Marginal (1 - old rotated crevasses, 2 - cracks of
new formation). (b) Transverse. (c) Splaying. (d) Terminal radial splaying. Arrow indicates the ice
movement direction (Sharp, 1960).
Figure 2.29. Bergschrund in the Aneto Glacier, Aragonese Pyrenees. Photo: F. Guti6rrez.
Glaciers 59
The bergschrund (Lliboutry et al., 1976) is the crevasse separating the glacier ice from
the rock in the upper part of the glacier head. Some ice adheres to the rock wall and the
rest of the glacier pulls away to form bergschrunds. They are located near very steep walls
(Fig. 2.29) and sometimes extend hundreds of metres. Frequently, instead of a single
crevasse, a system of bergschrund crevasses is developed.
Chapter 3
Glacial erosion
For a long time, the capacity of ice to substantially modify the preexisting relief has been
well known. Knowledge about its action, in spite of the large number of publications on
this issue, is not very extensive and this arises from the difficulty of studying the action of
the ice on its bed, because the more effective erosive mechanisms take place under a thick
ice cover. Some investigations directed to this aim by means of tunnel perforations have
taken place, mainly in cirques. To this observation difficulty must be added the deficiency
of knowledge on the preglacial relief, which can prevent a suitable quantitative estimation
of the glacial erosion. Glaciations have also been superposed in the course of time in many
areas, so that interpretations are more difficult (Price, 1973).
Glacial erosive action appears at the ice-rock (or ice-glacial deposits) interface. We
know that erosive effects of ice without either movement or rocky material (clean ice) are
practically negligible. For the same kind of ice in movement, if it affects fresh and little
jointed rocks, the erosive power will be null or very small, although if the bottom material
is loose it can substantially modify its morphology. The greater erosive effects occur under
the action of thick temperate glaciers causing the extraction of glacier bed fragments and
mobilizing and transporting them to a more-or-less distant place, being able in its transport
to erode the rocky or loose material at the interface.
1. Erosive processes
We have already indicated that knowledge of glacial erosion is not very extensive.
Normally we deduce processes by indirect methods by means of experimentation in a
laboratory or by applying the most common deductive method of the Earth sciences, in
which from the effects we try to classify the processes that were produced, in this case
glacial erosional landforms (Goudie, 2004a).
Many factors affect the intensity of glacial erosive processes. They can be subdivided
into three great groups: the ones inherent to the glacial system, those concerning the
lithological and structural characteristics of the rocky substratum, and the ones relative to
the geometry of the substrate (Sugden and John, 1976; Drewry, 1986).
The importance of the glacial system is given, on the one hand, by the basal ice
temperature. When the temperature is close to the melting point the erosive effects are
without a doubt much more important; in the case of glaciers with low basal temperature,
the erosive action is only effective if debris is present at the interface. In addition to the
influence of temperature, another factor of great significance is the basal speed of the ice
mass that affects the transported amount of debris and, therefore, their erosive action and
also the rock fracturing by means of differential pressure generation. A fundamental factor
within the glacial system is the ice mass thickness, because it affects the friction
mechanism at the ice-rock interface.
Characteristics of the rocky substrate, such as rock hardness and the presence of
discontinuity planes (stratification, joints, foliation, and so forth), undoubtedly affect the
action of erosive processes. Another important variable is the rocky bed permeability,
because it conditions the penetration of melt waters, with the consequent variation in its
erosive capacity and their possible absence in future regelation.
Finally, the geometry characteristics of the rocky bed, as to form, roughness and slope
have to be added to the previously indicated factors with regard to variations in the
intensity of the glacial erosive processes.
Not only the characteristics of the ice and the rocky substratum should be considered
but also the modifications that glacial erosion undergoes in the course of time within a
glacial age. Greater erosion intensity occurs in the first stages of the glacial advance
(Tricart and Cailleux, 1962), in which the weathering mantle is easily exported by the ice
masses in its first pulses. Once the regolith is evacuated, the ice needs a greater effort to be
able to erode the non-altered substratum.
Approximate calculations of glacial erosion rates have been made (Andrews, 1975).
On the one hand, the material transport (bottom, suspension and dissolution load) by
glacial melt waters next to the glacier terminus has been determined, supposing that this
material has originated in the glacial environment. Therefore, approximate calculations of
glacial erosion can be made. Average rates seem to vary between 1 and 5 m every 1000
years, but this is perhaps too high and it is necessary to divide them by a factor of 2 to 10.
In order to make estimations of glacial erosion rates for long time periods, theoretical
calculations are also made. They have been applied to cirques, glacier valleys and fjords.
The variables considered are volume, area and age of the analysed form. Volume is
divided by the area and the erosion or reduction for the time undergoing the action of ice is
obtained, to obtain the erosion rate. For cirques this value varies between 5 cm/1000 yr for
polar cirques and 40 cm/1000 yr for the cirques of western Scotland. In any case, these
calculations are to be considered as orientative only because true rates of glacial erosion
are a major research topic at the present time.
1.1. Process types
The erosion mechanisms of glaciers are various and the following ones can be
differentiated.
1.1.1. Abrasion
The wearing of rocks by the action of ice passing is clearly shown by its effects,
represented, among others, by the microforms of glacial erosion such as striations,
grooves, and other features. The abrasion process can be studied directly by means of
tunnels into the contact with the ice-rock interface, by direct observation in natural
subglacial cavities and also, indirectly, by laboratory experiments, analysing the behaviour
of the ice sliding on different rocks and circumstances (Embleton, 1979).
Glacial erosion 63
Because ice at 0~ has a hardness of 1.5 in the Mohs scale, it is difficult for it to scratch
or wear away the rocky surfaces over which it moves, because most minerals have higher
hardness numbers than it does, from which it is deduced that the abrasive action must be
due to the friction of rocky fragments held by the glacier in contact with the rocky
substrate (Sugden and John, 1976).
Calculations of the abrasion rate have taken place at numerous sites. Values vary
between 0.001 mm/yr for small glaciers and 36 mm/yr for a crystalline limestone substrate
located in Argenti6re Glacier (French Alps) under an ice thickness of 100 m and a velocity
of the glacier of 250 m/yr (Boulton, 1974). Perhaps this last number is too high, because
the dissolution process of limestone, which is known to be very important in cold waters,
has not been considered.
Factors affecting the abrasion process can be classified in two groups: those inherent to
transported fragments and characteristics of the rocky substrate and the ones related to the
particularities of the ice mass (Menzies, 1995b). Clean ice, no doubt due to its hardness,
does not make abrasive effects on the rocky bed; on the contrary, ice at the interface shows
scratches and cavities in it indicating the wearing action of the rocky bottom. The abrasion
process requires for its action, the presence of rocky fragments held in the glacier base. In
polar ice caps and glaciers the action of abrasion is very small, partly due to the paucity of
fragments in the ice mass. In some soundings of more than two thousand meters only a few
meters have been registered containing significant debris, although in soundings made in
Greenland (Herron and Langway, 1979) and in Antarctica (Gow et al., 1979) abundant
debris appears in the basal ice of these ice sheets. In contrast, the greater amount of
material transported in temperate glaciers produces a much more important abrasive
action, although in these glaciers the fragment content factor regarding the abrasion
intensity should not be the only one to be considered. Another important circumstance is
given by the relative hardness of the fragment with respect to that of the rocky substrate
(Drewry, 1986). It is obvious that a higher hardness of fragments becomes apparent in a
greater abrasive potential. To these factors it is necessary to add the morphologic
characteristics of the particles. The constant fragment interaction with each other and with
the rocky bed has as a consequence abrasion and reduction of angularity. This decreased
effectiveness of the process can be overcome with provision of a new particle or with the
breakage of previously worn fragments. The existing erosive processes at the ice-rock
interface produces a fine material that, in some cases, stands between the ice mass and the
rocky bed thus hindering further abrasion action; in these circumstances the presence of
subglacial water favours the export of this "glacial flour" film. It is also necessary to
consider the permeability of the rocky substrate, because the water presence at the
interface produces an increase of flow velocity, which is also an influencing factor of
abrasion, as discussed below.
The other group can be denoted as glaciological factors (Embleton, 1979). If the
existing temperatures in the basal ice are very low, an adhesion between bed and glacial
ice takes place, so that very strong stresses for sliding are necessary. If we add that under
these temperature regimes the fragment proportion is, as previously indicated, very small
or null, abrasive effects will practically be negligible. Basal flow speed is a factor of
considerable importance because it determines the number of particles running past one
point at the ice-rock contact; therefore, at higher speed a greater abrasive power is
expected. Ice mass thickness has an important influence. A particle located in the ice-rock
/
Zone A /
v = 100m yr -1/ t Zone B
x 1000--
I
E /
E
,o /
1
t--
0
-~ 100--
.Q
<
/
/
10
I I I I !
o 10 20 30 40 50
Effective normal pressure (bar)
Figure 3.1. Theoreticalabrasion rates and effective normal pressure for different ice speeds. In Zone A
abrasion increases with the rise of pressure and in Zone B it diminishes until approaching zero, producing
sedimentation of particles, k is a value depending on the relative hardness of fragments, the rocky substrate
and the debris amount (Boulton, 1974).
contact is put under a pressure produced by the weight of the ice column and fragments
that it supports. For a specific speed, abrasion intensity increases with rising pressure
(Boulton, 1974), or similarly with ice thickness, until a certain threshold where the
fragment-rocky substratum friction delays the advance of the fragment, so that ice flows
over it instead of dragging the particle at the same speed at which the ice moves (Fig. 3.1).
The chart shows the existence of two zones: one A, in which abrasion increases with rising
pressure and a B zone, in which abrasion diminishes with a new increase of pressure, until
it is null.
1.1.2. Fracturing
In this section we include the fractures generated by ice action as well as the ones existing
before the glacier passing. It is difficult, in some cases, to specify if the origin of the joints
is preglacial or is a process due strictly to the action of the ice.
We have already indicated that the existing rocky fragments in the base of the ice mass
can scratch and groove the bed, but can also produce fracturing in the substrate and extract
splinters when exerting pressure on the rocky material. This process becomes apparent
basically by the generation of friction cracks.
Average values of basal shear obtained range round 1 bar (Embleton, 1979). Fracturing
is exerted fundamentally by the action of basal fragments on the rocky bed. The
phenomena of basal fracturing are more evident in the reverse-slope zones of the glacial
flow. On the other hand, there is no doubt that the thermal rrgime of glaciers exerts a very
Glacial erosion 65
considerable influence on the glacial fracturing process. In the case of polar glaciers great
shear stresses are required, because the ice-rock contact remains as a unit due to its frozen
state.
A factor of undoubted importance is the state of the material before ice advances on it.
Thus the periglacial breaking of the bottom indicates the incidence of periglacial
processes on the preparation of easily exportable material with the later advance of the ice
(Boy6, 1949).
In the classic work carried out in the Yosemite Valley of California (Matthes, 1930) the
importance of sheeting in the evacuation of material by extraction was analysed. They
indicated that the optimal joint spacing for glacier mobilization ranges between 1.5 and
7.5 m. Sometimes it is difficult to determine if the origin of this topographic jointing is
preglacial or if it has been generated by load loss after the disappearance of an important
ice column. In some cases two sheet joint systems taking place at different times can cross
(Ollier, 1984), with the crossing zones being weakness areas (Fig. 3.2). In any case,
sheeting has a considerable influence on the control of the topographic form.
Existing joints, of one or another origin, are penetration routes for subglacial water; if
later the freezing of this interstitial water occurs it produces the congelifraction process
giving rise to rock breakage. Only weak temperature oscillations around the freezing point
are necessary for this mechanism to be effective. Congelifraction seems to take place in
glacier cirques where the melting water penetrates through the bergschrund. Obviously, on
the scale of climatic fluctuations, this process can be important during different glacier
oscillations.
Another circumstance of great interest, provided mainly by the investigators who worry
about the genesis of topographic modelling from the evacuation of thick weathering
Figure 3.2. Topography and sheeting in Vaiont glacial valley, Italy (Ollier, 1984).
profiles (Bakker, 1965; Thomas, 1994), is the presence of regolith of variable thickness
generated prior to the freezings. The existence of some regions of this material that are
easily eroded by the ice must be very carefully considered in the interpretation of glacial
erosion modelling and in the quantification of erosion by the ice. We consider that this
circumstance must be taken into account in the study of glacial sculpting in the Iberian
Central System, because the presence of thick weathering deposits is common in this area,
generated without a doubt in times prior to the Quaternary (Guti6rrez and Rodr/guez,
1978; Molina and Blanco, 1980).
In any case, fracturing of the rocky substratum, whether inherited or of glacial origin, is
a factor of extraordinary importance in glacial erosion, along with the availability of
preglacial loose material.
1.1.3. Evacuation of debris
When a glacier slides on its bed, the mobilization speed of particles depends
fundamentally on their size and forms and on the substrate roughness. For fragment
mobilization it is necessary for the traction force to exceed the frictional resistance
(Sugden and John, 1976). For the erosive action to be more effective it is also necessary to
evacuate the fragments contained in the substrate, generated by the previously indicated
processes. In this way the exposed rock can undergo glacial erosion mechanisms. It also
should be considered that a part of the debris export is made by subglacier melt waters
(Hallet, 1979; Drewry, 1986); this circumstance is significant in the case of temperate
glaciers and negligible in polar glaciers. Another form of bottom material mobilization in
temperate glaciers is by pressure of the ice mass on small size fragments and particles,
soaked in water and located at the rock-ice contact, flowing towards areas of lower
pressure.
2. Landform modelling by glacial erosion
The consequence of the action of glacial erosion processes is the generation of different
forms due to the action of ice masses. These are reflected in the classification of glacial
erosion forms of Table 3.1 (Sugden and John, 1976). It is difficult to carry out a
classification of this type, in which a certain dose of subjectivity becomes necessary. In
this classification, the three current process types are considered by distinguishing
unconfined areal ice flow, linear flow in rocky channels and a third process differentiation
corresponding to the interaction of glacial and periglacial activities. Another variable used
for the differentiation is based upon the height or depressed position of the resulting form.
Also as a classification norm the aligned or partially aligned morphology of the generated
form is used. All these criteria are located on the ordinate axis, and on the abscissa a
logarithmic size scale is placed.
As we will see, some of the forms are of doubtful interpretation, because strong
discrepancies exist if these have been generated by erosive action of ice or by subglacial
waters, or even by the combined action of specific processes of each one of these means,
intimately connected. It is therefore difficult in many cases to carry out a clear distinction
between glacial and fluvioglacial erosion.
Glacial erosion 67
E
~- O
EO o .E_ I o .o
O
O
(~ ,-. (~~
v
r~ L o
"0
E _c _~
J < J 7
~ 8q
I
~ Z
Eo
O
v
I
A A
Eot:l i
O
'-
I
! i cO') 7
0 ~ 13"
i ~ 0 .'-
~ 00
i
I'
C:
:::}
"-
"O
I
~
.a~
0O
9 1
' I ~ o n-
! .~,.., o o
~rr 9 ra
g g [ "" ,,
,-- ~" i
Eo -~
j=:
c
0
,
O
' E ; E
O @ '--
,.- ,.9.o
O
"O
r~
O I -r--
O
.,..~
~r~
~ E
.,.., v
"El '~ "0 "0
~ (D (D
E __E E E
O ~
~ "7<
WT-
rr0"~
t'- E t-
O O O
O U_ (/)
= I i i U.I ffl if) (/')
O -"i a - (D
>- c~
(D
121 a ID
.,~
c" ~ m
c- (i,)
(D
,._ 0
"7= '--~ (D "E
Most texts differentiate two overall categories for glacial erosion modelling, dividing
them into smaller and larger forms. This distinction is, however, not very realistic due to
the wide dimensional interval occupied by some of them.
2.1. Striations, grooves and glacial polishing
Friction of particles at the glacial base during movement over the rocky bed produces
erosive actions on it that become apparent as scratches, grooved forms and polishing of the
rocky substrate surface. This action implies extraction of particles from the rocky bottom
that incorporates them into the glacier.
One of the most common forms are striations or straie (Fig. 3.3), which are fine aligned
grooves generally not much more than a metre long and a few millimetres wide and deep.
Striations develop more easily in fine grained rocks and disappear when being exposed for
long to weathering agents. In limestones this disappearance is very fast due to dissolution.
So, on the slopes of the limestone pavement of Ingleborough in northern England,
striations cropping out by removal of a glacial till are removed after about 10 years
Figure 3.3. Glacial striations. Santa Elena Fort, Head of G~illego River, Aragonese Pyrenees.
Glacial erosion 69
(Sweeting, 1966). Not only the rocky substrate shows striations, they are also observed in
fragments in glacial tills. In this way, the existence of striated clasts can help with
determination of genesis, although with reservations in the interpretation of the origin of
problematic deposits (Tricart and Cailleux, 1962). Care must be taken, therefore, because
striations can be produced by other different processes, such as avalanches, mud flows,
subglacial currents, and so forth. Striations are more common in reverse-slope areas, but
they can even appear in vertical walls. Under the microscope, striations consist of
numerous increasing fractures (Iverson, 1995). They indicate direction, but not the
movement of glacial flow; sometimes, several directions or systems of striation
intercrossing occur. The morphologic irregularities of the substrate can explain the
directional variations; they can also indicate several glacial stages or phases of ice
advance. In any case, and mainly through the study of the great glacial ice sheets,
statistical analysis of hundreds of striations is fundamental to obtaining reliable data of the
glacial flow direction for a certain time.
Another modelling type is grooves (Fig. 3.4). They are furrows of variable dimensions
produced upon rocks greater than 1 metre long. In the Mackenzie River valley of
northwestern Canada, gigantic grooves appear whose size can reach 12 km in length, 30 m
depth and 100 m width (Smith, 1948). Their direction agrees with the glacial flow
directions deduced from other forms, although a strictly glacial genesis is difficult to
understand. Whereas striations seem to be formed by friction of silt size particles or sand
on the rock, the origin of grooves is due to the action of large individual or grouped
fragments. But grooves cannot only be generated this way, because many authors adduce
Figure 3.4. Groovesand glacial striations in Rongbuk Valley, Everest massif, Tibet. Photo: J. Lrpez-
Martinez.
Figure 3.5. Glacial polished surface and till. Pordoi Pass, Dolomitas, Italy.
the action of subglacial water currents or a similar preglacial morphology where these
characteristics are emphasized later by the ice.
The action of constant abrasion produces glacial polishing (Fig. 3.5). Observed with a
magnifying glass or under the microscope, a glacially polished rock has a surface furrowed
by fine scratches corresponding to small striations. This polishing is more evident on fine
grained rocks. It also disappears quickly when being exposed to weathering agents.
2.2. Friction cracks and P-forms
These are small forms related to glacial action appearing by fracturing of rock or by
concavities where extraction of rocky substratum material might have taken place or not.
The most classic works in which both differentiation and origin of the friction cracks are
considered are from Gilbert (1906), Harris (1943) and Dreimanis (1953).
Friction cracks are developed better on intermediate grain-sized rocks, in which
striations and polishing are less evident. Normally they appear on hard and fragile rocks
such as granite, basalt, quartzite, and so forth. Four types are distinguished by Embleton
and King (1975a) (Fig. 3.6):
(a) Lunate fracture (Fig. 3.6a), in which the horns point in the direction of the ice
movement. These features are formed by two fractures, one of low angle and another
subvertical; the intersection between them gives rise to a rock slab.
(b) Crescentic gouge (Figs 3.6b and 3.7), the horns point opposite to the ice flow. Its size
ranges between 2 cm and 2 m. They usually appear in two to six systems and generally
one behind another. They also have the two types of fractures mentioned above.
Glacial erosion 71
~ ,o
~iii!i!ii~i;i;i!!iiiii!~ili!ii~ii!;iiii:i:ilili;ilili!;iiiiii~ii!::ii!ii!!ii~
•ii!i•i•!i•ii!•i•i•i•i•!•i••iiii•i•i•iiii!i•••ii•i•ii•i•ii•i••i!ii!i! i•i•i••ii••ii•ii
(a)
::.:.:.:.:..:.: ::.. ========================================================================================
(b)
(c) (d)
-
!ii!!ii,,'iiiii!i,,'ii,'iiiiilZ!i'
iiiii i! i Sii!ii iii!iiiii!ii:i :i i!ii
Rock b
(e) (f)
Figure 3.6. Friction cracks and sichelwannen (according to several authors, in Embleton and King
(1975a). Horizontal arrow indicates the direction of the ice movement. (a) lunate fracture (plan view and
section); (b) crescentic gouge (plan view and section); (c) crescentic fractures (plan view and section); (d)
conchoidal fractures; (e) sichelwannen; (f) formation of a crescentic gouge; ade - original rock surface;
afdb - rock surface deformed by pressure transmitted from the block contained in the ice; cgd -
conchoidal fracture; fg - secondary fracture formed when the fgd rock wedge is broken.
Figure 3.7. Crescentic gouge. Threshold of Gerber cirque, Lerida Pyrenees.
(c) Crescentic fractures (Fig. 3.6c) are concave downstream to the glacier flow direction
and are constituted by subvertical fractures. Slab extraction, as in the two previous
examples, does not occur.
(d) Concoidalfractures (Fig. 3.6d) in which the fracture plane is concave upwards.
All these types are more frequently encountered in reverse-slope zones. As friction
cracks are larger than striations they are conserved more easily, because they resist
weathering and erosion effects longer. The perpendicular to half moon forms can indicate
ice flow direction. This can be determined by the inclination of the fracture of the smaller
angle diving downwards with the glacier. This rule, indicated by Gilbert (1906) and Harris
(1943), is not always fulfilled, because there are many opposite cases described in the
literature.
In general, it is agreed that friction cracks result from the pressure of a block on the
rocky substrate (Fig. 3.6f). This pressure is greater in areas of reverse-slope (Boulton,
1974), which is the reason for the greater abundance of this micromodelling in these zones.
It appears that at the beginning an elastic deformation followed by a breakage takes place,
generating a conchoidal fracture. A vertical fracture cutting through the fracture of low
inclination originates afterwards when the block advances. This type of crack has been
reproduced experimentally using a knife or a steel ball on glass.
P-forms, or plastically sculpted surfaces (Dahl, 1965), are on areas exposed to the
action of the ice. They are smaller forms never exceeding 20 m in size. The most abundant
form is the sichelwannen (Fig. 3.6e), a depression in the form of a half moon modelled on
hard rocks, whose horns aim downwards with the ice flow. They can be found in surfaces
with any inclination. Another type, the cavetto forms, are channels of steep edges (unlike
Glacial erosion 73
grooves) of up to half metre depth found on steep surfaces. In conjunction with these forms
appear grooves, basins, folds and curved and winding channels (Kor et al., 1991). Their
origin is very controversial. The most accepted hypothesis is the one explaining this
micromodelling as due to subglacial water action under pressure. Other authors, however,
defend the idea of abrasion processes (debris loaded basal ice), or erosive action produced
by the movement under pressure of water-saturated basal till (Gjessing, 1965; Gray, 1981).
2.3. Roches moutonn~es and other forms
Roches moutonn~es are aligned hills, generally grouped, asymmetric, with the minor slope
frequently polished and striated and the other one constituted by an irregular and
fragmented, sometimes steep, surface. This disposition stays constant indicating the
direction of the ice movement, moving from the minor slope side to the one of greater
inclination (Fig. 3.8). These forms are developed better on crystalline rocks and are very
common in areas covered by ice caps and also in thresholds of cirques. Their sizes are very
variable, from less than a metre to hundreds of metres, reaching, in the case offlyggberg, to
kilometric dimensions and heights of hundreds of metres (Sudgen et al., 1992). Within these
forms greater roches moutonn6es can be found, as well as others of smaller size.
The origin of these forms is not clear. In some cases (Matthes, 1930; Sugden et al.,
1992) the importance of joint spacing in the generation of this modelling type is indicated
(Fig. 3.9), although many cases do not conform to this rule. Others researchers, such as
Carol (1947), have explained the steep side of roches moutonn6es due to the action of the
Figure 3.8. Roches moutonn6es. Tollstringen, Trolls massif, Norway.

Figure 3.9. Longitudinal profile of a roches moutonn~e showing its form and the influence of joints.
Arrows indicate the direction of the ice pressure (Matthes, 1930).
congelifraction process, in frozen to melting temperature (Fig. 3.10). In these

circumstances, when passing over a protuberance, the pressure increases and ice turns
to a semiplastic state; pressure diminishes on the side of greater inclination and the melt
water is frozen again, breaking the rock. These observations were made in a subglacial
cavity. As indicated by Embleton and King (1975a,b), however, the question of whether or
not roches moutonn6es only form under temperate ice, as the subglacial congelifraction
hypothesis indicates, still stands open.
Another set of forms related or associated to roches moutonn6es are recognized.
Whalebacks, rock drumlins and streamlined spurs are more or less extended forms of
smoothed slopes. Sizes are reflected in Table 3.1. It seems that joints regulate the limit
between hills in whalebacks. Rocky drumlins sometimes appear associated with
elaborated fields of drumlins in glacial material. It has been questioned as to whether or
not streamlined spurs are due to glacial action, but the smoothness of their slopes along
with the parallelism of their direction with that of the glacial flow, inclines one to think that
this modelling is the result of glacial erosion.
All previously described forms constitute positive relief, but it is well known that a
landscape of glacial erosion is sprinkled with lakes of different sizes, that in most cases are
rock basins. Their morphometry is very variable and structure plays an important role as
much in their morphology as in their genesis. Their origin can be due to irregularities prior
Ice
Figure 3.10. Operatingprocesses in the genesis of roches moutonn~es (Carol, 1947).

Glacial erosion 75
to the passage of the ice, perpetuated and emphasized by its action; at other times, glacial
erosion processes are the ones that generate these closed depressions.
2.4. Cirques
Cirques are, along with glacial valleys, the most spectacular forms of glacial erosion. The
texts on glacial geomorphology indicate that the word "cirque" was used for the first time
in the Pyrenees by Jean de Charpentier in 1823. On the other hand, various denotations for
cirque in different languages and countries do exist. We can consider the glacial cirque to
be a semicircular or semielliptical depression dominated by steep slopes that is or has been
occupied by ice (Fig. 3.11). Some authors consider as a characteristic of their definition
the existence of a rocky basin, but not all cirques present this form (Fig. 3.11).
The presence of a rocky basin carries with it the existence of a threshold at the cirque's
exit, which is sometimes rocky or composed of glacial till. The walls of the cirque usually
appear fragmented and broken, unlike the basal surface, which is generally smoothed and
displays smaller forms of glacial erosion.
The dimensions of cirques are very varied and fluctuate between tens of meters and
kilometric widths. The biggest cirque known is the Walcott Cirque, in Antarctica, some
16 km wide with walls about 3000m height (Price, 1973). Size depends on many
variables, such as the duration of the glaciation, structural and lithological characteristics
of the rocky massif on which the cirque is located, and so forth (Embleton and King,
1975a). The form is also a function, among others, of this last characteristic.
Figure 3.11. Cirqueand morainic crest, developed on lower Cretaceous materials, head of Miera Valley,
Cantabria.
The morphometry of a cirque is d e t e r m i n e d with sufficient detail by m e a n s of a set of

characters that a p p e a r partially e x p r e s s e d in Figure 3.12 ( A n d r e w s and D u g d a l e , 1971). By
h i m s e l f A n d r e w s (1975) indicated another value, the one of the c o t a n g e n t of the angle
f o r m e d by the threshold of the cirque with the final wall that indicates the intensity of
glacial erosion. A n o t h e r index, the o c c l u s i o n d e g r e e (Evans, 1969) is defined as the
n u m b e r of d e g r e e s of the l o n g e r c o n t o u r and for 180 ~ indicating that the walls of the cirque
Chiarac'er3
are parallel. In addition to these values other p a r a m e t e r s have b e e n set out to describe the
7-/ . . . . '
razte
/ 12 Character5
/ /" / 1 z'~ /Characte, /
/////" ,/ /
""
// ea level
(a) Character 6
N
#
Character 10
Character 14
!
Character 7
(b)
Figure 3.12. Cirque characters (Andrews and Dugdale, 1971). (a) transverse section of a cirque, (b) plan
view of a cirque and its summit area. The different characters are the following ones: 1 - Maximum
mountain elevation in which the cirque is located. 2 - Maximum elevation of the cirque. 3 - Difference
between 1 and 2.4 - Elevation of the cirque threshold. 5 - Maximum vertical development of the cirque.
6 - Length of the cirque' s greater axis. 7 - Maximum width of the cirque perpendicular to the greater axis.
8 - Length/width relation. 9 - Length/height relation. 10 - Direction of the greater axis. 11 - Type of ice
mass: (a) empty, without ice; (b) snowfield; (c) ice plate; (d) cirque glacier. 12 - Final wall height of the
cirque. 13 - Relation between height of the final wall and the length of the cirque. 14 - Angle of the
regional slope.
Glacial erosion 77
morphometric characteristics of a cirque. In general, these values lack precision because

they are taken from topographic maps and, in some cases, for some parameters the
subjectivity of the investigator plays a role. As indicated by Andrews (1975), however,
"the morphologic analysis of cirques can provide great information about the erosion
processes in the cirque, although only of a deductive type: morphology is an answer to
processes...".
In order to understand better the processes that have given rise to its form, it is
necessary to understand, among others, the structure and movement of ice in glacial
cirques. The most precise observations were taken by McCall (1960) in the studies of a
small cirque glacier, Vesl-Skautbreen in Norway. Two tunnels were perforated (Fig. 3.13)
and among other observations, ice speeds at different points were measured and different
ablation surfaces were distinguished in summer. Observations showed that flow lines of
the ice dip 30 ~ in the upper zone of the glacier and in the lower one they incline 2 6 - 2 8 ~ but
in the opposite direction to the slope of the cirque, their inclination diminishing
downwards; this disposition indicates a synclinal structure for the ice layers. The
distribution of speeds indicates faster movements in the upper zone of the glacier and
slower ones in lower parts; in this area the different speed vectors reveal upward
movements of the ice. All these data provide evidence that movement takes place by
rotational sliding.
Erosive processes happening in a cirque are basically twofold and are due to glacial and
periglacial activity. At the ice-rock contact, abrasion carries out continuous wearing as
much in the walls as at the bottom and is responsible for the rocky basin existing in many
cirques (Richardson and Holmund, 1996). This depression is easily explained by rotational
sliding of the ice and this movement also produces the overcoming of the threshold and the
evacuation of debris. The other type of process, congelifraction, is very effective on
outcropping rocks over the glacier (Gardner, 1987). Here, it seems that the freeze-thaw
action is more effective in temperate glaciers than in polar ones. It is deduced that the
Observedvaluesof movement
.~er tunnel
Flowlines
\
lOOm
) Firnline
/
50m. l
~.....~Lo~,r tunnel
I /
I r , / ~ _~ ~ - " , ,
J . /
U / ~ ? ~ ZT--'ZT-"~'~,.. &~Debris~,
lOOm 200m
Figure 3.13. Sectionof Vesl-Skautbreen cirque glacier (Norway) in which the flow and speed lines are
indicated. After McCall (1960), simplified.
cirque enlarges fundamentally by the action of congelifraction and deepens by the effect of
abrasion. The speed of erosion in cirques obtained by different methods is of the order of
500 mm/ka (Benn and Evans, 1998). It seems that the backward movement is faster than
the excavation. On the other hand, the elevations of the bottoms of the cirques have
apparently not changed much during glaciation. Elevation, generally, is used to calculate
approximately the level of permanent snows, because it corresponds with the 0~ isotherm
in summer, in considering the lowest cirques of equal orientation (Flint, 1971).
The origin of cirques, according to most authors, is tied first to the snow accumulation
in a preexisting depression in which congelifraction processes and particle evacuation take
place by snow melting in summer (nivation) (Thorn and Hall, 1980; Evans, 2004c). This
brings as a consequence the widening of the depression and the generation of a nivation
niche. If excavation is deep enough, snow lasts from 1 year to another in the nivation
niche, becoming firn and also turning this material into ice. In this way an embryonic
cirque is formed. The nival accumulation stage is denoted as the "incipient phase" and
when snow begins to last, the "tim phase" (Tricart and Cailleux, 1962).
Once formed, if the circumstances are appropriate, the cirque continues its
development. Its evolution is influenced by different causes. The duration and number
of glaciations exert, without a doubt, a fundamental role in their development. Also, the
glacier has to evolve differently if it is located in temperate and subpolar latitudes, or in
polar areas. Another great incidence variable is the structure and lithology of the rocky
massif on which the cirque is based (Embleton, 1979). Sometimes cirques are located in
relief produced by lithological differences, which in a monoclinal series can give origin to
stepped cirques. The degree of fracturing and the spacing between joint planes play a
fundamental role in mechanical disintegration by gelifraction. Exfoliation can be of
importance in the form of the cirque. Another basic factor in the development of cirques is
their orientation. In middle latitudes of the northern hemisphere, most cirques presenting
good development face towards the north and east, because in that direction they are
protected from solar rays and the snow lasts more easily. One of the analysed factors is the
influence of climate in the variation of the cirque (Derbyshire and Evans, 1976). These
authors indicate that it is not easy to establish generalizations on this aspect, in spite of the
great number of regional studies on form and distribution of cirques.
The development of cirques also includes a backward movement through headward
erosion of its walls, which at the limits with other cirques (Fig. 3.14) can cause knife-
edged ridges known as arYtes. The disappearance by growth of the cirque of these ar~tes
can give rise to coalescent cirques (Fig. 3.15). The formation of pyramidal peaks or horns
(Figs 3.16 and 3.17), usually presenting three or four faces, is a result of the backward
movement of the multiple cirque walls.
2.5. Glacial valleys
Deep glacial valleys are one of the most characteristic forms of erosive activity of glaciers.
In contrast, in fluvial valleys, rivers are only in contact with a small part of the valley,
whereas the ice of a glacial valley occupies a greater part of the cross-sectional profile
(Price, 1973). These canalised ice masses excavate deep troughs to hundreds or even
thousands of metres of depth (Figs 3.18 and 3.19). This intense erosion is mainly carried
Glacial erosion 79
Figure 3.14. Aerial view of horns, cirques, ar~tes and glacial valleys. Eastern Pyrenees.
out in mountain areas located to windward, that receive great nival precipitation and also
feed thick glaciers running through preglacial valleys producing a considerable,
sometimes spectacular, modification of the relief.
According to Linton (1963) four types of glacial valleys are distinguished. The alpine
type has feeding that takes place from a cirque or set of cirques in high zones. The iceland
type has an ice cap escaping preglacial valleys that forms the glacier. The composed type
occurs where ice does not find sufficient previous valleys to unload all the ice, or similarly,
Figure 3.15. Coalescentcirques, ar~tes and a glacial valley (in Holmes (1965), Fig. 4.81 modified by Davis
(1906)).
Figure 3.17. Horn in central Andes, E1 Portillo, Chile.

Glacial erosion 81
Figure 3.19. U-shaped cross section in the glacial valley of Llanganuco, Cordillera Blanca, Peru. Photo
J. L6pez-Martfnez.
Fluvioglacial erosion and sedimentation l 17
250' A B C 250' D E . . . . . .~ F
1-- ----- 1 -- . . . .
~o0. . ........ ~oolt ..... :.~ . . . . . . . . . . . . . . . . . . . . .
1 0 0 ' - . . . . . . . . 1 0 0 ' ' ' "
250' G H I J _ K L M
o - - ~ 250'1
~ ~ I I I I I ~ I ~ ~ ~ ~ ~ ~ ~ I I I I I I I O I ~ ~ ~ ~ ~
200- 9 .
. . . . . . . ~200' 9 ~ ~ _ ,
,i ,~ 9 .'\ \'~\~ k" 9 " x \ \\ t .... ,
, ~ o ~ . m ,,.., o ~ . ~ . m ~ 1 7 6
100'J I 0 0 ' I~'*

'--"
9",, Crest of Esker
1948
V.E. = X 6 . 6 ~- " ' . Base of Esker
Crest of Esker
1963
~--~ Base of Esker
0 220 440 yds. - Lake of Esker 1948
9 I I
\\\\ Ice 1963
(a)
300' -
250' -
~_ I .,. : \'-," k ' ~ ' \ "~ " ",,. - ' , , ' \ " ~" ~'~,,. /
200' -
150' -
100' - "K. ~ ; -Z.., !z.,, .'.z- , ' ' / ~
50' --
,, ~" 9 Crest of Esker

1948
V.E. = X 6.6 9- - . Base of Esker
Crest of Esker
1963
~..~ Base of Esker
0 220 440 yds. \ \ \ \ Ice 1948
l ! I ///// Ice 1963
(b)
Figure 5.15. Profiles of Eskers in the Casement Glacier in Alaska (a) and in Breidamerkurj6kull in
Iceland (b) (Price, 1969).
Fluvioglacial deposits can be classified according to the sedimentation environment:

deposits due to ice melt channels, and deposits in confined waters, such as lakes and seas.
Also a classification can be established according to their position in relation to the ice
mass (Table 5.1). The proglacial sedimentary deposits occur at a certain distance from the
ice margin, generating smoothed morphologies, and the sedimentary structures are not
deformed. The deposits located in contact with the ice show distorted structures and their
morphologies are produced by the ice mass or by a later settling in relation to the melting
of buried ice. They are, in addition, classified according to the prevailing sediment and the
form of the accumulation.
Eskers (osar in Scandinavia) are ridges composed of stratified deposits with rounded
gravels and sands with some blocks. Their direction is parallel to the regional flow of the
ice. They are sinuous or straight, sometimes discontinuous, with sharpened or flattened
(a)
I(SE
(b)
Figure 5.16. Origin of kames. (a) Crevasses open until the ice-rock interface and development of a
cavity in which sediments accumulate. (b) Supraglacial and endoglacial cavities produce similar forms
(modified from Price, 1973).
Fluvioglacial erosion and sedimentation 119
crests, and lengths from a few metres to 400 km (Aylsworth and Shilts, 1989; Punkari,
1997) (Figure 5.13). Their height can reach 200 m, with 3 km in width. These latter
dimensions are related to the total length, in such a manner that the longest eskers are
commonly the widest and highest. They can also appear as unique ridges or develop an
interconnected network with confluent and diffluent eskers that join the main ridge. The
slopes of the eskers can have angles of about 30 ~ close to the angle of repose, though when
they are degraded the gradient is between 5 and 10~ (Embleton and King, 1975a). The
deposits dip toward the margins of the esker and can show distortions due to the under-
lying ice melt.
Eskers originate in subglacial, englacial, supraglacial, and ice-marginal channels
(Figure 5.14). They develop above all after the peak flow at the end of summer, when the
velocity of flow of the channels diminishes, and can be destroyed totally or partially with
new peak flows (Sugden and John, 1976; Menzies and Shilts, 1996). The deposits of the
bottom of the subglacial channels can accumulate between the ice walls, and the shape of
the cross section of the tunnels plays an important role in the final esker morphology
(Price, 1973). The hypothesis of subglacial channels hardly explains the existence of
tunnels hundreds of kilometres long. The origin from englacial channels must be treated
jointly with that of a supraglacial origin. Studies of aerial photos obtained in different
years made in areas of fast ice retreat of the Casement Glacier in southeast Alaska and the
Breidamerkurj6kull Ice Sheet in Iceland reveal that the profiles of the esker base and crest
have diminished in height (Price, 1966, 1969) (Figure 5.15). This diminution is due to the
melting of buried ice during the period between the aerial photos. This indicates an
englacial or supraglacial origin.
Figure 5.17. Kameterrace affected by badlands. Talon River, Jura of France. Photo: J.L. Pefia.
Figure 5.18. Grain-size distribution in some sandar. Skeidarfisandur is in Iceland, and the Scott and
Malaspina systems in Alaska (modified from Boothroyd and Nummedal, 1978).
Kames are small hills composed of deposits of stratified sands and gravels. Their size
varies from a few metres to conical hills of up to 50 m in height and 400 m in length. They
form by sediment accumulation in supraglacial and englacial cavities (Johnson and
Menzies, 1996) (Figure 5.16). The term kame is also used to describe forms built in a
particular position within or along the glacier. When a channel flows into a lake located on
the ice mass, it carries sediment deposited in the margins; after the ice melt, they result in a
kame delta (Figure 5.14). Kame terraces are formed by accumulation of channel deposits
that run between the valley wall and the lateral ice border (Figure 5.17). The side of the
deposit located close to the ice shows a steep border due to ice melt. A kame complex is
caused by accumulation of sediment in supraglacial depressions (karstic glacial
topography) and later settling by ice melt. The kames are modified by subsidence if
they have underlying ice, and can generate subcircular depressions called kettle holes.
When small hills are associated with depressions, then the term kame and kettle
topography is used, equivalent to the kame complex (Gray, 1991).
The discharge of large volumes of water from the ice margins causes the transport and
sedimentation of particles far away from the ice borders in different environments and
produces proglacial deposits (Table 5.1). In these areas, fluvioglacial sedimentation
prevails because of the decrease in the transport capacity of the ice melt water. The
discharge is associated with a network of braided channels of great lateral mobility,
depositing gravels and sands that build large depositional plains called outwash or sandur
(in plural sandar) (Maizels, 1995) (Figure 5.8). They are similar to alluvial fans, although
with a great influence of strong seasonal fluctuations. The grain sizes in the sandar
diminish from proximal to distal areas (Boothroyd and Nummedal, 1978) (Figure 5.18).
The longitudinal profiles are concave, whereas the cross sections are, in general, convex.
Sandur morphology is continually modified by extraordinary high-flow events as well as
daily variations in flow (Vandenberghe and Woo, 2002). The sandur valleys develop
where they are confined by glacial U-shaped valleys (Figure 5.19). Those developed in the
margins of ice sheets are called sandur plains (Figure 5.8). The proximal area in a sandur
has few ice-melt channels and is characterised by an area of gravels with a number of
closed depressions, resulting from the melting of buried ice; this is known as kettled
sandur or pitted outwash (Price, 1969) (Figure 5.20). Far away from the ice mass, the
Figure 5.19. Myrdalsj6kullIce Sheet in Iceland and sandur valley. Photo: F. Guti~rrez.
25. 1950
-~~~AGI~CIAL STREAM
20. ICE ~
{ 15.
NDUR
10. ICE
25 1951
2O
10
ICE
2oI 1952
o0
15
E
10
15
1955 KETTLED
__~ ~ ~ ~ . ~ SAyUR
~ ~ - - ~ . . . ~ - " ~ ~ ~ SANDUR
E
-~ ICE_~"
111 1960
KAME AND
~1 KETTLE
E
5 - ~ ICE
100m
Figure 5.20. Sketch showing the forms resulting from sedimentation by supraglacial and englacial
streams. The sequence has been obtained from aerial photos in Breidamerkurj6kull in Iceland (adapted
from Price, 1969, Figure 80).
number of braided channels increases. They are shallow (1 to 2 m in depth), change

frequently in position and commonly are flooded during high discharges. Finally, the distal
area has proglacial lakes. During the stages of glacial retreat, the proximal areas of the
sandar are eroded and new sandar develop from the previous ones.
The presence of lakes is very common in fluvioglacial environments, particularly in
marginal and proglacial positions. The discharge of materials in the lakes results in
glaciolacustrine beaches and deltas. If the lake is drained afterwards, then the bottom
deposits make up plains ofglaciolacustrine origin. Some of these deposits have a rhythmic
character and show seasonal variations. They are called varves and have been used in
absolute dating with diverse results. Each varve is composed of two layers, one of coarser
sand deposited in summer and the other of finer clay and silt that is deposited during the
rest of the year. The value of the method of counting varves for chronology construction
has been subject to criticism, because in some years more than a pair of layers can deposit,
whereas other years may not have a sedimentation record due to discharge variations or
deviations of the channels that flow into the lake (Smith et al., 1982).
Because tributary valleys are frequently dammed by the main valley glacier
(Figure 5.21), the moraine deposits can appear interbedded with fluvial and lacustrine
marginal deposits (Bordonau, 1993) (Figure 5.22). In glaciers close to the sea, the ice melt
water discharged in estuaries, fiords and bays causes the accumulation of a great quantity
of glacial and fluvioglacial sediment (Figure 5.23). If this sediment is uplifted as a result of
glacioisostatic rebound after a deglaciation, then it can result in raised beaches and deltas,
and exposed glaciomarine bottom muds, with similar morphologies to the glaciolacustrine
deposits. The identification of these deposits as marine is made by their fossil content.
Figure 5.21. Lateral moraine and dammed depressions in the Miera valley of Cantabria, Spain.
5
1800 4 rkk NW-SE
/~..~',~_
1500
Figure 5.22. Cross section of the Cerler lateral moraine complex, l~sera valley, Central Southern
Pyrenees, Spain. (3) Lateral moraine. (1 and 2) Older deposits. (4) Lacustrine deposits with interfingering
deposits as a result of glacial damming of a tributary valley. A later glacier advance in the tributary valley
deposited younger end moraine ridges. (5) All deposits related to the last glaciation. (Reprinted from
Bordonau, 1993).
Figure 5.23. Ice melting and sediment discharge in a glacier front, South Shetland Islands, Antarctica.
Photo: J. L6pez-Martfnez.
Chapter 6
Geomorphology applied to glacial regions
1. Introduction
Applied problems in glacial areas derive, on the one hand, from the activity of the actual
ice masses and, on the other hand, from the characteristics of the materials deposited by
the glaciers and meltwaters in past periods. Some of the functional processes that develop
from ice accumulations result in different geomorphic processes that will be analysed
below. From another point of view, these areas produce a number of benefits to human
communities. Those derived from tourism, such as landscape observation, skiing, trekking,
climbing, and other sports have undergone a considerable increase in the last few decades.
Summer ice melt in mountain areas leads to the exploitation of the water falling in order to
produce hydroelectric power. In addition, glacial melt water is a very important source for
irrigation in large arid zones of the world, such as the Thar Desert in northwest India,
irrigated by the Indira Gandhi and Great Bhakra Canals, the Peruvian-Chilean littoral
desert, and the areas of Mendoza and San Juan at the foot of the Argentinean Pre-
cordilleran Andes. All the discharges for these irrigation systems come from snow and ice
melt in the neighbouring mountain ranges.
Other important resources derive from the exploitation of the groundwater contained in
the Quaternary deposits of glacial origin. Thus, in North America, a large part of domestic
and industrial water is taken from fluvioglacial aquifers (Thornbury, 1954). One of the
most outstanding features of these accumulations is the variability in grain size and depth,
and consequently the different layers show different transmissivity, permeability and
storage capacity values (Derbyshire and Love, 1986). Till is a poor aquifer due to its clay
content, whereas gravel and sand deposits are excellent groundwater reservoirs (Lloyd,
1983) (Figure 6.1). The common lateral variations in the characteristics of these sediments
cause the occurrence of barriers for water flow, and they impede detailed knowledge of
the aquifer. In some cases, the presence of preglacial valleys buried and fossilized by these
sediments can result in areas of great water storage. For this reason, a precise knowledge
of the geomorphological history of any such region is necessary.
2. Hazards caused by the activity of ice masses
2.1. Icebergs
Icebergs are a consequence of the breaking of ice masses, especially ice sheets where
they arrive at the ocean (Figure 6.2). They are a risk for shipping and also for the oil drilling
platforms. This latter circumstance has encouraged studies on the forecasting of iceberg
-10
\
\
\ f
f
~20
'-30
-30
BOREHOLE
"40
%
9
LOCATION
GEOLOGICAL
f
f BOUNDARY
10 ROCKHEAD
J CONTOUR
SAND ISOPACHYTES
[
\
0-10m
10 -20 rn
TN
~ 20m+
/
0
,_
1
J
2
J km
/
5o
!
Figure 6.1. Structurecontour map of the top of the Sherwood Sandstone with isopach lines for overlying
glaciofluvial sands in the Fylde areas of Lancashire (Lloyd, 1983, in Eyles, 1983, Figure 15.8).
movements along the margins of the eastern Canadian coast (Marko et al., 1988). They can
travel for thousands of kilometres before melting. Their movement is controlled by the
surface oceanic currents, although it can be affected by deeper currents or by the wind.
Along the eastern coast of North America the icebergs come from the Greenland Ice
Sheet and travel with the Labrador Current. About one hundred icebergs per month cross
Geomorphology applied to glacial regions 127
Figure 6.2. Icebergs between Newfoundland and Iceland (Photo: F. Pueyo).
the latitudinal lines during the period from March to August, although they rarely go
far south of 50~ (Figure 6.3). In Antarctica the icebergs travel parallel to the continent
from east to west until they strike currents coming from the east that move away them from
the coast (Hambrey and Alean, 1992).
2.2. Ice avalanches
The study of ice avalanches is very important, because they affect villages, roads, ski
resorts, and hydropower plants. They occur in mountain areas and are much more rare than
snow avalanches, although in some places they can be common and are a tourist attraction,
such as the steep face of Balmhom Glacier in Switzerland (Hambrey and Alean, 1992).
Research carried out with aerial photos enables knowledge of their maximum runout
distances, which is most valuable information for evaluating the risk. The maximum
extent of an ice avalanche depends on the terrain roughness and the season (McClung and
Schaerer, 1993). Thus, the velocity is greater in winter because the avalanche displaces
over a snow cover that facilitates the displacement; the opposite occurs in summer, when
the friction increases and the movement is less. Ice avalanches are forecast by new cracks
and ice falls some weeks before the main event is triggered. In some cases, ice avalanches
rise hundred of metres on the opposite slope.
The oldest known ice avalanche occurred in the Valais Canton of Switzerland, on
31 August 1597, where an ice mass falling from the Balmen Glacier affected the village
of Eggen, which was buffed with 81 inhabitants (Table 6.1). In the same canton,
on 30 August 1965, an ice mass fell in the Mattmark-Saas valley during the construction
o
80 ~ 100 ~ 90 ~ 80 ~ 70 ~ 60 ~ 50 ~ 40 ~ 30 ~ 20 ~ 10 ~
~ ........ ,O.
:i:i::.i:i:ii:i" :::::::::::::::::::::::::::::
.....
~ii=======================================================
9
.....,.......-...-...'.'.'.'.'.'.''.:.:......-! 70 ~
::!;i!~ii i !i!:!:!:i!i;;!~;ii!!)i~~
...'.'.'.'.i.:.:.i.:.:.:-:.:.:.:-:....-...'.-~ii~i~ ~
'":':'::'.'.'.'.:-~.:.:.:-:'i.:'i""'"'"
i . . . . . . .- -...-. . . . . . . . .,.,....
70 ~
i;i:i!i:ii:iiii:iiiii!!iiiii;iiiiii
,:,:,:
.:.2.:-:.2.2.:-2-2":-2".'.-.'."
t!:!:i:!:i:i
!ili!i!iiiiii!!!ii!i,r
. . . . . . . . .
t
60 ~
60 ~ t
SEA
I
,iii!!i!iiii!iii!ii!iiiiii
'_
~ . . . .'.'.'. .'.'.]-i.>:.i..i.i.
! CANADA.i:!~i.".~:i~L~.~:i:!:i:
,
%
%
::::::::::::::::::::::::::::::::::::::::: %%
: ""t LABRADOR ~f.
::: :::::::::::::::::::: :::::::: :,;-.;-:.:.: : : : : ~ :: ..
"" 1
% 50 ~
9. . . . . . . . . . . . . - . . . . . . . . . . . . . . . . . . . : . . . . . . . . . ~ . ~
5 0 ~ :.... . . . . . . :. - . . . . . . ~# . ~ ~ ........
.i.;.~i::i::::i::~:::T~;i::~::i~.~.i.i.~.'.'.'.'.'.-.'.'.'-
..................... ~.~.i.~.~'~.'..~ 9
-J l
P ATLANTIC OCEAN
70 ~ 60 ~ 50 ~ 40 ~
II HIBERNIA
mlm iii MAXIMUM JUNE EXTENT
OF ICEBERGS
ICEBERG DRIFT PATHS
I MODER ICEBERG PRODUCER
& MAJOR ICEBERG PRODUCER
Figure 6.3. The pattern of iceberg circulation off eastern North America (Marko et al., 1988, en E1-Sabh
and Murty, T.S. (eds.), 1988, p. 437, Figure 1).
Table 6.1. Some disasters related with glaciers in Switzerland (excluding those affecting
mountaineers) (Hambrey and Alean, 1992).
Year Location Event Fatalities
1595 Glacier de Gi~tro Lake outburst 160

1597 Balmengletscher Ice avalanche 81
1636 Weisshorn Ice and snow avalanche 37
1782 Altels Ice avalanche 4
1965 Allalingletscher Ice avalanche 88
of a hydroelectric dam. The ice avalanche came from the Allalin Glacier tongue and
moved a million cubic metres of ice, burying the construction area and causing the death
of 88 workers (Hambrey and Alean, 1992; McClung and Schaerer, 1993). This disaster
encouraged research on ice avalanches, especially on their triggering mechanisms. Since
then, an estimation of the dam strength is made against the waves generated in the
reservoir as a consequence of avalanches. The ice avalanche that occurred on 14 August
1949 on the Tour Glacier in the French Alps is well known; it caused the fall of the lower,
very steep part of the glacier, displacing about 3 million cubic metres that led to the death
of six hikers. The causes are interpreted as a result of the weakening of the terminus of
the glacier due to alternating hot days and cold nights, resulting in important melting and
re-freezing (Vivian, 1979).
If ice avalanches are relatively common in the Alps, they reach greater volumes and
catastrophic effects in the Andes. In the Peruvian Cordillera Blanca some peaks contain
glaciers above 6000 m, and Huascarfin (6654 m) is the highest (Figure 6.4 and Figure 6.5).
The Rio Santa valley located in this area was affected by two large ice and rock
avalanches. The first, in 1962 caused 4000 deaths, and the second, of greater dimensions,
was triggered by an earthquake of a 7.7 magnitude on the Richter scale on 31 May 1979. In
this second case 50 million cubic metres of ice, rock, debris, and water were mobilized,
reaching velocities of around 280 km h-~, and covering 16 km of the valley bottom in
3 min. The village of Yungay was buried and between 18,000 and 20,000 people died
(Figure 6.6). This avalanche seems to have been the most important in historical times
because of its greater runout, velocity, and volume, and must be considered as an ice and
rock avalanche (Plafker and Ericksen, 1978).
2.3. Glacial dammed lakes
These marginal lakes can originate at the junction of two valley glaciers or when a glacial
tongue penetrates into an ice-free tributary valley. As a consequence of damming, the
lakes are progressively filled up, particularly during the summer. If the lake is completely
filled up, then the water overflows by supraglacial or ice-marginal channels. Another
possibility occurs when the ice dam suddenly floats upwards and breaks, resulting in flash
floods downstream. These breaks are relatively unpredictable.
Figure 6.4. Line drawing showing the ice and debris avalanche (shaded) in Huascarfin, Peruvian Andes,
on May 31, 1970. (a) Diagrammatic representation of the fall. (b) Air-launching of debris near terminus of
moraine ridges. (c) Splitting of the debris into the Yungay and Ranrahirca lobes (modified from Plafker
and Ericksen, 1978).
The dammed lakes used to be more common in the Polar and subpolar regions and
drained throughout overflowing channels, whereas in the temperate-ice dams sudden
breakings are relatively common (Figure 6.7). These lakes of the temperate latitudes are
smaller than those of the Polar Regions, but their proximity to inhabited areas makes the
hazard higher, with very important damages (Hambrey and Alean, 1992).
Ice-dammed lakes sometimes drain periodically, thus indicating that water depth and
their consequent hydrostatic pressure are the main factor controlling the drainage of the
lake. Several mechanisms of ice-dam breaking have been proposed. When the hydrostatic
pressure of the water in the lake is higher than the ice-dam pressure, then the basal
breaking of the ice body occurs and the lake drains through a subglacial outlet. In other
Figure 6.6. Area where the village of Yungay was located (Cordillera Blanca, Peru) and buried by a
catastrophic debris avalanche in 1970 (Photo: J. L6pez-Martfnez).
Figure 6.7. Lake dammed by the Perito Moreno Glacier of Argentina, where periodic slides and
overflows occur.
cases the breaking is linked to an enlargement of cracks due to the ice flow and the high
hydrostatic pressure. Finally, the channels located on the ice mass enlarge progressively,
weakening the strength of the ice dam. In addition, seismic movements increase the
instability of the system (Costa, 1988).
These sudden outputs of glacial waters result in extreme floods, called j6kulhlaups
in Iceland, which mobilize large quantities of sediment and build the sandar or gravel
and sand plains (Maizels, 1997). Commonly this activity is catastrophic. Two types of
hydrographs can be recognized in these floods (Costa, 1988a). One shows a sudden raising
limb, with an acute peak flow, almost impossible to be measured adequately, followed by
an abrupt decreasing limb. The time interval between the beginning of the discharge
increase and the end of the decrease ranges from some minutes to hours. This hydrograph
corresponds to a sudden breaking of the ice dam. The other type of hydrograph (Figure 6.8)
shows a progressive increase, with an acute peak flow and a sharp decrease, corresponding
to the reservoir exhaustion. The time interval between the beginning and the end of the
discharge ranges from hours to days. The characteristics of this hydrograph suggest the
occurrence of floods caused by the enlargement of englacial tunnels or overflowing of
lakes (Church, 1988).
Most classical, or the better-studied examples ofj6kulhlaups occur in Iceland, although
in this country the volcanic activity interacts with the glaciers. The most dramatic cases are
related to the united activity of both phenomena and will be analysed later. It is interesting
to note, however, that during the recent retreat of the glaciers, after the Little Ice Age
(Grove, 1988), j6kulhlaups have been more frequent but with a lower discharge, due to the
thinning of the ice dams.
200
180
Total discharge: 5.94 x 106 m 3
160
140
t~
E 120
12}
I,,..
t~
t-
O 100
a
80
60
40
20
12 16 20 24 04 08 12 16 20 24 04 08
20 July 21 July
Figure 6.8. J6kulhlaupof July 1967, in Ekalugad Fjord, Baffin Island, generated by overflow of an ice
dam (partially modified from Church, 1988).
A number of cases of glacier dam breakings are known in the Alps. Since the year 1600
floods have been recorded in the Oetz valley of the Austrian Tyrolean Alps (Figure 6.9).
The tributary Vernagt Glacier penetrates into the fiver, causing the development of a lake
that has broken many times (Grove, 1988; Gerrard, 1990). These breakings are also
recognized in the Allalin Glacier of Switzerland and in the Gietro Glacier, with catastrophic
floods throughout its history, with 140 deaths in 1595 and 50 in 1818 (Vivian, 1979).
Likewise, a wealth of documentation exists on sudden breakings within alpine glaciers.
The most outstanding known catastrophe occurred on 12 July 1892, as a consequence of
the breaking of a subglacial cavity in the T~te Rousse Glacier on the west-facing slope of
Mont Blanc; the floods caused 175 deaths and great material losses. These breakings are
also known in the Miage Glacier, with a number of floods in the last two centuries, and in
the Trient Glacier, where they occur with a periodicity of 3 to 5 years in July or August
(Vivian, 1979).
Other types of dammed lakes are caused by end moraines that are filled up by melt and
rain waters (Figure 6.10). In these lakes the glacial till is easily eroded during the fast ice
melt or rainstorms. The overflowing channel quickly erodes the deposit and consequently
a rapid increase occurs in the discharge. Holes are drilled to reach the lake bottom and
Figure 6.10. Glacial-origin lake, dammed by an terminal moraine arc. Cordillera Blanca, Peru (Photo:
J. L6pez-Mart~nez).
drain the lake in order to avoid the spilling and subsequent destruction of the moraine dam.
A catastrophic destruction of an end-moraine lake occurred in the Santa Ana river valley
of the Peruvian Andes on 15 March 1941. It resulted in a large flood that destroyed the
village of Huaraz, causing the death of 6000 people (Hambrey and Alean, 1992).
Numerous such examples are also known in the Himalayas as well.
2.4. Volcanism and glaciers
When the heat liberated by volcanic activity affects glaciers, a sudden ice melt can occur.
The resulting water, mixed with volcanic ash, results in lahars or volcanic mudflows that
move by gravity, affecting human settlements and activities located at the foot of the
volcano. If the melt waters achieve large volumes, then quite dramatic floods can be
triggered. Three examples from different parts of the world serve to illustrate the dramatic
development of lahars and their significant consequences in both human lives and
economic losses. The crater lake of Ruapehu volcano (2797 m), located on the North
Island of New Zealand, broke out along with significant ice and snow melt on Christmas
eve, 1953, causing a lahar that destroyed a railway bridge immediately before the passage
of a train. The locomotive and five cars fell into the river and 151 people died (Francis,
1993).
The Mount St. Helens volcano (2950 m), located in the state of Washington, in the
Cascades Range, USA, begun its eruption on 18 May 1980, though its activity had started
2 months before with earthquakes and small phreatic eruptions. The volcano was subject
to close vigilance in hopes of avoiding greater problems (Crandell et al., 1975; Wesson,
1983). The big eruption caused by the landslide failure of the north face of the mountain,
lowered the height of the volcano ---400 m and triggered large ice avalanches. The ash
resulting from the explosion covered the neighbouring glaciers, which advanced in the
following few years as a consequence of the protection against solar radiation. On the
other hand, large lahars of up to 30 km in length developed, destroying houses and forests.
Some dams restrained them, and in other cases the trees contributed to damming the waters
in the bridges, resulting in floods when the bridges broke. Fifty-seven people died as a
consequence of this eruption, and the economic losses were estimated to be about $1000
million dollars (Lipman and Mullineaux, 1981; Hickson and Peterson, 1990).
The Nevado del Ruiz volcano (5400 m), located in the Colombian Andes at 5~ north
latitude, has an ice field in its summit. On November 13th, 1985, it had a low magnitude
eruption, with the emission of a few pyroclastic products. The eruption, as in previous
cases, was forecast 1 year before due to seismic activity and one phreatic explosion. In
spite of the low energy of the eruption, ice and snow avalanches were triggered, and
the hot ash melted ice masses; the resulting waters carried a large quantity of recent and
old ash, causing lahars or volcanic mudflows running toward the densely inhabited valleys
(Figure 6.11). The lahars developed in two main stages. On the east slope they were
channelized by the Lagunillas fiver, with velocities of 30 to 40 km h-1, until at midnight
they reached the city of Armero, located 60 km away from the volcano. The village
was buried in a few minutes by the lahars and 22,000 people died. On the west-facing
slope in the locality of Chinchinfi, mudflows killed 2000 people (Parra and Cepeda,
1990; Voight, 1990).
Figure 6.11. Map showing volcanic hazards on Nevado del Ruiz of Colombia, compiled in November
1985. According to Parra and Cepeda (1990) and Voight (1990) (in Chester, 1993).
An increase of temperature at the base of temperate or polar glaciers due to geothermal

heat results in melting of the ice mass and, as a consequence, the development of
subglacial lakes. Probably most of them are located under the Antarctic Ice Sheet
(Hambrey and Alean, 1992). The largest of all known subglacial lakes is Lake Vostok,
below almost 4 km of ice at the Russian Antarctica Station of the same name. The lake
has an area of 15,000 km 2, an average depth of 125 m or more, and a considerable
accumulation of sediment in the bottom (Kapitsa et al., 1996). The lakes under the small
Iceland ice sheets have a well-documented history of volcanic activity under the glacial
ice and their consequent floods. Ten percent of these ice sheets cover volcanically active
rift areas.
Near the middle of the Vatnaj6kull Ice Sheet, the Grfmsv6tn subglacial lake is located
in an active geothermal area where a number of eruptions have occurred in historical times
(Thorarinsson, 1953). When the lake breaks very large floods (j6kulhlaups) are triggered
on the Skeidarfirsandur plains (1000 km 2 with a gradient of 1/200) at the south end of
the ice sheet, thus building the greatest of Icelandic sandur. In past times an average
of 12 floods every 10 years has occurred, and the mean volume of each flood has been 3 to
3.5 km 3, with their maximum discharge being 10,000 m 3 s -~. Each j6kulhlaup has carried
about 30 million tons of sediment (Bjornsson, 1979). These floods can be triggered by an
increase of geothermal heat or by an eruption, like that of 1934, wherein the total sediment
transported was 150 million tons. It is very difficult to plan a permanent road network in
the Skeidarfirsandur as a consequence of the high frequency of these floods.
Under the Myrdalsj6kull Ice Sheet the Karla volcano has an eruptive activity well
known since 1625. The last eruption was in 1918 and the discharge reached
200,000 m 3 s-1. The southern coast advanced 200 m as a consequence of this eruption,
and a progradation of 2.2 to 2.5 km has been estimated since 1660 (Bjornsson, 1979).
3. Problems derived from glaciotectonics
Deformations caused by glacial activity can produce locally severe disturbances in some
geological structures, resulting in folds, fractures, and genesis of breccia in the strati-
graphic sequences, or in a reduction of the mechanical resistance of rocks and sediment.
Therefore, it is very important to have a detailed knowledge of these deformations,
because they can have consequences to ore mineral surveys, mining works, civil engineer-
ing, and soil mapping.
In some mines of western Canada important glaciotectonic activity has been recognized
(Aber et al., 1989): about 12,000,000 m 3 of coal have been mobilized by overthrusting and
a large depression was generated, and later covered with sediments (Figure 6.12). Because
the coal has been relocated closer to the surface, it has oxidized, thus reducing its caloric
value, and in addition it has increased in water content, so for these reasons it is not
exploited.
In other cases the substrate can be affected by glaciotectonic deformation, with a
consequent decrease in shear strength. Under these circumstances, slides of weakened
materials can affect roads and bridges, as occurs with the Upper Cretaceous argilites in
the South Saskatchewan River valley of Canada (Krahn et al., 1979).
X X'
G l a c i e r Flow
J
T r a n s p o r t e d Coal
Bedrock . . . . . . . . . Coal S u b c r o p
._ Zone
,,,,2~
" " Till " "
Figure 6.12. Schematic cross section showing the depression formed by glaciotectonism and the hill
located downglacier that includes some of the coal transported from the depression (after Aber et al., 1989).
4. Some considerations about engineering geology in areas

previously occupied by ice
The most outstanding feature of glacial-origin sequences is their high variability in

grain size (Figure 6.13), inner structure, depth, geometry, and surface morphologies.
Most commonly these changes occur in short distances. Obviously, this variability is
responsible for a number of problems in civil engineering issues (Derbyshire and Love,
1986). For this reason, detailed geomorphological and geological maps are of extreme
importance (Strachan and Dearman, 1983). Likewise, drilling campaigns are necessary to
assess the depth and geometry of these glaciogenic deposits. In addition, soil mechanics
tests provide understanding of the geotechnical properties and predictable behaviour of
the materials.
The use of till deposits and fluvioglacial sands and gravels as material for constructing
linear engineering works is very common. On the other hand, glaciomarine and glacio-
lacustrine silts and clays are low resistance areas within the glacial sequences, and show
serious stability problems in steep slopes (Legget, 1974).
Glaciomarine clays that have been raised in postglacial periods by glacioisostasy,
can be extremely dangerous because of their tendency to fail, which is referred to as
remoulding in the engineering literature. These marine sediments have a high sodium
chloride content that increases cohesion of the clay particles. Where salts are dissolved
100
100 100
% sand
9 Subglacial tills o Glaciofluvial
Supraglacial and 9 Ice contact glaciofluvial

meltout tills
§ Lacustrine
Figure 6.13. Glaciogenic sediments and grain size distribution (according to Derbyshire and Love,
1986).
by freshwater, cohesion reduces, the clays liquefy and they can flow through long
distances, even at low gradients. They are called quick clays due to this change in
behaviour and subsequent flow, and are very common in Canada and Scandinavia
(Veder, 1981; Aune, 1983).
One of the most important failures linked to the activity of quick clays is that triggered
in the coastal city of Anchorage, Alaska, on 27 March 1964, which was triggered by
the greatest earthquake recorded in North America. The failure seems to be due to the
presence of water-saturated glaciomarine clays (Figure 6.14). The landslide went into the
valleys and inlets and destroyed 75 houses of a residential area. In this type of failure
Figure 6.15. Major placer mining areas in western Canada in relation to the maximum limits of
Pleistocene glaciers (compiled by Levson and Morison, 1996).
the greatest damages were related to graben areas in the upper part and in the push ridges
of the frontal part (Hansen, 1965).
5. Placer deposits in glacial environments
Placers are accumulations of heavy metals that have been extracted from the rock substrate
by weathering, water, wind, and glacial processes. They are important economic resources
and are ores of gold, diamonds, platinum, tin, and other minerals. Placers subject to ice
action are exploited in northern Asia, Canada, and Alaska, though most of these
exploitations are in environments not affected by glacial activity (Boyle, 1979).
Placer deposits (Figure 6.15) are located in a wide range of sedimentary situations,
such as alluvial, colluvial, glacial, fluvioglacial, coastal, and marine environments (Levson
and Morison, 1996). Alluvial placers are located in active or abandoned streams, such as
gullies, braided and meandering rivers, alluvial fans, and fan-deltas. Gold concentrates
essentially together with loose blocks and gravels and in a conglomeratic facies (Fayzullin,
1969).
Fluvial placers are the most frequent type and are located in recent alluvial deposits and
in lower terraces. In meandering rivers they locate in point bars and in the internal curve of
the meanders. In braided rivers the distribution of heavy metals is discontinuous; they
locate in transverse bars, curves of the channel, junctions of channels and stretches with
convergent flow (Smith and Minter, 1980).
In alluvial fans, placers locate in interbedded gravels and in debris-flow sediments.
These develop after glacial retreat, when the absence of plant cover facilitates slope
instability. Once they fail, they are subject to erosion by glacial meltwaters.
Chapter 7
The periglacial environment
1. Introduction
The term "periglacial" was introduced in 1909 by the Pole, Lozinski, to designate the
climatic conditions and geomorphic features of the areas peripheral to the Pleistocene ice-
sheets. Subsequently, the term has been extended to cover the characteristic processes and
landforms of cold climates, regardless of their proximity to glaciers or glaciated areas.
There are vast zones such as eastern Siberia unconnected to the glacial areas where
periglacial activity predominates (Jahn, 1975). Due to the lack of precision in the
definition, the term periglacial has had very wide use. Thus, the word "geocryology,"
commonly used by the Russian researchers, has been introduced as an approximately
equivalent, alternative term (Washburn, 1979). This science is commonly defined as the
study of the terrestrial materials with temperatures below 0~ that is to say permanently or
seasonally frozen grounds (Washburn, 1979; French, 1996).
Periglacial environments are characterized by freeze-thaw cycles and the presence of
permafrost or permanently frozen ground. Either or both of them occur in the overall
periglacial domain. In this way, some of the periglacial landforms are not associated with
the presence of permafrost. The periglacial environment is distributed in polar zones and
in medium and low latitude alpine areas of many mountain ranges of planet Earth (Harris,
1988). At the present time, the periglacial domain covers close to one fifth of the global
land surface and an additional 20% of the area has undergone periglacial conditions during
the Pleistocene cold periods (French and Karte, 1988).
Many palaeoclimatologists recognize the importance of periglacial phenomena for
palaeoclimatic reconstructions and of permafrost for palaeoclimatic modelling (Isarin,
1997). Ecologists and environmental scientists have pointed out the vulnerability of
present-day periglacial regions due to their increased anthropogenic use and the
implications of expected climate change (Vandenberghe and Thorn, 2002).
Periglacial environments have a wide range of climatic conditions. The mean annual
temperature may be close to or far below the freezing point, and the range between the
maximum and minimum yearly temperatures is generally large. According to Peltier
(1950), the annual precipitation varies between 130 and 1400 mm and between 50 and
1250 mm following Wilson (1969). Tricart (1967) has established a climatic classification
based on the temperature, precipitation, wind conditions, and their seasonal distribution.
Three types of periglacial climates are distinguished:
(a) Dry climates with cold winters. These conditions are found in subpolar zones of the
northern hemisphere. These areas have very low temperatures in winter, very short
summers, low precipitation, and strong winds. This climate propagates the existence
of a perigelisol (permafrost). The morphogenetic system linked to this climate type is
characterized by a strong frost action, a limited runoff influence and significant wind
activity.
(b) Cold and wet climates with strong winters. The arctic and mountain types are
differentiated within this group. The arctic type has large climatic irregularities due to
the oceanic influence and has permafrost. The temperatures are similar to climate
A with a lower annual range, precipitation is above 300 mm, and violent winds are
common. Consequently the frost action is less intense or persistent than in climate A,
the snow cover inhibits eolian processes, and run-off is relatively important. The
mountain type occurs in the alpine areas of the temperate zones. Temperatures are
similar to the arctic type with higher mean annuals and lower annual ranges.
The annual precipitation is higher than in the arctic type. The role of slope and aspect
are very important in these areas. Therefore, frost action is important, permafrost
is generally absent, the wind influence is weak, and run-off-related processes are
conspicuous.
(c) Climates with small annual range in temperature. The mean temperature is close to
0~ and the annual temperature range is around 10~ Two types are distinguished.
The climate of high latitude islands has marked weather instability, small temperature
range, and snow precipitation above 400 mm that inhibits the effects of the wind.
Freeze-thaw cycles are abundant and have limited penetration into the ground. The
low latitude mountain climate types lack seasonal temperature variations and the daily
range is very high and larger than the annual range. Precipitation is commonly high,
except in the arid mountains. These climatic conditions engender the existence of
permafrost, numerous freeze-thaw cycles, faint penetration of the frost action, and
insignificant wind activity except in arid mountains.
In addition, the periglacial domains present two main types of vegetation, the subarctic
or northern forest, and the arctic tundra. This allows distinguishing between the forest
periglacial environments and the woodland-free environments (French, 1996).
2. Permafrost
Permafrost, also called pergelisol or permanently frozen ground, is defined as a soil,

surficial deposit or rock mass that remains continuously below 0~ for more than 2 years,
regardless of other properties like the moisture content and lithology (Muller, 1945).
Glaciers and ice-sheets are excluded. Permafrost covers around 22% of the land surface in
the Northern Hemisphere and extensive areas in the Andes and Antarctica (Harris, 1986).
According to Washburn (1979), 14% of the land surface on the Earth is underlain by
permafrost. It extends as far as 50% of Canada and 80% of Alaska (Harry, 1988).
Permafrost also occurs in the bottom of shallow seas in high latitudes (submarine
permafrost). Two types of permafrost occur on the continents; polar permafrost in high
latitude areas and alpine permafrost. Alpine permafrost is developed in mountain
environments and varies depending on the latitude and elevation. The altitudinal lower
limit of the permafrost rises from high to low latitudes (Harris, 1988).
The continuous permafrost wedges out towards low latitudes and breaks up into
patches giving way to the discontinuous and sporadic permafrost (Harris, 1986)
The periglacial environment 147
Figure 7.1. Distribution of different permafrost types in the Northern Hemisphere (compiled by Harris,
1986).
Figure 8.7. Relict-sortedstripes. Sierra de Javalambre, Teruel Province, Iberian Range, Spain.
(Figure 8.7). They are generally straight and may reach 120 m in length (Washburn, 1969).
The clasts are usually on edge with the major axis parallel to the stripes. The size of the
clasts generally decreases with depth and the sorting penetrates up to 1 m. In slopes from
1~ the sorted polygons change progressively into sorted stripes (Figure 8.1) due to the
sealing by mass movements of the sides perpendicular to the slope gradient (Furrer, 1968).
Although attempts have been made to unravel the origin of the patterned ground in
periglacial environments, its genesis is not well understood. There is not doubt that they
are polygenic and that similar forms may be due to different processes. Besides the activity
of certain processes may give place to different landforms (Washburn, 1979). Despite the
abundant literature devoted to the topic, the origin of periglacial patterned ground is still
an unsolved challenge.
2. Ice-cored mounds
The ice-cored mounds comprise a group of small hills generated by frost action. The
numerous terms proposed for their differentiation in the abundant literature have created
some confusion. These mounds may be perennial like palsas and pingos, or seasonal, like
the hydrolaccoliths or blisters. The latter ones reach 8 m in diameter and 3 to 50 cm in
height and have been little studied because of their ephemeral development restricted
to the winter season when they are commonly covered by snow. Their genesis is related to
the freezing of groundwater above the permafrost (Pollard, 1988).
Periglacial landforms 177
2.1. Palsas
These mounds occur in boggy areas with discontinuous or sporadic permafrost (Zoltai,
1971). Their surface is criss-crossed with open fissures caused by dilation or desiccation
(Embleton and King, 1975b). At the present time they are occur in areas with a mean
annual temperature below 0~ as in Canada, Iceland, Sweden, and alpine regions
(Kershaw and Gill, 1979). Palsas may be isolated or form in groups and are differentiated
based on the morphology of dome-shaped, ribbon-shaped, and platform palsas. The former
ones are between 0.5 and 7 m in height, 10 to 30 m wide and 15 to 150 m long. The larger
palsas tend to be less conical than the smaller ones. The ribbon-shaped palsas have a lower
height and may be parallel or perpendicular to the contour lines, reaching several hundred
meters in length. The platform palsas rise 1 to 1.5 m above the surrounding bog and may
cover several square kilometres. Their generation is related to differential thawing.
Finally, the paisa complexes are formed by several types of palsas at different growth
stages and show numerous closed depressions generated by thawing (Sepp/il/i, 1988).
Some authors use the term string bogs to designate the association of different types of
palsas.
Based on their internal structure, two types of palsas are differentiated. The palsas with
a peat core have an outer layer of peat that thaws and dries in summer. Around 80 to 90%
of the core is composed ice in small crystals and lenses of segregated ice whose thickness
increases with depth from 5 to 10 mm to 5 to 10 cm. The base is generally formed by silt-
sized mineral particles. The silt core palsas also have an outer peat layer and a silt and clay
core with ice lenses and veins with a reticular arrangement (Sepp~il/i, 1988). Lithalsas are
similar mounds, but without any peat cover (Harris, 1998). We can see that no definitive
agreement exists with regard to terminology (Pissart, 2000).
The origin of the palsas is largely related to the lower heat conductivity of the dry peat
with respect to the saturated and frozen peat, with a conductivity around twenty times
higher (Washburn, 1979). In winter the peat wets and increases its heat conductivity
favouring the inhomogeneous penetration of the freezing and the formation of segregated
ice. The consequent differential cryostatic pressure gives place to elevations in the surface.
The outer peat layer that dries in summer with low heat conductivity has an insulating
effect preventing the thawing of the inner part of the paisa. The palsas may disintegrate
gradually from the margins due to a rise in temperature. Finally, they may become shallow
ponds and a new stage of peat accumulation may take place (Friedman et al., 1971).
The remnants of lithalsas present encircling ramparts (Pissart, 2000). Seasonal palsas
linked to oozes have been reported in the Paramera of Avila Province (Central Spain) at
1200 m in altitude (Molina and Pellitero, 1982). Although most of the palsas of the
Scandinavian countries formed 1000 to 3000 years ago, absolute datings indicate that
some of them were formed recently (Sepp~il~i, 1988).
2.2. Pingos
"Pingo" is the Eskimo term for mound. They are also called hydrolaccoliths and in Siberia
are called bulgunniaks. The pingos are domed perennial ice-cored mounds that generally
bulged out in large plains. Recently, submarine pingos have been recognised on some
subsea platforms. It is not known whether they have grown under the sea or have formed
on land and have been invaded latterly by marine waters. The highest documented pingo in
emerged land above sea level is the 48-m high Ibyuk pingo, located close to the
Mackenzie Delta in Canada. Most of the pingos are less than 20 m high and their diameters
range between 30 and 600 m. There is an inverse relationship between the diameter and
height of the pingos. In some cases they are elongated, more than a kilometre in length
and maximum height of 9 m. Because they keep a constant diameter, the gradient of the
slopes increases during the growth and never exceeds 45 ~ (Pissart, 1988). Pingos show two
types of cracks. Dilation radial cracks converge in the top and result from the upward push
produced by the ice core during its growth stage (Washburn, 1979). The opening of the
cracks may lead to the partial thawing of the underlying ice giving place to a crater-like
subsidence depression in the centre. The concentric cracks, not so conspicuous as the
former ones, result from the thawing of the ice core during the negative growth stage of
the pingo (Mtiller, 1959). In contrast to the palsas, pingos have a massive ice core that
penetrates several meters beneath the land surface (Lundquist, 1969). The cover material
is made up of loose sediments like gravel, sand and silt, but pingos with a rock cover of
sandstone and shale are also known. This cover layer may reach up to 14 m thick. The
growth rate varies from very low values to 1.5 m/yr (Mackay, 1973). On the other hand, all
the known pingos are younger than 10,000 years old and some of them are a few hundred
years old. Finally, pingos disintegrate progressively due to a rise in temperature giving
place to thaw lakes.
Two methods of pingo formation have been suggested. The so-called closed-system
origin explains the pingos of the Mackenzie Delta (Canada) and Central Yakutia (Siberia)
developed in areas with thick continuous permafrost. They form in relation to lake
depressions. During the development of the permafrost, the freezing of entrapped water
within a lake creates a massive ice core by downward percolation and aggradation. The
volume increase generates a cryostatic pressure that domes the lake sediments and may
even cause the extrusion of water towards the surface. If the water reaches the surface it
may give place to aufeis and when it contains gas it may produce explosive activity.
Generally, the water does not reach the surface but it freezes generating a massive ice core
(Mackay, 1979) (Figure 8.8). The cryostatic origin of the pingos in the Mackenzie Delta is
supported by the fact that 98% of the 1380 mapped pingos are located at the edge or inside
of contemporaneous or old lakes (Stager, 1956).
A great proportion of the open-system pingos are located in areas of relatively thin
permafrost of Alaska and Greenland. They are generated by groundwater flowing under
artesian pressure through a thin permafrost or in taliks within the permafrost. This water
freezes as it forces its way upwards forming an ice core that domes the surface. The
hydrostatic pressure result from height differences (Holmes et al., 1968) (Figure 8.9) so
that they form in topographic lows like valley bottoms or distal sectors of low angle
slopes (Mtiller, 1959).
3. Slope morphology and evolution
The slopes in periglacial environments show a wide range of morphologies because they
develop under variable conditions of temperature, moisture, lithology and vegetation
Water
I
9 . . . ~ , . , .
(a)
Water
9 , ' . 9 . " 9 9 . 9 9 9 9 9
(b)
Injected water lens
(c)
II
~2' "''" i'J Saturated unfrozen sand
[ J Water
l I 1111 ! I I Perma'rost(and ac,,ve ,a~er)

f]llFlllllllilill ,ce
Figure 8.8. Development of a closed-system pingo by cryostatic pressure (Mackay, 1979).
S 0 1000 feet N
t 9
~ 1 1 1 1 I !
~ r I
_ ~r-~-m,,,,,~,j,,,,,,llllllll
.... - c,,,,,,,,,,,,~]],,,,,,,,,,,,,,t
l i l l l ] l l l l l l l l i l l l l i l ! !
. . . . i l l l l l l l l l l l l l l l l l l ~ l ~ l ~ l ~
I l l l l l i l l i ~ l i i l l l l l l l l l l l l l i l l l l l l i l i l l l | ] ] i l l l l l l l l l l l l l l i l l
,,,,,,,,,,,n, ..... ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,lllllllll
• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • ] • • • • • • •
, l l , l l l l l l i l l l i l l l l l l l l l l l i , i l l l l l l i l i l l l l l l l , i l i l l l [ l l l l l l l l l l l l l l l l l , l l l l i . . . . l l l l l l ] l I
••••••••••••••I•••••••••••••••|••••••••••••••••••••••••••••••••••••••••••••••••••••••••••
I !i! !i!il ! I ' ~

Frozen ground Unfrozen Pingo ice Bedrock
alluvium
Figure 8.9. Origin of an open-system pingo by hydraulic pressure (Holmes et al., 1968).
Fluvioglacial erosion and sedimentation l 17
250' A B C 250' D E . . . . . .~ F
1-- ----- 1 -- . . . .
~o0. . ........ ~oolt ..... :.~ . . . . . . . . . . . . . . . . . . . . .
1 0 0 ' - . . . . . . . . 1 0 0 ' ' ' "
250' G H I J _ K L M
o - - ~ 250'1
~ ~ I I I I I ~ I ~ ~ ~ ~ ~ ~ ~ I I I I I I I O I ~ ~ ~ ~ ~
200- 9 .
. . . . . . . ~200' 9 ~ ~ _ ,
,i ,~ 9 .'\ \'~\~ k" 9 " x \ \\ t .... ,
, ~ o ~ . m ,,.., o ~ . ~ . m ~ 1 7 6
100'J I 0 0 ' I~'*

'--"
9",, Crest of Esker
1948
V.E. = X 6 . 6 ~- " ' . Base of Esker
Crest of Esker
1963
~--~ Base of Esker
0 220 440 yds. - Lake of Esker 1948
9 I I
\\\\ Ice 1963
(a)
300' -
250' -
~_ I .,. : \'-," k ' ~ ' \ "~ " ",,. - ' , , ' \ " ~" ~'~,,. /
200' -
150' -
100' - "K. ~ ; -Z.., !z.,, .'.z- , ' ' / ~
50' --
,, ~" 9 Crest of Esker

1948
V.E. = X 6.6 9- - . Base of Esker
Crest of Esker
1963
~..~ Base of Esker
0 220 440 yds. \ \ \ \ Ice 1948
l ! I ///// Ice 1963
(b)
Figure 5.15. Profiles of Eskers in the Casement Glacier in Alaska (a) and in Breidamerkurj6kull in
Iceland (b) (Price, 1969).
Fluvioglacial deposits can be classified according to the sedimentation environment:

deposits due to ice melt channels, and deposits in confined waters, such as lakes and seas.
Also a classification can be established according to their position in relation to the ice
mass (Table 5.1). The proglacial sedimentary deposits occur at a certain distance from the
ice margin, generating smoothed morphologies, and the sedimentary structures are not
deformed. The deposits located in contact with the ice show distorted structures and their
morphologies are produced by the ice mass or by a later settling in relation to the melting
of buried ice. They are, in addition, classified according to the prevailing sediment and the
form of the accumulation.
Eskers (osar in Scandinavia) are ridges composed of stratified deposits with rounded
gravels and sands with some blocks. Their direction is parallel to the regional flow of the
ice. They are sinuous or straight, sometimes discontinuous, with sharpened or flattened
(a)
I(SE
(b)
Figure 5.16. Origin of kames. (a) Crevasses open until the ice-rock interface and development of a
cavity in which sediments accumulate. (b) Supraglacial and endoglacial cavities produce similar forms
(modified from Price, 1973).
crests, and lengths from a few metres to 400 km (Aylsworth and Shilts, 1989; Punkari,
1997) (Figure 5.13). Their height can reach 200 m, with 3 km in width. These latter
dimensions are related to the total length, in such a manner that the longest eskers are
commonly the widest and highest. They can also appear as unique ridges or develop an
interconnected network with confluent and diffluent eskers that join the main ridge. The
slopes of the eskers can have angles of about 30 ~ close to the angle of repose, though when
they are degraded the gradient is between 5 and 10~ (Embleton and King, 1975a). The
deposits dip toward the margins of the esker and can show distortions due to the under-
lying ice melt.
Eskers originate in subglacial, englacial, supraglacial, and ice-marginal channels
(Figure 5.14). They develop above all after the peak flow at the end of summer, when the
velocity of flow of the channels diminishes, and can be destroyed totally or partially with
new peak flows (Sugden and John, 1976; Menzies and Shilts, 1996). The deposits of the
bottom of the subglacial channels can accumulate between the ice walls, and the shape of
the cross section of the tunnels plays an important role in the final esker morphology
(Price, 1973). The hypothesis of subglacial channels hardly explains the existence of
tunnels hundreds of kilometres long. The origin from englacial channels must be treated
jointly with that of a supraglacial origin. Studies of aerial photos obtained in different
years made in areas of fast ice retreat of the Casement Glacier in southeast Alaska and the
Breidamerkurj6kull Ice Sheet in Iceland reveal that the profiles of the esker base and crest
have diminished in height (Price, 1966, 1969) (Figure 5.15). This diminution is due to the
melting of buried ice during the period between the aerial photos. This indicates an
englacial or supraglacial origin.
Figure 5.17. Kameterrace affected by badlands. Talon River, Jura of France. Photo: J.L. Pefia.
Figure 5.18. Grain-size distribution in some sandar. Skeidarfisandur is in Iceland, and the Scott and
Malaspina systems in Alaska (modified from Boothroyd and Nummedal, 1978).
Kames are small hills composed of deposits of stratified sands and gravels. Their size
varies from a few metres to conical hills of up to 50 m in height and 400 m in length. They
form by sediment accumulation in supraglacial and englacial cavities (Johnson and
Menzies, 1996) (Figure 5.16). The term kame is also used to describe forms built in a
particular position within or along the glacier. When a channel flows into a lake located on
the ice mass, it carries sediment deposited in the margins; after the ice melt, they result in a
kame delta (Figure 5.14). Kame terraces are formed by accumulation of channel deposits
that run between the valley wall and the lateral ice border (Figure 5.17). The side of the
deposit located close to the ice shows a steep border due to ice melt. A kame complex is
caused by accumulation of sediment in supraglacial depressions (karstic glacial
topography) and later settling by ice melt. The kames are modified by subsidence if
they have underlying ice, and can generate subcircular depressions called kettle holes.
When small hills are associated with depressions, then the term kame and kettle
topography is used, equivalent to the kame complex (Gray, 1991).
The discharge of large volumes of water from the ice margins causes the transport and
sedimentation of particles far away from the ice borders in different environments and
produces proglacial deposits (Table 5.1). In these areas, fluvioglacial sedimentation
prevails because of the decrease in the transport capacity of the ice melt water. The
discharge is associated with a network of braided channels of great lateral mobility,
depositing gravels and sands that build large depositional plains called outwash or sandur
(in plural sandar) (Maizels, 1995) (Figure 5.8). They are similar to alluvial fans, although
with a great influence of strong seasonal fluctuations. The grain sizes in the sandar
diminish from proximal to distal areas (Boothroyd and Nummedal, 1978) (Figure 5.18).
The longitudinal profiles are concave, whereas the cross sections are, in general, convex.
Sandur morphology is continually modified by extraordinary high-flow events as well as
daily variations in flow (Vandenberghe and Woo, 2002). The sandur valleys develop
where they are confined by glacial U-shaped valleys (Figure 5.19). Those developed in the
margins of ice sheets are called sandur plains (Figure 5.8). The proximal area in a sandur
has few ice-melt channels and is characterised by an area of gravels with a number of
closed depressions, resulting from the melting of buried ice; this is known as kettled
sandur or pitted outwash (Price, 1969) (Figure 5.20). Far away from the ice mass, the
Figure 5.19. Myrdalsj6kullIce Sheet in Iceland and sandur valley. Photo: F. Guti~rrez.
25. 1950
-~~~AGI~CIAL STREAM
20. ICE ~
{ 15.
NDUR
10. ICE
25 1951
2O
10
ICE
2oI 1952
o0
15
E
10
15
1955 KETTLED
__~ ~ ~ ~ . ~ SAyUR
~ ~ - - ~ . . . ~ - " ~ ~ ~ SANDUR
E
-~ ICE_~"
111 1960
KAME AND
~1 KETTLE
E
5 - ~ ICE
100m
Figure 5.20. Sketch showing the forms resulting from sedimentation by supraglacial and englacial
streams. The sequence has been obtained from aerial photos in Breidamerkurj6kull in Iceland (adapted
from Price, 1969, Figure 80).
number of braided channels increases. They are shallow (1 to 2 m in depth), change

frequently in position and commonly are flooded during high discharges. Finally, the distal
area has proglacial lakes. During the stages of glacial retreat, the proximal areas of the
sandar are eroded and new sandar develop from the previous ones.
The presence of lakes is very common in fluvioglacial environments, particularly in
marginal and proglacial positions. The discharge of materials in the lakes results in
glaciolacustrine beaches and deltas. If the lake is drained afterwards, then the bottom
deposits make up plains ofglaciolacustrine origin. Some of these deposits have a rhythmic
character and show seasonal variations. They are called varves and have been used in
absolute dating with diverse results. Each varve is composed of two layers, one of coarser
sand deposited in summer and the other of finer clay and silt that is deposited during the
rest of the year. The value of the method of counting varves for chronology construction
has been subject to criticism, because in some years more than a pair of layers can deposit,
whereas other years may not have a sedimentation record due to discharge variations or
deviations of the channels that flow into the lake (Smith et al., 1982).
Because tributary valleys are frequently dammed by the main valley glacier
(Figure 5.21), the moraine deposits can appear interbedded with fluvial and lacustrine
marginal deposits (Bordonau, 1993) (Figure 5.22). In glaciers close to the sea, the ice melt
water discharged in estuaries, fiords and bays causes the accumulation of a great quantity
of glacial and fluvioglacial sediment (Figure 5.23). If this sediment is uplifted as a result of
glacioisostatic rebound after a deglaciation, then it can result in raised beaches and deltas,
and exposed glaciomarine bottom muds, with similar morphologies to the glaciolacustrine
deposits. The identification of these deposits as marine is made by their fossil content.
Figure 5.21. Lateral moraine and dammed depressions in the Miera valley of Cantabria, Spain.
5
1800 4 rkk NW-SE
/~..~',~_
1500
Figure 5.22. Cross section of the Cerler lateral moraine complex, l~sera valley, Central Southern
Pyrenees, Spain. (3) Lateral moraine. (1 and 2) Older deposits. (4) Lacustrine deposits with interfingering
deposits as a result of glacial damming of a tributary valley. A later glacier advance in the tributary valley
deposited younger end moraine ridges. (5) All deposits related to the last glaciation. (Reprinted from
Bordonau, 1993).
Figure 5.23. Ice melting and sediment discharge in a glacier front, South Shetland Islands, Antarctica.
Photo: J. L6pez-Martfnez.
Chapter 6
Geomorphology applied to glacial regions
1. Introduction
Applied problems in glacial areas derive, on the one hand, from the activity of the actual
ice masses and, on the other hand, from the characteristics of the materials deposited by
the glaciers and meltwaters in past periods. Some of the functional processes that develop
from ice accumulations result in different geomorphic processes that will be analysed
below. From another point of view, these areas produce a number of benefits to human
communities. Those derived from tourism, such as landscape observation, skiing, trekking,
climbing, and other sports have undergone a considerable increase in the last few decades.
Summer ice melt in mountain areas leads to the exploitation of the water falling in order to
produce hydroelectric power. In addition, glacial melt water is a very important source for
irrigation in large arid zones of the world, such as the Thar Desert in northwest India,
irrigated by the Indira Gandhi and Great Bhakra Canals, the Peruvian-Chilean littoral
desert, and the areas of Mendoza and San Juan at the foot of the Argentinean Pre-
cordilleran Andes. All the discharges for these irrigation systems come from snow and ice
melt in the neighbouring mountain ranges.
Other important resources derive from the exploitation of the groundwater contained in
the Quaternary deposits of glacial origin. Thus, in North America, a large part of domestic
and industrial water is taken from fluvioglacial aquifers (Thornbury, 1954). One of the
most outstanding features of these accumulations is the variability in grain size and depth,
and consequently the different layers show different transmissivity, permeability and
storage capacity values (Derbyshire and Love, 1986). Till is a poor aquifer due to its clay
content, whereas gravel and sand deposits are excellent groundwater reservoirs (Lloyd,
1983) (Figure 6.1). The common lateral variations in the characteristics of these sediments
cause the occurrence of barriers for water flow, and they impede detailed knowledge of
the aquifer. In some cases, the presence of preglacial valleys buried and fossilized by these
sediments can result in areas of great water storage. For this reason, a precise knowledge
of the geomorphological history of any such region is necessary.
2. Hazards caused by the activity of ice masses
2.1. Icebergs
Icebergs are a consequence of the breaking of ice masses, especially ice sheets where
they arrive at the ocean (Figure 6.2). They are a risk for shipping and also for the oil drilling
platforms. This latter circumstance has encouraged studies on the forecasting of iceberg
-10
\
\
\ f
f
~20
'-30
-30
BOREHOLE
"40
%
9
LOCATION
GEOLOGICAL
f
f BOUNDARY
10 ROCKHEAD
J CONTOUR
SAND ISOPACHYTES
[
\
0-10m
10 -20 rn
TN
~ 20m+
/
0
,_
1
J
2
J km
/
5o
!
Figure 6.1. Structurecontour map of the top of the Sherwood Sandstone with isopach lines for overlying
glaciofluvial sands in the Fylde areas of Lancashire (Lloyd, 1983, in Eyles, 1983, Figure 15.8).
movements along the margins of the eastern Canadian coast (Marko et al., 1988). They can
travel for thousands of kilometres before melting. Their movement is controlled by the
surface oceanic currents, although it can be affected by deeper currents or by the wind.
Along the eastern coast of North America the icebergs come from the Greenland Ice
Sheet and travel with the Labrador Current. About one hundred icebergs per month cross
Figure 6.2. Icebergs between Newfoundland and Iceland (Photo: F. Pueyo).
the latitudinal lines during the period from March to August, although they rarely go
far south of 50~ (Figure 6.3). In Antarctica the icebergs travel parallel to the continent
from east to west until they strike currents coming from the east that move away them from
the coast (Hambrey and Alean, 1992).
2.2. Ice avalanches
The study of ice avalanches is very important, because they affect villages, roads, ski
resorts, and hydropower plants. They occur in mountain areas and are much more rare than
snow avalanches, although in some places they can be common and are a tourist attraction,
such as the steep face of Balmhom Glacier in Switzerland (Hambrey and Alean, 1992).
Research carried out with aerial photos enables knowledge of their maximum runout
distances, which is most valuable information for evaluating the risk. The maximum
extent of an ice avalanche depends on the terrain roughness and the season (McClung and
Schaerer, 1993). Thus, the velocity is greater in winter because the avalanche displaces
over a snow cover that facilitates the displacement; the opposite occurs in summer, when
the friction increases and the movement is less. Ice avalanches are forecast by new cracks
and ice falls some weeks before the main event is triggered. In some cases, ice avalanches
rise hundred of metres on the opposite slope.
The oldest known ice avalanche occurred in the Valais Canton of Switzerland, on
31 August 1597, where an ice mass falling from the Balmen Glacier affected the village
of Eggen, which was buffed with 81 inhabitants (Table 6.1). In the same canton,
on 30 August 1965, an ice mass fell in the Mattmark-Saas valley during the construction
o
80 ~ 100 ~ 90 ~ 80 ~ 70 ~ 60 ~ 50 ~ 40 ~ 30 ~ 20 ~ 10 ~
~ ........ ,O.
:i:i::.i:i:ii:i" :::::::::::::::::::::::::::::
.....
~ii=======================================================
9
.....,.......-...-...'.'.'.'.'.'.''.:.:......-! 70 ~
::!;i!~ii i !i!:!:!:i!i;;!~;ii!!)i~~
...'.'.'.'.i.:.:.i.:.:.:-:.:.:.:-:....-...'.-~ii~i~ ~
'":':'::'.'.'.'.:-~.:.:.:-:'i.:'i""'"'"
i . . . . . . .- -...-. . . . . . . . .,.,....
70 ~
i;i:i!i:ii:iiii:iiiii!!iiiii;iiiiii
,:,:,:
.:.2.:-:.2.2.:-2-2":-2".'.-.'."
t!:!:i:!:i:i
!ili!i!iiiiii!!!ii!i,r
. . . . . . . . .
t
60 ~
60 ~ t
SEA
I
,iii!!i!iiii!iii!ii!iiiiii
'_
~ . . . .'.'.'. .'.'.]-i.>:.i..i.i.
! CANADA.i:!~i.".~:i~L~.~:i:!:i:
,
%
%
::::::::::::::::::::::::::::::::::::::::: %%
: ""t LABRADOR ~f.
::: :::::::::::::::::::: :::::::: :,;-.;-:.:.: : : : : ~ :: ..
"" 1
% 50 ~
9. . . . . . . . . . . . . - . . . . . . . . . . . . . . . . . . . : . . . . . . . . . ~ . ~
5 0 ~ :.... . . . . . . :. - . . . . . . ~# . ~ ~ ........
.i.;.~i::i::::i::~:::T~;i::~::i~.~.i.i.~.'.'.'.'.'.-.'.'.'-
..................... ~.~.i.~.~'~.'..~ 9
-J l
P ATLANTIC OCEAN
70 ~ 60 ~ 50 ~ 40 ~
II HIBERNIA
mlm iii MAXIMUM JUNE EXTENT
OF ICEBERGS
ICEBERG DRIFT PATHS
I MODER ICEBERG PRODUCER
& MAJOR ICEBERG PRODUCER
Figure 6.3. The pattern of iceberg circulation off eastern North America (Marko et al., 1988, en E1-Sabh
and Murty, T.S. (eds.), 1988, p. 437, Figure 1).
Table 6.1. Some disasters related with glaciers in Switzerland (excluding those affecting
mountaineers) (Hambrey and Alean, 1992).
Year Location Event Fatalities

1597 Balmengletscher Ice avalanche 81
1965 Allalingletscher Ice avalanche 88
of a hydroelectric dam. The ice avalanche came from the Allalin Glacier tongue and
moved a million cubic metres of ice, burying the construction area and causing the death
of 88 workers (Hambrey and Alean, 1992; McClung and Schaerer, 1993). This disaster
encouraged research on ice avalanches, especially on their triggering mechanisms. Since
then, an estimation of the dam strength is made against the waves generated in the
reservoir as a consequence of avalanches. The ice avalanche that occurred on 14 August
1949 on the Tour Glacier in the French Alps is well known; it caused the fall of the lower,
very steep part of the glacier, displacing about 3 million cubic metres that led to the death
of six hikers. The causes are interpreted as a result of the weakening of the terminus of
the glacier due to alternating hot days and cold nights, resulting in important melting and
re-freezing (Vivian, 1979).
If ice avalanches are relatively common in the Alps, they reach greater volumes and
catastrophic effects in the Andes. In the Peruvian Cordillera Blanca some peaks contain
glaciers above 6000 m, and Huascarfin (6654 m) is the highest (Figure 6.4 and Figure 6.5).
The Rio Santa valley located in this area was affected by two large ice and rock
avalanches. The first, in 1962 caused 4000 deaths, and the second, of greater dimensions,
was triggered by an earthquake of a 7.7 magnitude on the Richter scale on 31 May 1979. In
this second case 50 million cubic metres of ice, rock, debris, and water were mobilized,
reaching velocities of around 280 km h-~, and covering 16 km of the valley bottom in
3 min. The village of Yungay was buried and between 18,000 and 20,000 people died
(Figure 6.6). This avalanche seems to have been the most important in historical times
because of its greater runout, velocity, and volume, and must be considered as an ice and
rock avalanche (Plafker and Ericksen, 1978).
2.3. Glacial dammed lakes
These marginal lakes can originate at the junction of two valley glaciers or when a glacial
tongue penetrates into an ice-free tributary valley. As a consequence of damming, the
lakes are progressively filled up, particularly during the summer. If the lake is completely
filled up, then the water overflows by supraglacial or ice-marginal channels. Another
possibility occurs when the ice dam suddenly floats upwards and breaks, resulting in flash
floods downstream. These breaks are relatively unpredictable.
Figure 6.4. Line drawing showing the ice and debris avalanche (shaded) in Huascarfin, Peruvian Andes,
on May 31, 1970. (a) Diagrammatic representation of the fall. (b) Air-launching of debris near terminus of
moraine ridges. (c) Splitting of the debris into the Yungay and Ranrahirca lobes (modified from Plafker
and Ericksen, 1978).
The dammed lakes used to be more common in the Polar and subpolar regions and
drained throughout overflowing channels, whereas in the temperate-ice dams sudden
breakings are relatively common (Figure 6.7). These lakes of the temperate latitudes are
smaller than those of the Polar Regions, but their proximity to inhabited areas makes the
hazard higher, with very important damages (Hambrey and Alean, 1992).
Ice-dammed lakes sometimes drain periodically, thus indicating that water depth and
their consequent hydrostatic pressure are the main factor controlling the drainage of the
lake. Several mechanisms of ice-dam breaking have been proposed. When the hydrostatic
pressure of the water in the lake is higher than the ice-dam pressure, then the basal
breaking of the ice body occurs and the lake drains through a subglacial outlet. In other
Figure 6.6. Area where the village of Yungay was located (Cordillera Blanca, Peru) and buried by a
catastrophic debris avalanche in 1970 (Photo: J. L6pez-Martfnez).
Figure 6.7. Lake dammed by the Perito Moreno Glacier of Argentina, where periodic slides and
overflows occur.
cases the breaking is linked to an enlargement of cracks due to the ice flow and the high
hydrostatic pressure. Finally, the channels located on the ice mass enlarge progressively,
weakening the strength of the ice dam. In addition, seismic movements increase the
instability of the system (Costa, 1988).
These sudden outputs of glacial waters result in extreme floods, called j6kulhlaups
in Iceland, which mobilize large quantities of sediment and build the sandar or gravel
and sand plains (Maizels, 1997). Commonly this activity is catastrophic. Two types of
hydrographs can be recognized in these floods (Costa, 1988a). One shows a sudden raising
limb, with an acute peak flow, almost impossible to be measured adequately, followed by
an abrupt decreasing limb. The time interval between the beginning of the discharge
increase and the end of the decrease ranges from some minutes to hours. This hydrograph
corresponds to a sudden breaking of the ice dam. The other type of hydrograph (Figure 6.8)
shows a progressive increase, with an acute peak flow and a sharp decrease, corresponding
to the reservoir exhaustion. The time interval between the beginning and the end of the
discharge ranges from hours to days. The characteristics of this hydrograph suggest the
occurrence of floods caused by the enlargement of englacial tunnels or overflowing of
lakes (Church, 1988).
Most classical, or the better-studied examples ofj6kulhlaups occur in Iceland, although
in this country the volcanic activity interacts with the glaciers. The most dramatic cases are
related to the united activity of both phenomena and will be analysed later. It is interesting
to note, however, that during the recent retreat of the glaciers, after the Little Ice Age
(Grove, 1988), j6kulhlaups have been more frequent but with a lower discharge, due to the
thinning of the ice dams.
200
180
Total discharge: 5.94 x 106 m 3
160
140
t~
E 120
12}
I,,..
t~
t-
O 100
a
80
60
40
20
12 16 20 24 04 08 12 16 20 24 04 08
20 July 21 July
Figure 6.8. J6kulhlaupof July 1967, in Ekalugad Fjord, Baffin Island, generated by overflow of an ice
dam (partially modified from Church, 1988).
A number of cases of glacier dam breakings are known in the Alps. Since the year 1600
floods have been recorded in the Oetz valley of the Austrian Tyrolean Alps (Figure 6.9).
The tributary Vernagt Glacier penetrates into the fiver, causing the development of a lake
that has broken many times (Grove, 1988; Gerrard, 1990). These breakings are also
recognized in the Allalin Glacier of Switzerland and in the Gietro Glacier, with catastrophic
floods throughout its history, with 140 deaths in 1595 and 50 in 1818 (Vivian, 1979).
Likewise, a wealth of documentation exists on sudden breakings within alpine glaciers.
The most outstanding known catastrophe occurred on 12 July 1892, as a consequence of
the breaking of a subglacial cavity in the T~te Rousse Glacier on the west-facing slope of
Mont Blanc; the floods caused 175 deaths and great material losses. These breakings are
also known in the Miage Glacier, with a number of floods in the last two centuries, and in
the Trient Glacier, where they occur with a periodicity of 3 to 5 years in July or August
(Vivian, 1979).
Other types of dammed lakes are caused by end moraines that are filled up by melt and
rain waters (Figure 6.10). In these lakes the glacial till is easily eroded during the fast ice
melt or rainstorms. The overflowing channel quickly erodes the deposit and consequently
a rapid increase occurs in the discharge. Holes are drilled to reach the lake bottom and
Figure 6.10. Glacial-origin lake, dammed by an terminal moraine arc. Cordillera Blanca, Peru (Photo:
J. L6pez-Mart~nez).
drain the lake in order to avoid the spilling and subsequent destruction of the moraine dam.
A catastrophic destruction of an end-moraine lake occurred in the Santa Ana river valley
of the Peruvian Andes on 15 March 1941. It resulted in a large flood that destroyed the
village of Huaraz, causing the death of 6000 people (Hambrey and Alean, 1992).
Numerous such examples are also known in the Himalayas as well.
2.4. Volcanism and glaciers
When the heat liberated by volcanic activity affects glaciers, a sudden ice melt can occur.
The resulting water, mixed with volcanic ash, results in lahars or volcanic mudflows that
move by gravity, affecting human settlements and activities located at the foot of the
volcano. If the melt waters achieve large volumes, then quite dramatic floods can be
triggered. Three examples from different parts of the world serve to illustrate the dramatic
development of lahars and their significant consequences in both human lives and
economic losses. The crater lake of Ruapehu volcano (2797 m), located on the North
Island of New Zealand, broke out along with significant ice and snow melt on Christmas
eve, 1953, causing a lahar that destroyed a railway bridge immediately before the passage
of a train. The locomotive and five cars fell into the river and 151 people died (Francis,
1993).
The Mount St. Helens volcano (2950 m), located in the state of Washington, in the
Cascades Range, USA, begun its eruption on 18 May 1980, though its activity had started
2 months before with earthquakes and small phreatic eruptions. The volcano was subject
to close vigilance in hopes of avoiding greater problems (Crandell et al., 1975; Wesson,
1983). The big eruption caused by the landslide failure of the north face of the mountain,
lowered the height of the volcano ---400 m and triggered large ice avalanches. The ash
resulting from the explosion covered the neighbouring glaciers, which advanced in the
following few years as a consequence of the protection against solar radiation. On the
other hand, large lahars of up to 30 km in length developed, destroying houses and forests.
Some dams restrained them, and in other cases the trees contributed to damming the waters
in the bridges, resulting in floods when the bridges broke. Fifty-seven people died as a
consequence of this eruption, and the economic losses were estimated to be about $1000
million dollars (Lipman and Mullineaux, 1981; Hickson and Peterson, 1990).
The Nevado del Ruiz volcano (5400 m), located in the Colombian Andes at 5~ north
latitude, has an ice field in its summit. On November 13th, 1985, it had a low magnitude
eruption, with the emission of a few pyroclastic products. The eruption, as in previous
cases, was forecast 1 year before due to seismic activity and one phreatic explosion. In
spite of the low energy of the eruption, ice and snow avalanches were triggered, and
the hot ash melted ice masses; the resulting waters carried a large quantity of recent and
old ash, causing lahars or volcanic mudflows running toward the densely inhabited valleys
(Figure 6.11). The lahars developed in two main stages. On the east slope they were
channelized by the Lagunillas fiver, with velocities of 30 to 40 km h-1, until at midnight
they reached the city of Armero, located 60 km away from the volcano. The village
was buried in a few minutes by the lahars and 22,000 people died. On the west-facing
slope in the locality of Chinchinfi, mudflows killed 2000 people (Parra and Cepeda,
1990; Voight, 1990).
Figure 6.11. Map showing volcanic hazards on Nevado del Ruiz of Colombia, compiled in November
1985. According to Parra and Cepeda (1990) and Voight (1990) (in Chester, 1993).
An increase of temperature at the base of temperate or polar glaciers due to geothermal

heat results in melting of the ice mass and, as a consequence, the development of
subglacial lakes. Probably most of them are located under the Antarctic Ice Sheet
(Hambrey and Alean, 1992). The largest of all known subglacial lakes is Lake Vostok,
below almost 4 km of ice at the Russian Antarctica Station of the same name. The lake
has an area of 15,000 km 2, an average depth of 125 m or more, and a considerable
accumulation of sediment in the bottom (Kapitsa et al., 1996). The lakes under the small
Iceland ice sheets have a well-documented history of volcanic activity under the glacial
ice and their consequent floods. Ten percent of these ice sheets cover volcanically active
rift areas.
Near the middle of the Vatnaj6kull Ice Sheet, the Grfmsv6tn subglacial lake is located
in an active geothermal area where a number of eruptions have occurred in historical times
(Thorarinsson, 1953). When the lake breaks very large floods (j6kulhlaups) are triggered
on the Skeidarfirsandur plains (1000 km 2 with a gradient of 1/200) at the south end of
the ice sheet, thus building the greatest of Icelandic sandur. In past times an average
of 12 floods every 10 years has occurred, and the mean volume of each flood has been 3 to
3.5 km 3, with their maximum discharge being 10,000 m 3 s -~. Each j6kulhlaup has carried
about 30 million tons of sediment (Bjornsson, 1979). These floods can be triggered by an
increase of geothermal heat or by an eruption, like that of 1934, wherein the total sediment
transported was 150 million tons. It is very difficult to plan a permanent road network in
the Skeidarfirsandur as a consequence of the high frequency of these floods.
Under the Myrdalsj6kull Ice Sheet the Karla volcano has an eruptive activity well
known since 1625. The last eruption was in 1918 and the discharge reached
200,000 m 3 s-1. The southern coast advanced 200 m as a consequence of this eruption,
and a progradation of 2.2 to 2.5 km has been estimated since 1660 (Bjornsson, 1979).
3. Problems derived from glaciotectonics
Deformations caused by glacial activity can produce locally severe disturbances in some
geological structures, resulting in folds, fractures, and genesis of breccia in the strati-
graphic sequences, or in a reduction of the mechanical resistance of rocks and sediment.
Therefore, it is very important to have a detailed knowledge of these deformations,
because they can have consequences to ore mineral surveys, mining works, civil engineer-
ing, and soil mapping.
In some mines of western Canada important glaciotectonic activity has been recognized
(Aber et al., 1989): about 12,000,000 m 3 of coal have been mobilized by overthrusting and
a large depression was generated, and later covered with sediments (Figure 6.12). Because
the coal has been relocated closer to the surface, it has oxidized, thus reducing its caloric
value, and in addition it has increased in water content, so for these reasons it is not
exploited.
In other cases the substrate can be affected by glaciotectonic deformation, with a
consequent decrease in shear strength. Under these circumstances, slides of weakened
materials can affect roads and bridges, as occurs with the Upper Cretaceous argilites in
the South Saskatchewan River valley of Canada (Krahn et al., 1979).
X X'
G l a c i e r Flow
J
T r a n s p o r t e d Coal
Bedrock . . . . . . . . . Coal S u b c r o p
._ Zone
,,,,2~
" " Till " "
Figure 6.12. Schematic cross section showing the depression formed by glaciotectonism and the hill
located downglacier that includes some of the coal transported from the depression (after Aber et al., 1989).
4. Some considerations about engineering geology in areas

previously occupied by ice
The most outstanding feature of glacial-origin sequences is their high variability in

grain size (Figure 6.13), inner structure, depth, geometry, and surface morphologies.
Most commonly these changes occur in short distances. Obviously, this variability is
responsible for a number of problems in civil engineering issues (Derbyshire and Love,
1986). For this reason, detailed geomorphological and geological maps are of extreme
importance (Strachan and Dearman, 1983). Likewise, drilling campaigns are necessary to
assess the depth and geometry of these glaciogenic deposits. In addition, soil mechanics
tests provide understanding of the geotechnical properties and predictable behaviour of
the materials.
The use of till deposits and fluvioglacial sands and gravels as material for constructing
linear engineering works is very common. On the other hand, glaciomarine and glacio-
lacustrine silts and clays are low resistance areas within the glacial sequences, and show
serious stability problems in steep slopes (Legget, 1974).
Glaciomarine clays that have been raised in postglacial periods by glacioisostasy,
can be extremely dangerous because of their tendency to fail, which is referred to as
remoulding in the engineering literature. These marine sediments have a high sodium
chloride content that increases cohesion of the clay particles. Where salts are dissolved
100
100 100
% sand
9 Subglacial tills o Glaciofluvial
Supraglacial and 9 Ice contact glaciofluvial

meltout tills
§ Lacustrine
Figure 6.13. Glaciogenic sediments and grain size distribution (according to Derbyshire and Love,
1986).
by freshwater, cohesion reduces, the clays liquefy and they can flow through long
distances, even at low gradients. They are called quick clays due to this change in
behaviour and subsequent flow, and are very common in Canada and Scandinavia
(Veder, 1981; Aune, 1983).
One of the most important failures linked to the activity of quick clays is that triggered
in the coastal city of Anchorage, Alaska, on 27 March 1964, which was triggered by
the greatest earthquake recorded in North America. The failure seems to be due to the
presence of water-saturated glaciomarine clays (Figure 6.14). The landslide went into the
valleys and inlets and destroyed 75 houses of a residential area. In this type of failure
Figure 6.15. Major placer mining areas in western Canada in relation to the maximum limits of
Pleistocene glaciers (compiled by Levson and Morison, 1996).
the greatest damages were related to graben areas in the upper part and in the push ridges
of the frontal part (Hansen, 1965).
5. Placer deposits in glacial environments
Placers are accumulations of heavy metals that have been extracted from the rock substrate
by weathering, water, wind, and glacial processes. They are important economic resources
and are ores of gold, diamonds, platinum, tin, and other minerals. Placers subject to ice
action are exploited in northern Asia, Canada, and Alaska, though most of these
exploitations are in environments not affected by glacial activity (Boyle, 1979).
Placer deposits (Figure 6.15) are located in a wide range of sedimentary situations,
such as alluvial, colluvial, glacial, fluvioglacial, coastal, and marine environments (Levson
and Morison, 1996). Alluvial placers are located in active or abandoned streams, such as
gullies, braided and meandering rivers, alluvial fans, and fan-deltas. Gold concentrates
essentially together with loose blocks and gravels and in a conglomeratic facies (Fayzullin,
1969).
Fluvial placers are the most frequent type and are located in recent alluvial deposits and
in lower terraces. In meandering rivers they locate in point bars and in the internal curve of
the meanders. In braided rivers the distribution of heavy metals is discontinuous; they
locate in transverse bars, curves of the channel, junctions of channels and stretches with
convergent flow (Smith and Minter, 1980).
In alluvial fans, placers locate in interbedded gravels and in debris-flow sediments.
These develop after glacial retreat, when the absence of plant cover facilitates slope
instability. Once they fail, they are subject to erosion by glacial meltwaters.
Chapter 7
The periglacial environment
1. Introduction
The term "periglacial" was introduced in 1909 by the Pole, Lozinski, to designate the
climatic conditions and geomorphic features of the areas peripheral to the Pleistocene ice-
sheets. Subsequently, the term has been extended to cover the characteristic processes and
landforms of cold climates, regardless of their proximity to glaciers or glaciated areas.
There are vast zones such as eastern Siberia unconnected to the glacial areas where
periglacial activity predominates (Jahn, 1975). Due to the lack of precision in the
definition, the term periglacial has had very wide use. Thus, the word "geocryology,"
commonly used by the Russian researchers, has been introduced as an approximately
equivalent, alternative term (Washburn, 1979). This science is commonly defined as the
study of the terrestrial materials with temperatures below 0~ that is to say permanently or
seasonally frozen grounds (Washburn, 1979; French, 1996).
Periglacial environments are characterized by freeze-thaw cycles and the presence of
permafrost or permanently frozen ground. Either or both of them occur in the overall
periglacial domain. In this way, some of the periglacial landforms are not associated with
the presence of permafrost. The periglacial environment is distributed in polar zones and
in medium and low latitude alpine areas of many mountain ranges of planet Earth (Harris,
1988). At the present time, the periglacial domain covers close to one fifth of the global
land surface and an additional 20% of the area has undergone periglacial conditions during
the Pleistocene cold periods (French and Karte, 1988).
Many palaeoclimatologists recognize the importance of periglacial phenomena for
palaeoclimatic reconstructions and of permafrost for palaeoclimatic modelling (Isarin,
1997). Ecologists and environmental scientists have pointed out the vulnerability of
present-day periglacial regions due to their increased anthropogenic use and the
implications of expected climate change (Vandenberghe and Thorn, 2002).
Periglacial environments have a wide range of climatic conditions. The mean annual
temperature may be close to or far below the freezing point, and the range between the
maximum and minimum yearly temperatures is generally large. According to Peltier
(1950), the annual precipitation varies between 130 and 1400 mm and between 50 and
1250 mm following Wilson (1969). Tricart (1967) has established a climatic classification
based on the temperature, precipitation, wind conditions, and their seasonal distribution.
Three types of periglacial climates are distinguished:
(a) Dry climates with cold winters. These conditions are found in subpolar zones of the
northern hemisphere. These areas have very low temperatures in winter, very short
summers, low precipitation, and strong winds. This climate propagates the existence
of a perigelisol (permafrost). The morphogenetic system linked to this climate type is
characterized by a strong frost action, a limited runoff influence and significant wind
activity.
(b) Cold and wet climates with strong winters. The arctic and mountain types are
differentiated within this group. The arctic type has large climatic irregularities due to
the oceanic influence and has permafrost. The temperatures are similar to climate
A with a lower annual range, precipitation is above 300 mm, and violent winds are
common. Consequently the frost action is less intense or persistent than in climate A,
the snow cover inhibits eolian processes, and run-off is relatively important. The
mountain type occurs in the alpine areas of the temperate zones. Temperatures are
similar to the arctic type with higher mean annuals and lower annual ranges.
The annual precipitation is higher than in the arctic type. The role of slope and aspect
are very important in these areas. Therefore, frost action is important, permafrost
is generally absent, the wind influence is weak, and run-off-related processes are
conspicuous.
(c) Climates with small annual range in temperature. The mean temperature is close to
0~ and the annual temperature range is around 10~ Two types are distinguished.
The climate of high latitude islands has marked weather instability, small temperature
range, and snow precipitation above 400 mm that inhibits the effects of the wind.
Freeze-thaw cycles are abundant and have limited penetration into the ground. The
low latitude mountain climate types lack seasonal temperature variations and the daily
range is very high and larger than the annual range. Precipitation is commonly high,
except in the arid mountains. These climatic conditions engender the existence of
permafrost, numerous freeze-thaw cycles, faint penetration of the frost action, and
insignificant wind activity except in arid mountains.
In addition, the periglacial domains present two main types of vegetation, the subarctic
or northern forest, and the arctic tundra. This allows distinguishing between the forest
periglacial environments and the woodland-free environments (French, 1996).
2. Permafrost
Permafrost, also called pergelisol or permanently frozen ground, is defined as a soil,

surficial deposit or rock mass that remains continuously below 0~ for more than 2 years,
regardless of other properties like the moisture content and lithology (Muller, 1945).
Glaciers and ice-sheets are excluded. Permafrost covers around 22% of the land surface in
the Northern Hemisphere and extensive areas in the Andes and Antarctica (Harris, 1986).
According to Washburn (1979), 14% of the land surface on the Earth is underlain by
permafrost. It extends as far as 50% of Canada and 80% of Alaska (Harry, 1988).
Permafrost also occurs in the bottom of shallow seas in high latitudes (submarine
permafrost). Two types of permafrost occur on the continents; polar permafrost in high
latitude areas and alpine permafrost. Alpine permafrost is developed in mountain
environments and varies depending on the latitude and elevation. The altitudinal lower
limit of the permafrost rises from high to low latitudes (Harris, 1988).
The continuous permafrost wedges out towards low latitudes and breaks up into
patches giving way to the discontinuous and sporadic permafrost (Harris, 1986)
The periglacial environment 147
Figure 7.1. Distribution of different permafrost types in the Northern Hemisphere (compiled by Harris,
1986).
Desert surfaces: pavements, patterned ground, varnishes and crusts 265
(a) Coarse/Medium Sand
""-. ,.//
/ ',, ./ ....-" ;
-~........... f- ... ,,, . ..
......... - ,,,' ,.." ;, .,,,-_':_....
""-.. . . . . . - I ~. . . . . . . . . . / //
~. ........ ,/ /s s
............ -~'- ........... ~,.," s/ ~ ...........
~
iS ,/ ~ ..~"
........... .,e.. . . . . . . . . . . . . . ,," ___'~--" - .......
.,. .,- ,,, \ ---/.-"
I ,,,i \ . ..____..--
0 cm 5
! , J
(b) Clay
/'-" r162, - ' '"

.~.~1 t',.,
0 cm 5
i - I
Figure 12.6. Surface cracking pattern produced after 12 freeze-thaw cycles under experimental
conditions. (a) coarse-medium sand. (b) clay (Maizels, 1981, in Cooke et al., 1993).
All of these cracks form in hot deserts with flat topographies, which in the majority of
cases correspond to lacustrine depressions. The occurrence of locally developed giant
desiccation fissures (Figure 12.10) is also possible, with lengths of kilometres and depths
greater than 1 m. Their origin is very complex and many causes for their formation have
been put forward. They include cracking caused by the subsidence due to extraction of
subterranean water, fissurisation by hydrocompaction, contraction from desiccation,
seismic activity, and so forth. All of these processes may act alone or in combination.
In playa environments profuse nonorthogonal systems develop in salty clays, which
comprise mud plains at the margins of evaporite sedimentation (Figure 12.5). The same
system also occurs in gypsum deposits (Tucker, 1978), which on occasion exhibit thrusted
borders (Figure 12.11). In Death Valley, California, sorted polygons develop in gypsum-
rich material (Hunt and Washburn, 1966) (Figure 12.12), where cracking is generated by
processes of salt contraction as a consequence of drying or freezing, and expansion due
Figure 12.8. Nonorthogonal system of convex desiccation cracks with raindrop prints. Closed depression
of La Playa, Bujaraloz, Province of Zaragoza, Ebro Basin, Spain.
(a) (b) (c) (d)
NONORTHOGONAL REGULAR RANDOM IRREGULARRANDOM ORIRNTEDORTHOGONAL

ORTHOGONAL ORTHOGONAL
Figure 12.9. Common systems of desiccation cracks (after Lachenbruch, 1962, and Neal, 1965, in Cooke
and Warren, 1973, Figure 2.25).
.,o., .... ,....,.
,,"'('no coverage
" ' " " ",..
,' available) "'.
CERBAT p
iI
MOUNTAINS 9 i
,. ..... . s -p"
,s \
,,
,.," % ',,
1 ',,
|0 ",,. I , site I ,/,,,J/ '4 ~(
/ ~ fresh ~" t~ - , ~ 1 f l,~,k
~) fissures ~'~'~~ ~ . . _'~.," ,
l z ,~'~...... "'-~-,4;,~.~"5"," ~ "
I=_ "
iF / ~" "~fresh \
~ "X ~ fissures ~
// .
0 1.5
km ._,~',,o -'~ .
Figure 12.10. A generalised map of giant desiccation polygons, earth fissures and areas of recent
fissuring, Red Lake Arizona, based on the interpretation of aerial photographs. The fissures are attributed
to long-term desiccation and local subsidence due to water extraction and/or subsurface salt flow (after
Lister and Secrest, 1985, in Cooke et al., 1993, Figure 8.6).
Figure 12.11. Gypsum polygons with laterally expanded edges, Sabkha Matti, United Arab Emirates.
to hydration or heating. These processes, together with the mobilisation of salt by capillary
action, appear to be the fundamental reasons for the ordering and classification of gypsum-
rich soils. Equally, polygons developed over salt surfaces on both littoral and interior
sebkhas are very common (Figure 12.13). Such polygons may have suffered a complex
evolution.
3.2. Gilgai
Gilgai is an Australian vernacular term meaning pool, which is increasingly used in

scientific literature. It is used to describe very small mounds developed on clay-rich soils
Troughscontainingstones
Stonesin interiorof
polygonsreston
pedestals;some
9 s s
~ ~ j~/Roc~ ~ ~
RockSalt ~ ~ Gypsiferous
a ndGsYtPs!erol
ut
Wedge-shaped
massof hardened
gypsumaboveopen 0 metres 1
crackin rocksalt 9 . t
Figure 12.12. Sorted stone polygons related to ridges and cracks in an underlying layer of rock salt,
Death Valley, California (after Hunt and Washburn, 1960, in Cooke and Warren, 1973, Figure 2.30).
Figure 12.13. Polygonswith laterally expanded edges developed on a salt surface. E1 Melah Sebkha.
Zarzis, southern Tunisia.
and desert pavements (Hallsworth et al., 1955; Verger, 1964; Harris, 1968; Mabbutt, 1977;
Hubble et al., 1983; Dixon, 1994a).
Following Harris (1968) a gilgai can be differentiated into three parts on the basis of its
cross section (Figure 12.14). A channel, which constitutes a depression that can be 2.5 m
below the general ground level and can reach 12 m in length. The general level of the
microrelief is denoted as the shelf, above which small mounds of up to 50 m diameter stand
out. Some of the authors previously cited differentiate six gilgai types (Dixon, 1994a),
which in the main reflect the simple morphological classification of Verger (1964), based
on the form and mode of gilgai grouping (Figure 12.14).
Various processes and factors intervene in gilgai formation, which are developed in the
monographs of Cooke and Warren (1973) and Mabbutt (1977). Gilgai form in areas of
alternating humid and dry conditions. Dry periods, during which the surface cracks,
alternate with humid periods in which swelling dominates. They occur mainly in vertisol
areas (black expansive clay-rich soils) in tropical and subtropical zones. Gilgai microrelief
results from upward movement produced by swelling of the humid subsoil. This
expansivity, along with soil contraction, as a consequence of humidifying and drying
cycles, explains the genesis of circular and stepped gilgai (Ollier, 1966) (Figure 12.15).
Swelling is due to the presence of expansive clays, such as montmorillonite, and the
existence of elevated interchangeable sodium percentages. The gilgai amplitude increases
with increasing expansive clay and sodium content. The expansion which gives rise to
gilgai can be seen in the soil profile, where small fault surfaces and thrust planes are
relatively common. Similarly, stakes placed in the soil are found tilted or even expelled
Mound/ Mound/
Puff Depression/ Puff
~//Channel ''~--" Vertical
~ ~ e r v a l ~
GROUPINGS
BASIC Without Single
FORMS prefered prefered Several
orientation orientation orientation
a ~
a~~
0 #,#
"t3
t-
= b ~o
o
CIIIII
L
0
c~~
a 1
O~~ 0 aa e
o o0 0
t-
O
Q O o 0
0 0 O0 o ~
(/)
o ~ 0 0 e o O~
~0
~L
0
(D
0 0 ~%0 0 0
t- 0
t-
O,
I
I!
Figure 12.14. In the upper part, cross section of a gilgai (after Harris, 1968). In the middle-lower part,
classification of gilgai morphology (Verger, 1964).
after a year. Another way of producing gilgai is given by Hallsworth et al. (1955) and
expressed graphically by Verger (1964) (Figure 12.16). During a dry period, particles
fall into the interior of the soil cracks and this added material is believed to cause the
microrelief during humid periods, because it provokes significant pressure deep within the
soil. It is also important to distinguish between the various mechanisms due to the action of
distinct forces in diverse situations, in relation to cracking and the expansion-contraction
cycles (Knight, 1980), which helps in understanding gilgai genesis more adequately.
4. Desert varnish
A fine, dark film developed on exposed material commonly occurs in arid regions,
and this is called desert varnish or rock varnish. In some deserts, such as those in SW
USA, this covering occurs on 75% of the rocky surfaces (Engle and Sharp, 1958).
Figure 12.15. Developmentof circular and stepped gilgai in the central Australian Stony Desert (after
Oilier, 1966).
However, varnishes also occur in alpine, coastal, fluvial, and sub-surface environments, as
well as near springs (Dorn and Oberlander, 1982; Dorn, 1998). The dominant color is
black or brown, due to enrichment in Mn and Fe.
The origin of varnishes has been the subject of numerous discussions, often lacking
precise scientific foundations. With the development of various high-resolution techniques
(microscopy and electron microprobe), which permit detailed observations and spot
chemical analyses, the understanding of varnishes has progressed considerably. This work
started with Engle and Sharp (1958) and since then many advances have been made with
regards to composition, structure and origin (Dorn and Oberlander, 1981, 1982), and
important results were expected (Oberlander, 1994; Dorn, 1998).
4.1. Composition and age
Varnish covers rocky exposures (Figure 12.17) or surface stones (Figure 12.18). The color
tends to be black if it is rich in Mn, more orange if high in Fe and brown if both are
present in similar amounts. The varnish film is very thin and varies greatly in thickness
(0.002-0.5 mm), although it generally oscillates between 0.01 and 0.03 mm.
(a)
ii __
(b)
(c) . . . . . . .
-_-_ ;,- s --
Dry zone d Crackfillin9

l particles
Wetzone -~ Pressure
Figure 12.16. The formation of gilgai by wetting and drying (after Verger, 1964).
Figure 12.17. Bedrock surfaces covered by desert varnish. Anti-Atlas, Morocco. Photo by J.L. Pefia.
Figure 12.18. Varnishfilm on clasts of a desert pavement. Negev Desert, Israel.
It develops on all rock types, but mostly on silicates. Varnish is composed mainly of Si,
A1, Mn and Fe, with Ca, K, Na, Ba, Ti, Sr, and Cu as minor components. To these
components can be added many more in minor proportions, but which can reach
significant values locally (Oberlander, 1994). Varnish exhibits microlaminations with
different chemical compositions, mainly due to variations in Mn content. Many of these
elements are likely to originate from external sources, because they do not occur in the
supporting rock. Clay minerals constitute the major part of varnish, some 60 to 80%,
followed in importance by Fe, Mn, and Si in the form of amorphous oxihydroxides. Black
varnishes with elevated Si percentages also exist which can reach 0.01 mm in thickness
and form on siliceous rocks. This type of coveting is denoted desert glaze or silica glaze
(Dorn, 1998). It is thought to originate by chemical precipitation of monosilicic acid
Si(OH4) in the form of a gel (Krauskopf, 1956) or via biological activity from a source of
opaline phytoliths which dissolve and reprecipitate (Farr and Adams, 1984).
For many decades varnish development or degrees of patination have been used as an
indicator of relative age in geomorphological and archaeological investigations (Hunt and
Mabey, 1966; Tricart, 1969; Demangeot, 1981). These older studies indicated that varnish
did not normally appear on Holocene material, which implies that it needs many years for
its formation. Absolute age dating has been carried out using neutron activation analysis,
U series, palaeomagnetism, and AMS C- 14. For this it is necessary to collect varnish from
a surface area of 0.5 m 2 (Krauskopf, 1956), using what little organic material is present.
The application of some of these techniques has not been continuous and radiocarbon
dating has problems of contamination, through extracting part of the rock substrate along
with the varnish. Another way of obtaining relative ages has been suggested, based on
the distinct degree of leaching of the varnish constituents. A leaching index has been
proposed: Ca + K/Ti, but it presents many treatment and analytical determination
problems. All of these methods have been used in studies of fault activity, rate of sea cliff
erosion, Quaternary deposit ages and lithic artefacts (Oberlander, 1994).
4.2. Formation processes
Previously, investigators believed that the constituents of varnishes came from the
weathering of the supporting rock. Solutions rich in Fe and Mn were thought to rise to the
rock surface via capillary action, where they were deposited after humid and rainy periods.
Nowadays no one doubts that the constituents derive from external sources, through
aeolian dust transport or water transport in solution (Figure 12.19). Electron microscope
Figure 12.19. Rock varnish developed on a sandstone, probably due to the precipitation of waters with
silica on the surface. This sandstone is affected by cavernous weathering features (tafoni). Petra, Jordan.
Photo by J.L. Pefia.
investigations reveal that the varnish is superimposed on, and clearly separated from, the
rocky substrate, supporting the theory of an external origin of the constituents.
The most popular hypothesis is based on a biogenic origin, in which Mn enrichment is
due to fixing by mixotrophic bacteria which oxidise Mn 2+ to Mn 4+, along with clay
cementation which traps other elements. This then repeats itself, producing an accretion
and subsequent lamination of varnish, which in turn implies cycles of humidity and aridity.
At the present time the bacteria that fixes the Mn is believed to be Metallogenium
personatum, the oldest known living organism (Oberlander, 1994). Others propose fungi
and cyanobacterias as fixing organisms (Krumbein and Jens, 1981). Some advocate purely
physico-chemical processes (Engle and Sharp, 1958; Elvidge and Moore, 1979). Mn is
more soluble than Fe and in desert environments, Eh values are high, indicating elevated
oxidation, and pH is alkaline as a consequence of the low levels of leaching. According to
these authors, variations in oxidation-reduction and the basicity of the medium can give
rise to varnishes. It is clear that an intense debate exists concerning the origin of varnishes;
it is very possible that the conjunction of biotic activity with the physico-chemical
variations could be the route to follow in future experiments.
4.3. Environmental implications
The first studies of varnishes emphasised their climatic significance and indicated that a
more humid climate than the present was needed for their formation. This implied that
existing varnishes were all relict (see Dorn, 1998). The studies also argued for cycles of
patination, whereby the varnish is eroded and a new one created.
Currently, with modem high-resolution techniques, the varnish laminations which form
microlayers with distinct compositions are interpreted climatically. Periods of reduced Mn
accumulation correspond to arid or interpluvial phases. Periods where the Mn/Fe
relationship is increased indicate more humid, pluvial or lacustrine phases, where aeolian
deflation diminishes because of expansion of the lakes (Figure 12.20). In many places,
however, the varnish sections offer an unconvincing register of palaeoclimatic
oscillations, and are more likely to indicate that the processes of varnish formation are
very complex (Oberlander, 1994).
5. Duricrusts
5.1. Introduction
One of the characteristics of arid zones is the presence of surface and subsurface crusts of
differing chemical composition, such as calcareous, siliceous and gypsum crusts. It is also
possible to encounter laterites, formed in environments of much greater precipitation,
which are mostly found on desert margins. They extend over significant areas in arid zones
and some, thanks to their enhanced hardness, are resistant to erosion and are situated in
elevated areas. Furthermore, they are good indicators of palaeoenvironmental conditions
and an understanding of Quaternary crusts is very useful in the interpretation of past
geological stages (Goudie and Pye, 1983). They can also be of economic interest, for
2
it
1.080+ 60 ka A
1 i l l
0.5-
I i I 11111 I I I111 II1 i 1 | 1 1 1 I11 i 11 i 1| 1| i i 11 i l l l l l |il i11 iii 111 Iii 1| i i11 i II i l l l l l l l l I11 ll[I. llt lid !111 -- Ill Ill I l l [ l l l l l l l l l | l l l | I11 | I III I Ill
0 20 40 60 80 100 120 140 160 180 200 220 240 260

Depth(micrometers)
Figure 12.20. Ratiosof Mn/Fe through rock varnish on volcanics with K/Ar date of 1080 ___60 ka, Coso
area, California. Peaks and troughs are supported by multiple datapoints, but the record appears too
variable to be convincing as an index of climatic oscillations (Oberlander, 1994).
example, calcareous scales may contain uranium and other metals; they can also be used in
house construction and as road fill for motorways (Reeves, 1976).
In this section, we will try to analyse the nature, properties and significance of duricrust
scales, as well as the geochemical and biochemical implications of their formation, and
finally describe the models most commonly utilised in their description.
5.2. Cafiches
Caliche is a Spanish word derived from the Latin calix, which means lime or limestone,
and was used for the first time in 1719 (see Reeves, 1976). It is commonly used in the
Unites States and Mexico but not Spain, where the term costra calcdrea is more common,
most probably due to French (croute calcaire) and German (kalkruste) influences. At a
global level there still exists a certain confusion because the word has been used for other
types of deposits. Moreover, there are many locally-used terms in different areas and
countries (Goudie, 1973; Reeves, 1976). The term calcrete, however, used in English
literature, is gradually becoming more popular in the scientific community.
Following Goudie (1972a), "Calcrete is a term used for material of continental origin
formed dominantly, but not exclusively, of calcium carbonate, found from powdery and
nodular to strongly indurated in form and implies cementation of the soil, rock or
weathered material within a vadose zone. This definition, however, does not include
speleothems, travertines, cemented litoral deposits, or stromatolites in algae lakes."
Caliches are chalky in texture, occupy around 13% of the Earth's surface and occur in
areas with an annual precipitation between 400 and 600 mm (Goudie, 1983a,b). Other
authors, such as Rutte (1958) in his study of Spanish limestone scale, state that the limits
occur between 100 and 500 mm, with no caliches occurring above or below these values.
An optimum between 100 and 250 mm is, however, indicated. Although they have a wide
Figure 12.21. Caliche on cemented gravels, resistant to erosion, located on a wadi margin. Jaisalmer,
Thar Desert, India.
distribution, caliches are characteristic of arid and semi-arid zones (Figure 12.21) (Reeves,
1976). They develop fundamentally in surface pediment deposits and terraces, and also in
aeolian deposits. In the terraces of the River Cinca in the Ebro Depression of Spain, the
degree of calcification of the stepped alluvial levels increases with age, and is nonexistent
in the more modern levels (Sancho and Melrndez, 1992).
Several volumes have been published concerning caliches (Reeves, 1976; Bltimel, 1981;
Vogt, 1984; Wright and Tucker, 1991), studying their types, characteristics and genesis. As
given by Vogt (1984), the investigation of calcareous crust has to consider different scales or
levels. In the first place it is necessary to know the spatial distribution (at the km scale) and
its relation to the regional geomorphological context, such as how it may be related to
pediment and terrace systems. The next study has to be carried out at the outcrop level, by
means of analysing natural cuts and samples that provide sedimentological profiles, thus
allowing the caliches to be placed stratigraphically in the sequence. Finally, in the
laboratory, thin sections and electron microscopy can be used to analyse texture and
micromorphology, which provide data about the formation mechanisms. All of these
investigations should be complemented with chemical and mineralogical analyses that help
in understanding the generating processes.
5.2.1. Geochemistry and mineralogy
The chemical composition of caliches, calculated from some 300 samples from all over
the world (Goudie, 1972b, 1973, 1983a,b), is; CaCO3: 79.28% (CaO - 42.62%),
SiO2: 12.30%, MgO: 3.05%, A1203: 2.12% and Fe203: 2.03%. Those caliches rich in
magnesium are called dolocretes. The percentages of each distinct component vary
enormously from one place to another, as can be seen in the calcretes in southern Australia
(Hutton and Dixon, 1981; Dixon, 1994b). In a study of numerous chemical analyses of
caliches from Oklahoma and New Mexico (Aristarain, 1970) the calcium carbonate,
magnesium, and ferric iron contents were seen to diminish with depth, whereas the silica
content increased significantly in the middle part of the profile. Elsewhere, Dixon (1994b)
examined the role of topography in the chemistry variations in caliches and demonstrated a
higher percentage of calcium lower down in the toposequence and an increase in
magnesium in the upper part. In all studies the calcium content fell with depth.
Calcite and dolomite are the dominant carbonate minerals in caliches and are
accompanied by quartz, opal, and clay minerals. Caliches can also contain minor amounts
of soluble salts (gypsum), glauconite, phosphates, heavy minerals, etc. Mineralogical
investigations have centered around the study of the clay fraction and its origin and, as a
consequence, improving our understanding of the processes involved in the formation of
clay-beating caliches. The dominant clay minerals are palygorskite and sepiolite, although
illite, kaolinite, montmorillonite, inter-layered illite-montmorillonite, and chlorite have
all been recognised (Aristarain, 1970, 1971; Gardner, 1972; Reeves, 1976; Hay and
Wiggins, 1980). The origins of palygorskite and sepiolite have been attributed to
neoformation in an environment containing sufficient magnesium, provided by the
weathering of suitable lithologies (Hay and Wiggins, 1980; Hutton and Dixon, 1981;
Sancho et al., 1992). Other investigators maintain that they arise through the alteration of
montmorillonite and inter-layered illite-montmorillonite (Watts, 1980).
5.2.2. Morphology
Descriptions of caliche morphology are numerous and different classification schemes

have been proposed based on distinct criteria. Some are based on sequential evolution
(Netterberg, 1969), others on degree of hardness of the carbonates (Reeves, 1976). The
degree of calcification also has some bearing on differentiating between soil
characteristics (Gile et al., 1966; Machette, 1985). Some distinguish only a few
morphological types and others, on the contrary, numerous types (Reeves, 1976).
The fundamental macroscopic morphologies (Sancho and Mel6ndez, 1992) are calcified
gravels, which normally appear at the base of a caliche profile, though small levels can be
encountered in the interior of the profile. This is constituted of clasts with a vadose
carbonate coat and a matrix that may contain root casts. Chalky caliche (Figure 12.22) is
formed of loose, clay, and sand-sized carbonate particles with practically no host rock
material. It can be lightly cemented and have a leafy structure. Hardpan caliche occurs in
the upper parts of the profile (Figure 12.23). It consists of centimetre and decimetre scale
plates that stack laterally. They form ooids, basal speleothemic carbonates, root traces, and
dessication cracks. Hardpan caliche can also be found deformed, forming pseudo-
anticlines (Figure 12.23) generated by expansion and contraction as a consequence of
hydration-absorption of water to a minerals of highly soluble salts, periodic wetting and
drying, calcite crystallisation and subsequent arching. Deformation may also be due to
the presence of expandable clays (Reeves, 1976). Finally, nodular caliche can develop in
fine-grained material, in which the particles are cemented by carbonate and, in turn,
Figure 12.23. Hardpan and nodular caliche deformed by a pseudo-anticline. La Plana Negra, Province of
Zaragoza, Ebro Basin, Spain.
Figure 12.24. Nodularcaliche developed on loose fine deposits. Bawri, Thar Desert, India.
generate large quantities of nodules (Figure 12.24). It can also result from the fragmentation
of hardpan caliche, targeting loose and cemented nodules (Figure 12.25). A continuous
graduation between hardpan and nodular caliches can be distinguished.
The macroscopic caliche characteristics are frequently represented at the microscopic
scale: nodular, oolitic and speleothemic textures, carbonate coatings, traces of biologic
activity and so forth. All of the investigations directed towards a precise and detailed
understanding of these micromorphologies and textures have as an objective an improved
understanding of caliche formation processes.
5.2.3. Origin
The analysis of caliches reveals the presence of complex sequences, whereby the
superposition of various caliche horizons can be recognised. In addition, the majority of
the profiles exhibit diverse stages of precipitation, dissolution and reworking.
For caliche to form there must be a source of calcium carbonate, which may be in
the rocky substrate, plant remains, volcanic emissions, aeolian dust, and other factors
Figure 12.25. Nodularcaliche with nodule size coarsing upwards. Margalef, Province of Lleida, Ebro
Basin, Spain.
(Goudie, 1983b). For many authors, an origin of carbonate transported by wind and
dissolved in rainwater (Bltimel, 1981, 1982; Machette, 1985) is too generalised, because
caliches sometimes develop with great thickness over rocks that are practically devoid of
calcium.
In semiarid zones, where a water deficit exists, carbonates do not fully leach and tend to
move and precipitate from one place to another. The fundamental process implicated in
caliche genesis is the dissolution and precipitation of calcium carbonate in the presence of
carbon dioxide:
CaCO3 + C O 2 + H20 r 2HCO3 + Ca 2+
Carbonate precipitation takes place by the loss of CO2, evaporation, the common ion
effect, and biological activity. The biological mechanisms of fixing carbonates in soils can
be due to lichens, algae, bacterias and microrrhizae (Klappa, 1979).
Goudie (1973, 1983b) differentiated two kinds of caliche origin; one pedogenic, which
seems to be the most generalised, and the other nonpedogenic. In any case, it is very
difficult to distinguish between them, because nonpedogenic caliche may be modified by
pedogenic processes.
Pedogenic caliches result from the progressive, downward accumulation of carbonates
(the per descensum model of Goudie; Figure 12.26), through washing of the A horizon and
illuviation of the B horizon (petrocalcic horizon). As seen in young profiles, early genesis
is manifested by powdery filaments and nodules. In profiles of thousands of years of age,
however, the carbonates form a crust over stones and fill the spaces in between particles.
Dust and loess,

products of aeolian sand
grain abrasion etc
-..,
9 .5 - ..-. -:- .:....:.
" - ' 9
Volcanic dust and ash , , ,,

9 I II I I :
"'" I I I I iii
ifl~ ,: ~' t~ i I I ! iii I /
,.-~.. ~., z, ,,,
'..--..~-. , IR d,
~ ~ ~ i" .~" ~...,
.
~1~.: I I th roug IlllI
I ~"~2,~~',,,, ..~'e I I I I I
,,,
II i
.... . i I I i I
~ I/ .,,l~41~ltl IiIlI I i1.1

/ \ -.~z-~.~~.l/ ~~;'~', ,I ,,!,
/ \ Leaffall I ' : ' , " 'llil llillilllll I I
Shells,
1
organic
C
\L~ -,a,, 1, " . ' ' ~ II Accumulator plants ei
materials
Depth of rainfall {-.!7:{i7:fi-:.:7:{-{.7i{::{{7{-7}{-

.::!{ {7 ....:..}~.~7i..:...){{.7{.~7:...:..{..W..:.~:7:7..~::}:..~...~.7:{...{.i..?.:{.:-:]:::{..i.7..:{::..{.:.~{....:i~.}S:..:.~}{::/~:::y.i~7
--.,---..-:{.::i?{:,:!-.:{....-~I~i~.{..,:.-::..-..i.Zone of carbonate accumulation ,??-.-;-!.-:-:::
penetration i:~ii{i{~{!:7~?}~i.~::~:{.!;i..~i~:ii%?!?~:{~i~..}i{::~..{:7{..2.}.}:!:~.~.:{!i~{~:.~.~:{..i:{:~.}~..~{i?i.}.~{.
Upward flow of nutrients throughI ~ , / t

r~176 ~--/ ~7i ~Q-~'~- ?'" ~- Water-table
Figure 12.26. Representation of a model of calcrete formation by different inputs of carbonate from
above per descensum model (in Goudie, 1973b, Figure 44).
The presence of calcified filaments, fibrous calcites and fungus microfossils testify to a
pedogenic caliche origin (Vaniman et al., 1994). A problem with this hypothesis is that
the caliche thickness is generally much greater than that of the soil; this difficulty can be
explained by a slow rate of soil erosion during caliche production.
When carbonate solutions rise through capillary action and precipitate at the surface by
evaporation, nonpedogenic caliches form, corresponding to the per ascensum model of
Transpiration
"": " : 1"" "" " " "_L...J~~ v;tr : t:

E. poat!onofH20 :. :
.'......'......'.:0ard'pan.."
Leached zone
Zone of capilliary rise

and lime deposition
9~
,0e.~_.o__'.
~ --% .o
]
.~z.
-
-,__'~_.2~-'2ai..~_.. oi'o'_. 9.o. ~..~_~_o'__.,___CaCO3 - rich
groundwater level
CaCO3
Figure 12.27. A model of calcrete formation by capillarity rise from groundwater per ascensum model
(after Goudie, 1983b, Figure 4.5, reproduced with permission).
Goudie (Figure 12.27). They can form close to the phreatic level by evaporation in the
capillary zone and also beneath the phreatic level as a consequence of a fall in CO2 content
(Land, 1970). This type of caliche can arise through carbonate-rich flood waters, which
can reach thicknesses of more than 10 m. They may develop rapidly, because they occur in
very young alluvium (Machette, 1985), although this origin constitutes a special case.
5.3. Siliceous crust
Also known as silcrete, this is a grey, tan or green coloured, hard and fragile rock,
comprised of quartz grains cemented by a microcrystalline and amorphous matrix.
Common thicknesses vary between 1 and 3 m. It forms in arid environments, although it
can form in more humid zones. It is a product of replacement of surface materials
(regolith, sediment, soil) by silica, in which silification takes place at a low temperature
unrelated to volcanism, plutonism, or metamorphism (Summerfield, 1983). Siliceous
crusts abound in Australia and South Africa, although they have been observed elsewhere.
They can occupy the tops of hills due to their resistance to erosion. Precious opals can
occur in some silcretes.
In outcrop silcretes commonly exhibit a massive structure accompanied by columnar
joints. Plate, botroidal and pillowy structures can also be recognised (Wopfner, 1978).
The micromorphology reflects the characteristics of the host rock (grain size and shape,
mineralogy and fabric) and the silicification processes. This includes micro- and
cryptocrystalline quartz with boundaries indented in the mineral grains, which confirms
the idea of replacement (Milnes et al., 1991). Chemical analyses indicate that they are
usually more than 95% SiO2 by weight and contain minor amounts of aluminum, iron and
titanium. The aluminum is commonly associated with the clay minerals, whereas the
titanium appears to be related to the bedrock.
In silcrete formation it is necessary to consider the silica source and its dissolution,
transport and precipitation. The primary silica source can be derived from in situ
dissolution, silicate weathering and aeolian dust. In fluvial and lacustrine environments,
diatoms may play an important role. Dissolution and transport are produced by water at
alkaline pH, moving vertically and laterally (Summerfield, 1983). The lateral movement
model is used in fluvial and lacustrine environments. Vertical transfer has similar aspects
to the per descensum and ascensum models used by Goudie (1973b). Finally, precipitation
is produced by evaporation or a fall in pH.
Where siliceous crusts occur in great thicknesses, their age, obtained from diverse
places around the world, go back to Mesozoic times, and so are considered relict accumu-
lations, possibly generated in different climatic regimes (Cooke et al., 1993).
5.4. Gypsum crusts
These are found in numerous arid zones, but over much reduced areas (Coque, 1962).
They develop in deserts with annual precipitations of less than 250 mm (Watson, 1983).
They are defined as accumulations of gypsum found in the uppermost 10 m, having a
thickness of 0.1 to 5 m, a gypsum content of greater than 15% by weight and a minimum of
5% gypsum in the underlying substrate (Watson, 1985, 1989a).
Three types can be differentiated on the basis of their structure and stratigraphic
position (Watson, 1979). The first type is horizontally stratified crust containing 50 to
80% gypsum. The second is subsurface crust, formed of either large, lenticular crystals
1 to 0.5 m in diameter, commonly known as desert rose, or mesocrystalline material with
0.05 to 1 mm diameter crystals. They can reach up to 5 m in thickness and have a gypsum
content of 50 to 70%. The third type is the surface crust, which basically consists of
alabaster gypsum with crystals of less than 0.05 mm, occurring in columnar form on a
scale of 1 to 2 m, or as loose, dusty deposits. They contain approximately 90% gypsum.
Gypsum crusts can be differentiated on the basis of their generation via capillary rise
and evaporation of sulphate-rich waters, or per descensum through gypsum-rich aeolian
dust (Watson, 1985; Heine and Walter, 1996). These authors state that Tunisian dust
originates primarily from the sebkhas, and that dust in the Namib Desert originates from
coastal clouds. The nonpedogenic gypsum scale model is based on the precipitation of
gypsum in subterranean waters and lacustrine environments.
Chapter 13
The action of water in arid regions
1. Introduction
In the proceeding chapters we have discussed the importance of water in weathering

processes, its role in the generation of patterned ground, varnishes, and crusts. In these
cases the activity of water occurs in a thin film and its movement is clearly local. We now
turn our attention to the action of water activity connected to surface flow, organized in
fluvial systems. These are characterized (Schumm, 1977) by a headwater source area,
where water and sediments are derived, a transfer zone and, finally, a lower area or
sedimentation zone. This fluvial system has a group of independent and dependent
variables (Schumm and Lichty, 1965) that have different influences. Time, initial relief,
geology (lithological and structural) and climate are the independent variables most
significant in influencing the hydrology and erosive activity. The vegetation depends on
climate and lithology. The rest of the dependent variables are variation in relief, hydrology
(run-off, discharge, sediment production) and the morphology of the drainage network
and slopes. In arid regions the first three independent variables vary substantially between
different deserts, unlike climate that depends on precipitation and which is, in turn, the
cause of run-off and erosion that are conditioned by vegetation cover.
2. Precipitation, vegetation, and evapotranspiration
In the majority of the arid regions of the world precipitation data are rare as there are
few meteorological stations and generally, where they do exist, they are very isolated. As
a consequence, extrapolations under these conditions are likely to be incorrect. Precipita-
tion is a consequence of different atmospheric processes: storm fronts, tropical storms,
orographic effects, and convective cells (Graf, 1988). As most deserts are located in anti-
cyclonic belts within the global atmospheric circulation system, precipitation is scarce
and limited to the sporadic penetration of storm fronts that are rare and normally produce
little rainfall. Tropical storms (cyclones, hurricanes, and typhoons) are rarer than storm
fronts, but occasionally they can penetrate the interior of arid regions, such as in the
Sonora Desert, between Mexico and Arizona. Orographic effects occur due to a reduction
in pressure and a cooling of the rising air, conditions that favour condensation and
precipitation. When the air descends on the other side of a mountain range, it warms and,
because it has lost the majority of its humidity while rising, it creates a rain-shadow zone.
The final type of process is the convective storm that is a few kilometres in diameter and
results from warming of the air. When they affect small basins they can produce important
floods (such as that of the 28th July, 1990, which developed a delta at the mouth of the
lower Huerva River, a tributary of the Ebro in Spain.
Precipitation in arid regions varies significantly (Figure 10.4) and, as a result, greatly
influences changes in run-off and sediment production. For example, in the middle of the
Ebro Depression, at an experimental station for studying run-off and erosion at Lanaja
in Huesca Province, Spain (Figure 13.1), a 4-year record shows that the most important
precipitation corresponds to convective storms that occur primarily during spring and
autumn, with some in summer. The rest of the rainfall is from convective cells and is of
low magnitude (Guti~rrez et al., 1995). For this reason, the highest values of run-off and
erosion take place in spring and autumn. Intensities vary considerably within convective
cell precipitation. At times, the recorded rainfall from one storm exceeds the total mean
annual precipitation. Furthermore, experimental stations with a large quantity of rain
gauges show sharp spatial variations and steep gradients (Figure 13.2).
In the literature, a number of cases are cited where precipitation is related to the
advection of humid air masses supplying continuous rainfall to a semi-arid region (Inbar,
1972; in Starkel, 1976). The Mediterranean winter rainfall that reaches the mountains of
Lebanon, Israel, and Syria is an example. Rainfall during a 4-day period (20th to 23rd
January, 1969) varied between 120 and 360 mm, equivalent to a third of the mean annual
precipitation. Despite the low mean intensities of 2 to 3 mm/h, occasionally exceeding
8 mm/h, this rainfall corresponded to a return period of 100 years. During this event,
0.04 mm/yr of downcutting, six times the normal rate, was produced in the upper
Meshushim basin. This shows that storm precipitation can be of considerable importance
in semi-arid regions.
60
55
50
45
4O
E
E 35,
.=_ 30
rr 25
20
15,
If.
10,
Summe(
It[ l Jll Jl
[Autumn ! Winter I Spring lSummer! Autumn i Winter I Sprin~ i SummerJAutumn I Winter ! Spring
1991 ' 1992 ' I 1993 l 1994
Figure 13.1. Precipitation record at the Lanaja Experimental Station in Huesca Province, Spain, for the
period July 1991 to May 1994 (Guti~rrez et al., 1985).
The action of water in arid regions 287
1.0 1.0 1.0 0.8

o.~ 0.4 o.6 o . 8 ~ ~ .L
9 . 9 , . . , 9 , ~ ~ ~ , ~ ~ ~ ~
9 9 9 9 .. "~..-- ~0.6
%,%00 , 9 9 9 9 j,,,.,-N~ O.2

Q 9
02 -02 o-
1.0 0.6 0:4 0i2

0.6 ..,, . .. ,., 1.2 , [P.81 ~-v~L.,,&-,.,~ 9 ~ z,-.,,..,~M,~
0.2 0.4,1:0.,.,-i", 1.8 1,8
0.4 0 ~ ~ _ ~ ; ; ~
~~o.4 ~ ~ _ , , ~ 7/.~j (7 o:8

o.6 ~ "J,C'{'.//, #.2'f~Je' ~.o
~o~~.~ ~ - . ~ ~ o.8
0.6 1.84
Figure 13.2. The distribution of rainfall from thunderstorm cells over Walnut Gulch, Arizona: Above,
a single cell event. Dots represent rain gauges. (after Renard (1970), in Graf (1988), Figure 3.4).
Vegetation in deserts is sparse or even nonexistent. In environments below 100 mm of

precipitation the vegetation may only be developed in wadis (Figure 13.3) or closed
depressions. If precipitation varies between 100 and 500 mm, the natural vegetation is
arboreal. Generally, this has been degraded by human activities to a more open landscape
of shrubs and grasses that may also have been affected by overgrazing (Francis, 1994).
The scarcity of water and the elevated temperatures mean that plants must adapt to these
inhospitable environments by means of physiological or morphological adaptations and
procedures that can cope with the stress associated with an increase in aridity (Gfinster
et al., 1993).
We have already seen the role that flora plays in the biological alteration of rocks and
the influence that sparse organic material has on the processes of hydrological erosion.
There exists, in general, a good correlation between the percentage of vegetation cover and
the mean annual precipitation. Interception is a significant factor in hydrological erosion.
Nfivar and Bryan (1990), in a study on the influence of vegetation in a semi-arid region of
northeast Mexico, found that after precipitation of 230 mm, divided between 17 storms,
Figure 13.3. Shrub vegetation in the bed of the Nahal Zin in the south of the Negev Desert, Israel.
interception losses were 27%. On the other hand, infiltration is highly variable and
considerably greater in vegetated areas than in bare soil. In deserts there are some
environments that are more sensitive to vegetation changes and this is manifested in
sediment production. It increases on a slope with decreasing vegetation and reaches a
maximum on bare soils (Figure 13.4a). Erosion is barely affected, however, where the
vegetation cover is low (Figure 13.4b,c) (Schumm, 1977), although for De Ploey et al.
(1976) and Morgan et al. (1986) sediment production increased for low values of
vegetation cover (Figure 13.4d). Rogers and Schumm (1991) showed experimentally that,
on 10% slopes, erosion increases rapidly with a decrease in vegetation cover from 43 to
15% but below this last value erosion decreases. Consequently, with cover below
15%, vegetation does not curb sediment production.
Vegetation affects aeolian erosion by reducing the wind velocity and acting as an
obstacle, which generates small dunes, known as nebkhas, in the lee of the plant. One
observes a decrease in dune movement with an increase in vegetation (Ash and Wasson,
1983) that acts to stabilize them. The same function is carried out by biocrusts that stop
dune movement, as observed in the longitudinal dunes of Nizzana in the Negev Desert of
Israel (Veste, 1995) (Figure 11.13 and Figure 13.6).
The water that exists on the Earth's surface undergoes evaporation, which acts to dry
the surface materials. In desert regions drying, brought about by prolonged drought,
reaches a depth of various meters. Another type of water loss is due to water uptake
through plant roots. Transpiration is produced in the stomatal cells, where there is an
interchange of water vapor between the atmosphere and the leaves. Transpiration is
favoured by low humidity, high temperatures, and wind. In deserts, xerophyte plants adapt
r
.o_
0
III
(a) Vegetation ~ (b) Vegetation
t-
.o_
o~
C)
UJ
(c) Vegetation ~ (d) Vegetation
Figure 13.4. Hypothetical relations of erosion and sediment yield to vegetative cover. All show an
exponential increase of erosion and sediment yield at higher vegetation density but b, c, and d indicate a
decreased influence of vegetation at low values of vegetative cover (after Rogers and Schumm (1991)).
to this water shortage by reducing transpiration through the formation of hidden stomates
and by reducing leaf size, or losing them altogether as in the cactus. In addition to
xerophytic plants, in the wadis of arid regions, phreatic plants develop. The roots of these
plants penetrate deeply to reach the phreatic level, assuring the water source necessary
for their development. Evapotranspiration may reach values of 85 to 90% of total
precipitation (Renard, 1969), although there are differences between distinct vegetation
species. The majority of the evapotranspiration occurs soon after rainfall. The rates of
transpiration fluctuate between 2.3 and 10 mm/day and 0.1 ram/day during the driest
periods of the year (Thornes, 1994).
3. Characteristics of the fluvial network in deserts
Precipitation in arid regions is scarce and variable. This variability reflects itself in the
fluvial systems at the global level. One can distinguish between endogenous drainage
where precipitation occurs within the desert area and exogenous drainage that originates
outside the arid zone, but flows through it, such as the rivers Nile, Indus, Colorado in the
USA, and the Ebro in Spain. In deserts, an important part of the drainage is endorheic,
where the water does not reach the sea (Figure 13.7) due to infiltration and above all the
prevailing evaporation. Furthermore, there are large closed depressions that receive
FORMER COURSES OF
R. SARASWATHI IN
WESTERN THAR 0 AN U PGARH ~/'..=~=-- --=,~...
. .~. . at... ~ - ~ -- f ~ #
.. ~," .<,/ SAKHIo,,~z ~ / ,
.:- / I" '0'

.o" /
"t ~ 1t /~ " ~
28~
J
SUKKUR " TANCQ.T"
-~-_]HOTARU
9 kx BAP~
AHGA~.~~ POKARAN~ SCALE
2,"~ / 9 I 0 40 80 120 160 200
JAISALMER l i 1 I n i
KILOMETRES
II iI MIYAJLAR
"o .... ' '1
I
I
/ 73O0'E
1 I
.:.~;~" -- BARMER 9
'li UMARKOT
.y..- ~~ o_~,oo
r- - ~. kms
t 9 SALMEI~~
RANN OF KACHCHH
69O0'E 71~ '
~ ] EXlSTINGSTREAMS ~ FORMERSTREAMS ~ SALTMARSH
I,- ;q ~"~E~T
~ S S U~OEOF
~T~ T ~~ E~ ~ ~ES'S~' ~ T . ~ ~ESTE~,.~ T
~~L ~ ~ FLOW DIRECTION OF FORMER STREAMS
Figure 13.5. Former courses of Saraswati River in Western Thar, India (modified from Ghose et al.
(1979)).
superficial waters, such as the endorheic basins of Lake Chad, in the southern Sahara, and
Lake Eyre, in the Central Desert of Australia. Guilcher (in Demangeot, 1981) estimated
that in the majority of Saharan regions, the largest floods do not travel more than 300 km.
Areas with areic drainage are those that show no regular circulation of water and occupy
Figure 13.6. Biocrust developed in a system of longitudinal dunes affected by rilling, Nizzana
Experimental Station, Negev Desert, Israel.
important extents within desert zones. Some deserts have exoreic drainage such as the
meridional zones of the southern Atacama Desert of Chile that have a large water source
in the Andes.
The river courses of arid regions are usually ephemeral (Figure 13.8) reflecting the
scarce supply of water they receive. These rivers (or wadis), therefore, are permanently
dry most of the year (Figure 13.9). Perennial rivers also occur, however, if they receive an
important water supply from distant sources, usually from surrounding mountain ranges
where a contribution of water may come from snowmelt (Figure 13.10). At times, the most
distal part of the drainage network may not be currently active, as indicated by greater
hydrological activity in the past pluvial periods. This occurs in the central Sahara where
complete networks of wadis do not receive water during even the largest floods (Tricart,
1969). Also, the Niger and Senegal rivers that originate in the tropical mountains of
Guinea, flow northwards through a very dry region characterized by dunefields that
can only be traversed during pluvial periods (Tricart, 1969). The use of aerial photography
J
~.~. ~.~ '.~ , "
C',I
.o ~
0
t ~
9
t "~
.,..~
~ ." .._.
I ~. ~ ~l~o
S
~.
"
,,# .~ .~
"
'~'~ t , "
9
9
Q) l.~
._~ ,_ <~
~ .-~
.-~ ~
D| o,,~
Figure 8.7. Relict-sortedstripes. Sierra de Javalambre, Teruel Province, Iberian Range, Spain.
(Figure 8.7). They are generally straight and may reach 120 m in length (Washburn, 1969).
The clasts are usually on edge with the major axis parallel to the stripes. The size of the
clasts generally decreases with depth and the sorting penetrates up to 1 m. In slopes from
1~ the sorted polygons change progressively into sorted stripes (Figure 8.1) due to the
sealing by mass movements of the sides perpendicular to the slope gradient (Furrer, 1968).
Although attempts have been made to unravel the origin of the patterned ground in
periglacial environments, its genesis is not well understood. There is not doubt that they
are polygenic and that similar forms may be due to different processes. Besides the activity
of certain processes may give place to different landforms (Washburn, 1979). Despite the
abundant literature devoted to the topic, the origin of periglacial patterned ground is still
an unsolved challenge.
2. Ice-cored mounds
The ice-cored mounds comprise a group of small hills generated by frost action. The
numerous terms proposed for their differentiation in the abundant literature have created
some confusion. These mounds may be perennial like palsas and pingos, or seasonal, like
the hydrolaccoliths or blisters. The latter ones reach 8 m in diameter and 3 to 50 cm in
height and have been little studied because of their ephemeral development restricted
to the winter season when they are commonly covered by snow. Their genesis is related to
the freezing of groundwater above the permafrost (Pollard, 1988).
2.1. Palsas
These mounds occur in boggy areas with discontinuous or sporadic permafrost (Zoltai,
1971). Their surface is criss-crossed with open fissures caused by dilation or desiccation
(Embleton and King, 1975b). At the present time they are occur in areas with a mean
annual temperature below 0~ as in Canada, Iceland, Sweden, and alpine regions
(Kershaw and Gill, 1979). Palsas may be isolated or form in groups and are differentiated
based on the morphology of dome-shaped, ribbon-shaped, and platform palsas. The former
ones are between 0.5 and 7 m in height, 10 to 30 m wide and 15 to 150 m long. The larger
palsas tend to be less conical than the smaller ones. The ribbon-shaped palsas have a lower
height and may be parallel or perpendicular to the contour lines, reaching several hundred
meters in length. The platform palsas rise 1 to 1.5 m above the surrounding bog and may
cover several square kilometres. Their generation is related to differential thawing.
Finally, the paisa complexes are formed by several types of palsas at different growth
stages and show numerous closed depressions generated by thawing (Sepp/il/i, 1988).
Some authors use the term string bogs to designate the association of different types of
palsas.
Based on their internal structure, two types of palsas are differentiated. The palsas with
a peat core have an outer layer of peat that thaws and dries in summer. Around 80 to 90%
of the core is composed ice in small crystals and lenses of segregated ice whose thickness
increases with depth from 5 to 10 mm to 5 to 10 cm. The base is generally formed by silt-
sized mineral particles. The silt core palsas also have an outer peat layer and a silt and clay
core with ice lenses and veins with a reticular arrangement (Sepp~il/i, 1988). Lithalsas are
similar mounds, but without any peat cover (Harris, 1998). We can see that no definitive
agreement exists with regard to terminology (Pissart, 2000).
The origin of the palsas is largely related to the lower heat conductivity of the dry peat
with respect to the saturated and frozen peat, with a conductivity around twenty times
higher (Washburn, 1979). In winter the peat wets and increases its heat conductivity
favouring the inhomogeneous penetration of the freezing and the formation of segregated
ice. The consequent differential cryostatic pressure gives place to elevations in the surface.
The outer peat layer that dries in summer with low heat conductivity has an insulating
effect preventing the thawing of the inner part of the paisa. The palsas may disintegrate
gradually from the margins due to a rise in temperature. Finally, they may become shallow
ponds and a new stage of peat accumulation may take place (Friedman et al., 1971).
The remnants of lithalsas present encircling ramparts (Pissart, 2000). Seasonal palsas
linked to oozes have been reported in the Paramera of Avila Province (Central Spain) at
1200 m in altitude (Molina and Pellitero, 1982). Although most of the palsas of the
Scandinavian countries formed 1000 to 3000 years ago, absolute datings indicate that
some of them were formed recently (Sepp~il~i, 1988).
2.2. Pingos
"Pingo" is the Eskimo term for mound. They are also called hydrolaccoliths and in Siberia
are called bulgunniaks. The pingos are domed perennial ice-cored mounds that generally
bulged out in large plains. Recently, submarine pingos have been recognised on some
subsea platforms. It is not known whether they have grown under the sea or have formed
on land and have been invaded latterly by marine waters. The highest documented pingo in
emerged land above sea level is the 48-m high Ibyuk pingo, located close to the
Mackenzie Delta in Canada. Most of the pingos are less than 20 m high and their diameters
range between 30 and 600 m. There is an inverse relationship between the diameter and
height of the pingos. In some cases they are elongated, more than a kilometre in length
and maximum height of 9 m. Because they keep a constant diameter, the gradient of the
slopes increases during the growth and never exceeds 45 ~ (Pissart, 1988). Pingos show two
types of cracks. Dilation radial cracks converge in the top and result from the upward push
produced by the ice core during its growth stage (Washburn, 1979). The opening of the
cracks may lead to the partial thawing of the underlying ice giving place to a crater-like
subsidence depression in the centre. The concentric cracks, not so conspicuous as the
former ones, result from the thawing of the ice core during the negative growth stage of
the pingo (Mtiller, 1959). In contrast to the palsas, pingos have a massive ice core that
penetrates several meters beneath the land surface (Lundquist, 1969). The cover material
is made up of loose sediments like gravel, sand and silt, but pingos with a rock cover of
sandstone and shale are also known. This cover layer may reach up to 14 m thick. The
growth rate varies from very low values to 1.5 m/yr (Mackay, 1973). On the other hand, all
the known pingos are younger than 10,000 years old and some of them are a few hundred
years old. Finally, pingos disintegrate progressively due to a rise in temperature giving
place to thaw lakes.
Two methods of pingo formation have been suggested. The so-called closed-system
origin explains the pingos of the Mackenzie Delta (Canada) and Central Yakutia (Siberia)
developed in areas with thick continuous permafrost. They form in relation to lake
depressions. During the development of the permafrost, the freezing of entrapped water
within a lake creates a massive ice core by downward percolation and aggradation. The
volume increase generates a cryostatic pressure that domes the lake sediments and may
even cause the extrusion of water towards the surface. If the water reaches the surface it
may give place to aufeis and when it contains gas it may produce explosive activity.
Generally, the water does not reach the surface but it freezes generating a massive ice core
(Mackay, 1979) (Figure 8.8). The cryostatic origin of the pingos in the Mackenzie Delta is
supported by the fact that 98% of the 1380 mapped pingos are located at the edge or inside
of contemporaneous or old lakes (Stager, 1956).
A great proportion of the open-system pingos are located in areas of relatively thin
permafrost of Alaska and Greenland. They are generated by groundwater flowing under
artesian pressure through a thin permafrost or in taliks within the permafrost. This water
freezes as it forces its way upwards forming an ice core that domes the surface. The
hydrostatic pressure result from height differences (Holmes et al., 1968) (Figure 8.9) so
that they form in topographic lows like valley bottoms or distal sectors of low angle
slopes (Mtiller, 1959).
3. Slope morphology and evolution
The slopes in periglacial environments show a wide range of morphologies because they
develop under variable conditions of temperature, moisture, lithology and vegetation
Water
I
9 . . . ~ , . , .
(a)
Water
9 , ' . 9 . " 9 9 . 9 9 9 9 9
(b)
Injected water lens
(c)
II
~2' "''" i'J Saturated unfrozen sand
[ J Water
l I 1111 ! I I Perma'rost(and ac,,ve ,a~er)

f]llFlllllllilill ,ce
Figure 8.8. Development of a closed-system pingo by cryostatic pressure (Mackay, 1979).
S 0 1000 feet N
t 9
~ 1 1 1 1 I !
~ r I
_ ~r-~-m,,,,,~,j,,,,,,llllllll
.... - c,,,,,,,,,,,,~]],,,,,,,,,,,,,,t
l i l l l ] l l l l l l l l i l l l l i l ! !
. . . . i l l l l l l l l l l l l l l l l l l ~ l ~ l ~ l ~
I l l l l l i l l i ~ l i i l l l l l l l l l l l l l i l l l l l l i l i l l l | ] ] i l l l l l l l l l l l l l l i l l
,,,,,,,,,,,n, ..... ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,lllllllll
• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • ] • • • • • • •
, l l , l l l l l l i l l l i l l l l l l l l l l l i , i l l l l l l i l i l l l l l l l , i l i l l l [ l l l l l l l l l l l l l l l l l , l l l l i . . . . l l l l l l ] l I
••••••••••••••I•••••••••••••••|••••••••••••••••••••••••••••••••••••••••••••••••••••••••••
I !i! !i!il ! I ' ~

Frozen ground Unfrozen Pingo ice Bedrock
alluvium
Figure 8.9. Origin of an open-system pingo by hydraulic pressure (Holmes et al., 1968).
cover. Several processes may be involved in their genesis, such as gelifraction (frost-
shattering), mass movements, runoff or nivation. All of them may operate independently
or in a combined way with a highly variable intensity. On the other hand, many of the
resulting landforms are not specific to periglacial regions as they also generate in other
morphoclimatic zones. Besides, some of the periglacial slope morphologies were formed
in the past (relict) and may be out of balance with the currently existing conditions.
3.1. Gelifluction slopes
The differential movements related to the downslope displacement of colluvial deposits

and rocks caused by frost-creep and gelifluction processes give place to several landforms
whose differentiation is based on geometric criteria (Washburn, 1979). Gelifluction sheets
have a large lateral extent and a festooned lower edge. They develop on slopes with
gradients higher than 1 to 3~ Gelifluction benches are characterised by a terraced form and
the major dimension is commonly parallel to the contour lines. Gelifluction lobes with a
tongue-like shape develop, like the benches, on slopes with gradients up to 20 to 25 ~. The
lobate geometry is due to the higher displacement rate along the central axis of the lobe.
The gelifluction features with a pronounced linear geometry oriented parallel to the slope
gradient are called gelifluction streams. These mass movements, having some similarities
to debris flows, may flow on glacial ice. Their displacement is favoured by the presence of
water at the base where the ice surface is at the thawing point (Figure 8.10).
Figure 8.10. Gelifluctiondeposits fowing on glacial ice. Whisky Glacier, James Ross Island, east of the
Antarctic Peninsula. Photo: A. Martfn-Serrano.
All these landforms develop more easily on slopes with scarce vegetation and on sunny
facing slopes where the thawing is favoured by the higher solar radiation. The gelifluction
deposits are usually poorly sorted and may show a crude stratification. The clasts are
commonly angular and show an oriented fabric with the major axis oriented parallel to the
direction of movement. These characteristics may be also occur in deposits affected by
gelifluction in other morphoclimatic regions. Those, however, that develop in periglacial
conditions are more angular and are scarcely affected by chemical weathering.
3.2. Cryoplanation terraces and cryopediments
These are low-angle, planated surfaces developed in periglacial regions. They have
received numerous descriptive and genetic names in the geomorphological literature,
especially in areas such as Siberia (goletz) and Alaska (altiplanation) where such
landforms are most numerous. Cryoplanation terraces develop in the upper and middle
part of slopes on hills and mountains up to 3000 m high. They show a terraced profile and
outcrops of bedrock and tors (rocky residual relied (Czudek, 1964) (Figure 8.11). These
terraces occur on slopes with gradients lower than 25 ~ and at the foot of small scarps.
The enlargement of the highest terrace may eventually give place to a planation surface in
the summit. The width of the terraces ranges from 5 m to more than 1 km, and the length
from 30 m to more than 10 km. The gradient varies between 1 and 14~ and the height of the
scarp may reach 50 m. The scarps show an abrupt link with the terraces and may have an
accumulation of snow. The terraces are covered by gelifluction deposits up to 3 m thick
derived from gelifraction processes (Priesnitz, 1988). These benches may show a structural
control and their origin is related to nivation processes. These involve gelifluction and
overland flow derived from snow melt waters that mobilise the particles produced by
congelifraction (Bryan, 1946). These processes cause the retreat of the scarp with the
consequent enlargement of the terraces and increase in their retention capacity of snow
(Demek, 1969). The presence of patterned ground on the surface of cryoplanation bench
treads and on pediments has often been raised as a potential indicator of contemporary
inactivity (Thorn and Hall, 2002). Excellent examples of cryoplanation terraces have been
recognised in the Sierra Nevada (Betic Cordillera, Spain) and in the Pyrenees (G6mez
Ortiz, 1996) (Figure 8.12).
Cryopediments are low-angle erosional surfaces developed at the foot of the slopes in
valley margins. Generally, only one level is recognised although in some cases several
stepped levels may occur. They have larger dimensions than the altiplanation or
cryoplanation terraces, reaching lengths up to several tens of kilometres. The slope of
the cryopediments varies between 1 and 12~ and shows a straight or slightly concave
longitudinal profile. They are frequently covered by a thin veneer (< 2 m) of detrital
material and the bedrock may crop out locally. Gelifluction landforms are frequent in
the proximal sectors of the cryopediments and patterned ground commonly occurs in the
distal areas (Priesnitz, 1988). Their origin is related to frost action in the source area
where the detritus, as on the cryoplanation terraces, is mobilised by gelifluction and
overland flow. The latter process is more active in the cryopediments and in
consequence the transport of particles is longer. In the distal sectors the particles are
transported by shallow braided channels. The continuous production and transport of
1 Frost-riven cliffs
2 Tors
3 Cryoplan~tion terraces
960
3~1
m 880
l / ' " I i
0 200 400 600

m
1080
m 1060 3
1040 , -| !
1
0 180 360
m
740
700
3 ]:~1
660
620
360
m
Figure 8.11. Slope profiles with cryoplanation terraces and tors in the Hruby Jesenik Mountains, Czech
Republic (partially modified from Czudek, 1964).
Figure 8.12. Cryoplanationterrace covered with angular clasts. Planell de la Carabassa (2690 m). Sierra
de Colomer, Eastern Pyrenees, Spain. Photo: A. Grmez Ortiz.
particles leads to the slow retreat of slopes and the enlargement of the cryopediments
(Czudek and Demek, 1970a). From another point, cryopediments also show a clear
morphologic resemblance with pediments developed in hot deserts although they have a
quite different origin (Demek, 1969).
3.3. Talus slopes and cones
These are more or less continuous accumulations of angular clasts (scree) in slopes.
Although very common in periglacial environments and especially in alpine areas, they
also occur in other morphoclimatic zones, such as deserts (Steijn et al., 2002). A talus is an
approximately wedge-shaped heap of fragmented rock supplied by an upper cliff or
rock face that generally covers the middle and lower part of the slopes (Figure 8.13 and
Figure 8.14). Their thickness is highly variable and may reach 30 m (Brunner and
Scheidegger, 1974). These accumulations commonly have a concave longitudinal profile
and a higher gradient in the upper part. The slope of the talus corresponds to the repose
angle of the coarser particles and ranges between 25 and 40 ~ (French, 1996). The talus
commonly show a clear particle-size sorting along their longitudinal profile. The size of
the clasts increases towards the base of the accumulation as larger clasts have a higher
momentum and travel further down slope (Washburn, 1979). Besides, the larger clasts
tend to show their major axis oriented parallel to the slope gradient.
The fragmentation of the rocks in the upper rock face of the slope may be due to several
processes, although gelifraction is obviously the dominant process in periglacial regions.
Figure 8.14. Talus slopes and debris cones. Fluvioglacial deposits in the foreground. Punta Vacas, Main
Cordillera of the Andes, Mendoza Province, Argentina.
Figure 8.15. Talus cone in Punta Vacas, Main Cordillera of the Andes, Mendoza Province, Argentina.
The continuous supply of frost-shattered clasts leads to the retreat of the rock face which is
largely controlled by the rock type and structure. Retreat rates around 1 mm/yr have been
measured in Laponia and Spitzbergen (Rapp, 1957) and 1 to 3 mm/yr in Great Britain
(Ballantyne and Kirkbride, 1987). The clasts in the talus move by rolling, creep and small
slides. The displacement is usually restricted to the upper half meter and the velocity may
be highly variable with values ranging from 1 to 500 cm/yr (Washburn, 1979).
The frost-shattering processes may act preferentially along planes of weakness
generating steep channels in the rocky slopes. These channels or torrents (couloirs)
are commonly preferential paths for snow and rock avalanches. At the foot of these
channels the rock fragments loose confinement and generate talus cones, similar to
the talus slopes but with conspicuous convex transverse profiles (Figure 8.15). Several
adjacent cones may coalesce showing morphologies similar to those of the talus
slopes.
3.4. Block-fields, block-slopes and block-streams
These accumulations of large angular particles develop on or below outcrops of hard rocks
affected by systems of discontinuity planes with a dominantly decimetric and metric
spacing. In periglacial environments, the weathering processes acting on these materials,
mainly frost shattering caused by freeze-thaw cycles, give place to boulder-sized,
decimetric and metric clasts together with finer particles. These block accumulations may
have a finer interstitial matrix favouring the growth of vegetation. Where the matrix is
washed by interstitial water flows, the accumulation shows an openwork and clast-
supported texture. Based on their topographic position, these accumulations are called
block-fields (felsenmeers), block-slopes or block-streams. The latter developed in valley
bottoms and incisions excavated in the slopes (Washburn, 1979). These accumulations are
usually found in high-latitude alpine zones.
Block-fields are extensive accumulations of angular blocks embedded in a fine-grained
matrix that cover more than half of the surface. They develop in flat or gently sloping
surfaces and reflect a strong frost-shattering activity. A limit of 5 ~ is established to
differentiate between block-fields and block-slopes (Smith, 1953; Dahl, 1966).
Block-slopes (Figure 8.16) are relatively common accumulations of boulder-size
particles generated by frost-action on or below rocky slopes. The major axis of the clasts
are commonly oriented parallel to the maximum slope direction and show a vertical
sorting with the coarser particles at the top and the finer ones towards the base. An
increase in the proportion of finer particles and in the size of the blocks towards the foot
of the slope is also observed. Differential movements in the deposits produce lobate and
Figure 8.16. Block-slopecomposed of Silurian quartzite particles. Tremedal Paleozoic Massif, Sierra de
Albarracfn, Iberian Range Teruel Province, Spain.
bench-like shapes in the middle and lower parts of the slopes (Gutirrrez and Pefia, 1977)
(Figure 8.17).
Block-streams (Figure 8.18) are formed by angular frost-shattered blocks accumulated
in valley bottoms. They result from the supply of blocks coming from the valley slopes.
One of the most magnificent examples is found in the Tremedal Paleozoic Massif
(Figure 8.19) in the Iberian Range (NE Spain) where most of the main and lateral valleys
have block-streams (Gutirrrez and Pefia, 1977). The largest one reaches 2.6 km in length
and 0.25 km in width. The maximum thickness of these accumulations of quartzitic blocks
is not known although exposures up to 4 m thick have been measured. The longitudinal
profile of the valleys show highly variable slopes and locally show transverse steps that
could be structurally controlled. The surface of the block-streams has small closed
depressions up to several meters in diameter. Their origin could be due to the removal of
fines by subsurface wash and the consequent differential settlement of the blocks. Beneath
the upper blocks that show an openwork texture, the skeletal blocks are embedded in a
heterometric matrix. The lack of fines in the upper part of the deposit is attributed to the
wash-down of the matrix by percolation waters (Andersson, 1906; Smith and Smith, 1945;
Smith, 1953; Potter and Moss, 1968). This circumstance prevents the growth of
vegetation. The clasts in the block-streams of the Tremedal Massif do not show any clear
fabrics unlike other examples that show lobate forms and marked upslope dips in the major
axis of the blocks (Potter and Moss, 1968; Caine, 1972). The longitudinal displacement of
the blocks has been negligible because none of the block-streams extend beyond of the
valley margins. The blocks of these relict landforms are profusely covered by lichens and
are locally being colonised by vegetation. At the present time the water flows thought the
deposits as throughflow and emerges in the transverse steps. The processes involved in
the generation of these depositional landforms include gelifraction, frost-creep,
gelifluction, frost-sorting, and the wash down of interstitial fines (Washburn, 1979).
Gelifluction Gelifluction Block-

lobes benches streams
~ 9, 9 q
, . .. :
"' ,':'o.~ ," ,":'--~.~," ~.'" .
1 2
Figure 8.17. Blockdiagram showing depositional periglacial landforms in the slopes and valley bottoms
of the Tremedal Paleozoic Massif, Iberian Range, Teruel Province, NE Spain. 1: Lower Valentian
quartzites. 2: Valentian ampellitic shales (Gutirrrez and Pefia, 1977).
Figure 8.18. Block-streamcomposed of Silurian boulder-sized quartzite particles. Tremedal Paleozoic

Massif, Iberian Range, Teruel Province, NE Spain.
3.5. Rock glaciers
Rock glaciers are one of the most remarkable geomorphic features of the alpine
periglacial zones and a visible expression of mountain permafrost. Active rock glaciers
are defined by Barsch (1996) as lobate or tongue-shaped bodies of perennially frozen,
unconsolidated material, supersaturated with interstitial ice and ice lenses that move
downslope or downvalley by creep as a consequence of the deformation of the ice
contained in them (Figure 8.20). There is long debate about whether classifying the rock
glaciers as periglacial or glacial landform (Tricart and Cailleux, 1962), and indeed, there
is evidence that some rock-glacier forms are quite polygenetic (Giardino, 1983). Some
geomorphologists consider that rock glaciers may have an ice core of glacial origin
(Potter, 1972). At the present time, numerous authors consider that the ice-cored rock
glaciers form part of the glacial system, whereas those with interstitial ice are
characteristic of the periglacial systems with permafrost. There may be also some
confusion in differentiating between block-slopes, block-streams, certain landslides, and
rock glaciers since a continuous transition may be observed between these landforms
(Corte, 1976; Giardino, 1983).
Talus rock glaciers contain frost-shattered clasts from the talus slopes commonly fed by
rock walls of glacial cirques and valleys (Serrat, 1979; Guti6rrez and Pefia, 1981)
(Figure 8.21). According to their morphology, the tongue-shaped rock glaciers have a
length/width ratio higher than 1, the lobate rock glaciers have a length/width ratio smaller
Figure 8.20. Rock glaciers. Clot de Claror, Aranser Valley, Eastern Pyrenees. Photo: A. G6mez-Ortiz.
than 1, and the spatulate rock glaciers similar to the lobate rock glaciers with a broader
lower part (Wahrhaftig and Cox, 1959; Barsch, 1996) (Figure 8.22). The complex rock
glaciers are those formed by the aggregation of several individual rock glaciers
(Washburn, 1979; Barsch, 1996).
The width of the majority of rock glaciers range between 100 and 500 m and the length
may exceed 1 km. Regarding the texture of the detrital material, the rock glaciers are formed
by a 2 to 5 m thick upper bouldery layer underlain by a fine-grained ice supersaturated sandy
and silty material (Barsch, 1996). The total thickness may reach more than 50 m. Rock
glaciers show steep fronts, with inclinations between 35 and 45 ~ (Barsch, 1988). The side
slopes are also steep and the upper bouldery surface has a microrelief of ridges and furrows
which are bent downvalley (Figure 8.20 and Figure 8.23). The volume transported in a rock
glacier generally exceeds one million cubic meters, 40 to 50% correspond to detrital
material and the 50 to 60% to interstitial ice and ice lenses. The displacement of the rock
glaciers results from the plastic or creep deformation of the ice and is largely controlled
by factors like the slope, ice content, temperature, or grain size of the inner clasts. The
velocity of active rock glaciers, commonly higher in summer than in winter, varies from
a few centimetres per year (Barsch and Hell, 1975) (Figure 8.24) to 5 m/yr. The highest
values have been measured in the Obergurl rock glacier (Austrian Alps) at abrupt
changes in slope. The movement of the rock glaciers is quicker than in the glaciers and
slower than in the slope deposits affected by gelifluction.
Discharge of active rock glaciers is controlled by strong seasonal and diurnal
variations. Snowmelt water and summer storms water are released in a very rapid way,
provoking flooding events. The discharge of glaciers is very similar to that one of active
Figure 8.21. Geomorphological map of a glacial cirque in the Eastern Pyrenees, Bonaigua area.
(Guti6rrez and Pefia, 1981).
Figure 8.22. Morphologic types of rock glaciers developed in mountain polar environments (Modified
from Humlum, 1982, in Barsch, 1996, Figure 1.6).
rock glaciers. However, average yearly mean specific discharge in active rock glaciers is
much lower than water discharge released from glaciers (Krainer and Mostler, 2002).
According to their functionality, active rock glaciers are those that are moving at the
present time. Active rock glaciers moving at rates of 22 cm/yr have been recently reported
in the Pyrenees (Spain) (Serrano and Agudo, 1998). Inactive rock glaciers do not show
movement today but still contain frozen material (Barsch, 1988). Generally, the initial
Figure 8.23. Rockglacier with ridge and furrow topography in the headwaters of the Garona River Basin,
Eastern Pyrenees.
Figure 8.24. Zonationbased on the mean displacement rate of the rock glacier Murt61, (Corvatsch, Swiss
Alps) from measurements taken during the period 1932-1955 (23 years). The average estimated velocity
for the whole rock glacier is 7.1 cm/yr (Barsch and Hell, 1975, in Clark (Ed.), 1988, Figure 4.6).
evidences of inactivity correspond to the colonisation of the front by vegetation and the
development of a talus apron at the foot of the front. In the relict or fossil rock glaciers
(Barsch, 1977b) all the ice has been melted and consequently show subsidence features
and a subdued surface relief.
The rock glaciers develop preferably in mountain areas with continental and semiarid
climate since wetter conditions favour the generation of glacial bodies (H611erman, 1983).
Given of the scarcity of absolute datings, Barsch (1996) considers that an age of several
thousand years up to 10,000 years (postglacial) seems to be an appropriate temporal range
for active rock glaciers.
3.6. Grazes lit~es
The gr6zes lit6es are slope deposits made up of alternating layers of well-sorted angular
stones and finer material primarily composed of sand, silt, and clay material (Tricart, 1952)
(Figure 8.25). They are also called dboulis ordonn~s by some authors. The term groise
litdes is used when they contain a high proportion of boulder-sized particles. The coarse-
grained layers have a matrix-free, clast-supported texture. Where the deposit is composed
of calcareous particles, the less pervious fine-grained layer is cemented by calcium
carbonate. The boundary between layers is generally abrupt and channel-shaped bases
may be observed. The grazes lit6es may occur in slopes with any orientation although they
are more frequent in sunny-facing slopes. They are found on slopes with gradients ranging
from 7 to 45 ~ may reach 40 m in thickness and the layers thicken towards the foot of the
slope (Guillien, 1951). Grazes lit6es are relatively common in mid-latitudes but they have
not been documented in polar zones. Their genesis requires freeze-thaw cycles for the
production of frost-shattered clasts, enough water for the sorting of the particles, and
Figure 8.25. Grazes lit6es in Entrambasaguas, Sierra de Albarracfn, Iberian Range, Spain.
scarce or nonexistent vegetation cover (DeWolf, 1988). The stratification and sorting were
considered to be related to the action of melt waters from snow accumulations, responsible
for the sedimentation of the coarse material and the eluviation of the finer particles
(Guillien, 1951). Francou (1988) has studied stratified slope deposits in the Peruvian
Andes at 4400 to 4900 m in altitude. The slopes have an inclination between 33 and 35 ~
and show large solifluction lobes in the surface that move a few centimetres each year.
Cuttings in the front of these lobes reveal that their movement resembles the unrolling of a
carpet that buries the clasts that armour the lobes and generate a stratified deposit similar to
the gr6zes lit6es (Figure 8.26). The frost-heaving lifts the clasts separating them from the
fines. Besides, the clasts are displaced by frost-creep and pipkrake moving at a higher rate
Figure 8.26. Origin of the stratified slope deposits in periglacial environments (Francou, 1988).
than the fines that form part of the solifluction lobes. Equifinality can be the cause of many
difficulties in such interpretations (Steijn et al., 2002). Grazes lit6es are abundant in the
central Pyrenees, especially above heights of 600 to 800 m. It is thought that they have
formed during more than one phase although the main period corresponds to the
Late-Glacial (Pefia et al., 1998).
3.7. Nivation landforms
Nivation niches or nivation hollows develop in slopes beneath a snow cover (Figure 8.27).
The melt waters percolate through the ground beneath and adjacent to the snow patches.
When this water reaches the freezing point, mainly during the night, it causes frost
weathering of the substrate and the particles produced by this process may be transported
downslope by overland flow, gelifluction or frost-creep. Gelifraction due to freeze-thaw
cycles is the main process involved in the generation of the nivation niches. An
accumulation of detritus commonly develops at the lower edge of the niche (Thorn,
1988b). In addition, the eluviation of the fine particles from the niche may lead to the
development of periglacial pavements. Nivation processes are especially pronounced in
low latitude mountain environments, as in some sectors of the Andean Cordillera, where
more than 300 freeze-thaw cycles have been recorded (Troll, 1944). The continuous
activity of nivation processes results in the widening of the niches that start as small meter-
size nivation hollows and may reach 1 km by the coalescence of several niches. The
enlargement processes work preferentially out on the margins where the melting of the
snow and gelifraction are more active, whereas in the centre the insulating effect of the
Figure 8.27. Nivation niches in the Hardanger Massif (Norway).

Figure 8.28. Blockdiagram showing a nivation niche close to Knob Lake (Quebec): a: Bedrock outcrops
in the top of the hill; b: Detritus-covered slopes (10~ 15~ c: Scarped amphitheatre (30-40 ~ in colluvial
deposits with accumulations of blocks at its foot; Low gradient slope (approx. 5~ Swampy bottom on
gelifluction deposits (Henderson, 1956, in Embleton and King, 1975, Figure 5.3).
snow inhibits the freeze-thaw cycles. The niches that reach large dimensions are called
nivation cirques or thermocirques (Selby, 1985). The nivation niches have an
amphitheatre-like scarp with a gentle slope at its foot that facilitates the removal of
detritus (Henderson, 1956) (Figure 8.28). They develop more easily in rock outcrops than
in slopes covered with vegetated soils. The generation of nivation hollows in a few years
beneath snow patches has been reported. The niches may be embryonic cirques when
nivation is replaced by glacial processes as the accumulated snow persists enough time for
its transformation into ice (Tricart and Cailleux, 1962).
Protalus ramparts (Washburn, 1979) or nivation ridges (Butzer, 1976) are linear or
curvilinear ridges formed from the accumulation of frost-shattered debris at the foot of
lingering snow patches (Figure 8.29). The rock fragments derived from rock scarps above
snowfields or from the backwall of a nivation niche fall on top of the snow and roll and
slide to its lower edge. Once the snow melts, the resulting ridge becomes disconnected
from the previous slope profile. In some cases different parallel ridges are differentiated
reflecting the parallel retreat of the slope or the reduction of the snowfield (Sharpe, 1938)
(Figure 8.30).
3.8. Slope evolution
The slopes developed in periglacial environments are classified according to their shape
and morphogenetic processes. The main types are concave cliff-talus slopes with talus
accumulations, benched slopes that result from cryoplanation processes and the
Figure 8.29. Protalus ramparts in Turb6n calcareous Massif, Central Pyrenees, Spain. (Photo: J.L. Pefia).
c o n v e x - c o n c a v e slopes with a continuous cover of gelifluction material (French, 1996)

(Figure 8.31).
The cliff-talus slopes evolve by the parallel retreat of the free face and the
accumulation of detrital particles in the lower segment of the slope leading to the
increase in its length and the decrease of its inclination through time (Carson and
A B '~
Figure 8.30. Sketch showing a sequence of protalus ramparts generated by the accumulation of frost-
shattered debris at the foot of a retreating snowfield (Sharpe, 1938, in Fairbridge (Ed.), 1968, p. 377,
Figure 9).
Figure 8.31. Slopes on detrital Cretaceous sediments with a continuous cover of nonsorted deposit
affected by active gelifluction. The foot of the slope has been undermined by fluvial action during the thaw
season. James Ross Island, east of the Antarctic Peninsula. Photo: A. Martfn-Serrano.
Kirkby, 1972). Generally, scarp retreat in periglacial environments is slower than in other
morphoclimatic zones, although this process depends largely on the litho-structural
characteristics of the rock face and on microclimatic conditions determined by
factors such as the orientation (Souchez, 1966).
The benched-slopes result from cryoplanation processes that produce the flattening
and decrease in inclination of the profile. The erosion in the upper parts of the
slope and sedimentation in the lower ones, mainly in the valley bottoms, involve
parallel retreat of the scarps and widening of the benched slope segments. These
ideas were introduced by Peltier (1950) in his "cycle of periglacial erosion"
(Figure 8.32).
The convex-concave slopes are characterised by a profile without irregularities.
They are mantled by a continuous cover of frost-shattered clasts (Figure 8.33) that may
be affected by gelifluction, surface wash and filling in sectors with high gradient. The
incision of these slopes by gully systems may give place to tripartite slopes or
triangular slope facets (Bfidel, 1982) (Figure 8.34). The gullies responsible for the
dissection of the slopes develop small low-gradient debris cones in the lower part
of the slope. Once the slope facet is disconnected from the upper part of the slope, it
hardly receives overland flow and becomes a relict landform with a high preservation
potential. The alternation of accumulation and incision periods in these slope systems
gives place to sequences of tripartite slopes.
".'"7
s \-"- %
S
S
Frost-riven cliffs
Small surfaces of I ~ F ~rade~---a

cryoplanation ~ i valley floo~
s 9 sS
.,., d., ,.,' ""
Congelifractate-covered 1
surface of downwasting
,
ii
Figure 8.32. The periglacial cycle of erosion. (A) Youth, (B) Maturity, (C) Senility (following Peltier,
1950, in Small, 1970, Figure 6.5). (A) Initial surface, Scarps affected by frost shattering, Cryoplanation
surfaces, Aggradation valley. (B) Large cryoplanation surfaces. (C) Surface mantles with frost-shattered
clasts.
Figure 8.33. Convex and concave slopes and possibly relict slopes in Sierra de Javalambre, Teruel
Province, Iberian Range, NE Spain.
4. Fluvial landforms
The high-gradient, braided channels with a high bedload and large width-depth ratio
are the most common in periglacial environments. In summer, the high-discharge
flows produce significant mechanical erosion in the channel banks. The run-off waters
RELICT PLANA TION SURFACE cx B/S-U

cx B / S - U ~ . . . . . . . . . "l -- -~
,[. ~;.t.,( uPPER i\ / SLOPES S-/)" ', .'
,,,, ox B,s-~',.'e." t
,l~xa/S-M.\f// ,~ I9\ VY/- \\ / /// \i 1// , \ ~"-~
I I \Y / #ooL~ '~ V/'s~oPEs\<IT~,ANGU~4,~.i l r AcETs, \I
I / # 11 / /,, o ~ J "l I / / ~ ~ ,i, T # I

RELICTSLOPEand' / \ \ l, , LOWER SLOPES . ", , , ' l '. ' l '
TERRACEFRAGMENT.,..__%~.I" } / " - " ~ ' ~ / ', ' , , / / ,' ',, r , ,""'' ',' '" ' '
ut b a ~ ~
~ tl/V i-r'l /:)l~/J .,. ~ 9 --.mr'-
(braided river with gravel bars)
Figure 8.34. Tripartite slopes in the southeast of Spitsbergen. R: fills in the upper part of the slope, G:
gullies dissecting the middle and lower part of the slope (Modified from Btidel, 1982, Figure 31).
Figure 8.35. Thermalerosion on a river bank (Walker, 1973, in Fookes and Vaughan, 1986, Figure 8.7).
may thaw the permafrost and undercut the channel banks giving place to thermo-
erosion niches that may reach 3 m in height and 1 to 3 m in width (Czudek and
Demek, 1970b; Walker, 1973) (Figure 8.35). These niches eventually collapse
supplying additional bedload to the fiver. In Central Yakutia, frozen fiver banks are
Figure 14.4. Precambrian granite with folded pegmatites. The high degree of jointing confers a straight
tendency on the slope profile morphologies, Negev Desert, Israel.
Slopes in arid zones 321
Figure 14.5. Simple slope constituted for spectacular massive Neogene conglomerate cliffs, attached to
the southern Pyrenean border formations, Mallos of Riglos, Huesca Province, Spain.
Those slopes originating on the morphologies of cuestas or mesas show a great variety.
Schumm and Chorley (1966), in their classic study on the Colorado Plateaus,
differentiated different types of slopes that result from the combination of lithologic,
climatic and vegetation characteristics. Simple slopes are fundamentally constituted by
one rock type. It can be a clay material affected by water erosion that develops a badland
landscape. The other subtype is constituted by predominantly massive rocks, such as
sandstone and conglomerates that form important cliffs (Figure 14.5). Composite slopes
are crowned by resistant caprock below which are developed easily erodable rocks in
which can occur badlands or debris slopes that can start from the cornice (Figure 14.6).
Finally, complex slopes are formed from the alternation of hard and weak layers
(Figure 14.7). Simple and composite slopes are be analysed below, whereas complex
slopes are considered as a mixed form of the two previous types.
2.1. Simple slopes
In this section we treat the morphology and evolution of the slopes on stratified massive
rocks, basically sandstones, and in another section of this chapter badland forms will be
analysed, which result fundamentally from the activity of water erosion on easily erodable
materials.
Most of the research related to this theme has been carried out in the Colorado Plateaus
of the southwestern USA, although it has been also studied in other deserts (Sahara,
Australia, Atacama). The evolution of the simple slopes is related to the characteristics of
Figure 14.6. Slope composed of a scarp on limestones and a debris slope that almost totally covers a
mainly clayed substratum of Miocene age. The group is partially affected by rilling and gullying, Ballobar,
Huesca Province, Ebro Depression, Spain.
Figure 14.7. Complex slope corresponding to a crest front, worked in stratified material of different
resistance to the erosion, Dra Valley, south of Morocco. Photo: J.L. Pefia.
the rocks and the activity of numerous geomorphological processes, which act or have
acted jointly or separately, spatially as well as temporally.
Discontinuities in the sandstone, as with thin clay intercalations or beds with cross
stratification, constitute zones more favourable for erosion and generate manifest changes
in the profiles of the slopes. Joint systems and vertical fractures constitute weakness zones
and along them can form routes of major moisture. As a consequence, the production of
weathering particles is increased there and remains trapped in the bottom of the fracture
system (Doelling, 1985). Also, in zones of major fault density, slope retreat gives rise more
rapidly to successive indentations and salients (Nicholas and Dixon, 1986). It has been
already indicated that in massive rocks, such as those that produce these simple slopes, the
development of sheeting is common, which has been noted in the arkoses of the famous
monolith of Ayers Rock (Uluru), in the central Australian Desert (Twidale, 1978) and also
in the sandstones of the Colorado Plateau (Bradley, 1963). This produces a convexity in
the slopes and sheets of a few centimetres to a metre in thickness. The dominant
weathering processes in sandstones are salt weathering and wetting and drying, although
in areas with freezing periods, gelifraction can be significant. The composition of the
cement also is important. In sandstones with a carbonate cement, this dissolves more easily
in cold water and siliceous cements need a pH > 8 to accelerate the dissolution. The loss
of cement produces the disintegration of the rock grain by grain. In general, the activity of
the weathering processes creates convex forms by rounding of vertexes and edges (Baker,
1936) (Figure 14.8).
Figure 14.8. Roundedforms and tafoni in the Cambrian sandstone of Nubia, Petra, Jordan. Photo: J.L.
Pefia.
Figure 14.9. Change in slope form during its retreat (Schumm and Chorley, 1966).
In several areas of the Colorado Plateaus (Schumm and Chorley, 1966) modern rock
falls can be common. Nevertheless, some of the landslides that are recognized in the
deserts were produced at other times and, therefore, are relict forms. Some examples have
been studied by Grtinert and Busche (1980) in the central Sahara and Degraff (1978) in
Utah. The analysed landslides are commonly of a rotational type (slump) and were
generated in periods of major precipitation during the Pleistocene; moreover, these mass
movements can reactivate.
Other erosive process that can be recognized on the simple slopes include basal
undercutting produced by subterranean water that flows out in the contact of impermeable
layers under massive sandstones. Weathering and erosion by water develop shelters
(alcoves), which in their evolution give rise to rock detachments and slope retreat in
sandstones (Higgins, 1984; Laity and Malin, 1985; Baker et al., 1990; Higgins and
Osterkamp, 1990). These shelters or cavities are accompanied by alveoles and tafoni.
Simple slopes evolve slowly and at their beginnings can proceed from composite slopes
(Figure 14.9), which retreat parallel to themselves. In later stages the outcrop of the
underlying clays diminishes, caprocks become rounded and, finally, they convert to simple
slopes again (Schumm and Chorley, 1966). An enlargement of these ideas has been carried
out by Oberlander (1977a) in his study on the sandstone cliffs of Utah. It also starts in a
composite slope (Figure 14.10) with an easily erodable substrate, that when retreating
turns into a simple slope. The form of the slope is controlled by the internal structure of the
massive sandstones, with thin horizons of interstratificated clays. These produce changes
in the slope and the ends of these horizons appear as modifications in the profile. The cliffs
A B C D E F G H I J
ground level
Figure 14.10. Slopes developed on massive sandstones. In A there is a cliff over a fine stratified
substratum, which remains hidden in B. Afterwards the intraformational levels control the slope
morphology. The segments located above evolve by undercutting, while those located below become
rounded (Oberlander, 1977a).
produced evolve by basal undercutting, while the segments situated below the clay strata
become rounded.
2.2. Composite slopes
These types of slopes are constituted by a rocky cliff in the upper part and a group of easily
erodable layers below (Figure 14.6). The components of those slopes adjust to the ones
defined by Wood (1942) and Fair (1947, 1948) in their studies in South Africa, that where
extensively analysed and renamed by King (1957a,b). This author considered these slopes
as "normal" ones, instead of those indicated otherwise by numerous researchers in climatic
geomorphology. For King they are part of a universal model; these slopes from desert
environments are the most characteristic. The parts that he differentiated in a slope are
(Figure 14.11): (1) the free slope that is the upper part, sometimes with a convex profile, as
a consequence of weathering and creep; (2) the free slope that constitutes the outcrop of
the hardest rock and in which general erosive acting processes causes its retreat; (3) the
materials resulting from the destruction of the scarp form the debris slope, which are
basically mobilized by water erosion; and finally, (4) the piedmont, which constitutes an
important concave element that connects with the alluvial plain.
The free slope is rounded by creep, if it exists, weathering and also rainsplash action
and sheet erosion. This tendency to rounding increases with the dip of the resistant layer
that constitutes the free slope, and is emphasized if it is porous or affected by exfoliation
(Bradley, 1963). On the contrary, if retreat is caused by faults rounding disappears or
decreases.
Free slopes are part of the borders of mesas, structural platforms or cuestas. The
thickness of the layer that constitutes the free slopes is important because it controls the
cliff height and the length of the debris slope. The rock type is also fundamental, so that the
weathering processes are different that disaggregate, for example, sandstones or
ISLIP
k~ j SLIDING
_ MASS
MOVEMENT
WATER
ACTION
SLOPE
ELEMENTS
Figure 14.11. Elements of a "normal" slope, water action, and mass movements (King, 1957).
limestones. The cementation degree of the rock is also essential to its resistance. To
lithology has to be added the structure. If the free slope is formed by finely stratified or
jointed rocks, debris slopes show good development; on the other hand, blocks of a large
size will cover the slope.
It has been indicated that debris slopes come from the breaking of the resistant rock that
constitutes the free slope. This is what supplies most of the cover of a debris slope,
although other parts can come from the underlying substrate. Despite the great
development that they can undergo in most deserts, studies of them are scarce compared
with the ones carried out in Alpine areas (Oberlander, 1997b).
Schumm and Chorley (1966) indicated that the difference between one slope and others
could be explained quantitatively by the relation of weathering (W). The existence or
absence of debris at the foot of the scarp depends on the velocity of fragment generation in
the free slope (p) before the destruction at the foot of it (d). If this relation is higher than 1
it generates slope accumulation, alluvial fans and a great amount of debris in general. If W
is equal to 1, there is a balance between p and d. If W is less than 1, blocks when falling
disintegrate upon impact and the remainder is weathered before the next fall. This last is
common in the Colorado Plateaus.
This debris can totally or partially cover the slope and, at the same time, be affected by
rilling and gullying that gradually work down to a major extension of the rock substrate.
The extreme case results from the practical disappearance of the slope cover and, if clays
are the outcropping material a badland landform generates. In some cases, hard and weak
layers constitute the materials of the substrate; in this case a stepped form originates from
composite microslopes (free slope and debris slopes) along the main slope (Figure 14.7).
These are the characteristics of the complex slopes.
The fragments of the debris slope show a poor sorting. They are affected by sheet
erosion, which exports the fine particles to a lower part of the slope, with the bigger
fragments standing out. These also can show an important weathering, with development
of alveoles and tafoni in their walls. Finally, in some cases small mass movements can
occur that affect to the debris slope.
3. Talus flatirons
These landforms were originally referred to as talus flatirons (Koons, 1955), flatirons
(Everard, 1963), dreieckshiinge (Wirthmann, 1964; Btidel, 1970), versants tripartites
(Gossmann, 1976), tripartites slopes and triangular slope facets (Btidel, 1982), and talus
relicts (Gerson, 1982). It should be noted, however, that the term "flatiron" has been
previously designated only for upfolded, hard sedimentary rock, and "triangular facets"
for eroded fault scarps, so considerable care must be utilized to avoid sloppy terminology.
These talus flatirons are characteristic forms of arid and semi-arid environments, although
they have been also recognized, as seen before, in periglacial areas. They have been
studied in the Sahara-Arabian Desert, southwestern United States and in some
Mediterranean semi-arid zones.
Their origin initially corresponds to a composite slope, which afterwards is incised by
water erosion (rilling, gullies, piping), of the remaining portions of individualized relict
slopes (Koons, 1955) (Figure 14.12). These paleoslopes have a triangular form with their
9 ~ ., ~,
----_-_ \,~,,
Q1 ~ o ~
0 500 1000 Ft.

9 , A , ~11
Figure 14.12. Schematic section of a talus flatiron, western Grand Canyon, Arizona (Koons, 1955).
apex pointing to the scarp (Figure 14.13), but trapezoidal morphologies also exist. Talus
flatirons are separated from the scarp and are parallel to it. In the distal parts they connect
to piedmonts (Blume and Barth, 1972) and, finally, with fluvial or lacustrine terraces. The
talus flatirons are concave and are crowned by debris, in general are poorly sorted and
generally have a thickness less than 8 m. In the upper part they can reach angles close to
30 ~ whereas in the distal zones they vary between 2 and 5 ~ The inner slope of the talus
flatiron dips contrary to the first one and it is composed of the substrate materials. It is clear
Figure 14.13. Talus flatiron with triangular apex, developed at the foot of the hill of San Pablo. The area
is affected by an intense gullying, Villanueva de Huerva, Zaragoza Province, Ebro Depression, Spain.
that for generating a talus flatiron the alternation of stages is necessary in which
accumulation is dominant, followed by another where erosion is dominant. The succession
of several alternant stages gives rise to talus flatiron sequences, in which the oldest
flatirons are the more distant from the scarp (Figure 14.14 and Figure 14.15).
Talus flatirons are relict slope accumulations and the separation is caused by the
incision of the middle slope gullies, where a group of fills converge (Sancho et al., 1988).
Once gullies dissect the talus flatiron, they flow to the lower parts where they deposit their
sediment load in the form of alluvial fans. Once the talus flatiron is isolated, each one
functions in an independent way (Btidel, 1970).
The caprock that forms the free slope above the talus flatiron must not be very thick so
that the accumulation does not become excessive (Schmidt, 1987, 1989a), and the slope
erosion and scarp retreat will be relatively rapid. For reaching this, the substrate must be
constituted of easily erodable rocks. These circumstances appear in composite slopes but
not in complex slopes where the alternation of resistant and weak rocks inhibits the
development of talus flatirons (Gutifrrez et al., 1998a,b).
One of the formation models of these talus flatirons is due to Koons (1955). A part of a
big rock fall covered the slope, followed by a dissection stage that individualized the talus
flatiron (Figure 14.16). Nevertheless, the most accepted origin is related to climatic
oscillations (Everard, 1963; Gerson, 1982; Gerson and Grossman, 1987; Sancho et al.,
1988; Schmidt, 1989b, 1994, 1996; Arauzo et al., 1996a; Gutifrrez et al., 1998b). The
stages with more vegetation correspond to a major debris accumulation on slopes. A
decrease in vegetation cover can cause slope dissection creating a talus flatiron.
Nevertheless, in recent epochs human activity caused destruction of the vegetation cover,
Figure 14.14. Sequence of talus flatirons in the Chalamera region, Huesca province, Ebro Depression,
Spain.
Figure 14.15. Geomorphological map of the Chalamera zone (Huesca province) and idealized block-
diagram block of the four slope evolution stages, in which can be observed three phases of talus flatirons.
1 - Mesas and structural scarps. 2 - Infilled valleys. 3 - Gullies. 4, 5, and 6 - Talus flatirons $4, $3, and
Sa. 7 - Debris slope, $1.8, 9, and 10 - T3, T2 and T1 Terraces of the Cinca river. 11 - Alluvial fans
(Sancho et al., 1988).
(a) (b) (c)
Figure 14.16. Non-cyclic model of talus flatiron development, according to Koons (1955), in Schmidt
(1989a).
which can be the unrestrained factor for debris slope incision (Everard, 1963; Sancho et al.,
1988; Guti6rrez and Pefia, 1989, 1992, 1998).
4. Free face retreat velocity
To calculate this value it is necessary to know, on the one hand, the lineal dimension that
corresponds to the distance between two different moments of the free face, and on the
other hand, the time elapsed between these two points. In this way, scarp retreat rates can
be obtained (Figure 14.17). There is an important difficulty in obtaining these data,
which implies a large imaginative effort and the adoption of numerous and varied
assumptions to reach retreat values. As a consequence, the methodologies applied for the
Figure 14.17. Scheme of a talus flatiron in which is indicated the different parameters used to calculate
the scarp retreat rates (Guti6rrez et al., 1998a).
obtaining of these data are very variable (Oberlander, 1997a). On the South Sinai
Peninsula, Yair and Gerson (1974) calculated rates of 0.1 to 2 mm/yr from fault free face
retreats. The presence of a thin lava cover over a retreating free face in Northern
Arizona, dated by K/Ar, let Lucchitta (1975) obtain values of 6.7 mm/yr. Schmidt (1980,
1989b), analysing the geometry of the captured consequent valleys of the Colorado
Plateaus, obtained values of 0.3 mm/yr and 0.5 to 6.7 mm/yr. Other methodologies,
much criticized by several authors, are based on the pack-rat accumulations of Neotoma
existing in the mouths of Arizona caves and dated by C-14. The values obtained are
0.45 mm/yr (Cole and Mayer, 1982). Young (1985) studied the free face retreat from
the Lower Eocene erosion surface in northeastern Arizona, from a deduced position of
the scarp based on the location of the straight valleys and obtained values of 0.16 to
0.17 mm/yr. Schmidt (1987, 1988, 1996) calculated the retreat by graphical constructions
starting from talus flatiron profiles. In his 1996 publication he correlated the talus
flatirons with Illinoisian and Wisconsinan glaciations and the resulting values were 0.2 to
0.35 mm/yr. Most of these works were based upon supposed ages. Sancho et al. (1988)
fixed the date from archaeological remains in slopes and talus flatirons, obtaining free
face retreat rates of 0.3 mm/yr in the Tertiary formations from the Ebro Depression of
Spain. Temporal values are more accurate when different slope talus flatirons are dated
with C-14, in the central Ebro Depression, which have ash and carbon remains in their
contents (Arauzo et al., 1996b; Guti6rrez et al., 1998b). The rates obtained by these
authors are 0.9 to 1 mm/yr for the last 35,000 years. All these data reflect that the retreat
rates in arid zones of different world parts are about 0.1 to 7 mm/yr.
It has already been indicated that most of these values were obtained with simplified
indirect methods and with supposed ages. All together these constitute an important but
inconvenient method for correlation of these values, as long as there does not exist a clear
reliability of the methods and, therefore, in the results. Another problem is related to the
sinuosity of the free face. Generally, it supposes a straight front, which is more common as
much of the scarp thickness is; despite that it is not a universal rule. When using the
method of extrapolation of talus flatiron segments (Sancho et al., 1988) until their
intersection with the prolongation of the top of the free face, intersections are obtained at
different distances, probably due to the free face sinuosity (Figure 14.18). Nevertheless,
while having many intersection points, an arithmetic mean can be obtained that is much
closer to the actual retreat rate.
One of the most important characteristics, relative to the major or minor retreat
velocity, is related to the lithologic and structural features of the constituent material of the
free face, which at the same time controls the resistance degree to the erosive processes
(Schumm and Chorley, 1966; Nicholas and Dixon, 1986; Schmidt, 1989a). The
mineralogical composition of the rock is fundamental as it determines the reaction to
the weathering processes dominant in the area. The erosion rate is also influenced by the
dip of the layers (Howard and Selby, 1994), so that the higher the dip, the smaller the
volume of eroded rock, and vice versa. Equally, the thickness of the rock of the free face is
in inverse ratio to the retreat rate (Schumm and Chorley, 1966; Schmidt, 1987, 1989b).
The presence of important thicknesses of constituent rocks of the free face facilitates the
application of the acyclic model of the origin of talus flatirons of Koons (1955), supported
by Schipull (1980) and Schmidt (1987). With this origin, free face retreat takes place
starting from strong impulses, unconstrained by big rock falls, between which alternate the
....... ........:~"~ii,!i~:....,,ii'..-::::
9 :~111..... S4
.... ,..~.: .,,~,:~,,.. ...~....,: .....
d d
25 m.
25 ~n.
d ~ " d
Figure 14.18. Profiles of talus flatirons corresponding to different stages (51-54) drawn in unbroken line
and extrapolated curves to points. The altitude of the apex of the talus flatiron with respect to the current
scarp is h and d is the distance of the apex to the scarp, measured on aerial photography (Sancho et al.,
1988).
production of particles by the weathering processes present in the area. On the contrary, if
the free face materials are a little thick, free face retreat is much faster. The degree of
fracturing of the materials of the free face constitutes one of the most significant factors
with respect to the retreat rate and mainly controls the degree of resistance to erosion. In
this way, the rocks affected by a highly significant fracture density show higher values of
retreat rate. Evidently, the erosion intensity of the substrate also affects free face retreat,
because where it is composed of highly erodable rocks such as clays and marls, rills
progress toward the base of the scarp producing basal undercutting and rock falls
(Schipull, 1980; Gerson and Grossman, 1987).
Another extraordinarily important factor related to scarp retreat, is the intensity of the
physical, chemical, and biological weathering processes that produce rock disintegration
and accumulation on the debris slope. These processes vary as a function of the
lithological and structural characteristics of the rocks, as well as with the climate and its
changes, and also with the acting period of these processes.
It has been already indicated that there exists a balance between the supply rate of
fragments from the free face to the slope (p) and the erosion rate (d) of these
accumulations, which constitutes the weathering relation (W) of Schumm and Chorley
(1966). In some cases, p clearly surpasses d and the slope accumulation surpasses the free
face, fossilizing it. In these conditions free face retreat is paralysed, although chemical
weathering can continue to act (Figure 14.19). This balance of production-destruction can
occur without big oscillations because the changes do not significantly affect the system. It
can lead to major erosion, however, so that the debris slopes gradually remain suspended
or isolated. This breakdown of the balance is due to an environmental change and implies
the overcoming of a geomorphic threshold (Schumm, 1979). Most researchers estimate
Figure 14.19. Slopepartially covered with debris that fossilizes the scarp. Afterwards it is affected by an
intense filling that affects the red clay Miocene substratum, Hill of the Alfambra castle, Teruel Province,
Iberian range, Spain.
that the slopes have been subjected to numerous climatic changes in which in the humid/
cold stages accumulation dominates and in the dry/warm climate the erosion processes
prevail. Free face retreat takes place in both climatic types, but it is possible that it occurs
more in one climate that in another due to the weathering-intensity variation, generating
mass movements, increasing basal undercutting, and so forth.
Finally, it is necessary to keep in mind the speed of base-level decline (Howard and
Selby, 1994) that affects the drainage system, and which can also be due to climatic or
tectonic causes. Similarly, the proximity of the slope to a fiver with great erosive power
can break the slope profile in its lower part and unleash erosion that affects the free face
retreat, just as it appears in the Cinca fiver in the Chalamera area of Huesca Province in
Spain.
5. Badland slopes
These are areas of very intense water erosion, with high-drainage density (313-820
km/km2), lack vegetation, and having steep slopes that make transit across them difficult.
Badland slopes basically develop on labile materials, in arid and semi-arid environments,
commonly at the foot of the scarps (Figure 14.20). The morphology of the divides can be
rounded (Figure 13.14), sharp (Figure 13.15), and in their retrogression they can generate
turreted hoodoo forms (Figure 14.21) (Scheidegger et al., 1968). Some badland areas are
the consequence of the anthropic activities, such as mining, deforestation, construction,
exploitations, dumps, and so forth (Aghassy, 1973). As erosion rates are very fast, they offer
Figure 14.20. Badlandzone with slashed divides that affect clays and lenses of Miocene sandstones. The
top of the formation is eroded by a cemented accumulation glacis that is Pleistocene in age, Bardenas
Reales, Navarra, Ebro Depression, Spain.
a small laboratory with a complex microrelief in which the processes and the modifications
of the forms can be analysed in short periods of time. Nevertheless, one must be careful
when extrapolating these results to areas of major surfaces (Campbell and Honsaker, 1982;
Campbell, 1989).
Numerous processes affect badlands that control its development. Therefore, it is
necessary to know the physical, chemical and mineralogical properties of the materials
(Guti6rrez et al., 1995; Gallart et al., 2002) that constitute these zones of high erosion, to
try to understand the activity of the implied processes and their influence on the
morphology of the area. So, the dispersion index and swelling capacity are physical
properties of great importance in several processes. Chemical analysis permits the
obtaining of the sodium absorption ratio (SAR) or the sodium exchangeable percentage
(ESP), which are directly related to the susceptibility to piping of the materials. X-ray
diffraction analysis of the clay fraction, carried out on oriented aggregates, shows the
mineral species existent. The presence of swelling minerals, such as montmorillonite,
affects considerably the weathering and erosion processes while also producing successive
shrinking and swelling in the superficial materials.
In general, badlands develop under arid and semi-arid climates, where the droughts
are prolonged and the dominant rainfall is storm-like. Nevertheless, they can develop in
areas with 1075 mm rainfall, such as in Perth Amboy, New Jersey, although these
badlands are of anthropic origin (Schumm, 1956a). The role of rainfall is important from
the chemical point of view (Bryan and Yair, 1982), because in areas of scarce rainfall
salts of high solubility are not reached to dissolve. On the other hand, if water
Figure 14.21. Turreted morphology developed on Miocene clays and sandstone from the Tudela
formation, in an area of intensifying water erosion. Cabezo de Castildetierra, Bardenas Reales, Navara,
Ebro Depression, Spain.
availability is very abundant, calcium and magnesium will lixiviate. With intermediate
rainfall calcium predominates. These variations reflect upon swelling, the dispersion rate
of the clays in the regolith and, therefore, upon the rates of infiltration. At more detailed
scales, slope exposition influences water availability and, therefore, the activity degree of
several geomorphic processes (Yair et al., 1980). In zones subjected to frost action, as in
western Colorado, a variation in permeability is produced, converting a less permeable
rill surface, into a very permeable one covered with aggregates (Schumm and Lusby,
1963; Schumm, 1964). In Alberta, Canada, the freeze-thaw cycles affect the superficial
material, preparing it for a quick denudation by subsequent run-off (Campbell, 1974).
The surface outcropping materials are subjected to numerous fundamental physical
and chemical processes that destroy the primitive rock by making a formation of a
regolith. That of the arid regions is of limited thickness (20 to 30 cm) and the superficial
layer presents a polygonal cracking when the clays have a little swelling capacity. If it is
high, the shrinking and swelling cycles produce the breakdown of this superficial layer
Figure 14.22. Popcorn structure in which can be observed the fragmentation of the superficial layer of
the Tertiary clays. Cuesta del Viento-Colola, San Juan Province, Argentina, Andean Precordillera.
and convert it in fragments with big voids and popcorn structure (Figure 14.22). This
layer also exhibits an intense lixiviation of the soluble ions. Below occurs a 5 to 10 cm
thick layer that can be an amorphous dense crust (Gerits et al., 1987) or a granular
aggregate (Schumm and Lusby, 1963). The transition to fresh rock is represented by a
slightly weathered and fragmented layer. There is, however, a great diversity of regoliths
in terms of rock composition, the steepness, and the exposure of the slope. After rainfall,
if the regolith is dry, water infiltrates following the cracks, pipes and pores until it
saturates, so that run-off starts after a few minutes. The regolith swells when wetted, the
cracks begin to close, and the internal flux is restricted to the big cracks and to
micropipes (Figure 14.23) (Hodges and Bryan, 1982). Afterwards, the wetting front
descends gradually; infiltrating the water until it finally generates superficial flow and the
almost complete closing of the cracks.
The badland surface is affected by different water-erosion processes, such as the impact
and splash of rain drops (rainsplash), filling, gullying, and subsurface erosion (piping),
which has already been described before and which is a most important erosive agent.
Nevertheless, these areas are also subjected, on occasion, to mass movements.
Mass movements figure enormously in badland morphology. If regolith easily absorbs
water, as occurs with the Chadron Formation in South Dakota (Schumm, 1956a,b), slopes
evolve by creep and by small pellicular slides, which fundamentally affect the regolith
(Figure 14.24). It leads to the development of rounded divides. On the other hand, when
the absorption of water is small, as in the Brule Formation, filling is important and a high
STAGES IN THE INITIATION AND D E V E L O P M E N T

OF R U N - O F F FROM SHALE SLOPES
IA I N F A L L
INTERFLUVE LPRIMARY CRACK ~

1 ',
I J 1 [
,
PoPOO.N
:::. ":':':'-.".":,i~: .;": ':'
~,~t:>Iz~- ~ ' " ~ "P,P~

(a) DRY CHARACTERISTICS (b) WETTING :INCIPIENT RILL AND
PIPE FLOW
111 IL I 1
TRANSIT~
WETTING
,:,.:,,,,,~,,,~ ,,,,,,,, ,,, ,,, ~, ,, .... .,~!
o I'l I o 0101 1% i ,il iol
!lc~
~, SUSTAINEDRILL AND PIPE FLOW 9 (d) INCIPIENT CRACK FLOW 9
CRUST SEEPAGE MICRO-PIPING
11111 I I I I I
9 " .~-~=~ ..... I ~~~~,~,.~.. ,
(e) SUSTAINED
MICRO-PIPECRACK
FLOW FLOW" (f) OVERLANDFLOW-CRUSTFLOW
Figure 14.23. Stages in the initiation and development of the run-off on a shale slope (Hodges and Bryan,
1982).
Figure 14.24. Badlandzone with rounded divides, in which can be recognized numerous pellicular slides
and cicatrices with a half-moon beginning shape. The substratum corresponds to Miocene marine marls of
the Guadix-Baza depression. Experimental station of Las Dehesas, Granada, Betic Range, Spain.
drainage density, steep slopes and sharp divides, characterize the landform. Nevertheless,
on these slopes mudflows can be produced that are channelled along the existent rills
(Figure 14.25). In the Ebro Depression of Spain, the generation of mudflows originates
during the wintertime, during which rainstorms are the norm. Rainfall and the lower
evaporation, as a consequence of the lower temperatures, cause regolith saturation that can
reach liquid states and when flowing gives origin to the mud flows that fill the rills. During
convective rainfall these mud flows are re-eroded, starting a new cycle (Guti6rrez et al.,
1995).
These badland areas lacking vegetation show very high erosion rates. As a
consequence, microforms develop very fast and allow erosion monitoring over short
time intervals. To measure this soil loss, dynamic and volumetric methods can be used (De
Ploey and Gabriels, 1980). The first method measures the amount of transported sediment
by water erosion with collectors (Figure 14.26) or Gerlach troughs. The second measures
the denudation with erosion pins (Figure 14.27) and bidimensional (Figure 14.28) and
three-dimensional microprofilometers (Campbell, 1974, 1982). It is evident that the values
of soil loss obtained depend on the utilized technique (Yair et al., 1980) and the erosion
rates registered with volumetric techniques are, for some authors (Takei et al., 1981;
Rogowski et al., 1985; Hall, 1988), higher than those obtained with dynamic methods. On
the other hand, Sirvent et al., (1997) reached an opposite conclusion, which indicated that
the rates obtained with collectors were bigger than those recorded with erosion pins.
Nevertheless, one of the big problems is the conversion of the volumetric data, because it
Figure 14.25. Mud flow channelled in a previous rill. Miocene clay of the Hoya de Ayerbe, Los Corrales,
is necessary to know the bulk density of the regolith and this has varied through time
(Regfies et al., 1992).
Water erosion produces a lowering that has varied between 2 and 20 mm/yr for
different experiments carried out in different deserts (Campbell, 1989). The maximum
known rates of erosion are recorded in China where they reach values of 38,000 t/kmZ/yr.
Hadley and Schumm (1961) obtained values of about 13,400 t/km2/yr in the South Dakota
badlands. The values fluctuate considerably as a function of the material type, magnitude
and rainfall intensity, time of recording of the experimental stations, technique utilized,
and so forth.
As the erosion velocity is very rapid, we can at least have a partial knowledge of the
development and evolution of this landform. Many aspects of the major morphologies are
reproduced in miniature in the badlands (Figure 15.8), which furnishes information in
understanding their development (Scheidegger et al., 1968). Works carried out in Perth
Amboy by Schumm (1956a) on the drainage system developed in these badlands, reflect
Figure 14.26. Experimental Station Bardenas-2 for the study of water erosion, in which has been
installed fences and a collector device in a badland area over Miocene clays, Bardenas Reales, Navarra,
0 3m
I
Figure 14.27. Ground lowering contours recorded by erosion pins between November 1991 and
November 1993 in the experimental plot Lanaja-1. Huesca Province, Ebro Depression, Spain. Legend:
1 - rills; and 2 - ground lowering contours (mm).
Figure 14.28. Microtopographicprofilometer made up of 50 pins, that lets us obtain a profile between
two fixed points. Profiles are made every 6 months in order to study microtopographic variations.
that their evolution is similar to those of large fluvial systems. Although a dendritic
network can be randomly developed, the channel spacing and their distribution is
intimately related to the erodibility of the material and the run-off generated per surface
unit. The channels incised rapidly into the materials and generated steep slopes. As slopes
retreated they formed micropediments at the foot, which increased in size with the retreat
of the slopes. These piedmonts also finish eroding and slopes evolve (Schumm, 1962).
The investigations carried out in regions with different aridity degree show erosion
rates of 0.45 mm/yr in the northern Negev of Israel with a mean precipitation of 90 mm
(Yair et al., 1980). In the badlands of Alberta, Canada, rates of 4 mm/yr with a
precipitation of 330 mm are indicated (Campbell, 1982). All of this leads one to think that
in a certain badland zone, when analysing for long time intervals, erosion rates might have
been quite variable, because these arid and semi-arid regions have been subjected to
numerous climatic changes. In them, rainfall varies substantially from the more dry
(interpluvial) to more humid (pluvial) periods and, therefore, the erosion rates will be
lower or higher, respectively.
Chapter 15
The arid region piedmonts: glacis and alluvial fans
Piedmonts constitute the transition between the mountain areas, where erosion is the
predominant process, and low-lying areas where transport and sedimentation are the
dominant processes. Aridity controls weathering on the hillslopes and in this context being
the particle production from rock fragmentation, it is normally small. Nevertheless, within
mountain and basin deserts (Mabbutt, 1977), the climatic conditions are usually variable in
the uplands, but the contribution of rainfall or snow melting is important. In this way, as a
consequence of such external supply the piedmont zones will receive water flows with a
higher sediment load than those occurring in typical arid areas. The mobilized materials
within the mountain catchments are transported by confined streams, which emerge into
piedmont zones where the water flow is abruptly unconfined and then transported
sediments spread out radiating downslope. The unconfined water flow decreases its
velocity and transport capacity abruptly, giving rise to progressive sedimentation. On the
other hand, in the backing hillslopes the development of erosive-accumulative ramps is
typical, which constitute another characteristic landform of arid zone mountain fronts.
The landform assemblage generated within the piedmonts, therefore, can hold either an
erosive-nature pediment (glacis) or a depositional one (alluvial fans). Pediments and
alluvial fans can be intimately related, and sometimes may be difficult to differentiate
(Figure 15.1). Bull (1977) suggested that fans can be distinguished from pediments when
the thickness of the associated deposits is up to 1/100 times the axial length of the
depositional area. In addition, a fan can be levelled in response to an external base-level fall,
giving place to pediment-like surfaces, which in turn can be buried by renewed alluvial-fan
deposition. On the other hand, pediments tend to develop in low-relief areas (low and
narrow ranges and modest escarpments), whereas alluvial fans are usually linked to more
energetic topographic configurations (mountain fronts, valley junctions, and so forth).
1. Glacis and pediments: definition, terminology and geographic distribution
According to Ldpez-Bermfidez (1973) who carried out a good updated revision on glacis;
this term (which means the gentle sloping ramp of the ancient Roman fortifications) was
first used by Dresch (1938), to designate specific landforms with a gently concave
longitudinal profile and with slopes down to 6 ~ Tricart (1969) considered that glacis
constitute vast piedmont plains with a low gradient, and coated by a thin sedimentary
veneer, which eventually connected with the desert axial fluvial systems constituted by the
ephemeral stream channels of ouadis (wadis) (Figure 15.2). Important terminological
problems arise with the progressive incorporation of this term into different languages,
Figure 15.1. Piedmont deposits in which the discrimination between alluvial fan and pediment surfaces
proposed by Bull (1977) are not applicable. This is due to the absence of channel incisions which impede
the direct observation of the underlying bedrock. This picture can illustrate a mantled pediment, and also
coalescent alluvial fans. La Sierra del Cerro del Coronel, Province of San Juan, Argentina. Andean
Cordillera.
Figure 15.2. Pediment surfaces merging with the upper terrace of the Ouad Ziz. Central Atlas, Morocco.
Photo by F. Guti6rrez.
The arid region piedmonts: glacis and alluvial fans 345
which give rise to an extensive terminological and conceptual misunderstanding (Tator,

1952, 1953; Dumas, 1967; Tricart, 1969; P6csi and Sailfird, 1970; Cooke and Warren,
1973; Whitaker, 1973, 1979; Zuidam, 1976; Demangeot, 1981; Cooke et al., 1993). The
English writers usually use the word pediment, which was first used by Gilbert (1877), and
latter defined by Mabbutt (1977), as a flat piedmont surface cut in the bedrock and
separated from the backing hillslope by an abrupt knick-point or piedmont junction
(Figure 15.1). Where a thin veneer of debris coats the erosive surface of the pediment, it is
called mantled pediment. In the German literature the equivalent term is the so-called
fussfliiche (Mensching, 1958).
French authors make an extensive use of the term glacis, but either for clarification or
descriptive purposes, they label it with different adjectives generating a large list of terms
(Tricart, 1969). L6pez-Bermfidez (1973) compiled 28 different denotations, but Whitaker
(1973) managed a larger list of terms. Those glacis directly carved in the bedrock without
associated deposits (equivalent to pediments) have been denoted as bare glacis (Birot
and Dresch, 1966), ablation glacis (Dresch, 1970), erosion glacis (Dresch, 1938), or
denudation glacis (Tricart, 1969; Tricart et al., 1972). Different terms are also used
depending on the thickness of the associated detrital deposits. In consequence, Tricart
(1969) discriminated among denudation glacis (glacis de denudation) with a thin and
discontinuous sedimentary cover, sheet-flood glacis (glacis d'dpandage) with a
continuous and apparent (1 to 3 m thick) veneer of sediments and finally accumulation
glacis (glacis d'ennoyage) where the sedimentary cover is over 5 to 10 m thick. This
terminological overabundance joins with a significant conceptual confusion. For some
French authors the erosion glacis are only developed in soft rocks, like some limestones
and marls; Dresch (1957), for example, set aside the term "pediment" to characterize only
erosion surfaces in crystalline bedrock (Birot and Dresch, 1966). Birot (1960) considered
the definition of glacis as the main problem of arid zone geomorphology. Therefore, here
we propose that, to avoid any kind of terminological confusion, the term pediment should
be used for pure erosive surfaces carved in the bedrock, and mantled or covered pediment
for erosive surfaces fossilized by a thin veneer of alluvial sediments.
Pediments develop in a wide variety of climatic, lithologic, and tectonic contexts
(Dohrenwend, 1994). They occur from subpolar latitudes to arid or humid-tropical climates
(Whitaker, 1973, 1979) and therefore they can be considered plurizonal landforms
(Figure 15.4). The dictionary of Whitaker (1973) even quotes the term cryopediments in
relation to nivation processes occurring in periglacial areas. Pediments, however, develop
extensively throughout arid and semi-arid regions where they are characteristic landforms
(Figure 15.2). In arid zones, vegetation is scarce, thus reducing the impediments to the
erosive work of overland flows that generate the pediments. We understand that pediment
is synonymous for erosion glacis, as proposed by Dresch (1957). In consequence, we think
that most of the previously- listed terms are usually referred to arid and semi-arid climatic
conditions, or they can constitute ancient arid palaeo-landforms.
1.1. Pediment morphology
Pediments are commonly fed by small mountain catchments with limited relief. They can
emerge from mountain fronts, which constitute the junction between the catchment areas
t. "a L L . ~ (~+ ,'~
~- v ~.W" r- r, r- ~ ~ . . _....
.:. _~~
;~ ~ P, , >~'~-'..
,-(~ o..
Figure 15.3. Terminologyof pediments and related features of the desert piedmont zone. (1) Backing
hillslope. (2) Piedmont junction. (3) Piedmont angle. (4) Pediment. (5) Mantled pediment. (6) Mantle. (7)
Alluvial plain. (8) Alluvial fill. (9) Suballuvial floor (after Mabutt, 1977).
and the piedmont plains. Mountain fronts may have a linear or sinuous geometry
depending on the associated tectonic activity. In any case, they display spaced indentations
(irregularities) corresponding to the different valley outlets from the mountain catchment.
Valley-outlets are called "quebradas" in South America, where they usually correspond to
narrow rocky valleys with steep slopes. In the absence of tectonic activity, mountain-front
sinuosity increases with time (Bull and McFadden, 1977). On the other hand, linear
fronts develop in relation to active faulting or to lateral erosion of longitudinal stream
flows along the front-toe. Linear fronts develop well- defined, and sharp, knick-points.
Conversely, in other areas the piedmont junction is developed in a progressive way, giving
rise to a large concave hillslopes where it is difficult to identify the precise beginning of the
pediment surface (Figure 15.2 and Figure 15.3).
Few publications have been focused on the morphometric analysis of pediments
(Corbel, 1963; Mammerickx, 1964; Cooke, 1970b), consequently it is basic to list the
variables considered in these studies (pediment area and slope). Pediments present a
variable surface area, ranging between 2.5 and 650 km 2 in the Desert of Sonora in Arizona.
On the other hand, the fraction of the total surface area occupied by pediments in a specific
desert zone also may have great variability. In the Mojave Desert of California the portion
of area occupied by pediments is only the 6.7% of the total desert area, but in the Sonora
Desert, this proportion increases to 30%. Pediments surfaces usually develop from the
lower segments of the hillslopes down to an axial gully or sedimentation plain. The
longitudinal profiles of pediments may vary from concave to linear, and they everywhere
show a consistent downslope gradient reduction. When pediments border isolated relief, or
inselbergs, long-profiles display a characteristic radial form (Figure 15.5 and Figure 15.7).
These may be of two types: position inselberg (fernlinge, from the German authors),
which preservation is due to its watershed location, and resistant inselberg (hiirtlinge)
developed by the occurrence of differential erosion.
The pediment slopes range between 0.5 and 11 ~ (Tator, 1952), and they may vary along
their concave longitudinal profiles (Figure 15.6). Where pediment development occurs
~E,~, ~ , i ,~ , ' , ~ u,oZ.g~' ~,
~s
..." "'" " ," ,~ 9 i
0 ki/ometres 5000 ~
Figure 15.4. Map of the pediments mentioned in the literature (after Whitaker, 1973).
along active tectonic fronts their gradient is steeper than in areas of tectonic quiescence.
There are also relationships between slopegradient and clastsize, in such a way that
when the steeper the slope, larger clastsizes occur. For some authors lithology
is not a relevant factor (Mammerickx, 1964), while others point out its impact on the
Figure 15.5. Inselberg carved in Cambrian sandstones with the development of gently slope radial
pediments. Jodhpur-Jaisalmer region. Semi-arid zone of the Thar Desert (India).
Figure 15.6. S e g m e n t e d hillslope with a s u c c e s s i o n of c o n v e x and c o n c a v e s e g m e n t s . T h e r e l i e f

t e r m i n a t e s in a p e d i m e n t d e v e l o p e d f r o m a c o n c a v e k n i c k - z o n e . M a c h a k o s , K e n y a .
slopegradient (Cooke et al., 1993). Climate seems also to affect glacis gradients, with
smaller glacis slopes in the more arid regions (Mabbutt, 1977).
The channel network of pediments is usually entrenched in the proximal zone, but
channels radiate downslope towards the distal areas, where commonly overland flows
~ ~ ~ ' J _ t ~ " = - - - ~ ~, ~, -. . ~. . = . .. . . . . ~. . . " ' - . . . . 9 "~
. ,.~ , - ,.
9 i ,
.- ~, _
~, .
, ~ -. _
~_ -. -
,_ ~
~
= _
_ - - - ---- - - - - -
_ .
.
.
"
.
.
.
.
. . .. . . . . .. . . . . . . . /1/
/g."s
, - V -t'- ~. ,'.t, . . . . . . .... . . . . . . .
~ ' , , , . W , y - ~ : ; ~ , L~:-.~.: . . . . . . . . .
.. ~._.,,~, t ~-;-.~7-.~ ".." .e ~;.., ~ . , ,, ,
A B -C
9. .
. ."....--.E
. . . . . . :.-T.-:".
.. ..... ". . . . .
9. . . . . . . ". . . . . . . . . . . C
. '.................
~ , . ~ .- -"~'" ".......
.....
. '......
" - - ~
....... .9. _ - ~ " ~ - - " : : ~ - = _ - - ~ ~ ~ , - ~ - S ~ _ -T~_'.'-..-.L'. ".:.'..

......... T : . . :-- .....
.. ;---. ........ :.'""-'-~~:~,~",';~~.'_,u."m~.~- ~_.~.... "~-~.':: ...'. . . . .
I~ ..... -"- ---, . .. , 9 #., ~ S~7".~.- " '*' " ""'.k'~" ,'~" "~"~ "~"~ :'= -"'" ...........
C B A B C
Figure 15.7. Different e l e m e n t s of the p e d i m e n t . (A) D i s s e c t e d zone (up) and i n s e l b e r g s (down). (B)
P e d i m e n t w i t h residual reliefs. (C) M a n t l e d p e d i m e n t (Johnson, 1932a).
gradually diminish by infiltration. Most typical pediment channels have a braided pattern.
They are shallow channels subject to significant switching during extreme floods. During
periods of base-level fall, channel entrenchment can also propagate to distal locations
giving rise to subparallel drainage patterns (Cooke and Warren, 1973). In most cases,
pediments are planation surfaces directly carved in the bedrock, but mantled by a thin and
discontinuous alluvial cover. In particular cases some small relief emerges from the glacis
surface as residual landforms denoted by the French authors as "chicots" (Figure 15.7).
1.2. The role of geology and climate in pediment development
As aforementioned, some French authors differentiate between glacis developed on soft or

in hard rocks. This division attempts to emphasize the great influence exerted by lithology
and geological structure on glacis location and evolution. When pediments develop on
granites, or crystalline bedrock, they usually present a sharp knick-point attributed by
some Australian authors (Mabbutt, 1966; Twidale, 1967) to the occurrence of a more
intense weathering in the piedmont junction zone. As chemical weathering processes are
very limited in arid zones by water availability, the occurrence of sharp knick-points is
attributed to previous humid periods occurring during Quaternary, or even Tertiary times.
In the south of Arizona, where schists and basalts are the dominant bedrock lithology,
knick-point angles are significant (Kirkby and Kirkby, 1974). On the contrary, in more
erodible substrates constituted by several types of weak rocks, the transition between the
backing hillslope and the pediment surface is gradual, but marked by a large concave
knick-zone (Dohrenwend, 1994; Oberlander, 1997b). This latter is the case of the mantled
pediments developed in the Ebro Depression of Spain and throughout the Mediterranean
littoral.
The recent tectonic history of a region (i.e. neotectonic activity) exerts a primary
influence on pediment development. Tectonically stable areas are appropriate zones for
the initiation and development of glacis, which in addition are characterized by sinuous
mountain fronts. On the other hand, where tectonic activity is relevant, mountain fronts are
linear and alluvial-fan development prevails (Bull, 1977; Bull and McFadden, 1977). All
the published observations indicate that tectonic quiescence (crustal stability) is a very
important factor for pediment development.
In arid zones, pediment development is influenced by water availability for surface run-
off. In hyperarid areas (i.e. Atacama, Central Sahara) the deficient rainfall gives rise to the
nearly absent overland flow erosion (erosion paralysis of Oberlander, 1997b), and therefore
the existing pediments should derive from previous wet (pluvial) periods. Climatic changes
also affect the size and extent of the vegetation cover on the pediment surface, and
consequently to the relative interaction of the working erosion and depositional processes.
The role of climatic changes and their impact in pediment development is analysed in detail
in Section 15.1.5 (pediment evolution).
1.3. Dominant processes in pediment development
The processes involved in the generation of overland flows are may be the more important
ones in desert areas. Because the arid regions are specifically defined by a permanent
deficit of water (Tricart, 1969), however, the previous statement may be considered a
contradiction. Nevertheless, the typical desert rainfall events are rare high-magnitude
storm events during which energetic surface run-off processes (arroyada) are generated.
Overland flows can be unconfined generating sheet-floods, but eventually can be confined
in gullies.
In regions of tectonic stability, the pediment surface can be divided into a proximal
erosive area, an intermediate transport area, and a distal depositional area (Johnson,
1932b). During episodes of increasing sediment supply these different areas shift upward to
the proximal zone, and conversely when it diminishes, these different pediment segments
move downwards to distal areas (Cooke and Mason, 1973). The erosive and transport areas
are characterized by the occurrence of mobile shallow braided channels but also by sheet-
flow processes. In these two zones of the glacis, we can observe former pediment surfaces
slightly elevated over the functional ones. These are relict landforms in which the
dominant geomorphic processes are surface and subsurface weathering, eolian reworking,
development of desert pavements with varnish or special soils such as pedocals.
For some authors (mainly Australian ones), the subsurface weathering plays an
important role in pediment development, mainly on those generated on crystalline bedrock
(Ruxton, 1958; Mabbutt, 1966, 1977; Twidale, 1967, 1983). The upper regolith cover can
be easily removed by overland flows. This fact facilitates the exposure of the basal surface
of weathering, generating an etched plain or etch surface. This kind of process is common
in savannah areas, but they also occur in the semi-arid environment. The zones subjected
to deeper weathering are localized at the piedmont junction along the contact between
the bedrock and the regolith, which constitutes a linear weakness zone.
In softer bedrock the overland flows have higher erosive power. In the case of
calcareous rocks, planation is favoured and assisted by karstic corrosion (Nicod, 1992).
In the Iberian Range of Central Spain, mainly constituted of calcareous rocks, pediment
surfaces have an excellent development, giving rise to extensive pediplains.
The study of micro-topographic changes, carried out in small experimental parcels by
Schumm (1962), provides very valuable data about the variations of the micro-pediments
over an 8 year time-series. They were monitored with erosion pins in the gentle hillslopes
developed on the siltstones of the Badland National Monument, South Dakota. The micro-
pediment surfaces underwent a progressive lowering that increased with their slope
gradient, and the pediment junction experienced parallel retreat. On the other hand, in some
of the cases, active deposition took place in the distal areas (Figure 15.8). In that paper,
Schumm pointed out that the formation of such miniature pediments did not have to be
necessarily similar to the real case. Chorley et al. (1984), however, indicated that this
process may occur in many places at larger scales, suggesting that the analysis of badlands
can improve our understanding about the development of some type of pediments.
1.4. Hypothesis on the origin of pediments
Pediment surfaces occur preferentially in zones of low-relief mountain catchments.

When the catchments are more modest, the bigger will be the surface area occupied by
pediments. The main proposed hypothesis for pediment origin is based on the explanation
of those processes governing slope degradation and piedmont generation. The different
120
%, \ \75
. . . . . ~ . . . . . . , 4~
! 4 i 3 0~L . ]
..~ O0 30 CENTIMETERS
901",~ 6.0
37Q ~ 0 -5
~ ~ . . . . . . . . . . . . . . . . . .
6
2 5
~
_
5 I 4
4
-5
97~,- q ,
17.~ g ' 6.s 4.7 3.6 3 I 2.5 I 2 I
77,3~_~_ o , 5.5
1 7
5 ~
90
95 ~ _ _ _ 10
~ , ' ~ ~ , ~ " ~ 7 , ~-~-~ ~ , 5 I 3.5 ! 3 1
Figure 15.8. Micro-pediment profiles from the Badland National Monument, South Dakota. Lower
numbers indicate pediment slope (%) and upper numbers show the location of the different erosion pins
indicating the value of ground lowering (mm) produced by the erosion. Solid lines show pediment profiles
obtained on May 1961 and dotted lines those obtained on July 1953 (Schumm, 1962).
proposals can consider either: (a) the type of overland flows working at the piedmont
zone, or (b) the set of previous degradation processes preparing the bedrock for its later
planation. We have to consider both working hypotheses as complementary ones.
The most classic hypothesis is that of the sheet flood of McGee (1897), where in the
confined streams coming from the mountainous uplands, spread and merge downslope
from the piedmont angle, giving rise to a 0.5 m thick sheet of water and silts that spills out
in an unconfined turbulent flow. These water flows correspond to flood events during
which the erosive effect on the bedrock incorporates particles of very diverse size into the
running waters. Sheet-flood processes have been also recognized by Blair (1987) as active
processes in alluvial fans based on the analysis of aerial photos of the Rocky Mountains.
Tricart (1969) pointed out that a significant relief is necessary to concentrate the surface
waters into streams that later spill their sedimentary load on to the piedmont. This
indicates that with a very low relief the sheet-flood hypothesis does not seem applicable.
On the other hand, because the pediment feeding channels emerging from the mountain
front have some spacing, the intervening (inter-channel) areas are not commonly affected
by these sheet floods. Tricart (1969) also suggested that it is necessary for the occurrence
of a flat or gently topography, prior to the development of sheet floods. Lawson (1915)
introduced weathering processes and mass movements to explain the parallel retreat of
slopes during pediment development. This author also recognized that sheet flood is the
ultimate process in pediment development.
The hypothesis of lateral planation for pediment development was first proposed by
Gilbert (1875, 1877), later accepted by Bryan (1922) and improved by Johnson (1932a,b).
In the piedmont, the confined channel becomes a group of small channels with a braided
disposition. Each individual channel has decimetric widths and depths. These little
channels can find obstacles in their trajectory, such as bushes and large blocks that have to
be bypassed. The braided pattern of these channels is the same one as those developed in
the big rivers, which implies that lateral planation occurs in response to the continuum
divergence of these small channels. The erosive work exerted by the lateral shifting of the
channels gradually removes all the small existent reliefs and, eventually it develops the
pediment. The radial systems of glacis developed at the foot of the inselbergs can be
explained in this way, otherwise it is difficult to understand from the sheet-flood
hypothesis alone (Tricart, 1969). Some authors consider that lateral planation is more
effective when it affects the less resistant rocks (Sharp, 1940; Denny, 1967), whereas in the
granites or crystalline rocks the denudation is smaller (Howard, 1942). Although the
activity of the diffuse overland flows is verified in field studies, this hypothesis should not
be considered as a general explanation (Mabbutt, 1977).
Twidale (1962) and Corbel (1963) considered that the processes of chemical
weathering have some impact on pediment development. In shield and platform deserts,
the denudation of ancient weathering profiles can generate etchplains that later are
transformed pediment in response to planation processes (Mabbutt, 1977). On the other
hand, weathering is especially intense at piedmont angle locations (Twidale, 1962, 1967,
1976a,b), which collects all the overland flows generated on the adjacent hillslopes. This
author relates the origin of both pediment and flared slopes with this kind of selective
weathering. During humid periods the weathering front progresses downslope,
while during the dry periods active denudation removes the previously weathered
material (Figure 15.9).
1.5. Pediment evolution. Pediplains
The overland flows running downslope throughout the surface of the glacis eventually
pour into endoreic depressions, seas, oceans or, more frequently to a desert stream. If the
initial base level of the pediment changes, then different geomorphological processes
reshape the primitive pediment surface. In the first case, the endoreic basins undergo a
continuous aggradation, in response to the erosion of the basin borders which give rise
to a progressive reduction of the pediment slope. When sea-level falls, just as happens in
glacial periods, headward erosion propagates upslope dissecting, separating and isolating
the previous pediment surfaces, as is evidenced in numerous examples on the Mediterra-
nean fringe. On the other hand, when sea-level rises the area occupied by the pediments
decreases, being able to even be completely covered by the marine waters.
The most common cases of pediment evolution are those related to the wadis (desert
streams). The materials transported by the overland flows generate an erosion pediment in
the proximal areas, and accumulation ones in the mid and distal areas, where deposits left
by the desert streams may merge with the alluvial plain. In these zones of coalescence an
interfingering between both kinds of deposits occurs. Pediment deposits are generally
formed by angular clasts of generally the same composition, due to the shared sediment
source of the glacis channels, whereas the alluvial plain deposits are polymictic and
more rounded due to their longer longitudinal transport (Guti~rrez and Pefia, 1976). When
base-level falls in response to climatic or tectonic causes, the lateral alluvial systems incise
/ +
-t- +
.. 1 ...... a m9 desiccated during dry season
s + + +
-,3"" "Sr-7-" + +
~." + +
+
(a) / +
f
+ + +
+ + +
+ ./ + ~ +
(b) / + +
+ + +
j + + +
+ +
(C) + +
Figure 15.9. Stages in the development of flared and overhanging slopes by knick-point (scarpfoot)
weathering and subsequent differential erosion: (a) subsurface infiltration of moisture in the knick-point
area and lowering of the weathering front; (b) base-level fall and stripping of weathered debris resulting in
the exposure of the weathering front as a flared slope; (c) repetition of process and development of double
flare (Twidale, 1976a,b).
gradually into the previous pediment, resulting in the dissection and compartmentation of
the previous pediment surface and the generation of a terrace in the margins of the desert
channel (Figure 15.10). The development of similar cycles of base-level fall originates a
group of staircased pediments (Figure 15.11). These correspond to the glacis embfitds
(Joly, 1950; Dresch, 1957) of the French authors and to the stepped pediments or staged
pediments (Mabbutt, 1977) of the English writers. This system of staircased pediments
reveals the occurrence of successive phases of incision of the desert channels, with
intervening periods of glaciplanation or pedimentation. This model characterizes glacis
evolution, but mainly for soft rocks that evolve quickly in semi-arid climates (for example,
the Ebro Depression and inter mountainous basins of the Iberian ranges of Spain)
(Figure 15.12). On the other hand, in arid climates sensu stricto, where precipitation is
Figure 15.10. Geomorphologic map of a sector of the Alfambra river valley in the northern Teruel Basin,
Iberian Range, Spain. The upper mantled pediment (P3) interfingers with the upper fluvial terrace (T3) and
the lower mantled pediment (P2) with the corresponding lower terrace (T2). Interbedding of mantled
pediment and fluvial deposits occur at the connection zone. (Guti4rrez and Pefia, 1976).
Figure 15.11. Staircased pediments in which are evident the concave knick-zones with the backing
slopes of the Sierras Exteriores in the Pyrenees. Hoya de Ayerbe, Province of Huesca, Spain, Ebro
Depression.
Figure 15.12. Staircased pediments and fluvial terraces in the Sierra del Pobo, Escorihuela region,
Province of Teruel, Northern Teruel Basin. This picture covers a small zone of the area mapped in
Figure 15.10.
Figure 15.13. Mantled pediments that evolved to a desert pavement with varnish. Negev desert, Israel.
very scarce, the altemation between periods of planation and incision takes place within
large spans of time (Figure 15.10), because climatic oscillations of great magnitude are
needed to alter the dominant geomorphological processes. Finally, in hyper-arid areas
fluvial processes are almost completely inactive. In hard rocks, as the granitic ones, the
incision of the lateral tributaries is impeded by the own resistance of the rock, but a
previous weathering for pediment development seems necessary.
In areas of tectonic stability, when arid conditions persist over long periods of time,
the pediments continue their enlargement, as a consequence of the continuous parallel
slope retreat of the adjacent mountain catchment. Afterwards the pediment surface can
develop isolated inselbergs that eventually are also denuded. The final relief corresponds
to an extensive planation developed in arid climates, corresponding to those labelled by
King (1953) as pediplains in opposition to the concept of peneplains of Davis (1899),
which develop in humid climates. The pediplain is the final result of a pediplanation
cycle that is reactivated by surface uplift (Figure 15.14). King (1953) thought that
pediments were the fundamental landforms for all the worldwide landscapes.
Pediplanation gives rise to several consecutive pediplains within the continents. These
are normally assembled incised one into another generating an staircasement
(piedmonttreppen) of pediplains (King, 1962, 1976). These pediplains can be fossilized
or exhumed, as the pre-Triassic and intra-Miocene surfaces of the Iberian range, or may
outcrop, as the different erosion surfaces later developed in this same mountain system
(Guti~rrez and Gracia, 1997) (Figure 15.15). By means of the correlation of the erosion
and sedimentation records, King (1976) differentiated six global planation cycles from
the Jurassic to the Quaternary.
(a)! Uplifted Pediplain
V Base-level of erosion
(b)l
I
I
x\
,
.,.
, ,
ii
,
ii
,
ii
/
~
\
xx
~
x
~
,,\
, ,
it
,"
I
,
I" x \ 1 I I ! \ \ ~ \ i s i
I'~ '. '. , t ., i \ ~ , , / /- .,
I ,, ~ t / / I ~ \ x x / / i
I x_ x x ~ ~ / / \ x x. \ / t /
I% x xx X\ I / / / \ \ ~\ X t I /
I\ \ . ., r t ] \ x . ~ / / / /
(c)!
l/"\ -- i
iI \ [
I ~ i
Figure 15.14. The pediplanation cycle. (A) Youth. (B) Maturity. (C) Old age (Interpreted by Small,
1970).
2. Alluvial fans
2.1. Definition, antecedents, and terminology
Numerous known definitions are based on the geometry, sedimentation area, and
processes involved in fan development (Rachocki, 1981). The definition of Bull (1968),
one of the most excellent investigators in this field, points out that an alluvial fan is a
fluvial depositional body and the surface of which approaches the segment of a cone
radiating downslope from a point where a stream channel emerges from a mountain area
(Figure 15.16).
The history of the knowledge of alluvial fans, also called dejection cones (G6mez
Villar, 1996), goes back to the 19th century when these landforms were initially
brought to modest attention. Only in the late 19th century were some excellent papers
published (Gilbert, 1875; McGee, 1897) in relation to the exploration of the arid
territories of the southwestern USA. In the first half of the 20th century the
investigators started to inquire about the origin, sedimentation processes, reworking,
and tectonic disturbance of fan surfaces (Johnson, 1932a,b; Blackwelder, 1928, 1931).
The leading impulse to the analysis of these depositional alluvial-fan landforms,
Figure 15.15. Planation surface levelling an old folded cretaceous sedimentary sequence. This surface
corresponds to the pediplain elaborated during the middle and upper Miocene in central Spain, where due
to its outstanding outcropping area is denominated as the "Fundamental Erosion Surface of the Iberian
Range." The picture illustrates the Aliaga zone, of Teruel Province. northeastern Iberian Range.
however, took place during the early second half of the last century with the
publication of the key works of Blissenbach (1952, 1954), Bull (1963, 1964a,b), and
Denny (1967).
After that, alluvial fan analysis was focused on different topics, and their study also
extended to their identification and characterization in the ancient stratigraphical record.
The study of both ancient and modern fan deposits, took into account their geometry,
sedimentation and erosive processes, as well as the factors involved in fan development
and evolution. The specific works resulting from these studies will be cited and
commented in the following sections. Nevertheless, it is convenient to highlight some
papers and thematic monographs in which the bulk of knowledge about alluvial fans is
synthesized (Yazawa et al., 1971; Bull, 1972, 1977; Schumm, 1977; Colombo, 1979,
1989; Rachocki, 1981; Nilsen and Moore, 1984; Ashida, 1985; Lecce, 1990; Rachocki and
Church, 1990; Cooke et al., 1993; Blair and McPherson, 1994a,b; G6mez Villar, 1996;
Harvey, 1997a).
Alluvial fans have also been studied by means of simulations (Rachocki, 1981) and
experimental laboratory models. In spite of the considerable amount of problems
generated by simulations, very good results have been obtained (Blair and McPherson,
1994a). In this way, laboratory models have reproduced the development of sieve lobes in
sand-box analogues (Hooke, 1967). Moreover, using such sand and mud analogues, it has
been concluded that debris flow deposits predominate in the fan proximal zone, those
interfingered with channel deposits characterize middle fan locations, whereas at distal
Figure 15.16. Alluvialfans in the Sierra de Alhamilla piedmont, Almer/a Province. Betic Cordillera. The
distal zones are subject of intensive greenhouse agriculture.
areas fluvial deposits dominate (Hooke, 1967, 1987). Other laboratory analogues were
specifically projected for the establishment of the relationships among drainage basin
evolution, fan development and fan stratigraphy (Schumm, 1977; Hooke and Rohrer,
1979; Schumm et al., 1987).
Once the antecedents and trends on alluvial fan research are examined, it is necessary to
emphasize some relevant fan features, as well as to specify those terms commonly used
in the literature. Alluvial fans are characteristic landforms at mountain front locations
(Figure 15.16), but they also occur in intermontane basins and valley junctions
(Figure 15.17) throughout all the different morphoclimatic zones, and with very variable
sizes. An alluvial fan results from the sedimentary output of single or multiple drainage
basins and its morphology and stratigraphy record the process of basin excavation. The
water and sediment fluxes concentrate in the lowermost zone of the basin where a
mainstream channel issues from the mountain catchment. These mixed fluxes expand in an
unconfined flow at this kind of inland basin outlet where, as a consequence of the abrupt
slope and stream-power reduction, the transported load is deposited to build up the
alluvial-fan bodies.
In these settings individual fans frequently coalesce away from the mountain front
locations. In arid and semi-arid zones the alluvial fans constitute the most characteristic
landform on the piedmonts of basin and range desert zones, (Figure 15.18). Distally, the
coarse detrital fraction that constitutes the main fan-bodies, may grade to the thinner
sediments of playa-lake environments.
Figure 15.17. Alluvialfans dissected by the Mendoza River. Uspallata, Mendoza Province, Argentina.
Andean Cordillera.
Blissenbach (1954) divided the alluvial-fan bodies into three different parts
(Figure 15.19): the proximal area, or fanhead, located at the apex; the middle area,
between the fanhead and the lower parts of the fan; and the distal area constituted by the
zone more distant from the apex. Finally, it also included a lowermost area where fan
coalescence occurs.
The climatic classification of alluvial fans into "humid" and "dry" fans, or in other
words fans of arid zones formed by intermittent fluvial flows, and fans of temperate zones
generated by perennial flows (Schumm, 1977) has a positive acceptance. Nevertheless,
this classification is still controversial and it was enlarged by McGowen (1979) when
considering that in arid zones debris flows are more common than in humid areas, where
fluvial sedimentation prevails. Recently, Nilsen (1993) and Stanistreet and McCarthy
(1993) related the form and processes of the fan with the climate, concluding that their
semi-conical shape is built by debris flows in desert environments ("dry type"), whereas
other authors consider that this geometry is also developed in humid temperate regions
(Harvey, 1984a; Blair, 1987).
In dry regions fan development is favoured by different reasons (Harvey, 1997). The
vegetation cover is sparse or absent, intense storm rainfalls are frequent, and overland
flow processes predominate on the hillslopes. In consequence, there is a high rate of
sediment production during the storm conditions. The steep mountain streams with flashy
run-off regimes give high rates of sediment transport and delivery to mountain front
locations. Finally, the sediment transfer from the drainage basin to the fan body is
sporadic and basically linked to the aforementioned intense storm conditions.
Figure 15.18. Landsat image of the San Juan region of Argentina. Hills located to the northeast are the
Sierras Pampeanas, and those located to the northwest belong to the Andean foreland, which are separated
by the Bermejo Depression. The piedmont zone of these reliefs are constituted by coalescent alluvial fan
systems, which occasionally merge with playa-lake systems, locally called "barreales." San Juan City is
located to the south of the picture.
2.2. Alluvial fan morphology
The main morphological features of alluvial fan systems (Blair and McPherson, 1994a)
(Figure 15.20) are the following ones. The drainage basin constitutes the upland area,
generally with steep slopes, from which water and sediments are derived. The feeder
channel transfers the main fluid and solid fluxes to the fan apex and it corresponds to the
outlet point of the drainage basin in the mountain front, this apex is obvious if the front is
sharp. The incised channel is the downslope prolongation of the feeder channel on the fan,
and it may be divided into several channels. The incised channel usually terminates in
proximal or mid-fan areas, corresponding in its location to the intersection point. From this
point flows depart, become unconfined and expand downslope giving place to the active
depositional lobes. The headward-eroding gullies are common features in the distal part of
fans. These channels may progress upslope by headward erosion, giving eventual rise to
/ - - u w ~ r , a l , \
nt ~j-s'~gment X Zone of coalescence
1- with other fans
Apex
1 5 KM
| . . . . . 9
150 I" I
Apex ",,,. &~,
100 -
z
m
uJ ~ Radial profile A-A'

a
t-
n
t-
..J
< 50
01 > 02 ~ / . / ~
_
Vertical exaggeration x 53
.. i. . . . . . 1 .
5 10
SCALE IN KM
Figure 15.19. Typical segmentation and longitudinal profile of an alluvial fan. Example from the Tumey
Gulch fan, western Fresno County, California (Bull, 1964a).
the capture of the incised channel and promoting the downslope shift of the active
depositional lobes (Denny, 1967).
Because the depositional processes are strongly controlled by the sediment supply on to
the fan surface (Harvey, 1997), the bulk morphology of an individual fan should reflect the
main characteristics of its source area. Many of the studies on the morphology of alluvial
fans are based on the analyses of morphometric variables, such as fan shape, fan area and
fan gradient (slope).
Figure 15.20. Main geomorphological features of an alluvial fan system. Example from the Trail Canyon
fan, Death Valley, California (Blair and McPherson, 1994a).
The fan shape is semicircular, or a cone segment, with concave longitudinal profiles
(Figure 15.19) and convex traverse ones. Data extracted from detailed topographical maps
allow development of the following mathematical relationships (Troeh, 1965):
Z = P + SR + LR 2
where Z is the height of a particular point on the fan surface, P the height of the fan apex, S
the fan slope in point P, R the radial distance from P to Z, and L the average slope along its
longitudinal profile.
The fan area is the more studied variable and it has been demonstrated that there is a
simple relationship between fan area and drainage basin area that may by expressed by the
equation:
Af - pA~
Where Af and A b are fan area and drainage area (measured in km2), respectively. Studies of
different groups of fans have demonstrated similar values for the exponent q in different
zones ranging from 0.7 to 1. l, but the values for p show a wider range between 0.1 and 2.1.
These variations are interpreted as a consequence of differing fan age and history between
different regions, but also to the different rock resistance within the drainage basins
(Harvey, 1997). Figure 15.21 displays the relationships among the different morphometric
fan variables for different areas. Fans of the California Coastal Ranges have larger surface
area per unit drainage area than fans in the Death Valley region. The regression line for the
fans Spain of southeast has been obtained from 68 fans (Harvey, 1987a).
The fan gradient is usually taken as the axial fan surface slope in the upper part of the
fan and the most characteristic values range from 2 to 12~ Its relation to the fan is:
Gf = aAbc
Where Gf is the fan gradient. Figure 15.19 illustrates that the values of the exponent b
range between - 0 . 3 5 and -0.15, but values for the constant a show a greater range from
0.03 to 0.17, which is interpreted as linked to different sedimentary processes. Fans in
the Death Valley region are much steeper than the California Coastal Ranges, whereas
the SE Spain fans show an intermediate position. On the other hand, in the SE Spain fan
group of southeast Spain, those dominated by debris flows have steeper slopes than those
dominated by fluvial and/or sheet-flow processes (Figure 15.22), which reveals the
relevant impact of sedimentary processes on fan gradient (Harvey, 1984a).
2.3. Factors influencing alluvial-fan development
Fan deposition takes place if the sediment supply overcomes the transport capacity of the
feeder channel. This indicates that within the channel the critical stream power threshold
defined by Bull (1979) is surpassed. However, fan sedimentation is also affected by
different factors, the impact of which is variable depending on the different climatic and
topographic scenarios.
Topographic factors have a strong influence on sediment supply. In that way erosion
rates tend to be more important on steep slopes, favouring in addition the generation
of debris flows. These predominate in fans supplied by small steep drainage basins.
200 f
J
f
f
, Spanish Fans f
jJ
jJ
100 Californian Coast R a n g e s J
f
f t
P f
50 -I Death Valley Region ff /
/ f
i jJ
J j
f P
/ /
20 I
J /
J
f f
i J
I S
f
10 /
jpl j#
E
v J f
/
5 p
/
.J
p
t" pJ
p
J
J
u_ 2
0.5
p
/
0.2
0.1
I I I I I I I I
0.2 -
0.1 -
0.05
0.02 -
t'-
u.. 0.01 -
0.005 -
0.002 T ~ ~ i ~ s i ~ 1 '1
0.1 0.2 0.5 1 2 5 10 20 50 100 200
Drainage area (km 2)
Figure 15.21. Regression relationships of drainage area to fan area and fan gradient for three different
regions (Harvey, 1984a).
A Debris flow fans
0.5 --
o Intermediate and fluvial fans
0,2 --
r
(1)
O 9 9 9
0.1 - o A 9 9
03
x_
O')
9 O O
r A o=
03 o %0 ~ o
a_ 0.05 -
o o o
o o
0.02 -
0.01 I I I I I I I 1
0.1 0.2 0.5 1 2 5 10 20 50
Figure 15.22. Drainage are to fan gradient relationship for selected alluvial fans in SE Spain. Contrast
plotting position of steep slope debris flow fans with the gentle slope intermediate and fluvial fans (Harvey,
1987b).
On the other hand, fluvial processes tend to characterize the fans supplied by large and less
steep basins.
The clasts mobilized within the drainage basin of the fans came from the weathering,
collapse or removing of the outcropping bedrock substrate in their slopes. This lithologic
factor exerts a relevant influence on alluvial-fan development. The weathering processes,
mainly mechanical in arid zones, work- conditioned by the rock type, and consequently the
resultant sediments can be quite different. The resultant clast sizes mainly depend on the
granulometry (in detrital rocks) and on the spacing of discontinuity planes affecting the
bedrock. Therefore, when there is a wide spacing in the source rocks, they yield blocks or
coarse clasts. Because in arid zones chemical weathering is practically absent, the silt and
clay particle sizes are commonly scarce. Their abundance increases considerably,
however, where source rocks are fine grained, such as marls, mudstones, siltstones and
tephras. These fine materials are necessary to generate debris flows within the drainage
basins, just as Harvey (1984a, 1987) pointed out in different examples of alluvial fans of
SE Spain.
The climatic factor has both a direct and indirect influence on the weathering processes
conditioning the sediment production within the fan basins. Water availability directly
affects rock weathering, sediment transport, and also indirectly the vegetal cover
development. The intensity, duration, and frequency of strong rainfall events are the most
influencing factors in sediment production. The lack or scarcity of rainfall events promotes a
drastic reduction on the weathering and transport processes. Nevertheless, in areas such as
Death Valley, with a mean annual precipitation of 43 mm, the sediment suite accumulated
within the fan bodies is only derived from rainstorm events (Blair and McPherson, 1994b).
The temperature can also affect fan development because it may increase the rates of
chemical weathering. On the other hand, it is necessary to consider the common occurrence
of altitudinal coolness gradients, which may give place to congelifraction in the headwater
zones. Also, the amount of precipitation increases with altitude. These orographic gradients
may affect the vegetal cover, which in turn has a relevant impact on soil erosion and
slope stability. This is mainly driven by the development and penetration of the roots in
the ground, which is different for different types of plants. Consequently, when climatic
changes, or human activity (overgrazing, deforestation, fires, etc.), impact the expansion
and type of the vegetal cover, the affected hillslopes may became unstable.
Fans can develop in areas of tectonic quiescence, but also in active tectonic zones. The
latter are characterized by the occurrence of basin and range deserts such as those
developed in the semi-arid southwestern USA and central Asia. The tectonic activity
(tectonic factor) may condition the appropriate topographic settings for fan development,
such as active mountain fronts (Figure 15.23). At these locations, thick sequences of
alluvial fans may develop, recording an important depositional history all over the recent
Figure 15.23. Schematicblock diagram of an active mountain front showing the typical internal structure
of the thick alluvial sequence of the piedmont. The occurrence of triangular facets indicate recent faulting
along the mountain front (Bull, 1977).
Figure 15.25. Tectonic tilting and erosion of Pleistocene alluvial deposits in the Andean piedmont close
to the City of Mendoza, Argentina.
geological periods. Neotectonics constitutes the main control in fan location and
development in these areas. Many of the old sedimentary sequences of alluvial fans are
commonly interpreted from a tectonic point of view (Harvey, 1997). On the other hand, the
tectonic activity is not limited to the main range-front fault (Figure 15.24), but ground
deformations and faulting may also affect the entire fan system. These recent tectonic
ground dislocations may affect the feeder channels of the fans giving rise to relevant
channel offsets and displacement with the consequent shifting on the locus of the active
sedimentation (Figure 15.25).
Tectonic activity can also affect fan development by means of direct modification of
the fan slope. When the uplift rate of the mountainous front is greater than the incision
rate of the fan channel (Figure 15.26), fan slope increases and sedimentation occurs in the
proximal zone. On the other hand, if the channel incision rate exceeds the uplift rate the
active deposition shifts to the mid- or distal zones of the fan. Figure 15.26 illustrates how
during limited aggradation conditions the development of fanhead trench in the proximal
zone transfers the active deposition to distal locations (Bull, 1964a, 1977). This process
gives rise to segmented fan surfaces constituted by fan segments of different ages located
at different topographic positions. These different fan segments can display a different
degree of incision, as well as soil development (calcretes, desert pavements, and varnish,
etc.). These segmented fans (Figure 15.27) still have a very controversial origin, and
different hypothesis, based on recent extreme events, and complex responses to climatic
and tectonic changes, have been proposed (see Cooke et al., 1993).
Finally, Blair and McPherson (1994b) consider a set of extrinsic factors related to the
effect of neighbouring environments on fans. In this way, fans can be affected by aeolian
sedimentation, because windblown deposits can modify or even obstruct the fan channels.
The development of sandsheets or dunes in distal fan locations may occasionally promote
channel ponding. The fluvial channels developed in downfan locations can capture
the main fan channel, triggering an abrupt fall in the fan base-level and its subsequent
dissection. If the alluvial fans are bounded by either lacustrine or marine environments,
the subaerial fan flows may transform into fandeltas. In these cases, sea- or lake- level
changes directly affect the fan base-level influencing its erosional or sedimentary
behaviour. Finally, volcanic activity can generate lava flows that can interfere and modify
the internal drainage of fans. The occurrence of ashfall within the drainage basins favours
the generation of debris flows from the steep slopes during storm events.
2.4. Sedimentary processes
The sedimentary processes involved in fan development are of two types, primary and
secondary (Blair and McPherson, 1994a,b). The primary processes, or sediment supply
processes, are those responsible for the transportation of sediment from the drainage basin
to the mountain front locations where alluvial fans develop. The dominant transport
mechanism will be strongly conditioned by the water to sediment ratio in the fan flows, as
well as by the availability of fine, mud-sized materials within the drainage basin (Wells
and Harvey, 1987; Harvey, 1997). A low concentration of fines and high water/sediment
ratios favour the occurrence of fluvial transport of coarse clasts by traction. When the
content of fines increases the clasts can be matrix-supported and mobilized in a debris flow
Figure 15.26. Development of segmented alluvial fan in response to the different uplift rate of the
mountain front. In the upper sketch uplift and aggradation rates are bigger than channel incision. In the
lower sketch channel incision is bigger than uplift. In this case active depositional zone has shifted
downfan as consequence of channel entrenchment an intersection point development giving place to a
telescopic fan (Bull, 1977).
Figure 15.27. Segmentedfan surface. The upper segments are highly eroded by gulling, but also by the
dissection of the Jachal River. The picture was taken from a place located between Rodeos and San Jos6 de
Jachal, San Juan Province, Argentina.
movement by plastic flow, in which the solid and liquid phases move in a cohesive way.
The secondary processes, or reworking processes, lead to fan degradation and erosion
through the remobilisation of the sediments previously deposited by the primary
processes. Many of these processes work during prolonged spans of time on to the fan
surfaces, whereas the sediment supply events are generally triggered by low- frequency
and high- intensity storms.
2.4.1. Sediment supply processes
These may be divided into two classes: fluid-gravity and sediment-gravity flows (Blair and
McPherson, 1994a,b). The fluid-gravity flows can be described as typical fluvial flows,
whereas the sediment-gravity flows are mass movements on the colluvial slopes or on
fragmented bedrock within the drainage basins. These mass movements can be rockfalls
(i.e., gelifraction), rock avalanches (i.e., overturns), gravity slides (i.e., landslides, mud
flows) and debris flows.
The rocky materials coming from the upper parts of the drainage basin can reach the
proximal area of the fans. Rock avalanches take place by the abrupt fall of a large bedrock
mass that can disintegrate into boulders and clasts of diverse size, which move downslope
as a high-velocity granular flow. In the rock avalanche of the Monte Huascar~in (Cordillera
Blanca, Peruvian Andes), that took place in May 1970, a mean speed of 280 km/h was
reached. These granular flows may feed the fans (Nicoletti and Sorriso-Valvo, 1991) and
can reach run-out distances of several kilometres. Their resultant deposits, denoted as
"megabreccias" by some authors, display a characteristic coarsening-upwards bedding
with blocks up to 20 m at the surface, and matrix-rich lower layers. In plan view they
present a lobate shape, as in the case of the rock avalanches of the Sierras Pampeanas
in NW Argentina, studied by Fauque and Strecker (1988). They were 3 to 4 km long,
0.5 to 1.5 km wide, and displayed a front rim of 8 to 10 m high. Inside these lobes older
similar deposits were recognized.
Gravity slides (landslides) correspond to rapid downslope movements of the colluvial
slopes and/or the rocky substrate along a preferential shear surface. These include
rotational and translational slides. In these cases the mobilized material is a coherent and
homogeneous block, opposite to the granular flows involved in rock avalanches. These
processes are important because they directly or indirectly feed the fans. Also, many times
the generated landslide scars work as the nucleation zones for gully or stream channel
generation (Sorriso-Valvo, 1988). As in the case of rock avalanches, seismic shaking can
trigger most of these slope movements, whereas the debris flows are commonly generated
by the progressive addition of water to the colluvial slopes that eventually fail by plastic
flow.
Debris flows are one of the most characteristic deposits of arid and semi-arid zones in
both modem and ancient alluvial records. They constitute flows of sediments formed by
a mixture of coarse clasts embedded in a fine-grained matrix with a variable content of
water and air (Figure 15.28) (Johnson, 1970; Innes, 1983; Costa, 1984, 1988a,b; Johnson
and Rodine, 1984; Takahashi, 1991; Coussot and Meunier, 1996). Their source areas are
normally constituted by the colluvial slopes or high weathered zones of the drainage
basins. In the mountainous deserts, a complementary source area is constituted by
alluvial materials deposited during previous stages of fan activity. The water required to
generate the debris flows usually comes from high intensity storms, but sometimes also
from the rapid thaw or rainfall events on the snow covers located in the upper zones of
the drainage basins. These facts produce rapid overland flows that may generate debris
flows where they incorporate the heterometric fragments and silt/clay-size particles
mantling the hillslopes. Where the flow velocity increases and the internal shear
resistance reduces, they work as a viscous fluid (Johnson, 1970). The viscosity of debris
flows may vary considerably, and when they present a high viscosity and volume, the
channel-floor erosion can give rise to sharp levees along the channel banks. On the other
hand, with a low viscosity, their downslope movement over grassland slopes produces a
very limited erosion and smaller levees (Selby, 1993). All these processes are governed
by a flow of wet sediments moving downslope within the stream channels. The
morphology of debris flows is variable, but generally displays a set of common features.
Their most frequent morphologies are constituted by hummocky surfaces with lobate
Figure 15.28. Deposits faulting affecting to fan debris flow deposits (left) and bedded fluvial deposits
(right). Pleistocene alluvial fans of the Blanco fiver valley. Andean Precordillera, Mendoza Province,
Argentina.
frontal areas (Figure 15.29). In some cases the lateral and central area of the debris flow
lobes terminate in a frontal talus (Johnson and Rodine, 1984). Debris flows may exert a
relevant impact on the main fan channel, even overflowing it, and leaving deposits on
both banks. The internal structure is massive with large boulders randomly distributed
and embedded in a fine-grained matrix, but locally they may exhibit thin sand or silt
lenses deposited by intervening "fluvial" flows. These bedded deposits, denoted as
transitional deposits, can be generated during transitional stages between debris to fluvial
flows, apparently at the end of debris flow events (Wells and Harvey, 1987). These debris
flows are more frequent in the proximal areas of alluvial fans. Mud flows are mainly
constituted by silt-clay materials with occasional clasts. These clasts can display
imbricate fabrics parallel to the depositional flow, especially when clasts are elongated.
Mud flows, however, are rare deposits in the alluvial fans developed in arid and semi-
arid regions.
Two mechanisms have been traditionally considered for the origin of debris flows. The
more accepted idea considers the generation of debris flows from the downslope failure of
steep colluvial slopes (27 to 56 ~ triggered by the progressive wetting and saturation of the
ground during rainfall or snow-melt events. The water saturation of loose detrital materials
promotes the reduction of their shear stress assists their downslope movement as debris
flows (Campbell, 1975; Costa, 1988a,b). Other proposals suggest that debris flows are
generated by the mobilization of the loose materials mantling the hillslopes by single
Weathering processes and resulting forms 257
Figure 11.15. Gnammas in or pits, of unequal development, elaborated on Miocenes sandstones. These
small closed depressions are filled up with silt-clay material. Region of Alcafiiz, Ebro Depression, Teruel
Province, Spain.
slopes, with a triangular cross section and constituting an intermediate step towards tafoni
morphologies. Their origin is due to differential weathering along joints or especially
along intersection of weakness planes, although in some cases no structural control can be
observed. They can also develop under a soil or below a mantle of alteration (Figure 11.15),
2 3
4a 4b 5
Figure 11.16. Gnammas cross sections of the Alcafiiz region, Ebro Depression, Teruel Province, Spain.
(1) Gnamma with smooth borders. (2) Gnamma in or pit. (3) Gnamma in or pan. (4a) Gnamma in with
overhanging sidewall. (4b) Asymmetrical gnamma. (5) Gnamma in armchair or armchair (Guti6rrez and
Ibafiez, 1979).
Figure 11.17. Gnammain or pan partially filled of water, with an overflowing channel, developed on
Miocene calcareous sandstone. Region of Alcafiiz, Ebro Depression, Teruel Province, Spain.
in which later denudation is necessary to expose them. Exfoliation and granular

disintegration and in some cases, abundant lichen colonies can play a certain rule in their
generation.
The dominant processes are wetting and drying and salt weathering. In the case of
calcareous sandstones the dissolution of the carbonates is very important (Gutirrrez and
Ibfifiez, 1979), and if sandstones are siliceous, quartz lixiviation takes place if the pH is
over 8 (Benito et al., 1993). Once the closed depression is formed, water accumulates
inside it and chemical weathering can be predominant. The particles inside the gnamma
can be exported by deflation or by water overflowing when the depression fills up. In
some outcrops the presence of little channels of overflowing fitted into the rock is
common (Figure 11.17). The origin of the overhanging sidewall in some gnammas can
be due to a water prevalence in the bottom, for which weathering is major in this zone.
When gnammas and tafonis coexist in the same outcropping this can produce the
capture of one by the other, leaving the gnamma like an empty background buttonhole.
Also on occasion the headward erosion action of the fluvial network can reach in to
capture some gnammas.
Chapter 12
Desert surfaces: pavements, patterned ground,
varnishes and crusts
1. Introduction
The surface materials of arid zones and their existence at shallow depths are subjected
to numerous processes that act together or separately, depending upon lithology, climate,
relief, and so forth. The result of this activity leads to the generation of desert pavements,
patterned ground, varnishes, and crusts.
2. Desert pavements
2.1. General characteristics
Desert pavements consist of rounded or angular rocky fragments within a matrix of finer
sand, silt, and clay-sized material (Mabbutt, 1965b, 1977). They develop not only on low
relief surfaces such as alluvial fans, terraces and pediments, but also on slopes. They can
occur in other morphoclimatic settings, but it is in arid zones where they reach their
maximum expression, going on to occupy significant areas.
Locally they are given many names: gibber in Australia, sai" in central Asia and
hamada, reg, and serir in the Sahara, the Near East, and India. Hamadas are tabular relief
structures, pedimentation, or glaciplanation surfaces of negligible slope. The surface can
be devoid of fragments, in which case it is denoted a rocky hamada, or it can be covered in
angular blocks (boulder hamada) (Figure 12.1) that have not suffered any transport. In
some cases the blocks have suffered intense weathering in which a continuous flaking has
rounded the fragments (Evenari et al., 1971a,b). Reg or serir can develop, however, on
alluvial surfaces with a slight or negligible slope and in these cases the stones are rounded
as a consequence of their transport (Figure 12.2). At the present time many reg or serir
appear as relict surfaces that are remote or left away from the current drainage network.
A well-developed pavement consists of a stable surface, but when altered by
anthropomorphic action (e.g. all-terrain vehicle traffic in the SW USA and Saudi Arabia;
Webb and Wilshire, 1983) water and aeolian erosion can lead to substantial modification.
Similarly, the famous Nazca Trails, southem Peru (Figure 12.3), which date from
thousands of years ago, are desert pavements partially mobilised during their construction.
Elsewhere, the pavement's constituent stones can be prehistoric artefacts, and other times
the rocks are painted or carved (petroglyphs) (Brakenridge and Shuster, 1986),
circumstances by which it is possible to gain information about the relative age of desert
pavements.
Figure 12.1. Hamada covered by lag angular boulders. In the central part, there is a small closed hollow
(referred to as a "daya" in the Sahara) filled with fine deposits lighter in colour. This deposit may come
from eolian material, which has been reworked by run-off. Rann, Thar Desert, India.
Figure 12.2. Typical reg of rounded clasts developed on an alluvial plain. Bhojka, Thar Desert, India.
Figure 12.3. Paths of Nazca built on a desert pavement. Southern desert of Peru. Photo by C. Zazo.
2.2. Processes that participate in their formation
Desert pavements can originate via a single or various processes, which indicates a
problem of convergence of forms. Fundamentally the processes can be extrinsic in
character, produced by water and aeolian action, and intrinsic, in which the interior of the
soil is modified by differentiation processes.
The most universally applied process of pavement genesis is deflation, whereby fine
particles are exported by the wind, leaving a residue of coarse material that goes on to
constitute the pavement. Evidently, exportation is minimal and as a consequence the
surface reduction is slight. This is due to the fact that deflation decreases with increasing
surface roughness. It has been calculated that with a surface rock coverage of 50%,
exportation becomes ineffective (Sharon, 1962). Furthermore, a biocrust film sometimes
develops in the spaces occupied by the fine material, which impedes deflation and
stabilises the pavement, leaving the surface "sealed" (Campbell et al., 1989).
For some authors (Sharon, 1962) water erosion is more efficient in the generation of
pavements. In ground plot studies, surfaces with gradients of 5% exhibit a reduction of
5 to 50 cm over a period of 5 years. The exportation of fine material is accomplished via
rainsplash and surface wash, although the flow of water is made difficult by the presence of
stones. In the case of biocrust occurrence, splash is reduced considerably and the capacity
for infiltration diminishes markedly (Poessen, 1986), limiting water penetration to the
cracks that usually develop in soils.
Although less important than those mentioned previously, another process which can
also give rise to pavements is the upward migration of stones from the soil interior, which
then concentrate on the surface (Mabbutt, 1965a; Cooke, 1970a; Mabbutt, 1977). This
phenomenon of formation can be produced by freeze pushing in cold deserts. In hot deserts
the presence of expansive clays is necessary. These swell with an increase in humidity and
lead to a rise in ground level. During dry stages the soils contract and crack. Therefore,
pavement generation under these circumstances is favoured by enhanced humid/dry
cycles.
In general, mechanical surface weathering processes have been proposed for the
development of boulder hamadas, which break up the rocks and produce a residual
pavement of angular blocks. It is also believed that sub-surface weathering can play a part
in pavement development (Mabbutt, 1977). This author indicates that the higher sub-
surface humidity content, especially in the presence of chlorides and sulphates, leads to a
more efficient weathering and, therefore, coarse grains disintegrate gradually. On the
contrary, at the surface the water availability is much lower and the pavement stones
hardly weather. The end result is that the pavement persists.
2.3. Development
From the analysis of pavement-generating processes it can be deduced that genesis via
water or aeolian erosion is rapid, unlike pavements produced by movements in the soil
interior. Sub-aerial erosion produces a descent and levelling of the surface after long and
continual activity (Mabbutt, 1977). Moreover, as the pavement develops the availability of
fine material diminishes. The pavement is subjected to a slow evolution, over thousands of
years, during which time the stones weather and the material generated is partially
exported (Cooke, 1970a). This slowness in their development was shown in the study of a
group of terraces in the Dead Sea region (Hunt and Mabey, 1966), in which the stone-
cover percentage increases with terrace age. On the contrary, the oldest pavements in
Death Valley, California (Hunt and Mabey, 1966) are characterised by having a higher
percentage of fine particles and being more compact than the more recent pavements. In
this case the evolution is the inverse of the previous example, perhaps due to a more
intense surface weathering.
In addition, it has to remarked that as pavements evolve, their salt content can increase,
as the salts get moved by capillary action and stay trapped between the stones. Equally,
pavements can experience a continuous contribution of aeolian material. In any case, once
formed the pavement is relatively stable, as deflation decreases with an increase in stone
coverage, which tends to increase with water erosion. These circumstances mean that the
pavement is a fairly stable morphology in which both erosional and depositional processes
act, although with low intensities that can also counterbalance each other. This indicates
that pavements are resistant to erosion, because they protect the underlying material that is
prone to water and aeolian erosion.
3. Patterned ground
We have already indicated in the section related to periglacial morphoclimatic zones that
patterned ground (circles, polygons, nets, steps, stripes) reach a great development in both
Figure 12.4. Desertpavement with clasts veneered with a desert varnish. Note that the clasts are sorted
and structured forming circles. Negev Desert, Israel.
high-latitude and alpine areas (Washburn, 1956). The mechanisms invoked in their
generation are numerous and complex, practically all related to freeze-thaw processes.
Nevertheless these processes can also develop in arid zones, especially in areas of low
slope, (Cooke and Warren, 1973; Washburn, 1979), although in hot deserts the formation
mechanisms are very distinct from those in periglacial zones (Figure 12.4). This is, once
again, a situation of the convergence of forms or equifinality. These arid morphologies are
fundamentally related to humidity variations within the soils and, in an environment of salt
precipitation, they occur in relation to periodical inundation and desiccation. The
microforms generated can be ephemeral because on occasion they develop and disappear
between successive pluvial precipitations.
3.1. Desiccation cracks
Upon drying, humid, fine-grained sediments reach a limit in their contraction and they
develop surface cracks. This cracking tends to form polygonal systems (Figure 12.5). The
form that the polygons acquire depends on the physical, chemical, and mineralogical
characteristics of the material and on the environmental circumstances, as expressed by
temperature and humidity percentage. Horizontally, the cracks can be straight or curved
and the length, width, depth, and number of cracks, along with how they group, is highly
variable (Chico, 1963). The profile of the polygonal cells can be flat, concave, convex, or
irregular.
Figure 12.5. Polygonaldesiccation cracks on the side of a partially flooded sebkha. Sabknet Bou Jmel.
Zarzis region, southern Tunisia.
Most of the studies of crack morphology and genesis are related to experiments
concerned with freeze-thaw cycles (Lachenbruch, 1962; Corte and Higashi, 1964 in
Cooke and Warren, 1973; Maizels, 1987), and few studies exist related to the origins and
characteristics of cracks developed on clay and saline plains in playa environments.
Cracks produced in periglacial environments, however, resemble those originating in arid
zones, because the expansion and contraction coefficients of ice and salt are similar. The
few experiments reveal, amongst other conclusions, that clays develop 40 to 50 times more
cracks than sands and gravels (Figure 12.6). Equally, the cracks get shorter with time, due
to the development of new cracks which intersect with the older cracks. Furthermore, the
space between cracks increases with the rate of desiccation, clay content and percentage of
expansive clays.
Stones found on or near the surface act as points of crack division. Also, the polygonal
cell form gives information about the origin of these morphologies. Concave surfaces
indicate rapid drying of the uppermost material, leading to the production of mud curls on
the surface lamina (Figure 12.7). A convex morphology reveals the presence of salt in the
matrix (Figure 12.8) and, finally, a flat form suggests slow drying in the absence of salt.
On the other hand, Lachenbruch (1962) differentiated between orthogonal systems, in
which the cracks intersect at fight angles, and nonorthogonal systems where the
intersection angles vary around 120 ~ (Figure 12.9). Orthogonal systems are typical of
inhomogeneous media where the cracks can display preferred orientations. On the
contrary, nonorthogonal systems form in very homogeneous materials which dry
uniformly and the cracks develop instantaneously.
(a) Coarse/Medium Sand
""-. ,.//
/ ',, ./ ....-" ;
-~........... f- ... ,,, . ..
......... - ,,,' ,.." ;, .,,,-_':_....
""-.. . . . . . - I ~. . . . . . . . . . / //
~. ........ ,/ /s s
............ -~'- ........... ~,.," s/ ~ ...........
~
iS ,/ ~ ..~"
........... .,e.. . . . . . . . . . . . . . ,," ___'~--" - .......
.,. .,- ,,, \ ---/.-"
I ,,,i \ . ..____..--
0 cm 5
! , J
(b) Clay
/'-" r162, - ' '"

.~.~1 t',.,
0 cm 5
i - I
Figure 12.6. Surface cracking pattern produced after 12 freeze-thaw cycles under experimental
conditions. (a) coarse-medium sand. (b) clay (Maizels, 1981, in Cooke et al., 1993).
All of these cracks form in hot deserts with flat topographies, which in the majority of
cases correspond to lacustrine depressions. The occurrence of locally developed giant
desiccation fissures (Figure 12.10) is also possible, with lengths of kilometres and depths
greater than 1 m. Their origin is very complex and many causes for their formation have
been put forward. They include cracking caused by the subsidence due to extraction of
subterranean water, fissurisation by hydrocompaction, contraction from desiccation,
seismic activity, and so forth. All of these processes may act alone or in combination.
In playa environments profuse nonorthogonal systems develop in salty clays, which
comprise mud plains at the margins of evaporite sedimentation (Figure 12.5). The same
system also occurs in gypsum deposits (Tucker, 1978), which on occasion exhibit thrusted
borders (Figure 12.11). In Death Valley, California, sorted polygons develop in gypsum-
rich material (Hunt and Washburn, 1966) (Figure 12.12), where cracking is generated by
processes of salt contraction as a consequence of drying or freezing, and expansion due
Figure 12.8. Nonorthogonal system of convex desiccation cracks with raindrop prints. Closed depression
of La Playa, Bujaraloz, Province of Zaragoza, Ebro Basin, Spain.
(a) (b) (c) (d)
NONORTHOGONAL REGULAR RANDOM IRREGULARRANDOM ORIRNTEDORTHOGONAL

ORTHOGONAL ORTHOGONAL
Figure 12.9. Common systems of desiccation cracks (after Lachenbruch, 1962, and Neal, 1965, in Cooke
and Warren, 1973, Figure 2.25).
.,o., .... ,....,.
,,"'('no coverage
" ' " " ",..
,' available) "'.
CERBAT p
iI
MOUNTAINS 9 i
,. ..... . s -p"
,s \
,,
,.," % ',,
1 ',,
|0 ",,. I , site I ,/,,,J/ '4 ~(
/ ~ fresh ~" t~ - , ~ 1 f l,~,k
~) fissures ~'~'~~ ~ . . _'~.," ,
l z ,~'~...... "'-~-,4;,~.~"5"," ~ "
I=_ "
iF / ~" "~fresh \
~ "X ~ fissures ~
// .
0 1.5
km ._,~',,o -'~ .
Figure 12.10. A generalised map of giant desiccation polygons, earth fissures and areas of recent
fissuring, Red Lake Arizona, based on the interpretation of aerial photographs. The fissures are attributed
to long-term desiccation and local subsidence due to water extraction and/or subsurface salt flow (after
Lister and Secrest, 1985, in Cooke et al., 1993, Figure 8.6).
Figure 12.11. Gypsum polygons with laterally expanded edges, Sabkha Matti, United Arab Emirates.
to hydration or heating. These processes, together with the mobilisation of salt by capillary
action, appear to be the fundamental reasons for the ordering and classification of gypsum-
rich soils. Equally, polygons developed over salt surfaces on both littoral and interior
sebkhas are very common (Figure 12.13). Such polygons may have suffered a complex
evolution.
3.2. Gilgai
Gilgai is an Australian vernacular term meaning pool, which is increasingly used in

scientific literature. It is used to describe very small mounds developed on clay-rich soils
Troughscontainingstones
Stonesin interiorof
polygonsreston
pedestals;some
9 s s
~ ~ j~/Roc~ ~ ~
RockSalt ~ ~ Gypsiferous
a ndGsYtPs!erol
ut
Wedge-shaped
massof hardened
gypsumaboveopen 0 metres 1
crackin rocksalt 9 . t
Figure 12.12. Sorted stone polygons related to ridges and cracks in an underlying layer of rock salt,
Death Valley, California (after Hunt and Washburn, 1960, in Cooke and Warren, 1973, Figure 2.30).
Figure 12.13. Polygonswith laterally expanded edges developed on a salt surface. E1 Melah Sebkha.
Zarzis, southern Tunisia.
and desert pavements (Hallsworth et al., 1955; Verger, 1964; Harris, 1968; Mabbutt, 1977;
Hubble et al., 1983; Dixon, 1994a).
Following Harris (1968) a gilgai can be differentiated into three parts on the basis of its
cross section (Figure 12.14). A channel, which constitutes a depression that can be 2.5 m
below the general ground level and can reach 12 m in length. The general level of the
microrelief is denoted as the shelf, above which small mounds of up to 50 m diameter stand
out. Some of the authors previously cited differentiate six gilgai types (Dixon, 1994a),
which in the main reflect the simple morphological classification of Verger (1964), based
on the form and mode of gilgai grouping (Figure 12.14).
Various processes and factors intervene in gilgai formation, which are developed in the
monographs of Cooke and Warren (1973) and Mabbutt (1977). Gilgai form in areas of
alternating humid and dry conditions. Dry periods, during which the surface cracks,
alternate with humid periods in which swelling dominates. They occur mainly in vertisol
areas (black expansive clay-rich soils) in tropical and subtropical zones. Gilgai microrelief
results from upward movement produced by swelling of the humid subsoil. This
expansivity, along with soil contraction, as a consequence of humidifying and drying
cycles, explains the genesis of circular and stepped gilgai (Ollier, 1966) (Figure 12.15).
Swelling is due to the presence of expansive clays, such as montmorillonite, and the
existence of elevated interchangeable sodium percentages. The gilgai amplitude increases
with increasing expansive clay and sodium content. The expansion which gives rise to
gilgai can be seen in the soil profile, where small fault surfaces and thrust planes are
relatively common. Similarly, stakes placed in the soil are found tilted or even expelled
Mound/ Mound/
Puff Depression/ Puff
~//Channel ''~--" Vertical
~ ~ e r v a l ~
GROUPINGS
BASIC Without Single
FORMS prefered prefered Several
orientation orientation orientation
a ~
a~~
0 #,#
"t3
t-
= b ~o
o
CIIIII
L
0
c~~
a 1
O~~ 0 aa e
o o0 0
t-
O
Q O o 0
0 0 O0 o ~
(/)
o ~ 0 0 e o O~
~0
~L
0
(D
0 0 ~%0 0 0
t- 0
t-
O,
I
I!
Figure 12.14. In the upper part, cross section of a gilgai (after Harris, 1968). In the middle-lower part,
classification of gilgai morphology (Verger, 1964).
after a year. Another way of producing gilgai is given by Hallsworth et al. (1955) and
expressed graphically by Verger (1964) (Figure 12.16). During a dry period, particles
fall into the interior of the soil cracks and this added material is believed to cause the
microrelief during humid periods, because it provokes significant pressure deep within the
soil. It is also important to distinguish between the various mechanisms due to the action of
distinct forces in diverse situations, in relation to cracking and the expansion-contraction
cycles (Knight, 1980), which helps in understanding gilgai genesis more adequately.
4. Desert varnish
A fine, dark film developed on exposed material commonly occurs in arid regions,
and this is called desert varnish or rock varnish. In some deserts, such as those in SW
USA, this covering occurs on 75% of the rocky surfaces (Engle and Sharp, 1958).
Figure 12.15. Developmentof circular and stepped gilgai in the central Australian Stony Desert (after
Oilier, 1966).
However, varnishes also occur in alpine, coastal, fluvial, and sub-surface environments, as
well as near springs (Dorn and Oberlander, 1982; Dorn, 1998). The dominant color is
black or brown, due to enrichment in Mn and Fe.
The origin of varnishes has been the subject of numerous discussions, often lacking
precise scientific foundations. With the development of various high-resolution techniques
(microscopy and electron microprobe), which permit detailed observations and spot
chemical analyses, the understanding of varnishes has progressed considerably. This work
started with Engle and Sharp (1958) and since then many advances have been made with
regards to composition, structure and origin (Dorn and Oberlander, 1981, 1982), and
important results were expected (Oberlander, 1994; Dorn, 1998).
4.1. Composition and age
Varnish covers rocky exposures (Figure 12.17) or surface stones (Figure 12.18). The color
tends to be black if it is rich in Mn, more orange if high in Fe and brown if both are
present in similar amounts. The varnish film is very thin and varies greatly in thickness
(0.002-0.5 mm), although it generally oscillates between 0.01 and 0.03 mm.
(a)
ii __
(b)
(c) . . . . . . .
-_-_ ;,- s --
Dry zone d Crackfillin9

l particles
Wetzone -~ Pressure
Figure 12.16. The formation of gilgai by wetting and drying (after Verger, 1964).
Figure 12.17. Bedrock surfaces covered by desert varnish. Anti-Atlas, Morocco. Photo by J.L. Pefia.
Figure 12.18. Varnishfilm on clasts of a desert pavement. Negev Desert, Israel.
It develops on all rock types, but mostly on silicates. Varnish is composed mainly of Si,
A1, Mn and Fe, with Ca, K, Na, Ba, Ti, Sr, and Cu as minor components. To these
components can be added many more in minor proportions, but which can reach
significant values locally (Oberlander, 1994). Varnish exhibits microlaminations with
different chemical compositions, mainly due to variations in Mn content. Many of these
elements are likely to originate from external sources, because they do not occur in the
supporting rock. Clay minerals constitute the major part of varnish, some 60 to 80%,
followed in importance by Fe, Mn, and Si in the form of amorphous oxihydroxides. Black
varnishes with elevated Si percentages also exist which can reach 0.01 mm in thickness
and form on siliceous rocks. This type of coveting is denoted desert glaze or silica glaze
(Dorn, 1998). It is thought to originate by chemical precipitation of monosilicic acid
Si(OH4) in the form of a gel (Krauskopf, 1956) or via biological activity from a source of
opaline phytoliths which dissolve and reprecipitate (Farr and Adams, 1984).
For many decades varnish development or degrees of patination have been used as an
indicator of relative age in geomorphological and archaeological investigations (Hunt and
Mabey, 1966; Tricart, 1969; Demangeot, 1981). These older studies indicated that varnish
did not normally appear on Holocene material, which implies that it needs many years for
its formation. Absolute age dating has been carried out using neutron activation analysis,
U series, palaeomagnetism, and AMS C- 14. For this it is necessary to collect varnish from
a surface area of 0.5 m 2 (Krauskopf, 1956), using what little organic material is present.
The application of some of these techniques has not been continuous and radiocarbon
dating has problems of contamination, through extracting part of the rock substrate along
with the varnish. Another way of obtaining relative ages has been suggested, based on
the distinct degree of leaching of the varnish constituents. A leaching index has been
proposed: Ca + K/Ti, but it presents many treatment and analytical determination
problems. All of these methods have been used in studies of fault activity, rate of sea cliff
erosion, Quaternary deposit ages and lithic artefacts (Oberlander, 1994).
4.2. Formation processes
Previously, investigators believed that the constituents of varnishes came from the
weathering of the supporting rock. Solutions rich in Fe and Mn were thought to rise to the
rock surface via capillary action, where they were deposited after humid and rainy periods.
Nowadays no one doubts that the constituents derive from external sources, through
aeolian dust transport or water transport in solution (Figure 12.19). Electron microscope
Figure 12.19. Rock varnish developed on a sandstone, probably due to the precipitation of waters with
silica on the surface. This sandstone is affected by cavernous weathering features (tafoni). Petra, Jordan.
Photo by J.L. Pefia.
investigations reveal that the varnish is superimposed on, and clearly separated from, the
rocky substrate, supporting the theory of an external origin of the constituents.
The most popular hypothesis is based on a biogenic origin, in which Mn enrichment is
due to fixing by mixotrophic bacteria which oxidise Mn 2+ to Mn 4+, along with clay
cementation which traps other elements. This then repeats itself, producing an accretion
and subsequent lamination of varnish, which in turn implies cycles of humidity and aridity.
At the present time the bacteria that fixes the Mn is believed to be Metallogenium
personatum, the oldest known living organism (Oberlander, 1994). Others propose fungi
and cyanobacterias as fixing organisms (Krumbein and Jens, 1981). Some advocate purely
physico-chemical processes (Engle and Sharp, 1958; Elvidge and Moore, 1979). Mn is
more soluble than Fe and in desert environments, Eh values are high, indicating elevated
oxidation, and pH is alkaline as a consequence of the low levels of leaching. According to
these authors, variations in oxidation-reduction and the basicity of the medium can give
rise to varnishes. It is clear that an intense debate exists concerning the origin of varnishes;
it is very possible that the conjunction of biotic activity with the physico-chemical
variations could be the route to follow in future experiments.
4.3. Environmental implications
The first studies of varnishes emphasised their climatic significance and indicated that a
more humid climate than the present was needed for their formation. This implied that
existing varnishes were all relict (see Dorn, 1998). The studies also argued for cycles of
patination, whereby the varnish is eroded and a new one created.
Currently, with modem high-resolution techniques, the varnish laminations which form
microlayers with distinct compositions are interpreted climatically. Periods of reduced Mn
accumulation correspond to arid or interpluvial phases. Periods where the Mn/Fe
relationship is increased indicate more humid, pluvial or lacustrine phases, where aeolian
deflation diminishes because of expansion of the lakes (Figure 12.20). In many places,
however, the varnish sections offer an unconvincing register of palaeoclimatic
oscillations, and are more likely to indicate that the processes of varnish formation are
very complex (Oberlander, 1994).
5. Duricrusts
5.1. Introduction
One of the characteristics of arid zones is the presence of surface and subsurface crusts of
differing chemical composition, such as calcareous, siliceous and gypsum crusts. It is also
possible to encounter laterites, formed in environments of much greater precipitation,
which are mostly found on desert margins. They extend over significant areas in arid zones
and some, thanks to their enhanced hardness, are resistant to erosion and are situated in
elevated areas. Furthermore, they are good indicators of palaeoenvironmental conditions
and an understanding of Quaternary crusts is very useful in the interpretation of past
geological stages (Goudie and Pye, 1983). They can also be of economic interest, for
2
it
1.080+ 60 ka A
1 i l l
0.5-
I i I 11111 I I I111 II1 i 1 | 1 1 1 I11 i 11 i 1| 1| i i 11 i l l l l l |il i11 iii 111 Iii 1| i i11 i II i l l l l l l l l I11 ll[I. llt lid !111 -- Ill Ill I l l [ l l l l l l l l l | l l l | I11 | I III I Ill
0 20 40 60 80 100 120 140 160 180 200 220 240 260

Depth(micrometers)
Figure 12.20. Ratiosof Mn/Fe through rock varnish on volcanics with K/Ar date of 1080 ___60 ka, Coso
area, California. Peaks and troughs are supported by multiple datapoints, but the record appears too
variable to be convincing as an index of climatic oscillations (Oberlander, 1994).
example, calcareous scales may contain uranium and other metals; they can also be used in
house construction and as road fill for motorways (Reeves, 1976).
In this section, we will try to analyse the nature, properties and significance of duricrust
scales, as well as the geochemical and biochemical implications of their formation, and
finally describe the models most commonly utilised in their description.
5.2. Cafiches
Caliche is a Spanish word derived from the Latin calix, which means lime or limestone,
and was used for the first time in 1719 (see Reeves, 1976). It is commonly used in the
Unites States and Mexico but not Spain, where the term costra calcdrea is more common,
most probably due to French (croute calcaire) and German (kalkruste) influences. At a
global level there still exists a certain confusion because the word has been used for other
types of deposits. Moreover, there are many locally-used terms in different areas and
countries (Goudie, 1973; Reeves, 1976). The term calcrete, however, used in English
literature, is gradually becoming more popular in the scientific community.
Following Goudie (1972a), "Calcrete is a term used for material of continental origin
formed dominantly, but not exclusively, of calcium carbonate, found from powdery and
nodular to strongly indurated in form and implies cementation of the soil, rock or
weathered material within a vadose zone. This definition, however, does not include
speleothems, travertines, cemented litoral deposits, or stromatolites in algae lakes."
Caliches are chalky in texture, occupy around 13% of the Earth's surface and occur in
areas with an annual precipitation between 400 and 600 mm (Goudie, 1983a,b). Other
authors, such as Rutte (1958) in his study of Spanish limestone scale, state that the limits
occur between 100 and 500 mm, with no caliches occurring above or below these values.
An optimum between 100 and 250 mm is, however, indicated. Although they have a wide
Figure 12.21. Caliche on cemented gravels, resistant to erosion, located on a wadi margin. Jaisalmer,
Thar Desert, India.
distribution, caliches are characteristic of arid and semi-arid zones (Figure 12.21) (Reeves,
1976). They develop fundamentally in surface pediment deposits and terraces, and also in
aeolian deposits. In the terraces of the River Cinca in the Ebro Depression of Spain, the
degree of calcification of the stepped alluvial levels increases with age, and is nonexistent
in the more modern levels (Sancho and Melrndez, 1992).
Several volumes have been published concerning caliches (Reeves, 1976; Bltimel, 1981;
Vogt, 1984; Wright and Tucker, 1991), studying their types, characteristics and genesis. As
given by Vogt (1984), the investigation of calcareous crust has to consider different scales or
levels. In the first place it is necessary to know the spatial distribution (at the km scale) and
its relation to the regional geomorphological context, such as how it may be related to
pediment and terrace systems. The next study has to be carried out at the outcrop level, by
means of analysing natural cuts and samples that provide sedimentological profiles, thus
allowing the caliches to be placed stratigraphically in the sequence. Finally, in the
laboratory, thin sections and electron microscopy can be used to analyse texture and
micromorphology, which provide data about the formation mechanisms. All of these
investigations should be complemented with chemical and mineralogical analyses that help
in understanding the generating processes.
5.2.1. Geochemistry and mineralogy
The chemical composition of caliches, calculated from some 300 samples from all over
the world (Goudie, 1972b, 1973, 1983a,b), is; CaCO3: 79.28% (CaO - 42.62%),
SiO2: 12.30%, MgO: 3.05%, A1203: 2.12% and Fe203: 2.03%. Those caliches rich in
magnesium are called dolocretes. The percentages of each distinct component vary
enormously from one place to another, as can be seen in the calcretes in southern Australia
(Hutton and Dixon, 1981; Dixon, 1994b). In a study of numerous chemical analyses of
caliches from Oklahoma and New Mexico (Aristarain, 1970) the calcium carbonate,
magnesium, and ferric iron contents were seen to diminish with depth, whereas the silica
content increased significantly in the middle part of the profile. Elsewhere, Dixon (1994b)
examined the role of topography in the chemistry variations in caliches and demonstrated a
higher percentage of calcium lower down in the toposequence and an increase in
magnesium in the upper part. In all studies the calcium content fell with depth.
Calcite and dolomite are the dominant carbonate minerals in caliches and are
accompanied by quartz, opal, and clay minerals. Caliches can also contain minor amounts
of soluble salts (gypsum), glauconite, phosphates, heavy minerals, etc. Mineralogical
investigations have centered around the study of the clay fraction and its origin and, as a
consequence, improving our understanding of the processes involved in the formation of
clay-beating caliches. The dominant clay minerals are palygorskite and sepiolite, although
illite, kaolinite, montmorillonite, inter-layered illite-montmorillonite, and chlorite have
all been recognised (Aristarain, 1970, 1971; Gardner, 1972; Reeves, 1976; Hay and
Wiggins, 1980). The origins of palygorskite and sepiolite have been attributed to
neoformation in an environment containing sufficient magnesium, provided by the
weathering of suitable lithologies (Hay and Wiggins, 1980; Hutton and Dixon, 1981;
Sancho et al., 1992). Other investigators maintain that they arise through the alteration of
montmorillonite and inter-layered illite-montmorillonite (Watts, 1980).
5.2.2. Morphology
Descriptions of caliche morphology are numerous and different classification schemes

have been proposed based on distinct criteria. Some are based on sequential evolution
(Netterberg, 1969), others on degree of hardness of the carbonates (Reeves, 1976). The
degree of calcification also has some bearing on differentiating between soil
characteristics (Gile et al., 1966; Machette, 1985). Some distinguish only a few
morphological types and others, on the contrary, numerous types (Reeves, 1976).
The fundamental macroscopic morphologies (Sancho and Mel6ndez, 1992) are calcified
gravels, which normally appear at the base of a caliche profile, though small levels can be
encountered in the interior of the profile. This is constituted of clasts with a vadose
carbonate coat and a matrix that may contain root casts. Chalky caliche (Figure 12.22) is
formed of loose, clay, and sand-sized carbonate particles with practically no host rock
material. It can be lightly cemented and have a leafy structure. Hardpan caliche occurs in
the upper parts of the profile (Figure 12.23). It consists of centimetre and decimetre scale
plates that stack laterally. They form ooids, basal speleothemic carbonates, root traces, and
dessication cracks. Hardpan caliche can also be found deformed, forming pseudo-
anticlines (Figure 12.23) generated by expansion and contraction as a consequence of
hydration-absorption of water to a minerals of highly soluble salts, periodic wetting and
drying, calcite crystallisation and subsequent arching. Deformation may also be due to
the presence of expandable clays (Reeves, 1976). Finally, nodular caliche can develop in
fine-grained material, in which the particles are cemented by carbonate and, in turn,
Figure 12.23. Hardpan and nodular caliche deformed by a pseudo-anticline. La Plana Negra, Province of
Zaragoza, Ebro Basin, Spain.
Figure 12.24. Nodularcaliche developed on loose fine deposits. Bawri, Thar Desert, India.
generate large quantities of nodules (Figure 12.24). It can also result from the fragmentation
of hardpan caliche, targeting loose and cemented nodules (Figure 12.25). A continuous
graduation between hardpan and nodular caliches can be distinguished.
The macroscopic caliche characteristics are frequently represented at the microscopic
scale: nodular, oolitic and speleothemic textures, carbonate coatings, traces of biologic
activity and so forth. All of the investigations directed towards a precise and detailed
understanding of these micromorphologies and textures have as an objective an improved
understanding of caliche formation processes.
5.2.3. Origin
The analysis of caliches reveals the presence of complex sequences, whereby the
superposition of various caliche horizons can be recognised. In addition, the majority of
the profiles exhibit diverse stages of precipitation, dissolution and reworking.
For caliche to form there must be a source of calcium carbonate, which may be in
the rocky substrate, plant remains, volcanic emissions, aeolian dust, and other factors
Figure 12.25. Nodularcaliche with nodule size coarsing upwards. Margalef, Province of Lleida, Ebro
Basin, Spain.
(Goudie, 1983b). For many authors, an origin of carbonate transported by wind and
dissolved in rainwater (Bltimel, 1981, 1982; Machette, 1985) is too generalised, because
caliches sometimes develop with great thickness over rocks that are practically devoid of
calcium.
In semiarid zones, where a water deficit exists, carbonates do not fully leach and tend to
move and precipitate from one place to another. The fundamental process implicated in
caliche genesis is the dissolution and precipitation of calcium carbonate in the presence of
carbon dioxide:
CaCO3 + C O 2 + H20 r 2HCO3 + Ca 2+
Carbonate precipitation takes place by the loss of CO2, evaporation, the common ion
effect, and biological activity. The biological mechanisms of fixing carbonates in soils can
be due to lichens, algae, bacterias and microrrhizae (Klappa, 1979).
Goudie (1973, 1983b) differentiated two kinds of caliche origin; one pedogenic, which
seems to be the most generalised, and the other nonpedogenic. In any case, it is very
difficult to distinguish between them, because nonpedogenic caliche may be modified by
pedogenic processes.
Pedogenic caliches result from the progressive, downward accumulation of carbonates
(the per descensum model of Goudie; Figure 12.26), through washing of the A horizon and
illuviation of the B horizon (petrocalcic horizon). As seen in young profiles, early genesis
is manifested by powdery filaments and nodules. In profiles of thousands of years of age,
however, the carbonates form a crust over stones and fill the spaces in between particles.
Dust and loess,

products of aeolian sand
grain abrasion etc
-..,
9 .5 - ..-. -:- .:....:.
" - ' 9
Volcanic dust and ash , , ,,

9 I II I I :
"'" I I I I iii
ifl~ ,: ~' t~ i I I ! iii I /
,.-~.. ~., z, ,,,
'..--..~-. , IR d,
~ ~ ~ i" .~" ~...,
.
~1~.: I I th roug IlllI
I ~"~2,~~',,,, ..~'e I I I I I
,,,
II i
.... . i I I i I
~ I/ .,,l~41~ltl IiIlI I i1.1

/ \ -.~z-~.~~.l/ ~~;'~', ,I ,,!,
/ \ Leaffall I ' : ' , " 'llil llillilllll I I
Shells,
1
organic
C
\L~ -,a,, 1, " . ' ' ~ II Accumulator plants ei
materials
Depth of rainfall {-.!7:{i7:fi-:.:7:{-{.7i{::{{7{-7}{-

.::!{ {7 ....:..}~.~7i..:...){{.7{.~7:...:..{..W..:.~:7:7..~::}:..~...~.7:{...{.i..?.:{.:-:]:::{..i.7..:{::..{.:.~{....:i~.}S:..:.~}{::/~:::y.i~7
--.,---..-:{.::i?{:,:!-.:{....-~I~i~.{..,:.-::..-..i.Zone of carbonate accumulation ,??-.-;-!.-:-:::
penetration i:~ii{i{~{!:7~?}~i.~::~:{.!;i..~i~:ii%?!?~:{~i~..}i{::~..{:7{..2.}.}:!:~.~.:{!i~{~:.~.~:{..i:{:~.}~..~{i?i.}.~{.
Upward flow of nutrients throughI ~ , / t

r~176 ~--/ ~7i ~Q-~'~- ?'" ~- Water-table
Figure 12.26. Representation of a model of calcrete formation by different inputs of carbonate from
above per descensum model (in Goudie, 1973b, Figure 44).
The presence of calcified filaments, fibrous calcites and fungus microfossils testify to a
pedogenic caliche origin (Vaniman et al., 1994). A problem with this hypothesis is that
the caliche thickness is generally much greater than that of the soil; this difficulty can be
explained by a slow rate of soil erosion during caliche production.
When carbonate solutions rise through capillary action and precipitate at the surface by
evaporation, nonpedogenic caliches form, corresponding to the per ascensum model of
Transpiration
"": " : 1"" "" " " "_L...J~~ v;tr : t:

E. poat!onofH20 :. :
.'......'......'.:0ard'pan.."
Leached zone
Zone of capilliary rise

and lime deposition
9~
,0e.~_.o__'.
~ --% .o
]
.~z.
-
-,__'~_.2~-'2ai..~_.. oi'o'_. 9.o. ~..~_~_o'__.,___CaCO3 - rich
groundwater level
CaCO3
Figure 12.27. A model of calcrete formation by capillarity rise from groundwater per ascensum model
(after Goudie, 1983b, Figure 4.5, reproduced with permission).
Goudie (Figure 12.27). They can form close to the phreatic level by evaporation in the
capillary zone and also beneath the phreatic level as a consequence of a fall in CO2 content
(Land, 1970). This type of caliche can arise through carbonate-rich flood waters, which
can reach thicknesses of more than 10 m. They may develop rapidly, because they occur in
very young alluvium (Machette, 1985), although this origin constitutes a special case.
5.3. Siliceous crust
Also known as silcrete, this is a grey, tan or green coloured, hard and fragile rock,
comprised of quartz grains cemented by a microcrystalline and amorphous matrix.
Common thicknesses vary between 1 and 3 m. It forms in arid environments, although it
can form in more humid zones. It is a product of replacement of surface materials
(regolith, sediment, soil) by silica, in which silification takes place at a low temperature
unrelated to volcanism, plutonism, or metamorphism (Summerfield, 1983). Siliceous
crusts abound in Australia and South Africa, although they have been observed elsewhere.
They can occupy the tops of hills due to their resistance to erosion. Precious opals can
occur in some silcretes.
In outcrop silcretes commonly exhibit a massive structure accompanied by columnar
joints. Plate, botroidal and pillowy structures can also be recognised (Wopfner, 1978).
The micromorphology reflects the characteristics of the host rock (grain size and shape,
mineralogy and fabric) and the silicification processes. This includes micro- and
cryptocrystalline quartz with boundaries indented in the mineral grains, which confirms
the idea of replacement (Milnes et al., 1991). Chemical analyses indicate that they are
usually more than 95% SiO2 by weight and contain minor amounts of aluminum, iron and
titanium. The aluminum is commonly associated with the clay minerals, whereas the
titanium appears to be related to the bedrock.
In silcrete formation it is necessary to consider the silica source and its dissolution,
transport and precipitation. The primary silica source can be derived from in situ
dissolution, silicate weathering and aeolian dust. In fluvial and lacustrine environments,
diatoms may play an important role. Dissolution and transport are produced by water at
alkaline pH, moving vertically and laterally (Summerfield, 1983). The lateral movement
model is used in fluvial and lacustrine environments. Vertical transfer has similar aspects
to the per descensum and ascensum models used by Goudie (1973b). Finally, precipitation
is produced by evaporation or a fall in pH.
Where siliceous crusts occur in great thicknesses, their age, obtained from diverse
places around the world, go back to Mesozoic times, and so are considered relict accumu-
lations, possibly generated in different climatic regimes (Cooke et al., 1993).
5.4. Gypsum crusts
These are found in numerous arid zones, but over much reduced areas (Coque, 1962).
They develop in deserts with annual precipitations of less than 250 mm (Watson, 1983).
They are defined as accumulations of gypsum found in the uppermost 10 m, having a
thickness of 0.1 to 5 m, a gypsum content of greater than 15% by weight and a minimum of
5% gypsum in the underlying substrate (Watson, 1985, 1989a).
Three types can be differentiated on the basis of their structure and stratigraphic
position (Watson, 1979). The first type is horizontally stratified crust containing 50 to
80% gypsum. The second is subsurface crust, formed of either large, lenticular crystals
1 to 0.5 m in diameter, commonly known as desert rose, or mesocrystalline material with
0.05 to 1 mm diameter crystals. They can reach up to 5 m in thickness and have a gypsum
content of 50 to 70%. The third type is the surface crust, which basically consists of
alabaster gypsum with crystals of less than 0.05 mm, occurring in columnar form on a
scale of 1 to 2 m, or as loose, dusty deposits. They contain approximately 90% gypsum.
Gypsum crusts can be differentiated on the basis of their generation via capillary rise
and evaporation of sulphate-rich waters, or per descensum through gypsum-rich aeolian
dust (Watson, 1985; Heine and Walter, 1996). These authors state that Tunisian dust
originates primarily from the sebkhas, and that dust in the Namib Desert originates from
coastal clouds. The nonpedogenic gypsum scale model is based on the precipitation of
gypsum in subterranean waters and lacustrine environments.
Chapter 13
The action of water in arid regions
1. Introduction
In the proceeding chapters we have discussed the importance of water in weathering

processes, its role in the generation of patterned ground, varnishes, and crusts. In these
cases the activity of water occurs in a thin film and its movement is clearly local. We now
turn our attention to the action of water activity connected to surface flow, organized in
fluvial systems. These are characterized (Schumm, 1977) by a headwater source area,
where water and sediments are derived, a transfer zone and, finally, a lower area or
sedimentation zone. This fluvial system has a group of independent and dependent
variables (Schumm and Lichty, 1965) that have different influences. Time, initial relief,
geology (lithological and structural) and climate are the independent variables most
significant in influencing the hydrology and erosive activity. The vegetation depends on
climate and lithology. The rest of the dependent variables are variation in relief, hydrology
(run-off, discharge, sediment production) and the morphology of the drainage network
and slopes. In arid regions the first three independent variables vary substantially between
different deserts, unlike climate that depends on precipitation and which is, in turn, the
cause of run-off and erosion that are conditioned by vegetation cover.
2. Precipitation, vegetation, and evapotranspiration
In the majority of the arid regions of the world precipitation data are rare as there are
few meteorological stations and generally, where they do exist, they are very isolated. As
a consequence, extrapolations under these conditions are likely to be incorrect. Precipita-
tion is a consequence of different atmospheric processes: storm fronts, tropical storms,
orographic effects, and convective cells (Graf, 1988). As most deserts are located in anti-
cyclonic belts within the global atmospheric circulation system, precipitation is scarce
and limited to the sporadic penetration of storm fronts that are rare and normally produce
little rainfall. Tropical storms (cyclones, hurricanes, and typhoons) are rarer than storm
fronts, but occasionally they can penetrate the interior of arid regions, such as in the
Sonora Desert, between Mexico and Arizona. Orographic effects occur due to a reduction
in pressure and a cooling of the rising air, conditions that favour condensation and
precipitation. When the air descends on the other side of a mountain range, it warms and,
because it has lost the majority of its humidity while rising, it creates a rain-shadow zone.
The final type of process is the convective storm that is a few kilometres in diameter and
results from warming of the air. When they affect small basins they can produce important
floods (such as that of the 28th July, 1990, which developed a delta at the mouth of the
lower Huerva River, a tributary of the Ebro in Spain.
Precipitation in arid regions varies significantly (Figure 10.4) and, as a result, greatly
influences changes in run-off and sediment production. For example, in the middle of the
Ebro Depression, at an experimental station for studying run-off and erosion at Lanaja
in Huesca Province, Spain (Figure 13.1), a 4-year record shows that the most important
precipitation corresponds to convective storms that occur primarily during spring and
autumn, with some in summer. The rest of the rainfall is from convective cells and is of
low magnitude (Guti~rrez et al., 1995). For this reason, the highest values of run-off and
erosion take place in spring and autumn. Intensities vary considerably within convective
cell precipitation. At times, the recorded rainfall from one storm exceeds the total mean
annual precipitation. Furthermore, experimental stations with a large quantity of rain
gauges show sharp spatial variations and steep gradients (Figure 13.2).
In the literature, a number of cases are cited where precipitation is related to the
advection of humid air masses supplying continuous rainfall to a semi-arid region (Inbar,
1972; in Starkel, 1976). The Mediterranean winter rainfall that reaches the mountains of
Lebanon, Israel, and Syria is an example. Rainfall during a 4-day period (20th to 23rd
January, 1969) varied between 120 and 360 mm, equivalent to a third of the mean annual
precipitation. Despite the low mean intensities of 2 to 3 mm/h, occasionally exceeding
8 mm/h, this rainfall corresponded to a return period of 100 years. During this event,
0.04 mm/yr of downcutting, six times the normal rate, was produced in the upper
Meshushim basin. This shows that storm precipitation can be of considerable importance
in semi-arid regions.
60
55
50
45
4O
E
E 35,
.=_ 30
rr 25
20
15,
If.
10,
Summe(
It[ l Jll Jl
[Autumn ! Winter I Spring lSummer! Autumn i Winter I Sprin~ i SummerJAutumn I Winter ! Spring
1991 ' 1992 ' I 1993 l 1994
Figure 13.1. Precipitation record at the Lanaja Experimental Station in Huesca Province, Spain, for the
period July 1991 to May 1994 (Guti~rrez et al., 1985).
1.0 1.0 1.0 0.8

o.~ 0.4 o.6 o . 8 ~ ~ .L
9 . 9 , . . , 9 , ~ ~ ~ , ~ ~ ~ ~
9 9 9 9 .. "~..-- ~0.6
%,%00 , 9 9 9 9 j,,,.,-N~ O.2

Q 9
02 -02 o-
1.0 0.6 0:4 0i2

0.6 ..,, . .. ,., 1.2 , [P.81 ~-v~L.,,&-,.,~ 9 ~ z,-.,,..,~M,~
0.2 0.4,1:0.,.,-i", 1.8 1,8
0.4 0 ~ ~ _ ~ ; ; ~
~~o.4 ~ ~ _ , , ~ 7/.~j (7 o:8

o.6 ~ "J,C'{'.//, #.2'f~Je' ~.o
~o~~.~ ~ - . ~ ~ o.8
0.6 1.84
Figure 13.2. The distribution of rainfall from thunderstorm cells over Walnut Gulch, Arizona: Above,
a single cell event. Dots represent rain gauges. (after Renard (1970), in Graf (1988), Figure 3.4).
Vegetation in deserts is sparse or even nonexistent. In environments below 100 mm of

precipitation the vegetation may only be developed in wadis (Figure 13.3) or closed
depressions. If precipitation varies between 100 and 500 mm, the natural vegetation is
arboreal. Generally, this has been degraded by human activities to a more open landscape
of shrubs and grasses that may also have been affected by overgrazing (Francis, 1994).
The scarcity of water and the elevated temperatures mean that plants must adapt to these
inhospitable environments by means of physiological or morphological adaptations and
procedures that can cope with the stress associated with an increase in aridity (Gfinster
et al., 1993).
We have already seen the role that flora plays in the biological alteration of rocks and
the influence that sparse organic material has on the processes of hydrological erosion.
There exists, in general, a good correlation between the percentage of vegetation cover and
the mean annual precipitation. Interception is a significant factor in hydrological erosion.
Nfivar and Bryan (1990), in a study on the influence of vegetation in a semi-arid region of
northeast Mexico, found that after precipitation of 230 mm, divided between 17 storms,
Figure 13.3. Shrub vegetation in the bed of the Nahal Zin in the south of the Negev Desert, Israel.
interception losses were 27%. On the other hand, infiltration is highly variable and
considerably greater in vegetated areas than in bare soil. In deserts there are some
environments that are more sensitive to vegetation changes and this is manifested in
sediment production. It increases on a slope with decreasing vegetation and reaches a
maximum on bare soils (Figure 13.4a). Erosion is barely affected, however, where the
vegetation cover is low (Figure 13.4b,c) (Schumm, 1977), although for De Ploey et al.
(1976) and Morgan et al. (1986) sediment production increased for low values of
vegetation cover (Figure 13.4d). Rogers and Schumm (1991) showed experimentally that,
on 10% slopes, erosion increases rapidly with a decrease in vegetation cover from 43 to
15% but below this last value erosion decreases. Consequently, with cover below
15%, vegetation does not curb sediment production.
Vegetation affects aeolian erosion by reducing the wind velocity and acting as an
obstacle, which generates small dunes, known as nebkhas, in the lee of the plant. One
observes a decrease in dune movement with an increase in vegetation (Ash and Wasson,
1983) that acts to stabilize them. The same function is carried out by biocrusts that stop
dune movement, as observed in the longitudinal dunes of Nizzana in the Negev Desert of
Israel (Veste, 1995) (Figure 11.13 and Figure 13.6).
The water that exists on the Earth's surface undergoes evaporation, which acts to dry
the surface materials. In desert regions drying, brought about by prolonged drought,
reaches a depth of various meters. Another type of water loss is due to water uptake
through plant roots. Transpiration is produced in the stomatal cells, where there is an
interchange of water vapor between the atmosphere and the leaves. Transpiration is
favoured by low humidity, high temperatures, and wind. In deserts, xerophyte plants adapt
r
.o_
0
III
(a) Vegetation ~ (b) Vegetation
t-
.o_
o~
C)
UJ
(c) Vegetation ~ (d) Vegetation
Figure 13.4. Hypothetical relations of erosion and sediment yield to vegetative cover. All show an
exponential increase of erosion and sediment yield at higher vegetation density but b, c, and d indicate a
decreased influence of vegetation at low values of vegetative cover (after Rogers and Schumm (1991)).
to this water shortage by reducing transpiration through the formation of hidden stomates
and by reducing leaf size, or losing them altogether as in the cactus. In addition to
xerophytic plants, in the wadis of arid regions, phreatic plants develop. The roots of these
plants penetrate deeply to reach the phreatic level, assuring the water source necessary
for their development. Evapotranspiration may reach values of 85 to 90% of total
precipitation (Renard, 1969), although there are differences between distinct vegetation
species. The majority of the evapotranspiration occurs soon after rainfall. The rates of
transpiration fluctuate between 2.3 and 10 mm/day and 0.1 ram/day during the driest
periods of the year (Thornes, 1994).
3. Characteristics of the fluvial network in deserts
Precipitation in arid regions is scarce and variable. This variability reflects itself in the
fluvial systems at the global level. One can distinguish between endogenous drainage
where precipitation occurs within the desert area and exogenous drainage that originates
outside the arid zone, but flows through it, such as the rivers Nile, Indus, Colorado in the
USA, and the Ebro in Spain. In deserts, an important part of the drainage is endorheic,
where the water does not reach the sea (Figure 13.7) due to infiltration and above all the
prevailing evaporation. Furthermore, there are large closed depressions that receive
FORMER COURSES OF
R. SARASWATHI IN
WESTERN THAR 0 AN U PGARH ~/'..=~=-- --=,~...
. .~. . at... ~ - ~ -- f ~ #
.. ~," .<,/ SAKHIo,,~z ~ / ,
.:- / I" '0'

.o" /
"t ~ 1t /~ " ~
28~
J
SUKKUR " TANCQ.T"
-~-_]HOTARU
9 kx BAP~
AHGA~.~~ POKARAN~ SCALE
2,"~ / 9 I 0 40 80 120 160 200
JAISALMER l i 1 I n i
KILOMETRES
II iI MIYAJLAR
"o .... ' '1
I
I
/ 73O0'E
1 I
.:.~;~" -- BARMER 9
'li UMARKOT
.y..- ~~ o_~,oo
r- - ~. kms
t 9 SALMEI~~
RANN OF KACHCHH
69O0'E 71~ '
~ ] EXlSTINGSTREAMS ~ FORMERSTREAMS ~ SALTMARSH
I,- ;q ~"~E~T
~ S S U~OEOF
~T~ T ~~ E~ ~ ~ES'S~' ~ T . ~ ~ESTE~,.~ T
~~L ~ ~ FLOW DIRECTION OF FORMER STREAMS
Figure 13.5. Former courses of Saraswati River in Western Thar, India (modified from Ghose et al.
(1979)).
superficial waters, such as the endorheic basins of Lake Chad, in the southern Sahara, and
Lake Eyre, in the Central Desert of Australia. Guilcher (in Demangeot, 1981) estimated
that in the majority of Saharan regions, the largest floods do not travel more than 300 km.
Areas with areic drainage are those that show no regular circulation of water and occupy
Figure 13.6. Biocrust developed in a system of longitudinal dunes affected by rilling, Nizzana
Experimental Station, Negev Desert, Israel.
important extents within desert zones. Some deserts have exoreic drainage such as the
meridional zones of the southern Atacama Desert of Chile that have a large water source
in the Andes.
The river courses of arid regions are usually ephemeral (Figure 13.8) reflecting the
scarce supply of water they receive. These rivers (or wadis), therefore, are permanently
dry most of the year (Figure 13.9). Perennial rivers also occur, however, if they receive an
important water supply from distant sources, usually from surrounding mountain ranges
where a contribution of water may come from snowmelt (Figure 13.10). At times, the most
distal part of the drainage network may not be currently active, as indicated by greater
hydrological activity in the past pluvial periods. This occurs in the central Sahara where
complete networks of wadis do not receive water during even the largest floods (Tricart,
1969). Also, the Niger and Senegal rivers that originate in the tropical mountains of
Guinea, flow northwards through a very dry region characterized by dunefields that
can only be traversed during pluvial periods (Tricart, 1969). The use of aerial photography
J
~.~. ~.~ '.~ , "
C',I
.o ~
0
t ~
9
t "~
.,..~
~ ." .._.
I ~. ~ ~l~o
S
~.
"
,,# .~ .~
"
'~'~ t , "
9
9
Q) l.~
._~ ,_ <~
~ .-~
.-~ ~
D| o,,~
Figure 13.9. Cultivation in the channel of the Sabarmati River, Ahmadabad, northeast India, that
indicates the scarcity of superficial run-off that it receives.
Figure 13.10. River course in the piedmont zone of the Atlas Mountains that have significant snow
cover, Assif Mellah, southern Morrocco. Photo: J.L. Pefia.
and satellite images show that some desert areas had fluvial systems in the past, the courses
of which are currently buried beneath the desert plains, as is the case of the River
Saraswati in the Thar Desert of India (Ghose et al., 1979; Srivastava et al., 2001)
(Figure 13.5). Similarly, at the large endorheic Lake Chad, to the south of the Sahara, an
important part of the fluvial network has remained fossilized by aeolian sands (Grove and
Warren, 1968). Also, in the piedmont zones of some desert areas, alluvial fan systems
develop in which frequent channel shifting occurs. These alluvial fans can be affected by
aeolian erosion during periods of water shortage.
Run-off may be initiated in deserts through Hortonian overland flow (Horton, 1933),
produced when rainfall intensity is greater than the infiltration capacity of the soil or
rock, or when generated by saturation of the upper soil layers. During storm precipitation
in arid regions the rainfall exceeds the rate of infiltration and, as a consequence, the run-off
is produced by Hortonian flow.
Hydrographs that express the variation of discharge as a function of time show a very
sharp rising limb and a declining limb that is initially pronounced before smoothing off
(Figure 13.11). These sharp discharge variations over a short time period are typical of
arid regions. The curves illustrated in the figure are from the gauging station of Las
Tosquillas, a few kilometres downstream of Ugijar, in the Alpujarras of Granada Province
(Thornes, 1976). The hydrograph, corresponding to the record of the 24th to 31st August
1969, shows a practically instantaneous flood wave and an equally sudden declining limb,
typical of this area where floods are caused by intense summer convective rainfall
(Thornes, 1976).
80 2 4 - 31 August 1969
60
40
30
O
20-
10-
5"
E
3-
o.~:
0.6-
0.4-
0.2-
0.1
0
24 2s 2'6 2'7 d8 2~ d0 3'I
!
10- 1 5 - 22 October 1972 [

5-
3-
O
2-
o.~:
E 0.6-
O.4-
0.2-
0.1
0
15 16 1'7 1'8 " " i~ 2'0 2'i 2~
Figure 13.11. Selected traces of stage records from the Las Tosquillas station, Ugijar River, Alpujarras
region of Granada (after Thornes, 1976).
The floods in rivers of arid regions can be divided into four types (Graf, 1988): flash
floods, events with a single discharge peak, events with multiple peaks and seasonal
floods. The first are produced by storm rainfall that covers small areas of around 8 km
diameter and are, therefore, limited to basins of 100 km 2 or less. They are characterized
by reaching a peak from zero flow in a matter of minutes or hours and the response is
a wave that advances down the channel with a sharp, turbulent front loaded with debris.
The example described from Ugijar corresponds to this type of flood. The single peak
events have a longer duration than flash floods, from some hours to several days (Ward,
1978). They are produced by tropical storms or storm fronts that affect basins of thousands
of square kilometres in size. Those with multiple peaks are caused by precipitation from
tropical storms or fronts, like that recorded from 15 to 22 October in the Ugijar River in
southern Spain (Figure 13.11). They are produced by successive storms or shower systems
that affect a particular region.
In the river courses of arid regions a part of the flow is lost by infiltration and evapo-
ration. Floods in small ephemeral streams are lost over short distances, whereas in larger
rivers water is lost gradually, mainly through evaporation, with a clear reduction in
discharge and salt precipitation in marginal areas (Figure 13.12). This causes discharge
peaks to decrease considerably downstream. These transmission losses are illustrated
at two gauging stations located at Walnut Gulch, Tombstone, Arizona, that are 10.9 km
apart (Renard and Keppel, 1966). The loss in discharge is on the order of 57%
(Figure 13.13).
4. Run-off and sediment transport on slopes
Water from rainfall infiltrates the soil until it is saturated with water, exceeding its
infiltration capacity. The rate at which the water penetrates is determined by the rate of
infiltration that initially decreases rapidly then more slowly later. Infiltration is controlled
by a number of factors: type of rainfall (intensity and kinetic energy), texture and structure
of the soil, vegetation cover, and the slope of the area. The impact of raindrops causes an
increase in compaction that reduces infiltration (Morin and Benyami, 1977). On surfaces
with rock fragments studied in the laboratory and in the field with rainfall simulators,
infiltration is reduced with an increase in coverage of rocks (Poessen et al., 1990;
Abrahams and Parsons, 1991a). Infiltration also diminishes away from shrubs, around
which there is a substantial increase in the content of organic material and vegetative
detritus (Lyford and Qashu, 1969). The development of biocrusts (Talbot and Williams,
1978) and lichens (Alexander and Calvo, 1990) also reduces infiltration.
Figure 13.12. Evaporation loss in a wadi that has resulted in salt precipitation at the channel margin.
At the edge of the Sabkha Mutti, United Arab Emirates.
1500 -
I September 9, 1964
Flume 6 (upstream)
vol = 74.8 acre-feet
___ Flume 1 (downstream)

vol = 32.5 acre-feet
1000 -
!
i
o I
F
o~
L I
tl:i
c-
o I
O~ ! k
CI I\
500 -
I
I t
I I
I
\ I "\
0 ~ "F-v------a__ "~. . . . 4----~
4pro 8pm 12pm
Time (hours)
Figure 13.13. Transmissionlosses for a flood in Walnut Gulch, Arizona, represented by the hydrographs
for two flumes 10.9 km apart in the ephemeral-stream channel (after Renard and Keppel (1966), in Cooke
et al. (1993), Figure 11.3).
Initial run-off is produced when rainfall intensity exceeds infiltration capacity. Generally
in deserts run-off depends on rainfall intensity, as infiltration rate is a more important
factor than the state of soil saturation. In arid regions storm events predominate that
rapidly exceed infiltration rate. Furthermore, as the vegetation is sparse, interception is low.
On desert slopes practically all the run-off is produced by superficial flow, although in
some areas subsurface flow occurs, but only sporadically. Foster and Meyer (1975)
differentiated between flow and erosion on slopes produced in fills and inter-fill areas.
In both, the hydraulics and mechanisms of erosion are different. In the inter-rill areas
(Figure 13.14) the water is mobilized, in general, under sheetflow that diverges and
converges around objects in its path (for example, rock fragments and vegetation). As a
consequence, the velocity and depth of flow vary greatly over small distances, modifying
the type of flow (sheet, turbulent, transitional) (Abrahams et al., 1994). In arid regions
rills are one of the most characteristic features of slopes formed of soft materials
(Figure 13.15). Rills have longitudinal and transverse profiles that affect the velocity and
type of flow, which is commonly turbulent (Gilley et al., 1990).
In relation to water erosion there are little data regarding slopes in arid regions.
Saunders and Young (1983) in their review of rates of slope erosion indicate downcutting
exceeding 1 mm/yr in semi-arid climates, whereas in arid regions the values are on the
order of 0.01 mm/yr. In semi-arid zones, in areas composed of easily erodable material,
the rates of denudation are arguably the highest in the world.
Figure 13.14. Large inter-rill areas separated by small gullies formed in Quaternary clay materials.
The slopes are affected by intense cracking and unequal rills development. The Bardenas Reales of
Navarra, Spain, in the Ebro Basin.
Figure 13.15. Intense rill formation in Miocene clays similar to badlands. The Bardenas Reales, Navarra,
Spain, in the Ebro Basin.
Various factors can be recognized that control the intensity of erosion on desert slopes.
The effect of slope was analyzed by Abrahams and Parsons (1991 b) in a simulated rainfall
study at three experimental stations with distinct lithology and vegetation. At slope angles
below 12 ~ the run-off increased very slowly and sediment production increased with slope
due to the gravitational component (Figure 13.16). If the slopes were greater than 12 ~ run-
off decreased rapidly, by a value that increased with slope. This reduction compensated for
300 --
T"
(--
: ",
9 9
--El
~: 200 -- . 9 -- E2
~ , 9 ---. E3
E "
"1~
~" 100 - / e 9
'...
~ o~ ~
1~~ ~
o ! ! I i
(a) 0 0.2 0.4 0.6
80 -
tolq,
,j~o 9
m oo 99
6O
v n
40
8
S ~
#
20
I I 1 I"' I I
0 0.2 0.4 0.6
(b) Gradient, sin e
Figure 13.16. Curves fitted to graphs of (a) sediment yield and (b) run-off coefficient against gradient for
three sets of experiments denoted by E~, E2, and E3 at Walnut Gulch, Arizona. Experiments E~ and E2 were
conducted on plots underlain by Quaternary alluvium, with the ground vegetation being clipped for E] but
not for E2. Experiment E3 was performed on plots underlain by the Bisbee Formation (after Abrahams and
Parson ( 1991b )).
the increase due to the gravitational component, therefore, sediment production dropped.
The relationship between rock fragment cover and sediment production has been studied
in plots by Simanton et al. (1984). For similar slopes, negative correlations were obtained
due to a number of factors. In laboratory experiments, however, Poessen and Lavee (1991)
found positive correlations when the clasts were greater than 5 cm in diameter. On the
other hand, erosion increases after dry periods and decreases during humid ones due to
the effect of vegetation that protects the surface from raindrop impact and reduces
surface flow.
One of the most important processes that affects inter-fill areas is that of rainsplash
that provides material that can then be mobilized by sheetflow. When a raindrop
reaches the soil (Figure 13.17), the impact generates a small crater by compression and
the jolted particles are scattered over the surface by shearing. The greater the kinetic
energy of the raindrop, the more important is the effect of rainsplash. If there is a thin
film of water on the surface, particle dispersion is increased, reaching a maximum when
the thickness of the water lamina is the same as the diameter of the water drop (Palmer,
1963). On horizontal surfaces the distance that the material is ejected decreases
exponentially from the point of impact (Savat and Poesen, 1981; Torri et al., 1987).
Rainsplash is most effective when there is an increase in slope (Ellison, 1944) and wind
speed. This modifies the trajectory of the majority of the mobilized particles. In
addition, vegetation cover, the size and percentage of surface particles, organic content,
soil chemistry and aggregate stability, amongst other factors, all exert an influence on
this process (Scoging, 1989). The rainsplash is an important factor that affects the
supply of particles to sheetflow and their subsequent transport in rill and inter-rill areas
(Bryan, 1987).
5. Water erosion on slopes
Erosion by surface flow may occur by sheetflow or filling. According to Leopold et al.
(1966), 98% of sediment production during a period of 10 years in a semi-arid area of
New Mexico, came from slopes without rills. In environments of greater precipitation
in Kenya, Dunne (1980) found no filling over long slopes, confirming field experiments
using rainfall simulation (Dunne and Dietrich, 1980). Similarly, Emmett (1978)
indicated that some slopes in Wyoming not affected by rilling suffer uniform
degradation through sheetflow erosion. Whereas rill erosion has not received much
attention, inter-rill erosion has been intensively studied by a number of authors who
have focused their investigations on the interaction of splash and sheetflow erosion, both
in the field and in the laboratory. These studies have taken into account slope and the
physical, chemical and mineralogical properties of soils. They indicate that the processes
and rates of inter-fill erosion are extremely complicated and variable, depending on
extrinsic factors, such as rainfall intensity, raindrop size, and the presence or absence of
wind, and intrinsic factors such as soil texture, the nature of soil aggregates, surface
roughness, susceptibility to cracking, and the presence and density of organic detritus
(Bryan, 1987).
Rill erosion is predominantly produced by the mobilization of soil particles by
concentrated flow and affects a small part of the terrestrial surface. Rills may be developed
Figure 13.17. Sketchof the effect of a rainsplash (A1-Durrah and Bradford, 1982).
when the run-off is concentrated in small areas of irregular topography. Rill erosion,
however, does not occur until the erosive capacity of the flow exceeds the resistance to
mobilization of the soil particles, meaning that the run-off may flow a considerable
distance over the slope before rills develop (Meyer, 1986). Once the rill is formed, flow
increases rapidly and erosion is increased down the slope. The mechanisms that form rills
are very complex, as indicated by the fact they may also proceed backwards from the rill
head (Bryan and Poesen, 1989).
Dunne and Aubry (1986) believed that the dynamics of rill systems can be explained
by the balance between the intensity of filling and inter-rill processes or the processes of
sheetflow and rainsplash. When the latter do not occur, sheetflow becomes unstable, but
if it does and sets in motion the impacted particles of the inter-rill areas, the particles tend
to fill and eliminate the fills. This indicates various spatial and temporal variations that
reflect the changing effectiveness of these processes. At the experimental station of
Lanaja in Huesca Province, Spain, investigations of the temporal variations in rill cross
sections were carried out using microtopographical survey. Rills were infilled with
particles during periods of low precipitation as they were incapable of generating run-off
that could transport the loose material in the bottom of the fill (Figure 13.18). Conversely,
during periods dominated by convective rainfall of high magnitude and intensity, the run-
off generated possessed a greater erosive power. In a 2-year period, some fills, formed in
Miocene shales, entrenched 5 cm into the original bed (Sirvent et al., 1997). This indicates
the importance of temporal variations and the dependence of fills on inter-fill processes,
such as the variation in intensity of rill flow.
In addition, other distinct processes may be recognized in the genesis of rills. In shales,
subsurface micro-pipes may collapse and develop into fills (Figure 13.19) (Bryan et al.,
1978; Guti~rrez et al., 1988). In other situations rills form due to the intersection of
microrill cracks (Figure 13.20) (Haigh, 1978; Guti6rrez et al., 1988). In clay lithologies
with alternating clay layers with different expansion capacities, rills may develop in strata
with elevated contents of expandable clay. Expansion is accompanied by the development
of cracks separated by micro-humps, known as popcorn structure, which posses a high
Figure 13.18. Rill cross section taken several times by the profilometer for Lanaja 1 plot
(Sirvent et al., 1997).
Figure 13.19. Collapsed subsurface pipes and "bridges" in Holocene shales, that have formed rills or
small gullies during their evolution, Lupifi6n, Huesca Province, Spain, Ebro Basin.
infiltration capacity (Figure 13.21). In the Dinosaur Badlands, in a semi-arid area of

Alberta, Canada, fills form in micro-pediments as a response to the hydraulic conditions of
superficial flow (Hodges, 1982). In the same area Bowyer-Bower and Bryan (1986)
observed the development of fills by dissolution of resistant siltstones.
Piping or tunnelling (Figure 13.19) is common in semi-arid environments and can
be defined as the natural development of a subsurface drainage in clastic rocks that
consists of the mobilization of solid particles that are transported in suspension by water
(Parker, 1963). Piping occurs preferentially in areas whose climates show strong seasonal
contrasts, in terms of rainfall variability, such as in Mediterranean climates (Jones, 1981;
Bryan and Yair, 1982). Piping results from a complex combination of a number of factors
(Stocking, 1976; Bryan and Yair, 1982; Guti6rrez et al., 1997).
Cracking by desiccation is one of the most important factors in the generation of piping
(Parker et al., 1990) and reaches a high intensity with the presence of expandable clays
(Parker and Jenne, 1967). In summer, the surface materials shrink, storm rainfall fills
the cracks, and the clays can be dispersed (Marshall and Workman, 1977) and become
mobilized with a sufficient hydraulic gradient (Fletcher and Carrol, 1948; Jones, 1981).
Figure 13.21. Formation of rills in shales. At the lower strata rills disappear due to popcorn structures,
Custa del Viento, Colola, San Juan Province, Argentina. Andes Precordillera.
In this manner, the cracks enlarge and form pipes. These can occur in flat areas incised
by the fluvial network and on slopes. Where there is a large contrast in relief, large
galleries and pseudo-dolines (or sinkholes) can be formed by collapse, as observed in
the area of Lupifi6n, Huesca Province, Spain (Figure 13.22 and Figure 13.23), resembling
a karstic landscape (Guti6rrez et al., 1988). The continuous collapse of gallery ceilings
produced by piping may give rise to the formation of gullies, as testified by the presence of
natural arcs in detritic materials (Figure 13.24). These collapses also increase the rate of
headcutting in gullies (Crouch, 1983). Piping may also create instability on slopes and
contribute to the generation of landslides (Pierson, 1983).
We have already seen how the presence of expansive clays is an important factor in
the development of piping, but its presence is not essential for producing expansion
(Jones, 1981). In environments rich in sodium, illites and other clay minerals also expand
(Imeson et al., 1982), producing an expansion of 10 to 12% (Benito et al., 1993; Guti6rrez
et al., 1995). The principal property that controls susceptibility to piping is the content of
Na + in the exchange complex water with respect to the sum of the Ca 2+ and Mg 2+ cations
(Sherard and Decker, 1977). These values are obtained by calculating the quantities of
these cations in a water extract from saturated paste. Once known one can calculate the
exchangeable sodium percentage (ESP) or the sodium absorption relationship (SAR).
Clays with an ESP greater than 15 (Mclntyre, 1979), or a SAR above 5 (Aitchison
and Wood, 1965), contain high pH values (between 8 and 10) and are very susceptible
to piping. Furthermore, soils with elevated sodium concentrations facilitate clay
dispersion (Heede, 1971). Sherard's diagram (Sherard et al., 1972) establishes the
Figure 13.22. Grouping of pseudo-dolines formed by piping on thick slope deposits. Lupifi6n, Huesca
Province, Spain, Ebro Basin.
6 5 ~ ~ 6 0
65
60
40
15
b 10
0
i _ _
50m.
--J
PIPE COLLAPSE MAP
LEGEND
relativeheightof
--'--20----- contours in meters
Q funnel-shaped doline
(22) well-shaped doline

--~ subsurface channel
[":" .... ] coarse talus
gully
Miocene deposits
Figure 13.23. Map of pseudo-dolines developed in coarse talus (Guti~rrez et al., 1988).
different dispersive fields in relation to sodium content (Figure 13.25). The points
situated in the dispersion field of the graph are susceptible to piping (Benito et al.,
1993" Guti6rrez et al., 1997). The study of the physical, chemical, and mineralogical
properties of materials is important in civil engineering, especially in the construction
of earthen dams as some have collapsed due to piping (Jones, 1981; Parker et al.,
1990). If the ESP is elevated, it may be reduced by the addition of calcium, in the form
of gypsum or lime (Heede, 1971; Stocking, 1976) that makes the piping processes more
difficult.
Figure 13.24. Naturalbridge produced by piping processes. The gully has been developed by successive
collapses, Barranco del Tormillo or Barranco de la Clamor Vieja, E1 Tormillo, Huesca Province, Spain,
Ebro Basin.
6. Fluvial forms and processes
The principal processes that result from the action of concentrated surface run-off water
are channel erosion and the transport and sedimentation of particles. The modifications
in channel cross sections during floods are commonly spectacular (Figure 13.26), with
important downcutting and rapid accretion (Leopold et al., 1964). Again, semi-arid
environments constitute the areas of maximum erosion (Langbein and Schumm, 1958;
Fournier, 1960; Corbel, 1964), because these areas present sufficient precipitation and a
suitable percentage of vegetation cover that facilitates sediment production.
The energy in a fiver course is expended, in part, on the erosion of the channel.
The processes that act on bedrock river beds are chemical corrosion, or rock dissolution;
corrasion, or the mechanical fracturing and abrasion of the rock by transported materials,
that may generate rock bedforms; the mobilization of boulders by dragging or hydraulic
entrainment and, finally, cavitation, where the rock is broken by bubbles of low pressure
water vapour in high velocity currents that hit the rock and explode.
The channels may lengthen by erosion progressing upstream, with headcut erosion of
0.46 to 0.79 rn/yr having been measured (Leopold et al., 1964). They may also be widened
by erosion at the channel edge, commonly by bank undercutting, and by the detachment
or mass movement of saturated material (Figure 13.27). The channels, in addition to
lengthening and widening, also deepen their beds, at times producing deep gorges
(Figure 13.28). Some desert rivers display breaks in their longitudinal profile (knickpoints)
(Figure 13.29), where water drops in cascades, giving rise to small closed depressions and
100
t-- R9
(.9 R8
< $ R7
rr"
F--
C5 9 9 C4 BR2
X
I.U
Z
80 . . . . . . . . . . . . . . . .
R3 _~ R6
C3
BR1'~ 9
R5
o
-
~,, 9o BC3 BR4
DISPERSIVE C2 9 9 BR3
........... BR5 9 ,,, BC2
60 BC1
~Z +
, TRANSITION ~ o BA1
40 L. . . . . . . . . . . . . . . . . . . . . . .
a
o NONDISPERSIVE
I---
Z AC1
u.I
20 ~ .........
0
rr
i.u
I1. R2~iA
91 1 10 100 1000
TOTAL DISSOLVED SALTS IN SATURATION EXTRACT

meq/I (Ca+Mg+Na+K)
9 La Charca plot o Bardenas-1 plot

9 El Barranco plot 9 Bardenas-2 plot
Figure 13.25. Ion composition compared to Sherard's criterion (Sherard et al., 1972, 1976). Note that
majority of samples are in the dispersive field with high sodium percentage relative to the total dissolved
salts in the saturation extract (Guti~rrez et al., 1997).
undermining at the base that promotes collapse of the walls and the upstream migration of
the knickpoint (Bull, 1997).
As semi-arid areas have a high sediment production, fiver courses, where they are
active, transport a large sediment load that may reach values of 68% in weight, as shown
by Bondurant (1951) for the Puerco River, New Mexico. Sediment transport in fluvial
channels is carried out by suspension, saltation and traction processes. The latter two are
termed bed load. Perennial rivers transport elevated loads in suspension, whereas bed load
predominates in ephemeral channels, which in arid regions, such as the Nahal Yael in
southern Israel, reaches up to 87% of the total load.
One of the surprising characteristics of ephemeral channels is that during a flood a
certain amount of the bed is eroded and then is filled with approximately the same amount
of sediment during the waning of the flood. This occurred in the Arroyo de los Frijoles, in
New Mexico (Leopold et al., 1966) (Figure 13.30). The erosion and deposition was
produced in both sands and gravels and suggests a state of dynamic equilibrium (Chorley
and Beckinsale, 1980). The erosion results in an export of detritial material from the bed
that is transported by the run-off. It has been demonstrated that the distribution of the size
of the suspended load varies with the velocity of the water or with discharge in desert
rivers, contrary to that of perennial rivers (Frostick et al., 1983) (Figure 13.31). The study
of bedload is a complex problem. If sediment traps in the fiver bed are used (Figure 13.32)
15-- June 3, 1956-62,700 cfs ~' June 3, 1956-62,700 cfs ~__
I ~. . June 19, 1956-34,900 cfs l./

May 21' 1956-27'600 cfs V I ' I
10 -- Apr. 19, 1956-12,700 cfs _/ July 7, 1956-13,200 cfs _
/ / 2
~ Feb. 14,1956-4,760 cfs
_z 5--
~o
o I
-5
Low to High Flow High to Low Flow

-lO I I I -I ! I I I
0 100 200 300 400 0 100 200 300 400
(a) DISTANCE, IN FEET (b) DISTANCE, IN FEET
Figure 13.26. Scour and subsequent fill during flood passage, Colorado River at Lees Ferry, Arizona,
water year 1965. (a) Low to high flow. (b) High to low flow (after Leopold et al. (1964)).
Figure 13.27. Collapse of large limestone blocks by basal undercutting, Nahal Zin Canyon, Negev
Desert. Israel.
Figure 13.28. Todra Gorge, Central High Atlas, Morrocco. Photo: F. Guti6rrez.
they introduce disturbances into the transport processes. The use of painted particles
(Leopold et al., 1964) or magnetized pebbles that enable the clasts to be found with a metal
detector after the flow, allows the movement of these marked clasts to be studied. Leopold
et al. (1966) found movements of clasts up to 3 km downstream during a flood in a
predominantly sand-bed channel. Schick et al. (1987) used marked clasts and a detector
in a gravel channel to locate where they were buried. Over various floods the importance
of excavation and deposition in the transport of buried particles towards the surface and
vice versa (Figure 13.33) was apparent.
A typical model for ephemeral rivers was described by Bull (1991, 1997) for arroyos
of the southeastern USA. This author found a sequence that repeated itself along the
arroyo of alternating reaches dominated by processes of aggradation and degradation.
The sequence described is the following: a headcut reach (Figure 13.34) that concentrates
sheetflow; a channel with vertical walls then directs the flow towards the apex of a channel
fan; braided channels that end in an area of divergent sheetflow, and the sequence is closed
by convergent sheetflow that again drains towards a headcut reach (Figure 13.35).
Figure 13.29. Break in the longitudinal profile (knickpoint) caused by differential erosion of a thick bank
of Cretaceous limestones that are more resistant to erosion. One may also observe the basal undermining
caused by the waterfall, Nahel Zin Canyon, Negev Desert, Israel.
+1.0
... . . . . ~Fill .... Net c h a n g e " j
=" +0.5
0
..~
0
' I
(D
-0.5 "...,." '-..,." ~ /~
-o -1.0 - .. Flow of July 5, 1962 ,. _~
~ +1.0 / ...... ;' "'..,.-Fill , --7
,",, ,;" ""'.,, ..-,.
" ,, . . . . . . . . . . . . . . . . .. .."""" ',. ,:'' N et . .. . .. . . .. . . . -.. ....~,., ,,';",.': ',,..
=-
9 +0.5 ; ~ ......... j" ,,," ',,./ "~".,,-'" .,/ ,,#, ~ " . 9 ~ "-,... ........ , .... - .........
0
c
t-- -0.5
V "., ,." " ....... "" ,, . ',.~ : ',--'
-1.0 _~ '",, ..'". . . . . . Year 1962 "-,......... """ S c o u r ,,~ __~
= +1.0
-o
c
+0.5
0
-0.5
-1.0
0
(D 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
D i s t a n c e d o w n s t r e a m from w a t e r s h e d divide, t h o u s a n d s of feet
Figure 13.30. Location sketch map of Arroyo de los Frijoles near Santa Fe, New Mexico. Net change
of bed elevation during various periods (after Leopold et al. (1966)).
Figure 13.31. Concentration of suspended sediment by size fraction through three flash floods in the
II Kimere catchment, Kenya (after Frostick et al. (1983)).
Figure 13.32. Steel mesh trap for the collection of bed load in an ephemeral stream, Nahal Yael
Experimental Station, Eilat, Israel.
FLOOD FLOOD FLOOD

EVENT EVENT EVENT
1 2 3
9 9 9
'el
jjjj+:j 35
ill0
14
l li ,0 ~176 W j,,,lla
P%/// ~///////// ////////~/~ Y/I/I/Ill/
'~l134,,,i ~ ' ~ I I 391, 1,,, t l]

16
65
6 6 6
9 m3 s-1 33 m3 s-1 18 m3 s-1
Figure 13.33. Vertical exchange of coarse particles within the scour layer as a result of flood events
in Nahal Hebron. Numbers are percentage of the original 282 particles, whose weight ranged between 199
and 3200 g, and which were placed on the channel bed. In this analysis, burial was defined as a state in
which at least 50% of the particle was covered (Shick et al., 1987).
The sedimentation of the load transported by fluvial courses in arid regions not only
occurs, as in other morphoclimatic zones, in straight, meandering and braided channels,
but also by unconfined flows in glacis, pediments and alluvial fans that will be analyzed
in another chapter.
7. Valley fills and downcutting channels
One of the most characteristic features of semi-arid zones is the existence of alluvial
fills with gullies entrenched within them. This supposes an alternation between stages of
accumulation followed by downcutting, interpreted in many cases as a consequence of
climate change (Leopold, 1994). When these erosion/accumulation stages are repeated
terrace systems are developed (Figure 13.36). Incision takes place where the stream power
of the concentrated water flow exceeds the resistance of the materials over which the fiver
Figure 13.34. The headcut reach of an ephemeral stream eroded in a valley fill over a gypsum bedrock.
Barranco de Villafranca, Villafranca de Ebro, Zaragoza Province, Ebro Basin, Spain.
flows (Schumm et al., 1984). This downcutting may form continuous or discontinuous
gullies (Figure 13.37). Discontinuous gullies occur in relation to a slope threshold at which
the gully does not have sufficient transport capacity to continue forming (Schumm and
Hadley, 1957). In the southwestern USA, gullies with vertical walls are termed arroyos
(or ramblas in Spain). These morphological features have been the subject of numerous
studies in this region (Cooke and Reeves, 1976), with the aim of determining the causes of
~L==:E?=~, ,, --~ Smallfans and ~ II "lo"e -"

~ E E ~ , sheetflo'wreach Braided ~,, .va ey ~ ~ ......
~ , ~ ~ ~ ~ A p e x o f r, " ~ r u n k stream channel

Headcut " ~
reach ~
Figure 13.35. Characteristic erosional and depositional reaches of a diagrammatic channel fan of a
discontinuous ephemeral stream (Bull, 1997).
Figure 13.37. Small discontinuous gullies on a slope between Rodeos and San Jos6 de Jachal, San Juan
Province, Argentina, Andes Precordillera.
incision. Climate changes with a tendency towards greater aridity (Antevs, 1952; Leopold,
1976, 1994) and rainfall of elevated intensity and low frequency have been identified
as important. The decrease in percentage vegetation cover in semi-arid regions, due to
overgrazing and construction work such as the building of communication routes, is one of
the fundamental causes of incision. The colonization of the western USA constitutes one
of the most significant and most studied examples (Cooke and Reeves, 1976). The
decrease in vegetation cover, combined with important storms, was the cause of some
gullies incising more than 50 m. This erosive process took place between 1850 and 1920,
especially during the period 1870 to 1890, and as a result of this new environmental
problem many of the colonists were forced to move on to more favourable areas.
Chapter 14
Slopes in arid zones
In arid regions the production of particles by weathering processes is generally smaller

than those mobilized by superficial run-off or by the wind. Evidently, vegetation cover,
tightly related to the climate, exerts a manifest influence in this balance. Therefore, it can
be said that the degree of aridity controls the balance between supply and evacuation of
particles. In the cold mountainous deserts, however, where frost action is the dominant
process, the generation of frost shattering is higher than exportation (Figure 8.13), and
moreover if aridity is strong and, as a consequence, water erosion is not as important, it is
unable to mobilize all the clasts produced.
Although the influence of climate and vegetation is evident in the evolution of the
desert slopes, the lithologic composition, along with the frequency of the existing planes of
discontinuity, are the most fundamental factors in the development of the modelling of
the slopes under these arid environments. These geologic characteristics directly affect the
activity and intensity of the weathering processes. These factors, therefore, determine
the character and make possible the differentiation of three types of slopes developed on
crystalline rocks (Figure 14.1), stratified formations, and highly erodable materials
(Mabbutt, 1977). Some authors refer to this first slope type as "massive rocks" and include
in it sandstones and conglomerates (Howard and Selby, 1994).
1. Slopes on crystalline rocks
The chemical composition of the magma exerts a direct influence on the morphology of
the expelled volcanic materials. The basic products produce extensive basaltic flows,
which alternate with pyroclastic deposits. The differential erosion of these formations
produces stepped slopes (trap), characteristic of the big plateau basalts outcroppings, of
different ages, existent in the world. Similarly, morphologies of this type occur in the rift
valleys of East Africa. The emission of magmas of acid composition, as a consequence of
their greater viscosity, generates explosive eruptions that are accompanied by domes and
needles, as in the trachytes and phonolites of the Hoggar massif, in the central region of the
Sahara.
In plutonic rocks two types of forms are recognized. Some are controlled by sheeting
and other by cross-jointing systems. Sheeting develops in massive rocks that are
fundamentally crystalline, is formed by curved joints parallel to the surface, and is denoted
also as topographical jointing. Their origin is attributed to decompression or unloading of
rocks that have being subjected to major force inside of the Earth's crust. Later where the
rocks ascend to the surface and lose load, they relax and fracture of the rock massif takes
Figure 14.1. Slopesdeveloped on Precambrian granite and on Cretaceous limestones and marls, Negev
Desert, Eilat, Israel.
place (Gilbert, 1904). Nevertheless, other origins are also adduced in relation to other
strength types (Howard and Selby, 1994). In rocks with numerous discontinuity planes
sheeting does not propagate, because the unloading forces dissipate along these
discontinuities.
Sheeting occurs in almost all the climatic types and is relatively common in crystalline
rocks of the African and Australian Deserts (Twidale, 1981, 1982a,b; Howard and Selby,
1994). Sheeting creates domal or hemispheric forms that were described for the first time
by the German geologist Bornhardt (1900). Willis (1936) named these domal inselbergs
that stand out on the plains, bornhardts, in honour to their discoverer. Many of these
bornhardts are the result of the denudation of alteration profiles generated, in past times, in
tropical humid climates (B~idel, 1957; Ollier, 1988a,b, 1991; Thomas, 1989a,b, 1994;
Twidale, 1990). The climatic change towards more arid conditions unleashes water
erosion processes, which brings the rock substrate to outcrop.
Specific types of rock domes are those in form of a mushroom (Figure 14.2) that
outcrop in the north of the Eyre Peninsula, in Southern Australia (Twidale, 1962). They are
composed of Precambrian granites controlled by curved joints and present in their walls
some tafoni and numerous weathering grooves. The concavity is interpreted as originating
by subsurface weathering, but in some places in the profile double concavities occur that
can be explained as remainders of earlier subsurface weathering levels.
Slopes developed on crystalline rocks affected by jointing systems, seen from afar seem
to show rectilinear profiles, but upon closer inspection there can be observed in general a
group of rock steps. In his study of the morphology of the igneous rocks in southwestern
FEET
Figure 14.2. Slope profiles from the western extreme of Pidappa Hill, Australia (Twidale, 1962).
Arizona, Bryan (1927) differentiated a steep slope constituted by blocks or tors and castle-
like forms (castle kopje or koppie) (Figure 14.3). At the foot was developed a slope
covered by detritus, formed by in situ weathered blocks; these circumstances can be
deduced by the lack of blocks on the lower part of the slope. Nevertheless, on some slopes
mobilized blocks do occur. If the joint spacing is appropriate, the slope can show a stepped
profile (Carson, 1971); on the other hand the morphology is more rectilinear (Figure 14.4).
In the lower parts a wash slope develops that is constituted by fine particles, which are the
result of the block weathering and which have been mobilized to the lower zones. Also it is
necessary to keep in mind that some slopes have been modelled from weathering profiles
in plutonic rocks and, therefore, their morphological and evolutionary characteristics will
be different (Oberlander, 1972).
2. Slopes on stratified rocks
The platform deserts (Mabbutt, 1977), which cover cratonic areas of the arid zones of the
northern Sahara, eastern Arabia, Thar in the Indian Desert and the Colorado Plateau of the
USA, develop slopes on near horizontal stratified formations of different ages. These
slopes occur in chains of intermediate type, as those of Israel-Lebanon, and in Tertiary
continental basins, such as those existing in arid Spain and in the United Arab Emirates.
Figure 14.4. Precambrian granite with folded pegmatites. The high degree of jointing confers a straight
tendency on the slope profile morphologies, Negev Desert, Israel.
Figure 14.5. Simple slope constituted for spectacular massive Neogene conglomerate cliffs, attached to
the southern Pyrenean border formations, Mallos of Riglos, Huesca Province, Spain.
Those slopes originating on the morphologies of cuestas or mesas show a great variety.
Schumm and Chorley (1966), in their classic study on the Colorado Plateaus,
differentiated different types of slopes that result from the combination of lithologic,
climatic and vegetation characteristics. Simple slopes are fundamentally constituted by
one rock type. It can be a clay material affected by water erosion that develops a badland
landscape. The other subtype is constituted by predominantly massive rocks, such as
sandstone and conglomerates that form important cliffs (Figure 14.5). Composite slopes
are crowned by resistant caprock below which are developed easily erodable rocks in
which can occur badlands or debris slopes that can start from the cornice (Figure 14.6).
Finally, complex slopes are formed from the alternation of hard and weak layers
(Figure 14.7). Simple and composite slopes are be analysed below, whereas complex
slopes are considered as a mixed form of the two previous types.
2.1. Simple slopes
In this section we treat the morphology and evolution of the slopes on stratified massive
rocks, basically sandstones, and in another section of this chapter badland forms will be
analysed, which result fundamentally from the activity of water erosion on easily erodable
materials.
Most of the research related to this theme has been carried out in the Colorado Plateaus
of the southwestern USA, although it has been also studied in other deserts (Sahara,
Australia, Atacama). The evolution of the simple slopes is related to the characteristics of
Figure 14.6. Slope composed of a scarp on limestones and a debris slope that almost totally covers a
mainly clayed substratum of Miocene age. The group is partially affected by rilling and gullying, Ballobar,
Figure 14.7. Complex slope corresponding to a crest front, worked in stratified material of different
resistance to the erosion, Dra Valley, south of Morocco. Photo: J.L. Pefia.
the rocks and the activity of numerous geomorphological processes, which act or have
acted jointly or separately, spatially as well as temporally.
Discontinuities in the sandstone, as with thin clay intercalations or beds with cross
stratification, constitute zones more favourable for erosion and generate manifest changes
in the profiles of the slopes. Joint systems and vertical fractures constitute weakness zones
and along them can form routes of major moisture. As a consequence, the production of
weathering particles is increased there and remains trapped in the bottom of the fracture
system (Doelling, 1985). Also, in zones of major fault density, slope retreat gives rise more
rapidly to successive indentations and salients (Nicholas and Dixon, 1986). It has been
already indicated that in massive rocks, such as those that produce these simple slopes, the
development of sheeting is common, which has been noted in the arkoses of the famous
monolith of Ayers Rock (Uluru), in the central Australian Desert (Twidale, 1978) and also
in the sandstones of the Colorado Plateau (Bradley, 1963). This produces a convexity in
the slopes and sheets of a few centimetres to a metre in thickness. The dominant
weathering processes in sandstones are salt weathering and wetting and drying, although
in areas with freezing periods, gelifraction can be significant. The composition of the
cement also is important. In sandstones with a carbonate cement, this dissolves more easily
in cold water and siliceous cements need a pH > 8 to accelerate the dissolution. The loss
of cement produces the disintegration of the rock grain by grain. In general, the activity of
the weathering processes creates convex forms by rounding of vertexes and edges (Baker,
1936) (Figure 14.8).
Figure 14.8. Roundedforms and tafoni in the Cambrian sandstone of Nubia, Petra, Jordan. Photo: J.L.
Pefia.
Figure 14.9. Change in slope form during its retreat (Schumm and Chorley, 1966).
In several areas of the Colorado Plateaus (Schumm and Chorley, 1966) modern rock
falls can be common. Nevertheless, some of the landslides that are recognized in the
deserts were produced at other times and, therefore, are relict forms. Some examples have
been studied by Grtinert and Busche (1980) in the central Sahara and Degraff (1978) in
Utah. The analysed landslides are commonly of a rotational type (slump) and were
generated in periods of major precipitation during the Pleistocene; moreover, these mass
movements can reactivate.
Other erosive process that can be recognized on the simple slopes include basal
undercutting produced by subterranean water that flows out in the contact of impermeable
layers under massive sandstones. Weathering and erosion by water develop shelters
(alcoves), which in their evolution give rise to rock detachments and slope retreat in
sandstones (Higgins, 1984; Laity and Malin, 1985; Baker et al., 1990; Higgins and
Osterkamp, 1990). These shelters or cavities are accompanied by alveoles and tafoni.
Simple slopes evolve slowly and at their beginnings can proceed from composite slopes
(Figure 14.9), which retreat parallel to themselves. In later stages the outcrop of the
underlying clays diminishes, caprocks become rounded and, finally, they convert to simple
slopes again (Schumm and Chorley, 1966). An enlargement of these ideas has been carried
out by Oberlander (1977a) in his study on the sandstone cliffs of Utah. It also starts in a
composite slope (Figure 14.10) with an easily erodable substrate, that when retreating
turns into a simple slope. The form of the slope is controlled by the internal structure of the
massive sandstones, with thin horizons of interstratificated clays. These produce changes
in the slope and the ends of these horizons appear as modifications in the profile. The cliffs
A B C D E F G H I J
ground level
Figure 14.10. Slopes developed on massive sandstones. In A there is a cliff over a fine stratified
substratum, which remains hidden in B. Afterwards the intraformational levels control the slope
morphology. The segments located above evolve by undercutting, while those located below become
rounded (Oberlander, 1977a).
produced evolve by basal undercutting, while the segments situated below the clay strata
become rounded.
2.2. Composite slopes
These types of slopes are constituted by a rocky cliff in the upper part and a group of easily
erodable layers below (Figure 14.6). The components of those slopes adjust to the ones
defined by Wood (1942) and Fair (1947, 1948) in their studies in South Africa, that where
extensively analysed and renamed by King (1957a,b). This author considered these slopes
as "normal" ones, instead of those indicated otherwise by numerous researchers in climatic
geomorphology. For King they are part of a universal model; these slopes from desert
environments are the most characteristic. The parts that he differentiated in a slope are
(Figure 14.11): (1) the free slope that is the upper part, sometimes with a convex profile, as
a consequence of weathering and creep; (2) the free slope that constitutes the outcrop of
the hardest rock and in which general erosive acting processes causes its retreat; (3) the
materials resulting from the destruction of the scarp form the debris slope, which are
basically mobilized by water erosion; and finally, (4) the piedmont, which constitutes an
important concave element that connects with the alluvial plain.
The free slope is rounded by creep, if it exists, weathering and also rainsplash action
and sheet erosion. This tendency to rounding increases with the dip of the resistant layer
that constitutes the free slope, and is emphasized if it is porous or affected by exfoliation
(Bradley, 1963). On the contrary, if retreat is caused by faults rounding disappears or
decreases.
Free slopes are part of the borders of mesas, structural platforms or cuestas. The
thickness of the layer that constitutes the free slopes is important because it controls the
cliff height and the length of the debris slope. The rock type is also fundamental, so that the
weathering processes are different that disaggregate, for example, sandstones or
ISLIP
k~ j SLIDING
_ MASS
MOVEMENT
WATER
ACTION
SLOPE
ELEMENTS
Figure 14.11. Elements of a "normal" slope, water action, and mass movements (King, 1957).
limestones. The cementation degree of the rock is also essential to its resistance. To
lithology has to be added the structure. If the free slope is formed by finely stratified or
jointed rocks, debris slopes show good development; on the other hand, blocks of a large
size will cover the slope.
It has been indicated that debris slopes come from the breaking of the resistant rock that
constitutes the free slope. This is what supplies most of the cover of a debris slope,
although other parts can come from the underlying substrate. Despite the great
development that they can undergo in most deserts, studies of them are scarce compared
with the ones carried out in Alpine areas (Oberlander, 1997b).
Schumm and Chorley (1966) indicated that the difference between one slope and others
could be explained quantitatively by the relation of weathering (W). The existence or
absence of debris at the foot of the scarp depends on the velocity of fragment generation in
the free slope (p) before the destruction at the foot of it (d). If this relation is higher than 1
it generates slope accumulation, alluvial fans and a great amount of debris in general. If W
is equal to 1, there is a balance between p and d. If W is less than 1, blocks when falling
disintegrate upon impact and the remainder is weathered before the next fall. This last is
common in the Colorado Plateaus.
This debris can totally or partially cover the slope and, at the same time, be affected by
rilling and gullying that gradually work down to a major extension of the rock substrate.
The extreme case results from the practical disappearance of the slope cover and, if clays
are the outcropping material a badland landform generates. In some cases, hard and weak
layers constitute the materials of the substrate; in this case a stepped form originates from
composite microslopes (free slope and debris slopes) along the main slope (Figure 14.7).
These are the characteristics of the complex slopes.
The fragments of the debris slope show a poor sorting. They are affected by sheet
erosion, which exports the fine particles to a lower part of the slope, with the bigger
fragments standing out. These also can show an important weathering, with development
of alveoles and tafoni in their walls. Finally, in some cases small mass movements can
occur that affect to the debris slope.
3. Talus flatirons
These landforms were originally referred to as talus flatirons (Koons, 1955), flatirons
(Everard, 1963), dreieckshiinge (Wirthmann, 1964; Btidel, 1970), versants tripartites
(Gossmann, 1976), tripartites slopes and triangular slope facets (Btidel, 1982), and talus
relicts (Gerson, 1982). It should be noted, however, that the term "flatiron" has been
previously designated only for upfolded, hard sedimentary rock, and "triangular facets"
for eroded fault scarps, so considerable care must be utilized to avoid sloppy terminology.
These talus flatirons are characteristic forms of arid and semi-arid environments, although
they have been also recognized, as seen before, in periglacial areas. They have been
studied in the Sahara-Arabian Desert, southwestern United States and in some
Mediterranean semi-arid zones.
Their origin initially corresponds to a composite slope, which afterwards is incised by
water erosion (rilling, gullies, piping), of the remaining portions of individualized relict
slopes (Koons, 1955) (Figure 14.12). These paleoslopes have a triangular form with their
9 ~ ., ~,
----_-_ \,~,,
Q1 ~ o ~
0 500 1000 Ft.

9 , A , ~11
Figure 14.12. Schematic section of a talus flatiron, western Grand Canyon, Arizona (Koons, 1955).
apex pointing to the scarp (Figure 14.13), but trapezoidal morphologies also exist. Talus
flatirons are separated from the scarp and are parallel to it. In the distal parts they connect
to piedmonts (Blume and Barth, 1972) and, finally, with fluvial or lacustrine terraces. The
talus flatirons are concave and are crowned by debris, in general are poorly sorted and
generally have a thickness less than 8 m. In the upper part they can reach angles close to
30 ~ whereas in the distal zones they vary between 2 and 5 ~ The inner slope of the talus
flatiron dips contrary to the first one and it is composed of the substrate materials. It is clear
Figure 14.13. Talus flatiron with triangular apex, developed at the foot of the hill of San Pablo. The area
is affected by an intense gullying, Villanueva de Huerva, Zaragoza Province, Ebro Depression, Spain.
that for generating a talus flatiron the alternation of stages is necessary in which
accumulation is dominant, followed by another where erosion is dominant. The succession
of several alternant stages gives rise to talus flatiron sequences, in which the oldest
flatirons are the more distant from the scarp (Figure 14.14 and Figure 14.15).
Talus flatirons are relict slope accumulations and the separation is caused by the
incision of the middle slope gullies, where a group of fills converge (Sancho et al., 1988).
Once gullies dissect the talus flatiron, they flow to the lower parts where they deposit their
sediment load in the form of alluvial fans. Once the talus flatiron is isolated, each one
functions in an independent way (Btidel, 1970).
The caprock that forms the free slope above the talus flatiron must not be very thick so
that the accumulation does not become excessive (Schmidt, 1987, 1989a), and the slope
erosion and scarp retreat will be relatively rapid. For reaching this, the substrate must be
constituted of easily erodable rocks. These circumstances appear in composite slopes but
not in complex slopes where the alternation of resistant and weak rocks inhibits the
development of talus flatirons (Gutifrrez et al., 1998a,b).
One of the formation models of these talus flatirons is due to Koons (1955). A part of a
big rock fall covered the slope, followed by a dissection stage that individualized the talus
flatiron (Figure 14.16). Nevertheless, the most accepted origin is related to climatic
oscillations (Everard, 1963; Gerson, 1982; Gerson and Grossman, 1987; Sancho et al.,
1988; Schmidt, 1989b, 1994, 1996; Arauzo et al., 1996a; Gutifrrez et al., 1998b). The
stages with more vegetation correspond to a major debris accumulation on slopes. A
decrease in vegetation cover can cause slope dissection creating a talus flatiron.
Nevertheless, in recent epochs human activity caused destruction of the vegetation cover,
Figure 14.14. Sequence of talus flatirons in the Chalamera region, Huesca province, Ebro Depression,
Spain.
Figure 14.15. Geomorphological map of the Chalamera zone (Huesca province) and idealized block-
diagram block of the four slope evolution stages, in which can be observed three phases of talus flatirons.
1 - Mesas and structural scarps. 2 - Infilled valleys. 3 - Gullies. 4, 5, and 6 - Talus flatirons $4, $3, and
Sa. 7 - Debris slope, $1.8, 9, and 10 - T3, T2 and T1 Terraces of the Cinca river. 11 - Alluvial fans
(Sancho et al., 1988).
(a) (b) (c)
Figure 14.16. Non-cyclic model of talus flatiron development, according to Koons (1955), in Schmidt
(1989a).
which can be the unrestrained factor for debris slope incision (Everard, 1963; Sancho et al.,
1988; Guti6rrez and Pefia, 1989, 1992, 1998).
4. Free face retreat velocity
To calculate this value it is necessary to know, on the one hand, the lineal dimension that
corresponds to the distance between two different moments of the free face, and on the
other hand, the time elapsed between these two points. In this way, scarp retreat rates can
be obtained (Figure 14.17). There is an important difficulty in obtaining these data,
which implies a large imaginative effort and the adoption of numerous and varied
assumptions to reach retreat values. As a consequence, the methodologies applied for the
Figure 14.17. Scheme of a talus flatiron in which is indicated the different parameters used to calculate
the scarp retreat rates (Guti6rrez et al., 1998a).
obtaining of these data are very variable (Oberlander, 1997a). On the South Sinai
Peninsula, Yair and Gerson (1974) calculated rates of 0.1 to 2 mm/yr from fault free face
retreats. The presence of a thin lava cover over a retreating free face in Northern
Arizona, dated by K/Ar, let Lucchitta (1975) obtain values of 6.7 mm/yr. Schmidt (1980,
1989b), analysing the geometry of the captured consequent valleys of the Colorado
Plateaus, obtained values of 0.3 mm/yr and 0.5 to 6.7 mm/yr. Other methodologies,
much criticized by several authors, are based on the pack-rat accumulations of Neotoma
existing in the mouths of Arizona caves and dated by C-14. The values obtained are
0.45 mm/yr (Cole and Mayer, 1982). Young (1985) studied the free face retreat from
the Lower Eocene erosion surface in northeastern Arizona, from a deduced position of
the scarp based on the location of the straight valleys and obtained values of 0.16 to
0.17 mm/yr. Schmidt (1987, 1988, 1996) calculated the retreat by graphical constructions
starting from talus flatiron profiles. In his 1996 publication he correlated the talus
flatirons with Illinoisian and Wisconsinan glaciations and the resulting values were 0.2 to
0.35 mm/yr. Most of these works were based upon supposed ages. Sancho et al. (1988)
fixed the date from archaeological remains in slopes and talus flatirons, obtaining free
face retreat rates of 0.3 mm/yr in the Tertiary formations from the Ebro Depression of
Spain. Temporal values are more accurate when different slope talus flatirons are dated
with C-14, in the central Ebro Depression, which have ash and carbon remains in their
contents (Arauzo et al., 1996b; Guti6rrez et al., 1998b). The rates obtained by these
authors are 0.9 to 1 mm/yr for the last 35,000 years. All these data reflect that the retreat
rates in arid zones of different world parts are about 0.1 to 7 mm/yr.
It has already been indicated that most of these values were obtained with simplified
indirect methods and with supposed ages. All together these constitute an important but
inconvenient method for correlation of these values, as long as there does not exist a clear
reliability of the methods and, therefore, in the results. Another problem is related to the
sinuosity of the free face. Generally, it supposes a straight front, which is more common as
much of the scarp thickness is; despite that it is not a universal rule. When using the
method of extrapolation of talus flatiron segments (Sancho et al., 1988) until their
intersection with the prolongation of the top of the free face, intersections are obtained at
different distances, probably due to the free face sinuosity (Figure 14.18). Nevertheless,
while having many intersection points, an arithmetic mean can be obtained that is much
closer to the actual retreat rate.
One of the most important characteristics, relative to the major or minor retreat
velocity, is related to the lithologic and structural features of the constituent material of the
free face, which at the same time controls the resistance degree to the erosive processes
(Schumm and Chorley, 1966; Nicholas and Dixon, 1986; Schmidt, 1989a). The
mineralogical composition of the rock is fundamental as it determines the reaction to
the weathering processes dominant in the area. The erosion rate is also influenced by the
dip of the layers (Howard and Selby, 1994), so that the higher the dip, the smaller the
volume of eroded rock, and vice versa. Equally, the thickness of the rock of the free face is
in inverse ratio to the retreat rate (Schumm and Chorley, 1966; Schmidt, 1987, 1989b).
The presence of important thicknesses of constituent rocks of the free face facilitates the
application of the acyclic model of the origin of talus flatirons of Koons (1955), supported
by Schipull (1980) and Schmidt (1987). With this origin, free face retreat takes place
starting from strong impulses, unconstrained by big rock falls, between which alternate the
....... ........:~"~ii,!i~:....,,ii'..-::::
9 :~111..... S4
.... ,..~.: .,,~,:~,,.. ...~....,: .....
d d
25 m.
25 ~n.
d ~ " d
Figure 14.18. Profiles of talus flatirons corresponding to different stages (51-54) drawn in unbroken line
and extrapolated curves to points. The altitude of the apex of the talus flatiron with respect to the current
scarp is h and d is the distance of the apex to the scarp, measured on aerial photography (Sancho et al.,
1988).
production of particles by the weathering processes present in the area. On the contrary, if
the free face materials are a little thick, free face retreat is much faster. The degree of
fracturing of the materials of the free face constitutes one of the most significant factors
with respect to the retreat rate and mainly controls the degree of resistance to erosion. In
this way, the rocks affected by a highly significant fracture density show higher values of
retreat rate. Evidently, the erosion intensity of the substrate also affects free face retreat,
because where it is composed of highly erodable rocks such as clays and marls, rills
progress toward the base of the scarp producing basal undercutting and rock falls
(Schipull, 1980; Gerson and Grossman, 1987).
Another extraordinarily important factor related to scarp retreat, is the intensity of the
physical, chemical, and biological weathering processes that produce rock disintegration
and accumulation on the debris slope. These processes vary as a function of the
lithological and structural characteristics of the rocks, as well as with the climate and its
changes, and also with the acting period of these processes.
It has been already indicated that there exists a balance between the supply rate of
fragments from the free face to the slope (p) and the erosion rate (d) of these
accumulations, which constitutes the weathering relation (W) of Schumm and Chorley
(1966). In some cases, p clearly surpasses d and the slope accumulation surpasses the free
face, fossilizing it. In these conditions free face retreat is paralysed, although chemical
weathering can continue to act (Figure 14.19). This balance of production-destruction can
occur without big oscillations because the changes do not significantly affect the system. It
can lead to major erosion, however, so that the debris slopes gradually remain suspended
or isolated. This breakdown of the balance is due to an environmental change and implies
the overcoming of a geomorphic threshold (Schumm, 1979). Most researchers estimate
Figure 14.19. Slopepartially covered with debris that fossilizes the scarp. Afterwards it is affected by an
intense filling that affects the red clay Miocene substratum, Hill of the Alfambra castle, Teruel Province,
Iberian range, Spain.
that the slopes have been subjected to numerous climatic changes in which in the humid/
cold stages accumulation dominates and in the dry/warm climate the erosion processes
prevail. Free face retreat takes place in both climatic types, but it is possible that it occurs
more in one climate that in another due to the weathering-intensity variation, generating
mass movements, increasing basal undercutting, and so forth.
Finally, it is necessary to keep in mind the speed of base-level decline (Howard and
Selby, 1994) that affects the drainage system, and which can also be due to climatic or
tectonic causes. Similarly, the proximity of the slope to a fiver with great erosive power
can break the slope profile in its lower part and unleash erosion that affects the free face
retreat, just as it appears in the Cinca fiver in the Chalamera area of Huesca Province in
Spain.
5. Badland slopes
These are areas of very intense water erosion, with high-drainage density (313-820
km/km2), lack vegetation, and having steep slopes that make transit across them difficult.
Badland slopes basically develop on labile materials, in arid and semi-arid environments,
commonly at the foot of the scarps (Figure 14.20). The morphology of the divides can be
rounded (Figure 13.14), sharp (Figure 13.15), and in their retrogression they can generate
turreted hoodoo forms (Figure 14.21) (Scheidegger et al., 1968). Some badland areas are
the consequence of the anthropic activities, such as mining, deforestation, construction,
exploitations, dumps, and so forth (Aghassy, 1973). As erosion rates are very fast, they offer
Figure 14.20. Badlandzone with slashed divides that affect clays and lenses of Miocene sandstones. The
top of the formation is eroded by a cemented accumulation glacis that is Pleistocene in age, Bardenas
Reales, Navarra, Ebro Depression, Spain.
a small laboratory with a complex microrelief in which the processes and the modifications
of the forms can be analysed in short periods of time. Nevertheless, one must be careful
when extrapolating these results to areas of major surfaces (Campbell and Honsaker, 1982;
Campbell, 1989).
Numerous processes affect badlands that control its development. Therefore, it is
necessary to know the physical, chemical and mineralogical properties of the materials
(Guti6rrez et al., 1995; Gallart et al., 2002) that constitute these zones of high erosion, to
try to understand the activity of the implied processes and their influence on the
morphology of the area. So, the dispersion index and swelling capacity are physical
properties of great importance in several processes. Chemical analysis permits the
obtaining of the sodium absorption ratio (SAR) or the sodium exchangeable percentage
(ESP), which are directly related to the susceptibility to piping of the materials. X-ray
diffraction analysis of the clay fraction, carried out on oriented aggregates, shows the
mineral species existent. The presence of swelling minerals, such as montmorillonite,
affects considerably the weathering and erosion processes while also producing successive
shrinking and swelling in the superficial materials.
In general, badlands develop under arid and semi-arid climates, where the droughts
are prolonged and the dominant rainfall is storm-like. Nevertheless, they can develop in
areas with 1075 mm rainfall, such as in Perth Amboy, New Jersey, although these
badlands are of anthropic origin (Schumm, 1956a). The role of rainfall is important from
the chemical point of view (Bryan and Yair, 1982), because in areas of scarce rainfall
salts of high solubility are not reached to dissolve. On the other hand, if water
Figure 14.21. Turreted morphology developed on Miocene clays and sandstone from the Tudela
formation, in an area of intensifying water erosion. Cabezo de Castildetierra, Bardenas Reales, Navara,
availability is very abundant, calcium and magnesium will lixiviate. With intermediate
rainfall calcium predominates. These variations reflect upon swelling, the dispersion rate
of the clays in the regolith and, therefore, upon the rates of infiltration. At more detailed
scales, slope exposition influences water availability and, therefore, the activity degree of
several geomorphic processes (Yair et al., 1980). In zones subjected to frost action, as in
western Colorado, a variation in permeability is produced, converting a less permeable
rill surface, into a very permeable one covered with aggregates (Schumm and Lusby,
1963; Schumm, 1964). In Alberta, Canada, the freeze-thaw cycles affect the superficial
material, preparing it for a quick denudation by subsequent run-off (Campbell, 1974).
The surface outcropping materials are subjected to numerous fundamental physical
and chemical processes that destroy the primitive rock by making a formation of a
regolith. That of the arid regions is of limited thickness (20 to 30 cm) and the superficial
layer presents a polygonal cracking when the clays have a little swelling capacity. If it is
high, the shrinking and swelling cycles produce the breakdown of this superficial layer
Figure 14.22. Popcorn structure in which can be observed the fragmentation of the superficial layer of
the Tertiary clays. Cuesta del Viento-Colola, San Juan Province, Argentina, Andean Precordillera.
and convert it in fragments with big voids and popcorn structure (Figure 14.22). This
layer also exhibits an intense lixiviation of the soluble ions. Below occurs a 5 to 10 cm
thick layer that can be an amorphous dense crust (Gerits et al., 1987) or a granular
aggregate (Schumm and Lusby, 1963). The transition to fresh rock is represented by a
slightly weathered and fragmented layer. There is, however, a great diversity of regoliths
in terms of rock composition, the steepness, and the exposure of the slope. After rainfall,
if the regolith is dry, water infiltrates following the cracks, pipes and pores until it
saturates, so that run-off starts after a few minutes. The regolith swells when wetted, the
cracks begin to close, and the internal flux is restricted to the big cracks and to
micropipes (Figure 14.23) (Hodges and Bryan, 1982). Afterwards, the wetting front
descends gradually; infiltrating the water until it finally generates superficial flow and the
almost complete closing of the cracks.
The badland surface is affected by different water-erosion processes, such as the impact
and splash of rain drops (rainsplash), filling, gullying, and subsurface erosion (piping),
which has already been described before and which is a most important erosive agent.
Nevertheless, these areas are also subjected, on occasion, to mass movements.
Mass movements figure enormously in badland morphology. If regolith easily absorbs
water, as occurs with the Chadron Formation in South Dakota (Schumm, 1956a,b), slopes
evolve by creep and by small pellicular slides, which fundamentally affect the regolith
(Figure 14.24). It leads to the development of rounded divides. On the other hand, when
the absorption of water is small, as in the Brule Formation, filling is important and a high
STAGES IN THE INITIATION AND D E V E L O P M E N T

OF R U N - O F F FROM SHALE SLOPES
IA I N F A L L
INTERFLUVE LPRIMARY CRACK ~

1 ',
I J 1 [
,
PoPOO.N
:::. ":':':'-.".":,i~: .;": ':'
~,~t:>Iz~- ~ ' " ~ "P,P~

(a) DRY CHARACTERISTICS (b) WETTING :INCIPIENT RILL AND
PIPE FLOW
111 IL I 1
TRANSIT~
WETTING
,:,.:,,,,,~,,,~ ,,,,,,,, ,,, ,,, ~, ,, .... .,~!
o I'l I o 0101 1% i ,il iol
!lc~
~, SUSTAINEDRILL AND PIPE FLOW 9 (d) INCIPIENT CRACK FLOW 9
CRUST SEEPAGE MICRO-PIPING
11111 I I I I I
9 " .~-~=~ ..... I ~~~~,~,.~.. ,
(e) SUSTAINED
MICRO-PIPECRACK
FLOW FLOW" (f) OVERLANDFLOW-CRUSTFLOW
Figure 14.23. Stages in the initiation and development of the run-off on a shale slope (Hodges and Bryan,
1982).
Figure 14.24. Badlandzone with rounded divides, in which can be recognized numerous pellicular slides
and cicatrices with a half-moon beginning shape. The substratum corresponds to Miocene marine marls of
the Guadix-Baza depression. Experimental station of Las Dehesas, Granada, Betic Range, Spain.
drainage density, steep slopes and sharp divides, characterize the landform. Nevertheless,
on these slopes mudflows can be produced that are channelled along the existent rills
(Figure 14.25). In the Ebro Depression of Spain, the generation of mudflows originates
during the wintertime, during which rainstorms are the norm. Rainfall and the lower
evaporation, as a consequence of the lower temperatures, cause regolith saturation that can
reach liquid states and when flowing gives origin to the mud flows that fill the rills. During
convective rainfall these mud flows are re-eroded, starting a new cycle (Guti6rrez et al.,
1995).
These badland areas lacking vegetation show very high erosion rates. As a
consequence, microforms develop very fast and allow erosion monitoring over short
time intervals. To measure this soil loss, dynamic and volumetric methods can be used (De
Ploey and Gabriels, 1980). The first method measures the amount of transported sediment
by water erosion with collectors (Figure 14.26) or Gerlach troughs. The second measures
the denudation with erosion pins (Figure 14.27) and bidimensional (Figure 14.28) and
three-dimensional microprofilometers (Campbell, 1974, 1982). It is evident that the values
of soil loss obtained depend on the utilized technique (Yair et al., 1980) and the erosion
rates registered with volumetric techniques are, for some authors (Takei et al., 1981;
Rogowski et al., 1985; Hall, 1988), higher than those obtained with dynamic methods. On
the other hand, Sirvent et al., (1997) reached an opposite conclusion, which indicated that
the rates obtained with collectors were bigger than those recorded with erosion pins.
Nevertheless, one of the big problems is the conversion of the volumetric data, because it
Figure 14.25. Mud flow channelled in a previous rill. Miocene clay of the Hoya de Ayerbe, Los Corrales,
is necessary to know the bulk density of the regolith and this has varied through time
(Regfies et al., 1992).
Water erosion produces a lowering that has varied between 2 and 20 mm/yr for
different experiments carried out in different deserts (Campbell, 1989). The maximum
known rates of erosion are recorded in China where they reach values of 38,000 t/kmZ/yr.
Hadley and Schumm (1961) obtained values of about 13,400 t/km2/yr in the South Dakota
badlands. The values fluctuate considerably as a function of the material type, magnitude
and rainfall intensity, time of recording of the experimental stations, technique utilized,
and so forth.
As the erosion velocity is very rapid, we can at least have a partial knowledge of the
development and evolution of this landform. Many aspects of the major morphologies are
reproduced in miniature in the badlands (Figure 15.8), which furnishes information in
understanding their development (Scheidegger et al., 1968). Works carried out in Perth
Amboy by Schumm (1956a) on the drainage system developed in these badlands, reflect
Figure 14.26. Experimental Station Bardenas-2 for the study of water erosion, in which has been
installed fences and a collector device in a badland area over Miocene clays, Bardenas Reales, Navarra,
0 3m
I
Figure 14.27. Ground lowering contours recorded by erosion pins between November 1991 and
November 1993 in the experimental plot Lanaja-1. Huesca Province, Ebro Depression, Spain. Legend:
1 - rills; and 2 - ground lowering contours (mm).
Figure 14.28. Microtopographicprofilometer made up of 50 pins, that lets us obtain a profile between
two fixed points. Profiles are made every 6 months in order to study microtopographic variations.
that their evolution is similar to those of large fluvial systems. Although a dendritic
network can be randomly developed, the channel spacing and their distribution is
intimately related to the erodibility of the material and the run-off generated per surface
unit. The channels incised rapidly into the materials and generated steep slopes. As slopes
retreated they formed micropediments at the foot, which increased in size with the retreat
of the slopes. These piedmonts also finish eroding and slopes evolve (Schumm, 1962).
The investigations carried out in regions with different aridity degree show erosion
rates of 0.45 mm/yr in the northern Negev of Israel with a mean precipitation of 90 mm
(Yair et al., 1980). In the badlands of Alberta, Canada, rates of 4 mm/yr with a
precipitation of 330 mm are indicated (Campbell, 1982). All of this leads one to think that
in a certain badland zone, when analysing for long time intervals, erosion rates might have
been quite variable, because these arid and semi-arid regions have been subjected to
numerous climatic changes. In them, rainfall varies substantially from the more dry
(interpluvial) to more humid (pluvial) periods and, therefore, the erosion rates will be
lower or higher, respectively.
Chapter 15
The arid region piedmonts: glacis and alluvial fans
Piedmonts constitute the transition between the mountain areas, where erosion is the
predominant process, and low-lying areas where transport and sedimentation are the
dominant processes. Aridity controls weathering on the hillslopes and in this context being
the particle production from rock fragmentation, it is normally small. Nevertheless, within
mountain and basin deserts (Mabbutt, 1977), the climatic conditions are usually variable in
the uplands, but the contribution of rainfall or snow melting is important. In this way, as a
consequence of such external supply the piedmont zones will receive water flows with a
higher sediment load than those occurring in typical arid areas. The mobilized materials
within the mountain catchments are transported by confined streams, which emerge into
piedmont zones where the water flow is abruptly unconfined and then transported
sediments spread out radiating downslope. The unconfined water flow decreases its
velocity and transport capacity abruptly, giving rise to progressive sedimentation. On the
other hand, in the backing hillslopes the development of erosive-accumulative ramps is
typical, which constitute another characteristic landform of arid zone mountain fronts.
The landform assemblage generated within the piedmonts, therefore, can hold either an
erosive-nature pediment (glacis) or a depositional one (alluvial fans). Pediments and
alluvial fans can be intimately related, and sometimes may be difficult to differentiate
(Figure 15.1). Bull (1977) suggested that fans can be distinguished from pediments when
the thickness of the associated deposits is up to 1/100 times the axial length of the
depositional area. In addition, a fan can be levelled in response to an external base-level fall,
giving place to pediment-like surfaces, which in turn can be buried by renewed alluvial-fan
deposition. On the other hand, pediments tend to develop in low-relief areas (low and
narrow ranges and modest escarpments), whereas alluvial fans are usually linked to more
energetic topographic configurations (mountain fronts, valley junctions, and so forth).
1. Glacis and pediments: definition, terminology and geographic distribution
According to Ldpez-Bermfidez (1973) who carried out a good updated revision on glacis;
this term (which means the gentle sloping ramp of the ancient Roman fortifications) was
first used by Dresch (1938), to designate specific landforms with a gently concave
longitudinal profile and with slopes down to 6 ~ Tricart (1969) considered that glacis
constitute vast piedmont plains with a low gradient, and coated by a thin sedimentary
veneer, which eventually connected with the desert axial fluvial systems constituted by the
ephemeral stream channels of ouadis (wadis) (Figure 15.2). Important terminological
problems arise with the progressive incorporation of this term into different languages,
Figure 15.1. Piedmont deposits in which the discrimination between alluvial fan and pediment surfaces
proposed by Bull (1977) are not applicable. This is due to the absence of channel incisions which impede
the direct observation of the underlying bedrock. This picture can illustrate a mantled pediment, and also
coalescent alluvial fans. La Sierra del Cerro del Coronel, Province of San Juan, Argentina. Andean
Cordillera.
Figure 15.2. Pediment surfaces merging with the upper terrace of the Ouad Ziz. Central Atlas, Morocco.
Photo by F. Guti6rrez.
which give rise to an extensive terminological and conceptual misunderstanding (Tator,

1952, 1953; Dumas, 1967; Tricart, 1969; P6csi and Sailfird, 1970; Cooke and Warren,
1973; Whitaker, 1973, 1979; Zuidam, 1976; Demangeot, 1981; Cooke et al., 1993). The
English writers usually use the word pediment, which was first used by Gilbert (1877), and
latter defined by Mabbutt (1977), as a flat piedmont surface cut in the bedrock and
separated from the backing hillslope by an abrupt knick-point or piedmont junction
(Figure 15.1). Where a thin veneer of debris coats the erosive surface of the pediment, it is
called mantled pediment. In the German literature the equivalent term is the so-called
fussfliiche (Mensching, 1958).
French authors make an extensive use of the term glacis, but either for clarification or
descriptive purposes, they label it with different adjectives generating a large list of terms
(Tricart, 1969). L6pez-Bermfidez (1973) compiled 28 different denotations, but Whitaker
(1973) managed a larger list of terms. Those glacis directly carved in the bedrock without
associated deposits (equivalent to pediments) have been denoted as bare glacis (Birot
and Dresch, 1966), ablation glacis (Dresch, 1970), erosion glacis (Dresch, 1938), or
denudation glacis (Tricart, 1969; Tricart et al., 1972). Different terms are also used
depending on the thickness of the associated detrital deposits. In consequence, Tricart
(1969) discriminated among denudation glacis (glacis de denudation) with a thin and
discontinuous sedimentary cover, sheet-flood glacis (glacis d'dpandage) with a
continuous and apparent (1 to 3 m thick) veneer of sediments and finally accumulation
glacis (glacis d'ennoyage) where the sedimentary cover is over 5 to 10 m thick. This
terminological overabundance joins with a significant conceptual confusion. For some
French authors the erosion glacis are only developed in soft rocks, like some limestones
and marls; Dresch (1957), for example, set aside the term "pediment" to characterize only
erosion surfaces in crystalline bedrock (Birot and Dresch, 1966). Birot (1960) considered
the definition of glacis as the main problem of arid zone geomorphology. Therefore, here
we propose that, to avoid any kind of terminological confusion, the term pediment should
be used for pure erosive surfaces carved in the bedrock, and mantled or covered pediment
for erosive surfaces fossilized by a thin veneer of alluvial sediments.
Pediments develop in a wide variety of climatic, lithologic, and tectonic contexts
(Dohrenwend, 1994). They occur from subpolar latitudes to arid or humid-tropical climates
(Whitaker, 1973, 1979) and therefore they can be considered plurizonal landforms
(Figure 15.4). The dictionary of Whitaker (1973) even quotes the term cryopediments in
relation to nivation processes occurring in periglacial areas. Pediments, however, develop
extensively throughout arid and semi-arid regions where they are characteristic landforms
(Figure 15.2). In arid zones, vegetation is scarce, thus reducing the impediments to the
erosive work of overland flows that generate the pediments. We understand that pediment
is synonymous for erosion glacis, as proposed by Dresch (1957). In consequence, we think
that most of the previously- listed terms are usually referred to arid and semi-arid climatic
conditions, or they can constitute ancient arid palaeo-landforms.
1.1. Pediment morphology
Pediments are commonly fed by small mountain catchments with limited relief. They can
emerge from mountain fronts, which constitute the junction between the catchment areas
t. "a L L . ~ (~+ ,'~
~- v ~.W" r- r, r- ~ ~ . . _....
.:. _~~
;~ ~ P, , >~'~-'..
,-(~ o..
Figure 15.3. Terminologyof pediments and related features of the desert piedmont zone. (1) Backing
hillslope. (2) Piedmont junction. (3) Piedmont angle. (4) Pediment. (5) Mantled pediment. (6) Mantle. (7)
Alluvial plain. (8) Alluvial fill. (9) Suballuvial floor (after Mabutt, 1977).
and the piedmont plains. Mountain fronts may have a linear or sinuous geometry
depending on the associated tectonic activity. In any case, they display spaced indentations
(irregularities) corresponding to the different valley outlets from the mountain catchment.
Valley-outlets are called "quebradas" in South America, where they usually correspond to
narrow rocky valleys with steep slopes. In the absence of tectonic activity, mountain-front
sinuosity increases with time (Bull and McFadden, 1977). On the other hand, linear
fronts develop in relation to active faulting or to lateral erosion of longitudinal stream
flows along the front-toe. Linear fronts develop well- defined, and sharp, knick-points.
Conversely, in other areas the piedmont junction is developed in a progressive way, giving
rise to a large concave hillslopes where it is difficult to identify the precise beginning of the
pediment surface (Figure 15.2 and Figure 15.3).
Few publications have been focused on the morphometric analysis of pediments
(Corbel, 1963; Mammerickx, 1964; Cooke, 1970b), consequently it is basic to list the
variables considered in these studies (pediment area and slope). Pediments present a
variable surface area, ranging between 2.5 and 650 km 2 in the Desert of Sonora in Arizona.
On the other hand, the fraction of the total surface area occupied by pediments in a specific
desert zone also may have great variability. In the Mojave Desert of California the portion
of area occupied by pediments is only the 6.7% of the total desert area, but in the Sonora
Desert, this proportion increases to 30%. Pediments surfaces usually develop from the
lower segments of the hillslopes down to an axial gully or sedimentation plain. The
longitudinal profiles of pediments may vary from concave to linear, and they everywhere
show a consistent downslope gradient reduction. When pediments border isolated relief, or
inselbergs, long-profiles display a characteristic radial form (Figure 15.5 and Figure 15.7).
These may be of two types: position inselberg (fernlinge, from the German authors),
which preservation is due to its watershed location, and resistant inselberg (hiirtlinge)
developed by the occurrence of differential erosion.
The pediment slopes range between 0.5 and 11 ~ (Tator, 1952), and they may vary along
their concave longitudinal profiles (Figure 15.6). Where pediment development occurs
~E,~, ~ , i ,~ , ' , ~ u,oZ.g~' ~,
~s
..." "'" " ," ,~ 9 i
0 ki/ometres 5000 ~
Figure 15.4. Map of the pediments mentioned in the literature (after Whitaker, 1973).
along active tectonic fronts their gradient is steeper than in areas of tectonic quiescence.
There are also relationships between slopegradient and clastsize, in such a way that
when the steeper the slope, larger clastsizes occur. For some authors lithology
is not a relevant factor (Mammerickx, 1964), while others point out its impact on the
Figure 15.5. Inselberg carved in Cambrian sandstones with the development of gently slope radial
pediments. Jodhpur-Jaisalmer region. Semi-arid zone of the Thar Desert (India).
Figure 15.6. S e g m e n t e d hillslope with a s u c c e s s i o n of c o n v e x and c o n c a v e s e g m e n t s . T h e r e l i e f

t e r m i n a t e s in a p e d i m e n t d e v e l o p e d f r o m a c o n c a v e k n i c k - z o n e . M a c h a k o s , K e n y a .
slopegradient (Cooke et al., 1993). Climate seems also to affect glacis gradients, with
smaller glacis slopes in the more arid regions (Mabbutt, 1977).
The channel network of pediments is usually entrenched in the proximal zone, but
channels radiate downslope towards the distal areas, where commonly overland flows
~ ~ ~ ' J _ t ~ " = - - - ~ ~, ~, -. . ~. . = . .. . . . . ~. . . " ' - . . . . 9 "~
. ,.~ , - ,.
9 i ,
.- ~, _
~, .
, ~ -. _
~_ -. -
,_ ~
~
= _
_ - - - ---- - - - - -
_ .
.
.
"
.
.
.
.
. . .. . . . . .. . . . . . . . /1/
/g."s
, - V -t'- ~. ,'.t, . . . . . . .... . . . . . . .
~ ' , , , . W , y - ~ : ; ~ , L~:-.~.: . . . . . . . . .
.. ~._.,,~, t ~-;-.~7-.~ ".." .e ~;.., ~ . , ,, ,
A B -C
9. .
. ."....--.E
. . . . . . :.-T.-:".
.. ..... ". . . . .
9. . . . . . . ". . . . . . . . . . . C
. '.................
~ , . ~ .- -"~'" ".......
.....
. '......
" - - ~
....... .9. _ - ~ " ~ - - " : : ~ - = _ - - ~ ~ ~ , - ~ - S ~ _ -T~_'.'-..-.L'. ".:.'..

......... T : . . :-- .....
.. ;---. ........ :.'""-'-~~:~,~",';~~.'_,u."m~.~- ~_.~.... "~-~.':: ...'. . . . .
I~ ..... -"- ---, . .. , 9 #., ~ S~7".~.- " '*' " ""'.k'~" ,'~" "~"~ "~"~ :'= -"'" ...........
C B A B C
Figure 15.7. Different e l e m e n t s of the p e d i m e n t . (A) D i s s e c t e d zone (up) and i n s e l b e r g s (down). (B)
P e d i m e n t w i t h residual reliefs. (C) M a n t l e d p e d i m e n t (Johnson, 1932a).
gradually diminish by infiltration. Most typical pediment channels have a braided pattern.
They are shallow channels subject to significant switching during extreme floods. During
periods of base-level fall, channel entrenchment can also propagate to distal locations
giving rise to subparallel drainage patterns (Cooke and Warren, 1973). In most cases,
pediments are planation surfaces directly carved in the bedrock, but mantled by a thin and
discontinuous alluvial cover. In particular cases some small relief emerges from the glacis
surface as residual landforms denoted by the French authors as "chicots" (Figure 15.7).
1.2. The role of geology and climate in pediment development
As aforementioned, some French authors differentiate between glacis developed on soft or

in hard rocks. This division attempts to emphasize the great influence exerted by lithology
and geological structure on glacis location and evolution. When pediments develop on
granites, or crystalline bedrock, they usually present a sharp knick-point attributed by
some Australian authors (Mabbutt, 1966; Twidale, 1967) to the occurrence of a more
intense weathering in the piedmont junction zone. As chemical weathering processes are
very limited in arid zones by water availability, the occurrence of sharp knick-points is
attributed to previous humid periods occurring during Quaternary, or even Tertiary times.
In the south of Arizona, where schists and basalts are the dominant bedrock lithology,
knick-point angles are significant (Kirkby and Kirkby, 1974). On the contrary, in more
erodible substrates constituted by several types of weak rocks, the transition between the
backing hillslope and the pediment surface is gradual, but marked by a large concave
knick-zone (Dohrenwend, 1994; Oberlander, 1997b). This latter is the case of the mantled
pediments developed in the Ebro Depression of Spain and throughout the Mediterranean
littoral.
The recent tectonic history of a region (i.e. neotectonic activity) exerts a primary
influence on pediment development. Tectonically stable areas are appropriate zones for
the initiation and development of glacis, which in addition are characterized by sinuous
mountain fronts. On the other hand, where tectonic activity is relevant, mountain fronts are
linear and alluvial-fan development prevails (Bull, 1977; Bull and McFadden, 1977). All
the published observations indicate that tectonic quiescence (crustal stability) is a very
important factor for pediment development.
In arid zones, pediment development is influenced by water availability for surface run-
off. In hyperarid areas (i.e. Atacama, Central Sahara) the deficient rainfall gives rise to the
nearly absent overland flow erosion (erosion paralysis of Oberlander, 1997b), and therefore
the existing pediments should derive from previous wet (pluvial) periods. Climatic changes
also affect the size and extent of the vegetation cover on the pediment surface, and
consequently to the relative interaction of the working erosion and depositional processes.
The role of climatic changes and their impact in pediment development is analysed in detail
in Section 15.1.5 (pediment evolution).
1.3. Dominant processes in pediment development
The processes involved in the generation of overland flows are may be the more important
ones in desert areas. Because the arid regions are specifically defined by a permanent
deficit of water (Tricart, 1969), however, the previous statement may be considered a
contradiction. Nevertheless, the typical desert rainfall events are rare high-magnitude
storm events during which energetic surface run-off processes (arroyada) are generated.
Overland flows can be unconfined generating sheet-floods, but eventually can be confined
in gullies.
In regions of tectonic stability, the pediment surface can be divided into a proximal
erosive area, an intermediate transport area, and a distal depositional area (Johnson,
1932b). During episodes of increasing sediment supply these different areas shift upward to
the proximal zone, and conversely when it diminishes, these different pediment segments
move downwards to distal areas (Cooke and Mason, 1973). The erosive and transport areas
are characterized by the occurrence of mobile shallow braided channels but also by sheet-
flow processes. In these two zones of the glacis, we can observe former pediment surfaces
slightly elevated over the functional ones. These are relict landforms in which the
dominant geomorphic processes are surface and subsurface weathering, eolian reworking,
development of desert pavements with varnish or special soils such as pedocals.
For some authors (mainly Australian ones), the subsurface weathering plays an
important role in pediment development, mainly on those generated on crystalline bedrock
(Ruxton, 1958; Mabbutt, 1966, 1977; Twidale, 1967, 1983). The upper regolith cover can
be easily removed by overland flows. This fact facilitates the exposure of the basal surface
of weathering, generating an etched plain or etch surface. This kind of process is common
in savannah areas, but they also occur in the semi-arid environment. The zones subjected
to deeper weathering are localized at the piedmont junction along the contact between
the bedrock and the regolith, which constitutes a linear weakness zone.
In softer bedrock the overland flows have higher erosive power. In the case of
calcareous rocks, planation is favoured and assisted by karstic corrosion (Nicod, 1992).
In the Iberian Range of Central Spain, mainly constituted of calcareous rocks, pediment
surfaces have an excellent development, giving rise to extensive pediplains.
The study of micro-topographic changes, carried out in small experimental parcels by
Schumm (1962), provides very valuable data about the variations of the micro-pediments
over an 8 year time-series. They were monitored with erosion pins in the gentle hillslopes
developed on the siltstones of the Badland National Monument, South Dakota. The micro-
pediment surfaces underwent a progressive lowering that increased with their slope
gradient, and the pediment junction experienced parallel retreat. On the other hand, in some
of the cases, active deposition took place in the distal areas (Figure 15.8). In that paper,
Schumm pointed out that the formation of such miniature pediments did not have to be
necessarily similar to the real case. Chorley et al. (1984), however, indicated that this
process may occur in many places at larger scales, suggesting that the analysis of badlands
can improve our understanding about the development of some type of pediments.
1.4. Hypothesis on the origin of pediments
Pediment surfaces occur preferentially in zones of low-relief mountain catchments.

When the catchments are more modest, the bigger will be the surface area occupied by
pediments. The main proposed hypothesis for pediment origin is based on the explanation
of those processes governing slope degradation and piedmont generation. The different
120
%, \ \75
. . . . . ~ . . . . . . , 4~
! 4 i 3 0~L . ]
..~ O0 30 CENTIMETERS
901",~ 6.0
37Q ~ 0 -5
~ ~ . . . . . . . . . . . . . . . . . .
6
2 5
~
_
5 I 4
4
-5
97~,- q ,
17.~ g ' 6.s 4.7 3.6 3 I 2.5 I 2 I
77,3~_~_ o , 5.5
1 7
5 ~
90
95 ~ _ _ _ 10
~ , ' ~ ~ , ~ " ~ 7 , ~-~-~ ~ , 5 I 3.5 ! 3 1
Figure 15.8. Micro-pediment profiles from the Badland National Monument, South Dakota. Lower
numbers indicate pediment slope (%) and upper numbers show the location of the different erosion pins
indicating the value of ground lowering (mm) produced by the erosion. Solid lines show pediment profiles
obtained on May 1961 and dotted lines those obtained on July 1953 (Schumm, 1962).
proposals can consider either: (a) the type of overland flows working at the piedmont
zone, or (b) the set of previous degradation processes preparing the bedrock for its later
planation. We have to consider both working hypotheses as complementary ones.
The most classic hypothesis is that of the sheet flood of McGee (1897), where in the
confined streams coming from the mountainous uplands, spread and merge downslope
from the piedmont angle, giving rise to a 0.5 m thick sheet of water and silts that spills out
in an unconfined turbulent flow. These water flows correspond to flood events during
which the erosive effect on the bedrock incorporates particles of very diverse size into the
running waters. Sheet-flood processes have been also recognized by Blair (1987) as active
processes in alluvial fans based on the analysis of aerial photos of the Rocky Mountains.
Tricart (1969) pointed out that a significant relief is necessary to concentrate the surface
waters into streams that later spill their sedimentary load on to the piedmont. This
indicates that with a very low relief the sheet-flood hypothesis does not seem applicable.
On the other hand, because the pediment feeding channels emerging from the mountain
front have some spacing, the intervening (inter-channel) areas are not commonly affected
by these sheet floods. Tricart (1969) also suggested that it is necessary for the occurrence
of a flat or gently topography, prior to the development of sheet floods. Lawson (1915)
introduced weathering processes and mass movements to explain the parallel retreat of
slopes during pediment development. This author also recognized that sheet flood is the
ultimate process in pediment development.
The hypothesis of lateral planation for pediment development was first proposed by
Gilbert (1875, 1877), later accepted by Bryan (1922) and improved by Johnson (1932a,b).
In the piedmont, the confined channel becomes a group of small channels with a braided
disposition. Each individual channel has decimetric widths and depths. These little
channels can find obstacles in their trajectory, such as bushes and large blocks that have to
be bypassed. The braided pattern of these channels is the same one as those developed in
the big rivers, which implies that lateral planation occurs in response to the continuum
divergence of these small channels. The erosive work exerted by the lateral shifting of the
channels gradually removes all the small existent reliefs and, eventually it develops the
pediment. The radial systems of glacis developed at the foot of the inselbergs can be
explained in this way, otherwise it is difficult to understand from the sheet-flood
hypothesis alone (Tricart, 1969). Some authors consider that lateral planation is more
effective when it affects the less resistant rocks (Sharp, 1940; Denny, 1967), whereas in the
granites or crystalline rocks the denudation is smaller (Howard, 1942). Although the
activity of the diffuse overland flows is verified in field studies, this hypothesis should not
be considered as a general explanation (Mabbutt, 1977).
Twidale (1962) and Corbel (1963) considered that the processes of chemical
weathering have some impact on pediment development. In shield and platform deserts,
the denudation of ancient weathering profiles can generate etchplains that later are
transformed pediment in response to planation processes (Mabbutt, 1977). On the other
hand, weathering is especially intense at piedmont angle locations (Twidale, 1962, 1967,
1976a,b), which collects all the overland flows generated on the adjacent hillslopes. This
author relates the origin of both pediment and flared slopes with this kind of selective
weathering. During humid periods the weathering front progresses downslope,
while during the dry periods active denudation removes the previously weathered
material (Figure 15.9).
1.5. Pediment evolution. Pediplains
The overland flows running downslope throughout the surface of the glacis eventually
pour into endoreic depressions, seas, oceans or, more frequently to a desert stream. If the
initial base level of the pediment changes, then different geomorphological processes
reshape the primitive pediment surface. In the first case, the endoreic basins undergo a
continuous aggradation, in response to the erosion of the basin borders which give rise
to a progressive reduction of the pediment slope. When sea-level falls, just as happens in
glacial periods, headward erosion propagates upslope dissecting, separating and isolating
the previous pediment surfaces, as is evidenced in numerous examples on the Mediterra-
nean fringe. On the other hand, when sea-level rises the area occupied by the pediments
decreases, being able to even be completely covered by the marine waters.
The most common cases of pediment evolution are those related to the wadis (desert
streams). The materials transported by the overland flows generate an erosion pediment in
the proximal areas, and accumulation ones in the mid and distal areas, where deposits left
by the desert streams may merge with the alluvial plain. In these zones of coalescence an
interfingering between both kinds of deposits occurs. Pediment deposits are generally
formed by angular clasts of generally the same composition, due to the shared sediment
source of the glacis channels, whereas the alluvial plain deposits are polymictic and
more rounded due to their longer longitudinal transport (Guti~rrez and Pefia, 1976). When
base-level falls in response to climatic or tectonic causes, the lateral alluvial systems incise
/ +
-t- +
.. 1 ...... a m9 desiccated during dry season
s + + +
-,3"" "Sr-7-" + +
~." + +
+
(a) / +
f
+ + +
+ + +
+ ./ + ~ +
(b) / + +
+ + +
j + + +
+ +
(C) + +
Figure 15.9. Stages in the development of flared and overhanging slopes by knick-point (scarpfoot)
weathering and subsequent differential erosion: (a) subsurface infiltration of moisture in the knick-point
area and lowering of the weathering front; (b) base-level fall and stripping of weathered debris resulting in
the exposure of the weathering front as a flared slope; (c) repetition of process and development of double
flare (Twidale, 1976a,b).
gradually into the previous pediment, resulting in the dissection and compartmentation of
the previous pediment surface and the generation of a terrace in the margins of the desert
channel (Figure 15.10). The development of similar cycles of base-level fall originates a
group of staircased pediments (Figure 15.11). These correspond to the glacis embfitds
(Joly, 1950; Dresch, 1957) of the French authors and to the stepped pediments or staged
pediments (Mabbutt, 1977) of the English writers. This system of staircased pediments
reveals the occurrence of successive phases of incision of the desert channels, with
intervening periods of glaciplanation or pedimentation. This model characterizes glacis
evolution, but mainly for soft rocks that evolve quickly in semi-arid climates (for example,
the Ebro Depression and inter mountainous basins of the Iberian ranges of Spain)
(Figure 15.12). On the other hand, in arid climates sensu stricto, where precipitation is
Figure 15.10. Geomorphologic map of a sector of the Alfambra river valley in the northern Teruel Basin,
Iberian Range, Spain. The upper mantled pediment (P3) interfingers with the upper fluvial terrace (T3) and
the lower mantled pediment (P2) with the corresponding lower terrace (T2). Interbedding of mantled
pediment and fluvial deposits occur at the connection zone. (Guti4rrez and Pefia, 1976).
Figure 15.11. Staircased pediments in which are evident the concave knick-zones with the backing
slopes of the Sierras Exteriores in the Pyrenees. Hoya de Ayerbe, Province of Huesca, Spain, Ebro
Depression.
Figure 15.12. Staircased pediments and fluvial terraces in the Sierra del Pobo, Escorihuela region,
Province of Teruel, Northern Teruel Basin. This picture covers a small zone of the area mapped in
Figure 15.10.
Figure 15.13. Mantled pediments that evolved to a desert pavement with varnish. Negev desert, Israel.
very scarce, the altemation between periods of planation and incision takes place within
large spans of time (Figure 15.10), because climatic oscillations of great magnitude are
needed to alter the dominant geomorphological processes. Finally, in hyper-arid areas
fluvial processes are almost completely inactive. In hard rocks, as the granitic ones, the
incision of the lateral tributaries is impeded by the own resistance of the rock, but a
previous weathering for pediment development seems necessary.
In areas of tectonic stability, when arid conditions persist over long periods of time,
the pediments continue their enlargement, as a consequence of the continuous parallel
slope retreat of the adjacent mountain catchment. Afterwards the pediment surface can
develop isolated inselbergs that eventually are also denuded. The final relief corresponds
to an extensive planation developed in arid climates, corresponding to those labelled by
King (1953) as pediplains in opposition to the concept of peneplains of Davis (1899),
which develop in humid climates. The pediplain is the final result of a pediplanation
cycle that is reactivated by surface uplift (Figure 15.14). King (1953) thought that
pediments were the fundamental landforms for all the worldwide landscapes.
Pediplanation gives rise to several consecutive pediplains within the continents. These
are normally assembled incised one into another generating an staircasement
(piedmonttreppen) of pediplains (King, 1962, 1976). These pediplains can be fossilized
or exhumed, as the pre-Triassic and intra-Miocene surfaces of the Iberian range, or may
outcrop, as the different erosion surfaces later developed in this same mountain system
(Guti~rrez and Gracia, 1997) (Figure 15.15). By means of the correlation of the erosion
and sedimentation records, King (1976) differentiated six global planation cycles from
the Jurassic to the Quaternary.
(a)! Uplifted Pediplain
V Base-level of erosion
(b)l
I
I
x\
,
.,.
, ,
ii
,
ii
,
ii
/
~
\
xx
~
x
~
,,\
, ,
it
,"
I
,
I" x \ 1 I I ! \ \ ~ \ i s i
I'~ '. '. , t ., i \ ~ , , / /- .,
I ,, ~ t / / I ~ \ x x / / i
I x_ x x ~ ~ / / \ x x. \ / t /
I% x xx X\ I / / / \ \ ~\ X t I /
I\ \ . ., r t ] \ x . ~ / / / /
(c)!
l/"\ -- i
iI \ [
I ~ i
Figure 15.14. The pediplanation cycle. (A) Youth. (B) Maturity. (C) Old age (Interpreted by Small,
1970).
2. Alluvial fans
2.1. Definition, antecedents, and terminology
Numerous known definitions are based on the geometry, sedimentation area, and
processes involved in fan development (Rachocki, 1981). The definition of Bull (1968),
one of the most excellent investigators in this field, points out that an alluvial fan is a
fluvial depositional body and the surface of which approaches the segment of a cone
radiating downslope from a point where a stream channel emerges from a mountain area
(Figure 15.16).
The history of the knowledge of alluvial fans, also called dejection cones (G6mez
Villar, 1996), goes back to the 19th century when these landforms were initially
brought to modest attention. Only in the late 19th century were some excellent papers
published (Gilbert, 1875; McGee, 1897) in relation to the exploration of the arid
territories of the southwestern USA. In the first half of the 20th century the
investigators started to inquire about the origin, sedimentation processes, reworking,
and tectonic disturbance of fan surfaces (Johnson, 1932a,b; Blackwelder, 1928, 1931).
The leading impulse to the analysis of these depositional alluvial-fan landforms,
Figure 15.15. Planation surface levelling an old folded cretaceous sedimentary sequence. This surface
corresponds to the pediplain elaborated during the middle and upper Miocene in central Spain, where due
to its outstanding outcropping area is denominated as the "Fundamental Erosion Surface of the Iberian
Range." The picture illustrates the Aliaga zone, of Teruel Province. northeastern Iberian Range.
however, took place during the early second half of the last century with the
publication of the key works of Blissenbach (1952, 1954), Bull (1963, 1964a,b), and
Denny (1967).
After that, alluvial fan analysis was focused on different topics, and their study also
extended to their identification and characterization in the ancient stratigraphical record.
The study of both ancient and modern fan deposits, took into account their geometry,
sedimentation and erosive processes, as well as the factors involved in fan development
and evolution. The specific works resulting from these studies will be cited and
commented in the following sections. Nevertheless, it is convenient to highlight some
papers and thematic monographs in which the bulk of knowledge about alluvial fans is
synthesized (Yazawa et al., 1971; Bull, 1972, 1977; Schumm, 1977; Colombo, 1979,
1989; Rachocki, 1981; Nilsen and Moore, 1984; Ashida, 1985; Lecce, 1990; Rachocki and
Church, 1990; Cooke et al., 1993; Blair and McPherson, 1994a,b; G6mez Villar, 1996;
Harvey, 1997a).
Alluvial fans have also been studied by means of simulations (Rachocki, 1981) and
experimental laboratory models. In spite of the considerable amount of problems
generated by simulations, very good results have been obtained (Blair and McPherson,
1994a). In this way, laboratory models have reproduced the development of sieve lobes in
sand-box analogues (Hooke, 1967). Moreover, using such sand and mud analogues, it has
been concluded that debris flow deposits predominate in the fan proximal zone, those
interfingered with channel deposits characterize middle fan locations, whereas at distal
Figure 15.16. Alluvialfans in the Sierra de Alhamilla piedmont, Almer/a Province. Betic Cordillera. The
distal zones are subject of intensive greenhouse agriculture.
areas fluvial deposits dominate (Hooke, 1967, 1987). Other laboratory analogues were
specifically projected for the establishment of the relationships among drainage basin
evolution, fan development and fan stratigraphy (Schumm, 1977; Hooke and Rohrer,
1979; Schumm et al., 1987).
Once the antecedents and trends on alluvial fan research are examined, it is necessary to
emphasize some relevant fan features, as well as to specify those terms commonly used
in the literature. Alluvial fans are characteristic landforms at mountain front locations
(Figure 15.16), but they also occur in intermontane basins and valley junctions
(Figure 15.17) throughout all the different morphoclimatic zones, and with very variable
sizes. An alluvial fan results from the sedimentary output of single or multiple drainage
basins and its morphology and stratigraphy record the process of basin excavation. The
water and sediment fluxes concentrate in the lowermost zone of the basin where a
mainstream channel issues from the mountain catchment. These mixed fluxes expand in an
unconfined flow at this kind of inland basin outlet where, as a consequence of the abrupt
slope and stream-power reduction, the transported load is deposited to build up the
alluvial-fan bodies.
In these settings individual fans frequently coalesce away from the mountain front
locations. In arid and semi-arid zones the alluvial fans constitute the most characteristic
landform on the piedmonts of basin and range desert zones, (Figure 15.18). Distally, the
coarse detrital fraction that constitutes the main fan-bodies, may grade to the thinner
sediments of playa-lake environments.
Figure 15.17. Alluvialfans dissected by the Mendoza River. Uspallata, Mendoza Province, Argentina.
Andean Cordillera.
Blissenbach (1954) divided the alluvial-fan bodies into three different parts
(Figure 15.19): the proximal area, or fanhead, located at the apex; the middle area,
between the fanhead and the lower parts of the fan; and the distal area constituted by the
zone more distant from the apex. Finally, it also included a lowermost area where fan
coalescence occurs.
The climatic classification of alluvial fans into "humid" and "dry" fans, or in other
words fans of arid zones formed by intermittent fluvial flows, and fans of temperate zones
generated by perennial flows (Schumm, 1977) has a positive acceptance. Nevertheless,
this classification is still controversial and it was enlarged by McGowen (1979) when
considering that in arid zones debris flows are more common than in humid areas, where
fluvial sedimentation prevails. Recently, Nilsen (1993) and Stanistreet and McCarthy
(1993) related the form and processes of the fan with the climate, concluding that their
semi-conical shape is built by debris flows in desert environments ("dry type"), whereas
other authors consider that this geometry is also developed in humid temperate regions
(Harvey, 1984a; Blair, 1987).
In dry regions fan development is favoured by different reasons (Harvey, 1997). The
vegetation cover is sparse or absent, intense storm rainfalls are frequent, and overland
flow processes predominate on the hillslopes. In consequence, there is a high rate of
sediment production during the storm conditions. The steep mountain streams with flashy
run-off regimes give high rates of sediment transport and delivery to mountain front
locations. Finally, the sediment transfer from the drainage basin to the fan body is
sporadic and basically linked to the aforementioned intense storm conditions.
Figure 15.18. Landsat image of the San Juan region of Argentina. Hills located to the northeast are the
Sierras Pampeanas, and those located to the northwest belong to the Andean foreland, which are separated
by the Bermejo Depression. The piedmont zone of these reliefs are constituted by coalescent alluvial fan
systems, which occasionally merge with playa-lake systems, locally called "barreales." San Juan City is
located to the south of the picture.
2.2. Alluvial fan morphology
The main morphological features of alluvial fan systems (Blair and McPherson, 1994a)
(Figure 15.20) are the following ones. The drainage basin constitutes the upland area,
generally with steep slopes, from which water and sediments are derived. The feeder
channel transfers the main fluid and solid fluxes to the fan apex and it corresponds to the
outlet point of the drainage basin in the mountain front, this apex is obvious if the front is
sharp. The incised channel is the downslope prolongation of the feeder channel on the fan,
and it may be divided into several channels. The incised channel usually terminates in
proximal or mid-fan areas, corresponding in its location to the intersection point. From this
point flows depart, become unconfined and expand downslope giving place to the active
depositional lobes. The headward-eroding gullies are common features in the distal part of
fans. These channels may progress upslope by headward erosion, giving eventual rise to
/ - - u w ~ r , a l , \
nt ~j-s'~gment X Zone of coalescence
1- with other fans
Apex
1 5 KM
| . . . . . 9
150 I" I
Apex ",,,. &~,
100 -
z
m
uJ ~ Radial profile A-A'

a
t-
n
t-
..J
< 50
01 > 02 ~ / . / ~
_
Vertical exaggeration x 53
.. i. . . . . . 1 .
5 10
SCALE IN KM
Figure 15.19. Typical segmentation and longitudinal profile of an alluvial fan. Example from the Tumey
Gulch fan, western Fresno County, California (Bull, 1964a).
the capture of the incised channel and promoting the downslope shift of the active
depositional lobes (Denny, 1967).
Because the depositional processes are strongly controlled by the sediment supply on to
the fan surface (Harvey, 1997), the bulk morphology of an individual fan should reflect the
main characteristics of its source area. Many of the studies on the morphology of alluvial
fans are based on the analyses of morphometric variables, such as fan shape, fan area and
fan gradient (slope).
Figure 15.20. Main geomorphological features of an alluvial fan system. Example from the Trail Canyon
fan, Death Valley, California (Blair and McPherson, 1994a).
The fan shape is semicircular, or a cone segment, with concave longitudinal profiles
(Figure 15.19) and convex traverse ones. Data extracted from detailed topographical maps
allow development of the following mathematical relationships (Troeh, 1965):
Z = P + SR + LR 2
where Z is the height of a particular point on the fan surface, P the height of the fan apex, S
the fan slope in point P, R the radial distance from P to Z, and L the average slope along its
longitudinal profile.
The fan area is the more studied variable and it has been demonstrated that there is a
simple relationship between fan area and drainage basin area that may by expressed by the
equation:
Af - pA~
Where Af and A b are fan area and drainage area (measured in km2), respectively. Studies of
different groups of fans have demonstrated similar values for the exponent q in different
zones ranging from 0.7 to 1. l, but the values for p show a wider range between 0.1 and 2.1.
These variations are interpreted as a consequence of differing fan age and history between
different regions, but also to the different rock resistance within the drainage basins
(Harvey, 1997). Figure 15.21 displays the relationships among the different morphometric
fan variables for different areas. Fans of the California Coastal Ranges have larger surface
area per unit drainage area than fans in the Death Valley region. The regression line for the
fans Spain of southeast has been obtained from 68 fans (Harvey, 1987a).
The fan gradient is usually taken as the axial fan surface slope in the upper part of the
fan and the most characteristic values range from 2 to 12~ Its relation to the fan is:
Gf = aAbc
Where Gf is the fan gradient. Figure 15.19 illustrates that the values of the exponent b
range between - 0 . 3 5 and -0.15, but values for the constant a show a greater range from
0.03 to 0.17, which is interpreted as linked to different sedimentary processes. Fans in
the Death Valley region are much steeper than the California Coastal Ranges, whereas
the SE Spain fans show an intermediate position. On the other hand, in the SE Spain fan
group of southeast Spain, those dominated by debris flows have steeper slopes than those
dominated by fluvial and/or sheet-flow processes (Figure 15.22), which reveals the
relevant impact of sedimentary processes on fan gradient (Harvey, 1984a).
2.3. Factors influencing alluvial-fan development
Fan deposition takes place if the sediment supply overcomes the transport capacity of the
feeder channel. This indicates that within the channel the critical stream power threshold
defined by Bull (1979) is surpassed. However, fan sedimentation is also affected by
different factors, the impact of which is variable depending on the different climatic and
topographic scenarios.
Topographic factors have a strong influence on sediment supply. In that way erosion
rates tend to be more important on steep slopes, favouring in addition the generation
of debris flows. These predominate in fans supplied by small steep drainage basins.
200 f
J
f
f
, Spanish Fans f
jJ
jJ
100 Californian Coast R a n g e s J
f
f t
P f
50 -I Death Valley Region ff /
/ f
i jJ
J j
f P
/ /
20 I
J /
J
f f
i J
I S
f
10 /
jpl j#
E
v J f
/
5 p
/
.J
p
t" pJ
p
J
J
u_ 2
0.5
p
/
0.2
0.1
I I I I I I I I
0.2 -
0.1 -
0.05
0.02 -
t'-
u.. 0.01 -
0.005 -
0.002 T ~ ~ i ~ s i ~ 1 '1
0.1 0.2 0.5 1 2 5 10 20 50 100 200
Figure 15.21. Regression relationships of drainage area to fan area and fan gradient for three different
regions (Harvey, 1984a).
A Debris flow fans
0.5 --
o Intermediate and fluvial fans
0,2 --
r
(1)
O 9 9 9
0.1 - o A 9 9
03
x_
O')
9 O O
r A o=
03 o %0 ~ o
a_ 0.05 -
o o o
o o
0.02 -
0.01 I I I I I I I 1
0.1 0.2 0.5 1 2 5 10 20 50
Figure 15.22. Drainage are to fan gradient relationship for selected alluvial fans in SE Spain. Contrast
plotting position of steep slope debris flow fans with the gentle slope intermediate and fluvial fans (Harvey,
1987b).
On the other hand, fluvial processes tend to characterize the fans supplied by large and less
steep basins.
The clasts mobilized within the drainage basin of the fans came from the weathering,
collapse or removing of the outcropping bedrock substrate in their slopes. This lithologic
factor exerts a relevant influence on alluvial-fan development. The weathering processes,
mainly mechanical in arid zones, work- conditioned by the rock type, and consequently the
resultant sediments can be quite different. The resultant clast sizes mainly depend on the
granulometry (in detrital rocks) and on the spacing of discontinuity planes affecting the
bedrock. Therefore, when there is a wide spacing in the source rocks, they yield blocks or
coarse clasts. Because in arid zones chemical weathering is practically absent, the silt and
clay particle sizes are commonly scarce. Their abundance increases considerably,
however, where source rocks are fine grained, such as marls, mudstones, siltstones and
tephras. These fine materials are necessary to generate debris flows within the drainage
basins, just as Harvey (1984a, 1987) pointed out in different examples of alluvial fans of
SE Spain.
The climatic factor has both a direct and indirect influence on the weathering processes
conditioning the sediment production within the fan basins. Water availability directly
affects rock weathering, sediment transport, and also indirectly the vegetal cover
development. The intensity, duration, and frequency of strong rainfall events are the most
influencing factors in sediment production. The lack or scarcity of rainfall events promotes a
drastic reduction on the weathering and transport processes. Nevertheless, in areas such as
Death Valley, with a mean annual precipitation of 43 mm, the sediment suite accumulated
within the fan bodies is only derived from rainstorm events (Blair and McPherson, 1994b).
The temperature can also affect fan development because it may increase the rates of
chemical weathering. On the other hand, it is necessary to consider the common occurrence
of altitudinal coolness gradients, which may give place to congelifraction in the headwater
zones. Also, the amount of precipitation increases with altitude. These orographic gradients
may affect the vegetal cover, which in turn has a relevant impact on soil erosion and
slope stability. This is mainly driven by the development and penetration of the roots in
the ground, which is different for different types of plants. Consequently, when climatic
changes, or human activity (overgrazing, deforestation, fires, etc.), impact the expansion
and type of the vegetal cover, the affected hillslopes may became unstable.
Fans can develop in areas of tectonic quiescence, but also in active tectonic zones. The
latter are characterized by the occurrence of basin and range deserts such as those
developed in the semi-arid southwestern USA and central Asia. The tectonic activity
(tectonic factor) may condition the appropriate topographic settings for fan development,
such as active mountain fronts (Figure 15.23). At these locations, thick sequences of
alluvial fans may develop, recording an important depositional history all over the recent
Figure 15.23. Schematicblock diagram of an active mountain front showing the typical internal structure
of the thick alluvial sequence of the piedmont. The occurrence of triangular facets indicate recent faulting
along the mountain front (Bull, 1977).
Figure 15.25. Tectonic tilting and erosion of Pleistocene alluvial deposits in the Andean piedmont close
to the City of Mendoza, Argentina.
geological periods. Neotectonics constitutes the main control in fan location and
development in these areas. Many of the old sedimentary sequences of alluvial fans are
commonly interpreted from a tectonic point of view (Harvey, 1997). On the other hand, the
tectonic activity is not limited to the main range-front fault (Figure 15.24), but ground
deformations and faulting may also affect the entire fan system. These recent tectonic
ground dislocations may affect the feeder channels of the fans giving rise to relevant
channel offsets and displacement with the consequent shifting on the locus of the active
sedimentation (Figure 15.25).
Tectonic activity can also affect fan development by means of direct modification of
the fan slope. When the uplift rate of the mountainous front is greater than the incision
rate of the fan channel (Figure 15.26), fan slope increases and sedimentation occurs in the
proximal zone. On the other hand, if the channel incision rate exceeds the uplift rate the
active deposition shifts to the mid- or distal zones of the fan. Figure 15.26 illustrates how
during limited aggradation conditions the development of fanhead trench in the proximal
zone transfers the active deposition to distal locations (Bull, 1964a, 1977). This process
gives rise to segmented fan surfaces constituted by fan segments of different ages located
at different topographic positions. These different fan segments can display a different
degree of incision, as well as soil development (calcretes, desert pavements, and varnish,
etc.). These segmented fans (Figure 15.27) still have a very controversial origin, and
different hypothesis, based on recent extreme events, and complex responses to climatic
and tectonic changes, have been proposed (see Cooke et al., 1993).
Finally, Blair and McPherson (1994b) consider a set of extrinsic factors related to the
effect of neighbouring environments on fans. In this way, fans can be affected by aeolian
sedimentation, because windblown deposits can modify or even obstruct the fan channels.
The development of sandsheets or dunes in distal fan locations may occasionally promote
channel ponding. The fluvial channels developed in downfan locations can capture
the main fan channel, triggering an abrupt fall in the fan base-level and its subsequent
dissection. If the alluvial fans are bounded by either lacustrine or marine environments,
the subaerial fan flows may transform into fandeltas. In these cases, sea- or lake- level
changes directly affect the fan base-level influencing its erosional or sedimentary
behaviour. Finally, volcanic activity can generate lava flows that can interfere and modify
the internal drainage of fans. The occurrence of ashfall within the drainage basins favours
the generation of debris flows from the steep slopes during storm events.
2.4. Sedimentary processes
The sedimentary processes involved in fan development are of two types, primary and
secondary (Blair and McPherson, 1994a,b). The primary processes, or sediment supply
processes, are those responsible for the transportation of sediment from the drainage basin
to the mountain front locations where alluvial fans develop. The dominant transport
mechanism will be strongly conditioned by the water to sediment ratio in the fan flows, as
well as by the availability of fine, mud-sized materials within the drainage basin (Wells
and Harvey, 1987; Harvey, 1997). A low concentration of fines and high water/sediment
ratios favour the occurrence of fluvial transport of coarse clasts by traction. When the
content of fines increases the clasts can be matrix-supported and mobilized in a debris flow
Figure 15.26. Development of segmented alluvial fan in response to the different uplift rate of the
mountain front. In the upper sketch uplift and aggradation rates are bigger than channel incision. In the
lower sketch channel incision is bigger than uplift. In this case active depositional zone has shifted
downfan as consequence of channel entrenchment an intersection point development giving place to a
telescopic fan (Bull, 1977).
Figure 15.27. Segmentedfan surface. The upper segments are highly eroded by gulling, but also by the
dissection of the Jachal River. The picture was taken from a place located between Rodeos and San Jos6 de
Jachal, San Juan Province, Argentina.
movement by plastic flow, in which the solid and liquid phases move in a cohesive way.
The secondary processes, or reworking processes, lead to fan degradation and erosion
through the remobilisation of the sediments previously deposited by the primary
processes. Many of these processes work during prolonged spans of time on to the fan
surfaces, whereas the sediment supply events are generally triggered by low- frequency
and high- intensity storms.
2.4.1. Sediment supply processes
These may be divided into two classes: fluid-gravity and sediment-gravity flows (Blair and
McPherson, 1994a,b). The fluid-gravity flows can be described as typical fluvial flows,
whereas the sediment-gravity flows are mass movements on the colluvial slopes or on
fragmented bedrock within the drainage basins. These mass movements can be rockfalls
(i.e., gelifraction), rock avalanches (i.e., overturns), gravity slides (i.e., landslides, mud
flows) and debris flows.
The rocky materials coming from the upper parts of the drainage basin can reach the
proximal area of the fans. Rock avalanches take place by the abrupt fall of a large bedrock
mass that can disintegrate into boulders and clasts of diverse size, which move downslope
as a high-velocity granular flow. In the rock avalanche of the Monte Huascar~in (Cordillera
Blanca, Peruvian Andes), that took place in May 1970, a mean speed of 280 km/h was
reached. These granular flows may feed the fans (Nicoletti and Sorriso-Valvo, 1991) and
can reach run-out distances of several kilometres. Their resultant deposits, denoted as
"megabreccias" by some authors, display a characteristic coarsening-upwards bedding
with blocks up to 20 m at the surface, and matrix-rich lower layers. In plan view they
present a lobate shape, as in the case of the rock avalanches of the Sierras Pampeanas
in NW Argentina, studied by Fauque and Strecker (1988). They were 3 to 4 km long,
0.5 to 1.5 km wide, and displayed a front rim of 8 to 10 m high. Inside these lobes older
similar deposits were recognized.
Gravity slides (landslides) correspond to rapid downslope movements of the colluvial
slopes and/or the rocky substrate along a preferential shear surface. These include
rotational and translational slides. In these cases the mobilized material is a coherent and
homogeneous block, opposite to the granular flows involved in rock avalanches. These
processes are important because they directly or indirectly feed the fans. Also, many times
the generated landslide scars work as the nucleation zones for gully or stream channel
generation (Sorriso-Valvo, 1988). As in the case of rock avalanches, seismic shaking can
trigger most of these slope movements, whereas the debris flows are commonly generated
by the progressive addition of water to the colluvial slopes that eventually fail by plastic
flow.
Debris flows are one of the most characteristic deposits of arid and semi-arid zones in
both modem and ancient alluvial records. They constitute flows of sediments formed by
a mixture of coarse clasts embedded in a fine-grained matrix with a variable content of
water and air (Figure 15.28) (Johnson, 1970; Innes, 1983; Costa, 1984, 1988a,b; Johnson
and Rodine, 1984; Takahashi, 1991; Coussot and Meunier, 1996). Their source areas are
normally constituted by the colluvial slopes or high weathered zones of the drainage
basins. In the mountainous deserts, a complementary source area is constituted by
alluvial materials deposited during previous stages of fan activity. The water required to
generate the debris flows usually comes from high intensity storms, but sometimes also
from the rapid thaw or rainfall events on the snow covers located in the upper zones of
the drainage basins. These facts produce rapid overland flows that may generate debris
flows where they incorporate the heterometric fragments and silt/clay-size particles
mantling the hillslopes. Where the flow velocity increases and the internal shear
resistance reduces, they work as a viscous fluid (Johnson, 1970). The viscosity of debris
flows may vary considerably, and when they present a high viscosity and volume, the
channel-floor erosion can give rise to sharp levees along the channel banks. On the other
hand, with a low viscosity, their downslope movement over grassland slopes produces a
very limited erosion and smaller levees (Selby, 1993). All these processes are governed
by a flow of wet sediments moving downslope within the stream channels. The
morphology of debris flows is variable, but generally displays a set of common features.
Their most frequent morphologies are constituted by hummocky surfaces with lobate
Figure 15.28. Deposits faulting affecting to fan debris flow deposits (left) and bedded fluvial deposits
(right). Pleistocene alluvial fans of the Blanco fiver valley. Andean Precordillera, Mendoza Province,
Argentina.
frontal areas (Figure 15.29). In some cases the lateral and central area of the debris flow
lobes terminate in a frontal talus (Johnson and Rodine, 1984). Debris flows may exert a
relevant impact on the main fan channel, even overflowing it, and leaving deposits on
both banks. The internal structure is massive with large boulders randomly distributed
and embedded in a fine-grained matrix, but locally they may exhibit thin sand or silt
lenses deposited by intervening "fluvial" flows. These bedded deposits, denoted as
transitional deposits, can be generated during transitional stages between debris to fluvial
flows, apparently at the end of debris flow events (Wells and Harvey, 1987). These debris
flows are more frequent in the proximal areas of alluvial fans. Mud flows are mainly
constituted by silt-clay materials with occasional clasts. These clasts can display
imbricate fabrics parallel to the depositional flow, especially when clasts are elongated.
Mud flows, however, are rare deposits in the alluvial fans developed in arid and semi-
arid regions.
Two mechanisms have been traditionally considered for the origin of debris flows. The
more accepted idea considers the generation of debris flows from the downslope failure of
steep colluvial slopes (27 to 56 ~ triggered by the progressive wetting and saturation of the
ground during rainfall or snow-melt events. The water saturation of loose detrital materials
promotes the reduction of their shear stress assists their downslope movement as debris
flows (Campbell, 1975; Costa, 1988a,b). Other proposals suggest that debris flows are
generated by the mobilization of the loose materials mantling the hillslopes by single
Figure 15.29. Idealized sketch of a debris flow lobe, showing the surface debris waves (Johnson and
Rodine, 1984).
overland flows (Johnson, 1970; Johnson and Rodine, 1984). These authors denote this type
of starting process, as the "hosepipe effect." Whatever the case, the production of debris
flows needs the occurrence offlash floods originated by high intensity storm events. These
energetic rainfall events are unusual, and the recurrence interval for debris flow generation
has been estimated to range from 300 to 10,000 years (Costa, 1988a,b), but locally return
periods may be shorter.
In fluvial processes sediments and water remain as separate phases during transport,
and the detrital particles are mobilized by means of turbulent flows as bed or suspended
load. When the solid load constitutes an important fraction (20 to 47%) hyper-
concentrated flows are generated (Costa, 1988a,b). Two types of fluvial processes can be
differentiated; sheet floods and incised channel flows (Blair and McPherson, 1994a,b).
Sheet floods are the most common process in the sediment input of alluvial fans. These
are broad unconfined flows that expand downslope over the piedmont surface. From aerial
photo analysis Blair (1987) described the catastrophic sheet flood event occurred in the
Roaring River fan (Rocky Mountain National Park, USA), as a continuous sheet of water
of 0.5 m depth flowing over the 320-mlong, active depositional lobe of this fan. Sheet
flooding is produced by strong storms events of low frequency and high intensity that can
even promote catastrophic flooding (Guti6rrez et al., 1998). These sporadic events are
separated by periods of rare fan activity, during which the fan surface is reworked by
secondary processes. The deposits can be constituted by a varied suite of blocks, gravels,
sands, and silts. These usually are arranged in interbedded laminar couplets 10 to 30 cm
thick, with depositional slopes of 2 to 8~ parallel to the fan surface (Blair and McPherson,
1994a). The longitudinal axes of the bigger blocks may have an orientation perpendicular
to the slope, but sometimes they display imbricated fabrics. Wavy gravel ribbons
transverse (transverse fibs) to the fan slope have been also described as products of sheet
flooding.
The incised channels dissect the fan apex and can be several metres deep. They
constitute paths for the transfer of sediments between the proximal and distal zones of
the fan, and consequently the feeding channels of the active depositional lobes. These
channels also contribute to fan sediment reworking by means of erosive work on their
incised channel beds. The resulting sedimentary facies are characteristic of confined flows
including fluvial sand and gravels mobilized during flooding events.
A distinctive sedimentation variety of alluvial fans are the so-called sieve lobes or sieve
deposits which were described by (Hooke, 1967) after laboratory studies with sand and
granule features resembling fans. Characteristic lobate deposits build up from the points
where the water flow loses its transport capacity due to its rapid infiltration into the
permeable fan substrate, commonly in the mid-fan intersection points. The coarsest
material is accumulated in the frontal sector of the depositional lobe, but upslope finer
sand-sized sediments prevail. Sieve deposits commonly develop during flooding events
and are mainly deposited in mid- and distal fan zones. They can also originate from
previously deposited debris flow lobes where overland flows sieve the finer material.
Commonly the result is a matrix-free, clast-supported fabric. In this last case, the origin of
the sieve lobes is secondary.
The three-dimensional analysis of alluvial fans provides evidence for the occurrence
of numerous and complex graded bedding features, which can be interpreted as climatic
or tectonic responses. Typically particle size decreases downslope with the coarser
material in the proximal areas and the finer material in the distal ones. The channelized
fluvial blocks and gravels, and debris flow deposits prevail in the proximal areas
(Figure 15.30c,d) (Harvey, 1997). According to Miall (1978) if the dominant sedimentary
facies are fluvial ones we have "Scott"- type sedimentary sequences. On the other hand,
the profusion of debris flow deposits distinguishes the so-called "Trollheim-type"
sequences (Figure 15.30a,b). In mid-fan areas sieve deposits can be developed, but sheet-
flood deposits commonly extend towards the distal parts of the fans. On the other hand,
strong changes in the distribution of sedimentary facies can be occur between untrenched
and proximally trenched fans; in the latter the more recent deposits are shifted towards the
distal areas of the fans (Figure 15.30e,f).
2.4.2. Reworking processes
As previously mentioned, alluvial fans mainly develop in response to high intensity

storm events, but the return period of these energetic events is important. In consequence,
the fan surface is exposed to a variety of reworking processes during the long periods
of sedimentary inactivity. During these periods the depositional surfaces created by the
primary sedimentary processes are largely modified. In addition, in active tectonic areas,
fan surfaces can be affected by ductile and brittle deformations, such as ground folding,
faulting (Figure 15.30f) and warping. These reworking processes are diverse and they
have been analysed by different authors (see Blair and McPherson, 1994b).
The most relevant reworking processes are the overland flows derived from ordinary
rainfall episodes that dominate the long periods of time between strong storm events.
Figure 15.30. Variation of sedimentary facies in alluvial fans: Vertical facies variations: (a) "Trollheim"
type and (b) "Scott"- type sedimentary sequences (Miall, 1978). Proximal-distal facies relationships:
(c) Schematic model (Rust, 1979). (d) Downfan variations in proportion offacies exposed in fan section on
Ceporro fan, southeast Spain (Harvey, 1984b). Spatial facies variations: (e) Apex area of a proximally
aggrading fan, Trollheim fan, Deep Springs valley, California (Hooke, 1967). (f) Proximally entrenched
and distally aggrading fan, Hanaupah Canyon fan, Death Valley, California (Hunt and Mabey, 1966).
(After Harvey, 1997).
These processes result in the winnowing and transference of the existing fine material
towards the distal fan areas. Erosive flows may result in the generation of a net of rills
and distributary channels radiating downslope from the fan apex to distal areas where
wash channels dominate (Denny, 1967). It is also necessary to consider the activity of
the subsurface groundwater within the fan bodies, which work as important aquifers.
Groundwater availability impacts plant growth rates and, consequently on subsequent
processes of overland flow erosion. On the other hand, water infiltration can favour the
calcium carbonate cementation of fan surfaces. In rich-carbonate semi-arid zones these

processes can give rise to the occurrence of petrocalcic soils.
Wind action is an important reworking process in arid regions. Windflow produces the
pick-up and winnowing of fine particles from the fan surface that can eventually give
place to the development of desert pavements. The exposed gravels undergo aeolian
abrasion resulting in the generation of ventifacts sometimes coated with desert varnish.
The sand-sized particles transported by windflows can also be deposited on to the fan
surface giving place to continuous sand-sheets or generate dune- type accumulations
around obstacles, such as plants. These aeolian deposits can attain an important thickness
in favourable "leeside-slopes" of the fans giving place even to relevant disturbances on
fan channels.
The weathering of fan surfaces includes the action of mechanical, chemical and
biological processes. Mechanical weathering, such as salt crystal growth, mineral
hydration-dissecation cycles and gelifraction, disintegrate and fragment the clasts and
they are conditioned by the abundance in water and their salt content. Biological
weathering, or bioturbation, is generated by the burrowing and material removal produced
by the activity of plants and fauna. Calcium carbonate precipitation near the fan surface
produces the development of petrocalcic soils (Gile and Hawley, 1966; Machette, 1985;
Harvey, 1987a). The cementation of fan sediments generates resistant calcrete crusts,
which prevent the erosion and reduce the infiltration of water on fan surfaces (Harvey,
1990, 1997).
2.5. Dynamics of alluvial-fan development
The alluvial fans are the landforms of faster development rates within arid areas
(Mabbutt, 1977). Average accumulation rates in alluvial fans of the Death Valley (USA)
are of 1 m every 1000 years (Hooke, 1972). Alluvial fans are dynamic geomorphic
systems showing progressive morphologic changes during their development (Harvey,
1997). This author proposed a model for fan evolution in which the relationships between
the fan- surface and fan- channel profiles control the balance of aggradation an incision
in mid and distal fan areas (Harvey, 1987a, 1997) (Figure 15.31). Types A and B are
characterized by the occurrence of fanhead entrenchment and distal aggradation from
the mid-fan intersection point. In Type C, scour rather than deposition takes place at
the intersection point, and in Types D and E this scour is sufficiently effective to create a
series of headcuts, which may shift to distal areas as a consequence of a progressive
downfan migration of successive intersection points. These successive headcuts can
eventually link together as a mid-distal fan trench, dissecting the fan throughout its length
(Type F). Fan dissection can also occur by base level-induced headward erosion from
the fan-toe (Type G). This can be considered a key model for the understanding of fan
aggradation and dissection dynamics, in which tectonic disturbances, climatic changes,
human activity, and extreme storm events play a relevant role in the modification of the
dominant erosive or depositional processes.
There are other models of fan development described in the classic literature, such as
that proposed by Davis (1905) based on their cyclical conception of landscape evolution.
This author, proposed a juvenile fan stage in which the higher relief continuously provides
AGGRADING FAN
'AMERICAN'TYPE
LIMITED DISTAL
AGGRADATION
c
INTERSECTION POINT
D SCOUR
~ , ~ INTERSECTIONPOINT
E ~ EADCUT DEVELOPMENT
DISTALTRENCHING
F
TOTAL DISSECTION
BASE-LEVEL INDUCED
DISTAL DISSECTION
~ ' ~ Fan surface
Channel profile
Intersection point
Figure 15.31. Alluvial-fan development model based on the relationships between fan- surface and fan-
channel longitudinal profiles and their impact on the dissectional or aggradational behaviour of the fan
(Harvey, 1987a).
sediments to the piedmont. The progressive relief reduction gives rise to a dominance of
sheet flooding and laminar erosion during the final stages of the cycle, where pediments
are generated.
Sorriso-Valvo (1988) proposed an evolutionary model for fans based on mass
movement occurrence in the Calabria region of Italy. This author proposed four main
stages for fan development (Figure 15.32). The first stage, labelled as creep-stage, is
characterized by shallow landslides located downslope, whereas upslope creep processes
are dominant during long periods of time. In stage B, or collapse-stage, landslides and
debris avalanches dominate upslope where important gravitational scars are produced,
whereas in the piedmont the collapsed material is accumulated as colluvial cones. The
collapsed material can be removed throughout the main channel during high intensity and
low frequency rainfall events. Both the weakened bedrock and the debris remaining in
the scar, are mobilized by subsequent mass movements (reactivation of the previous
Figure 15.32. Cyclic evolution of an hypothetical landslide-related fan. Arrows indicate the possible
successive stages. The shape is typical of bedrock with ductile behaviour. In brittle and jointed bedrocks
the main flow stage B may take the form of a high-speed debris avalanche. (Sorriso-Valvo, 1988).
ones) and/or debris flows. These processes can build a fan in a relatively short term.
This stage C corresponds to the alluvial fan construction. Finally, the fan destruction, or
stage D, takes place when the slided material has been removed and stream flowing from
the scar attain a higher erosional power generating the erosion of the fan, which progress
from the apex. For this author this model is cyclic and considers the succession of stages
A-B-C-D-B-C-D... as the most common evolutionary trend on fan evolution.
cobbly, pebbly, or sandy sheetfloods or

clast-poor debris flows
/
\co,,uv,u . . . .
Precursor Stage Stage 1 Stages 2 & 3
Figure 15.33. Schematic diagram of the increasing slopes during the different evolutionary stages of fan
development (vertical exaggeration x 2). The different stages present various associations of dominant
sedimentary processes. During fan development depositional slopes diminish, fan radii lengthen, and
knick-points decrease in elevation (Blair and McPherson, 1994a,b).
However, the model of Sorriso-Valvo (1988) basically considers the contribution of mass
movements to the generation of alluvial fans, but undervalue the contribution of true
fluvial processes.
Blair and McPherson (1994a,b) proposed a similar model with a four-
stage evolutionary scenario (Figure 15.33), initiated by mass movements and
development of talus cones at mountain front locations (Figure 15.34). In stage 1 the
incipient drainage basins hold an abrupt relief, where first order channels prevail and
landslides and debris flows are the dominant processes. During stage 2 the gravitational
mobilization of these deposits buildsup the alluvial fan and enlarge the drainage basin,
which acquires an amphitheater shape. The area of the fan is also increased during this
stage and fan radial length may reach values of 1 to 5 km. On the other hand, fan slope
reduces during this stage displaying characteristic values between 5 and 15~ During
stage 3 the expansion of the fan system continues, whereas in the drainage area the
number of first-order channels increases and 4 or 5 channel orders develop. Mass
movements are still characteristic within the drainage basin, but fan sedimentation is
driven by debris flow and sheet-flood processes (Figure 15.35). In this stage the fan
Figure 15.34. Debriscones in Punta Vacas, Mendoza Province, Argentina. Andean Cordillera.
Figure 15.35. Alluvial fan and main fan channel (right). Debris cone with incipient drainage basin
modelled in amphitheatre shape. East of Punta Vacas, Mendoza Province, Argentina. Andean Cordillera.
radial length may enlarge from 3 to 10 km and the fan slope diminishes to 2 to 8 ~
Finally, this last stage can be maintained during long periods of time until rivers or
glaciers will eventually dissect the fans, nevertheless the fan system can be reactivated
by tectonic activity.
Chapter 16
Desert lakes: playa lakes and sebkhas
1. Main features
Large areas of desert regions are characterized by endorheic drainage and, consequently,
waters collect in the lower topographic areas creating lakes of different sizes and origins.
Lakes can last if water input (surface run-off, ground waters, rivers) is higher than output
(mostly evaporation). If solute discharge increases, the lake becomes saline. Lakes from
arid zones can dry out during long periods of time, although in most cases, they show a
seasonal regime, drying out during the summer season. Only the Aral, Caspian, and Dead
Sea (Figure 16.1) are permanent water bodies, as their main water input is derived from
nearby mountains or far-away areas. The Dead Sea, fed by the Jordan River, outlet of the
Sea of Galilee, has a very high salinity (300,000ppm TDS), which impedes the
development of life in its waters. Saline lakes with a salinity of about 5000 mg/1 are
relatively common in arid and semi-arid areas (Shaw and Thomas, 1997). This figure,
mainly based on biological tolerance, is considered the boundary between fresh and saline
water bodies (Euster and Hardie, 1978).
There is a very wide terminology referring to closed depressions developed in arid
environments, largely due to the use of vernacular names (Tricart, 1969; Cooke and
Warren, 1973; Neal, 1975; Cooke et al., 1993; Currey, 1994; Rosen, 1994; Shaw
and Thomas, 1997; Briere, 2000; Guti6rrez, 2001). The term playa has a Spanish origin
and was introduced in the English language from the exploration of the North American
southwest by the Spaniards (Guti6rrez, 2001; Guti~rrez et al., 2002) (Figure 16.2).
Recently, Briere (2000) presented a thorough compilation of names used to designate
these depressions. This author proposes the use of the term playa for intracontinental
depressions with a negative water balance that remain dry more than 75% of the year,
and playa lake, a transition between playa and lake, for depressions that host water from
25 to 75% of the year. In many cases these definitions may be difficult to apply since the
hydrologic regime of the playas may change substantially in successive years.
Tchakerian (1999a), in the Encyclopaedia of the Deserts, defined playas as depressions
typical of arid and semi-arid zones, affected by flooding by ephemeral surficial waters or
fluctuations in the water table. The definition of pan given by Dregne (1999) is very
similar to the previous one. Several authors consider both terms as synonyms (Shaw and
Thomas, 1997; Goudie, 2004b). On the other hand, the word sabkha has clear geographic
meaning restricted to depressions located in coastal plains (Tchakerian, 1999b; Briere,
2OOO).
Other synonyms are sebkha, an Arabic term used in northern Africa and sabkha,
common in the Arabian Peninsula (Figure 16.3). In Iran, kavir denotes depressions rich
in salts. Another Arabic word, chott, means a saline basin with groundwater input, and is
Figure 16.1. The northerncentral sector of the Dead Sea from the western margin, Israel.
mainly used in the Sahara region (Chott E1 Jerid, Tunisia). Takir is a clay-rich depression
in Central Asia, somewhat similar to the barreal in the Argentinean Andes (Figure 16.4).
In the South American deserts, salar, salina, and bols6n are used for depressions with
clay-rich or saline sedimentary deposits.
Figure 16.2. Saline closed depression with old salt exploitation facilities in the foreground, Laguna La
Playa, Central Sector of the Ebro Depression, Bujaraloz, Zaragoza Province, Spain.
Desert lakes: playa lakes and sebkhas 385
Figure 16.4. Barreal Blanco (whitish lacustrine mud) composed of carbonate-rich silts transported from
the calcareous Cambrian-Ordovician formation outcrops in the watershed, Andean Precordillera, San
Juan Province, Argentina.
Figure 16.5. Large halite accumulations in the central areas of the littoral sebkha, E1 Melah, Zarzis,
Tunisia.
Glennie (1970) distinguished between interior and littoral sebkhas (Figure 16.5). The
latter are saline plains located in littoral arid areas, affected by marine water penetration
and with tidal-related deposits. High-soluble salts are commonly precipitated in these
littoral sebkhas because they are located in areas with high evaporation rates. Some
modern examples occur in the Persian Gulf and south of Tunisia.
Most of the closed desert depressions are small. The largest one is the Salar de Uyuni
located in the Bolivian Altiplano at 3653 m a.s.1 (Figure 10.9 and Figure 16.6) with a
surface area up to 9000 k i n 2 (Rettig et al., 1980). During the Pleistocene period, the Salar
was a lake (Minchin Lake) that reached more than 43,000 k i n 2 (Blair, 1986). Lacustrine
depressions represent less than 1% of the total desert area (Shaw and Thomas, 1997),
although in some regions of South Africa they cover up to 20% of the surface (Goudie
and Wells, 1995). In NW Texas there are between 30,000 and 37,000 lacustrine basins
(Osterkamp and Wood, 1987). With only the exception of tectonic lakes, most basins
located in arid zones do not exceed 15 m water depth at the maximum lake level (Mabbutt,
1977). Tectonic lakes are commonly elongated, parallel to the regional tectonic faults and
structures. Deflation, or wind-erosion, can also originate elongated lakes following the
direction of the prevalent winds. In most cases, however, closed desert depressions are
approximately circular in shape. If water and sediment input is low, the bottom of the lake
is usually irregular. With increasing clastic deposition, small topographic bottom
irregularities disappear. Waves, currents, and other dynamic agents produce bars, barrier
islands, spits, and other geomorphic features in the saline lakes, as in the Laguna de
Gallocanta in Zaragoza and Teruel provinces in Spain (Figure 16.7, Gracia, 1995).
Figure 16.7. Depositional littoral morphologies during the dry season (September 1996) in the Laguna de
Gallocanta, Zaragoza and Teruel provinces, Spain, Central Iberian Range. Photo: B. Ler~inoz.
From an ecological point of view, lacustrine habitats are limited to water bodies deeper
than 2 m; water bodies less than 2 m deep are considered as palustrine or paludal (swamp)
habitats (Cowardin et al., 1979, in Currey (1994)). Lacustrine environments contain
numerous plant- (blue green algae and reds) and animal-salinity-tolerant species (Eugster
and Hardie, 1978). Lakes with high salinity are mined for chlorides, nitrates, sodium
carbonates, borax, and other minerals of lithium and potassium. Those lakes with large flat
areas can be used as airports, or for sport and military activities. Archaeological sites have
been found in the littoral areas of some saline lakes where ancient people settled during
pluvial periods (Demangeot, 1981).
2. Origin of closed depressions in arid environments
Desert lakes have a variety of origins. Hutchinson (1957) distinguished two main types
(structural and geomorphic) and enumerated 13 processes that led to 75 different types of
closed depressions. Mabbutt (1977) identified four main lake basins: tectonic, erosive,
depositional, and crater lakes (volcanic and meteoritic). Each one of these main types is
subdivided into more categories. Shaw and Thomas (1997) followed this classification and
added some new cases.
Tectonic basins originate as a result of fault-bounded depressions or large sag basins.
Examples in the East African Rift Valley system are the Dead Sea in Israel and Jordan,
Lake Turkana or Rudolph in Kenya, and Lake Natron in Tanzania. Lake Victoria and Lake
Chad in Africa, and Lake Eyre in Australia are related to large, tectonic intracontinental
sag basins. Examples of landscapes composed of mountains and depressions due to
tectonics are the Basin and Range province in the western US deserts (Figure 16.8) and in
8 0ri i Gr voon0s n L kec y "Paya0eoosts

Figure 16.8. Blockdiagram of a tectonic basin with playa, typical of the southwest United States (from
Motts (1970) in Mabbutt (1977), Figure 55).
Figure 16.9. Map showing depressions, sand seas, and lines of dunes in the Egyptian Desert (Holmes,
1944).
the Argentinean Andean Pre-cordillera. Small depressions fed by ground waters typically
occur along strike-slip faults.
Erosive basins in arid regions are essentially originated by deflation and wind erosion
and they are very common in most world deserts. The largest deflation basin is the Qattara
Depression (Figure 16.9) in NW Egypt with a surface area of 18,000 klTl 2 and the lowest
point located 134 m below sea level. The sediments eroded form a large dune system
located SW of this depression (Glennie, 1970). Erosive basins of this size are rare, but
basins of hundreds of metres of diameter are quite common, and excellent examples occur
in the central areas of the Duero River Basin of Spain (Guti6rrez et. al., 2005). Absence of
vegetation is a requirement for formation of deflation basins.
Evolution of deflation basins can be halted if particle size of the exposed beds is too
coarse to be transported by the wind (Tricart, 1969). A similar effect is caused by a thick salt
layer. A groundwater table close to the surface also limits the depth of deflation (Shaw and
Thomas, 1997). Deflation can also form sebkhas between the dunes (Figure 16.3). The
hamadas, large plains in the Sahara and India, contain some closed shallow depressions,
denoted as dayas, that have been interpreted as a result of mechanical disaggregation and
wind erosion (Tricart, 1969), or karstic activity according to other authors (see Shaw and
Thomas, 1997, Guti6rrez et. al., 2002). The deflated sediment particles may be deposited
along the edge of the closed depressions and form aeolian build-ups shaped as half-moons
(lunette (Hills, 1940) in English literature and bourrelet in French literature) (Figure 16.10).
These aeolian deposits are composed of clay to sand sized particles. They occur in areas
with annual rainfall between 100 and 700 mm. They may be a few kilometres long and
,,>-
9 )"" i
n~
] .
4-
#
,..'," I t~ >
, .s:"
x
\ o.,
.... .~_,, IN .~
.,. .................. ?,':
", .... ,"" ;.h". . . . . . . . . . . . "N
", ......... ,.Q''-,l~
. . . . ,., _/..-, .......

.,'+' . . . . . /. '"""7''" .................. " " .............
~.+,i; . . . . . . ~ , , , , . . .~
.,, i
...... - ....
.r" f,.,.i
o,+ '~
.~, ........... k 9 ' ' " ' " " " " I"%
! .... j, ) ,i~ t
.................~.......(~
. (i I
: . ,,, f ,.. ~.~... 9 ~-,

o |
~9 ~. ,","
) , .... -
: ,.,, ,,
oo ~
o.o ~
Ni~II 1 4
/
2 ~-----~ 5 .1
~ 7
0
Figure 16.10. E1 Kourzia sebkha and its associated lunette. 1 - Small cliff surrounding the sebkha's
edge. 2 - Carbonate-rich silt area. 3 - Alluvial fans from the main tributaries. 4 - Area of
micropolyhedra. 5 - Carbonate-rich silt area with thin gypsum lenses. 6 - Area of black mud with summer
halite crust. 7 - Lunette of el Kourzia Sebkha. In black: older lunette (modified from Perthuisot and
Jauzein (1975)).
more than 6 0 m height. L a w s o n and T h o m a s (2002) suggested that g r o u n d w a t e r

fluctuations may be important in controlling sediment supply to lunettes in South Africa.
Within this category of erosive basins, we also include the closed depressions caused by
karstic dissolution of surface and subsurface rocks, particularly evaporites. One e x a m p l e is
the Bujaral6z region, in the central Ebro Basin of Spain, which contains a large n u m b e r
of karstic depressions, seasonally flooded with water (Figure 16.11 and Figure 16.2).
t~
~
Figure 16.11. G e o m o r p h o l o g i c a l m a p o f the B u j a r a l o z salt lakes. 1 - Structural l a n d f o r m s on g y p s u m . 2 - P l a t f o r m s and c u e s t a s on l i m e s t o n e s . 3 - Karst

d e p r e s s i o n s and d o l i n e s . 4 - Salt lakes. 5 - A r e a s o f t e m p o r a l s w a m p . 6 - R e c e n t g l a c i s and s l o p e d e p o s i t s . 7 - Infilled v a l l e y s . 8 - G u l l i e s . 9 - T o p o g r a p h i c
point. 10 - D i p .
3 Nov 1949 Sept 1950 I Aug 1951
1 Nov 1951 20 Nov 1951 7 Apr 1952
(b) m
1II 1 I I
]PP
-30 ~ .... 1 I, I I ' i . 20.000 .-,
!
c,D
iSalJnaty (%0) o
-32 " ~ 15.000 e,~
\. I1[
E
i,,,.
-34 . . . . . . . . .
a 10.0o0
- -36 -
' '11:"
li 132 - 5.000 m
0
~3 E
-38 . . . . . . -,~ . . ~
o
2.000
m
0
>
-40 - 1
~ 320
. . . . . . SALT_ ......... LEVEL ~.o~_._ _
-42
S O N D
l J F M AMjJA
III l S 0 N D J F M A
1950 1951 1952
8000 7000 3000 2000 1000
L l 1 l |
Area of water (km 2)
Figure 16.12. Drying cycle of Lake Eyre North, Australia, 1949 to 1952. (a) Extent of surface water.
(b) Relationship between level, volume, area and salinity of water (from Boython and Mason, 1953, in
Mabbutt (1977), Figure 54).
The largest one is called La Playa, and it is more than 4 km in length. The dissolution of
the underlying Miocene formations composed of alternating beds of limestone and
gypsum is responsible for the genesis of many small depressions. The basins are elongated
with the main axis following a NW direction, which corresponds to the prevailing
winds ("cierzo") and the main tectonic lineaments and fractures (Quirantes, 1965; Pueyo,
1978 to 1979, Sancho and Guti6rrez, 1993).
Piping is another process leading to the appearance of closed depressions in semi-arid
regions (Figure 13.23). In Lupifien in the Huesca Province of Spain in the central sector of
the Ebro Depression, piping occurs along the contact between the loose, very-porous, thick
(up to 15 m) slope deposits and the Miocene Formations. The enlargement of the tunnels
and pipes decreases the mechanical resistance of the ceilings and the cavities finally
collapse. On the surface, the collapsed areas are indicated as pseudolines up to 35 m wide
and 20 m deep and with funnel or cylinder shapes (Figure 13.22 and Figure 13.23)
(Guti~rrez et al., 1988).
Animal activity provides some examples of erosive basins. Erosive processes may be
accelerated because of increased grazing by large herbivores and by soil compaction
caused by large herds. Elephants excavate with their legs and trunks closed depressions
up to 23 m in diameter and 1.5 m deep (Thomas, 1988).
Sedimentary depressions can be formed after a drainage system is abandoned or
changed by another fluvial network. Maizels (1987) described a complex series of
paleochannels in Oman that during the Plio-Pleistocene alternation of humid and dry
periods caused closed depressions. Also in floodplains some lacustrine depressions can be
formed. Dunes can impound creeks and rivers, as it happens south and east of Timbuktu in
Mali. Lava flows may produce the same effects. Littoral evolution causes the appearance
of beaches and bars, and the formation of littoral sebkhas (Glennie, 1970). Some of the
albuferas in the Spanish Mediterranean coast, such as the Pego and Elche Albuferas, have
the same origin (Sanjaume, 1985).
Finally, volcanic activity builds craters that could be partially filled up with water as
with Zuni Salt Lake in New Mexico. Meteoritic impacts in arid zones form impact craters;
Meteor Crater in Arizona is the most outstanding example of this type of closed
depression.
3. General features of hydrology of lakes from arid zones
Lake level fluctuations depend upon water input and output balance. Inputs are surface
run-off, groundwater, and secondarily, direct rainfall over the lake. Outputs are
evaporation, infiltration and, when the lake basin overflows, the flow of the outlet.
When input and output are not equal, the lake level fluctuates and shorelines retreat and
advance. In arid regions these fluctuations are very frequent, except in very deep lakes
such as the East Africa rift valley lakes. Desert climates are characterized by scarce
rainfall and intense evaporation, and consequently, lakes dry out very often.
Lake water volume depends on the relative altitude of the watershed boundaries in
relationship to the lake bottom and the lake bathymetry. With time, watershed highs are
eroded and bottom irregularities are covered with sediments, and, consequently, water
storage capacity of the basins decreases. Height and width can be modified by aeolian
deflation, which usually tends to widen and deepen the closed depression. The shape and
the storage capacity of the lake basin can be modified by tectonics in active regions.
Volcanism plays a similar role. This is the case of the numerous volcanoes located in the
East Africa rift valley, which in more active periods altered the morphology of the lakes
located in these large tectonic structures.
Some complex lake basins are composed of connected sub-basins, and water, solutes
and sediments are transferred and mixed from one sub-basin to another. Sediment and
solute transport occurs along the surface channels or during flooding episodes when one
sub-basin overflows (Currey, 1994).
Lake level decrease in desert lakes is mostly controlled by evaporation which can easily
account for 2 to 3 m loss per year. Near Timbuktu in Mali, 2.5 to 3 m of water evaporated
from the diffluent Niger lakes was recorded during an 8-month period (Tricart, 1969). As
precipitation in arid zones is very irregular, after a long drying-out period, intense rains
may flood the lakes for several years. These large flooding episodes are characteristic of
large basins, because small basins commonly dry out in less than a year.
Lake Eyre in the Simpson Desert of Australia has a surface area of 8000 km 2, a
watershed about 100 times larger, and an annual rainfall between 100 and 500 mm
(Bonython and Mason, 1953). During the mid-20th century, the lake was flooded after a
200 to 400 mm rainfall in the southern area of the lake during February and March of 1949
(Figure 16.12a). After more intense rainfall during the next year, particularly about
500 mm during the January to April period, the lake filled up in September 1950. The lake
level desiccation phase started the following month, with a salinity increase and lake level,
water volume, and lake surface area decreasing (Figure 16.12b). In 1953 the lake had
completely dried out. The average salt composition is NaC1 (96.14%) and CaSO4 (2.38%).
Lake Eyre fills up completely about 12 times per century.
From a geomorphological point of view, lacustrine processes are more active in the
shallow, littoral environments, where the water column is mixed by turbulent flow, and
low-energy waves are active transport agents towards the coast and along the littoral zone.
Beaches, bars, and other morphologies are common littoral features.
Detailed studies of littoral morphology and evolution, and analyses of stratigraphic
records have allowed reconstructions of raising and lowering lake levels of many
lacustrine systems during the Pleistocene and Holocene ages. These reconstructions are
the hydrograph of the lake during those periods (Figure 16.13).
4. Sedimentation in desert lakes
Sediment deposition in closed depressions located in arid zones is mostly controlled by the
hydrological regime. In dry playa lakes only affected by flooding, sediments are mostly
silts and clays transported by the water inflows. During higher energy floods, sand can be
transported into the playa and deposited in small channels. This kind of clastic
sedimentation occurs mostly along the margins of the basin; alluvial fans may also develop
in these areas. Waters reaching the central areas of the basins are subjected to strong
evaporation, and consequently, salinity increases and salts precipitate after chemical
saturation is reached. In wet playa lakes, saline ground waters are dominant and sediment
deposits comprise salt-imbedded clastics and evaporites (Mabbutt, 1977).
1295 Danger ,, J
I "' I Cave I I I I I I I I I I
t _ . Post-Bonneville (Younger Dryas)
hstand of Great Salt Lake
m
1292 j-
m
1289
oo~
..~176176 9~
~176176176
I ? I.:.:.:.:.:.:.:.:.:.:.:,:.:.:.:.:.:.:.:,:. Holocene hlghstand ..~ _ r~
1286
E
c:
. m
1283
~~176176176176176176
1280
.,.,.....,.....,...............,,.........,.....,.......,,., ,................... :
1277
, i: i i: i: !: i~~i!!::!: i~i: i: !: i: ~!::!: !: i: !: i~i:.i?::i!~:i:!i:!i:!:i!::i!?~:i!!i?:!i:!i:!i:!i:!~i:!:~!:~.i:!:!:!:!:i:!:i:!:i!:i:!:!:?:!:!:!:!~:!:!:!: : .:i: :!:;~:!:!:!:?:i:?:
13 12 11 10 9 8 7 6 5 4 3 2 1 0
Age in years BP X 1000
Figure 16.13. Hydrograph of the Great Salt Lake Basin, Utah during the last 13,000 years (Murchinson, 1989, in Currey (1994)). First records of human activity
have been dated between 10 and 11 ka in Danger Cave.
In mountain deserts and tectonic basins, playa lake models are commonly characterized
by a topographically high margin, which corresponds to the raised block (Figure 16.8). At
the foothills, alluvial fans develop, with coarse gravel and sand proximal facies grading
into silty and clayey distal facies, commonly associated with salts. Finally, in the
topographic lows, a shallow lake develops and evaporites form (Eugster and Hardie, 1975;
Hardie et al., 1978; Anad6n, 1989). Most of these depressions are located in tectonically-
active areas, with high subsidence rates, and consequently, thick sediment series
accumulate. Many examples occur in the western United States deserts.
Ephemeral lakes located in platform and shield deserts commonly have very low
topographic gradients and thin sediment sequences composed of sand, silt, and clay with
some evaporites. In many cases, the rocky substrate crops out in the middle of the
depressions. These lakes show large surface area expansions during the recharge periods.
During dry periods, aeolian activity is the dominant process and, if the basin is large
enough, dune systems can develop in different areas, as has happened in Lake Chad
(Central Africa).
4.1. Clayey closed depressions
These are flat areas dominated by clay sediments with minor amounts of sands and silts.
They occur at the bottom of the lake or surrounding the inner evaporite-dominated areas of
the lake (Figure 16.8). Finer particles remain in suspension and deposit more slowly,
whereas coarser silt and clay particles only reach the inner part of the depression during
flooding episodes. Surface waters have commonly low solute concentrations and dissolved
solutes are transported into the saline core of the lake or are mixed with the finer clay
fraction. Desiccation of the muddy sediments form polygon systems characterized by
concave cells if they lack salts, or with convex cells if salts precipitate with the fine clastic
sediments.
An interesting example of formation small clayey closed depressions has been described
by Osterkamp and Wood (1987) in the Texas and New Mexico High Plains. First, small
depressions form by deflation and are partially filled with water. Waters percolate and
slowly dissolve underlying calcrete horizons, which causes subsidence and a widening of
the depressions. These small basins are filled with fine clastic material without salts.
4.2. Saline lakes
Saline lakes are natural laboratories where many active processes can be studied. They
also provide useful modern analogue environments to interpret fossil sedimentary records
(Reeves, 1968). From a geological point of view, they have been intensively studied
during the last two decades, mostly because of their economic significance through their
association with valuable oil and evaporite minerals. They also contain a detailed archive
of past climatic, hydrological, sedimentological, geochemical, and biochemical conditions
(Eugster and Kelts, 1983).
One of the first problems to solve in saline lakes studies is the origin of the salts. In the
Senegal Delta, the marine spray caused by waves is the main salt source for some of the
sebkhas. Sprayed waters evaporate and small salt crystals precipitate and remain floating
in suspension forming a whitish fog in the littoral areas. Winds transport this fog inland
and the salt crystals are finally deposited during rainy events. Daily temperature changes
provide another mechanism for crystal deposition. At night, increased relative humidity
causes water molecules to adhere to the hygroscopic salt crystals, increasing their size and
forcing them to settle. During stronger rain events, the salt crystals disperse in the surface
dissolve and are transported to the interior of the sebkhas. Finally, waters evaporate and
salts are precipitated (Tricart, 1956). Marine spray is a very important solute source in
littoral arid regions with strong wave action. The amount of salt of marine origin rapidly
decreases at distances farther away than 100 km from the coast (Mabbutt, 1977).
Another significant solute source is chemical weathering. One example of such types of
chemical reactions is the dissolution of salt-beating rocks, such as the Triassic and Tertiary
formations of some Spanish basins. Salt recycling has been common through geological
time, as it has been shown using isotopic signatures of some of the elements. The final
water compositions after chemical weathering reactions take place depends on the existing
rock minerals, their relative abundance, the degree of weathering, the type of chemical
reactions, and the specific drainage conditions (Eugster and Hardie, 1978).
Finally, another salt source is saline crusts from beaches and littoral sebkhas subjected
to deflation. Wind transport salt mineral particles considerable distances.
Chemical composition and concentration of lake waters from desert lakes are very
variable. The actual mineral composition of the precipitated salt depends on the brine
composition, but other factors are also important. Such factors include for example, brine
concentration, gas diffusion, particularly CO2, mixing of water of different composition,
and temperature changes. The interplay of all these processes on different brine types is
responsible for the precipitation of different mineral assemblages (Table 16.1) (Eugster
and Hardie, 1978).
With increasing chemical concentration in water, the precipitation sequence commonly
starts with Ca-carbonates, and is followed by Ca-Mg-carbonates with increasing Mg 2+
content. Sulphates are the next to precipitate, and as a result, the Mg/Ca ratio increases, and
the brine becomes more conducive to the formation of protodolomite and magnesite. After
carbonate and sulphate precipitation, water concentrations increase by evaporation until
they reach saturation in the most soluble salt minerals (chlorides). As chemical
concentration increases, the different solubility of the salt minerals controls the
precipitation and, as a result, a concentric pattern is common in many saline deposits
(Figure 16.14) (Hunt et al., 1966). Calcium and magnesium carbonates occur in the
outermost areas, sulphates the intermediate, and the inner core is occupied by chlorides
(Figure 11.8).
Saline deposits alternate with fine-grained sediments, as fluvial and aeolian sand, silt,
clay, and organic-rich beds (kerogens) (Figure 16.15). Saline lacustrine sequences reflect
the complex sedimentary evolution of the basins, with flooding and desiccation periods
and a large spatial and temporal variability of the depositional processes (aeolian, fluvial).
The changes in water content in playa lakes define a saline cycle, composed of a
flooding stage, evaporative concentration, and desiccation phases (Bryant et al., 1994).
Other consequences of flooding events are the reworking of the playa lake surfaces, re-
dissolution of salts, and mobilization of fine particles.
The micromorphology of the saline lake surface comprises a large number of
microforms, mostly controlled by the chemical composition and the depth of the water
Table 16.1. The major saline minerals of the different brine types (Eugster and Hardie, 1978).
Brine type Saline minerals
Ca-Mg-Na-(K)-C1 Antarcticite CaC12.6H20

Bischofite MgC12.6H20
Carnallite KC1.MgClz.6H20
Halite NaC1
Sylvite KC1
Tachyhydrite CaClz.2MgCI2.12H20
Na-(Ca)-SO4-C1 Gypsum CaSOa-2H20
Glauberite CaSO4.NazSO4
Halite NaC1
Mirabilite NazSO4-10H20
Thenardite Na2SO4
Mg-Na-(Ca)-SO4-C1 Bischofite MgC12.6H20
Bloedite Na2SO4.MgSO4.4H20
Epsomite MgSO4.7H20
Glauberite CaSO4.Na2SO4
Gypsum CaSO4.2H20
Halite NaC1
Hexahydrite MgSO4.6H20
Kieserite MgSO4.H20
Mirabilite NazSO4.10H20
Thenardite NazSO4
Na-CO3-C1 Halite NaCI
Nahcolite NaHCO3
Natron Na2CO3.10H20
Thermonatrite Na2CO3"H2~
Trona NaHCO3.Na2CO3.2H20
Na-CO3-SO4-C1 Burkeite NazCO3.2NazSO4
Halite NaC1
Mirabilite NazSO4.10H20
Nahcolite NaHCO3
Natron NazCO3.10H20
Thenardite NazSO4
Thermonatrite NazCO3.H20
table. Depositional processes and deflation, both influenced by the water table depth, also
play also a significant role in the genesis of micromorphological features (Neal, 1965).
Microforms very common in the sulphate-rich zones of saline lakes are hemispherical
mounds, named domes or tumuli (Figure 11.8). These morphologies also appear in other
environments in Spain, such as the gypsum-rich Tertiary formations in the Ebro
Depression (Artieda, 1993) and in the Sorbas region of Almeria (Pulido-Bosch, 1986).
The mounds are hemispherical, with diameters about 1 m long, and heights less than
0.5 m. The central part of the domes is commonly collapsed, which allows assessment of
the thickness of the deformed bed (about 20 cm) and the hollow inside of the mound
(Figure 16.16). Several hypotheses have been proposed to explain the origin of these
structures. Artieda (1993) considered that in situ dissolution and precipitation of surface
Figure 16.14. Salinityzones in a playa surface. Cottonball Basin, Death Valley, California (from Hunt
et al. (1966)).
gypsum are the main factors. Secondary gypsum crystals filling the existing pores increase
the volume and the crystallization pressure is responsible for lateral gypsum expansion
and the appearance of dome morphologies. Calaforra (1996) described three hypotheses:
(1) the existence of compressive tectonic forces; (2) the volume increase during the
transformation of anhydrite to gypsum; and (3) the crystallization of a gypsum-rich matrix
within the interstitial space between the macro-crystalline gypsum that would provoke
Figure 16.15. Sectionshowing the halite surface crust and the underlying organic-rich mud, Sebkha el
Melah, Zarzis, Tunisia.
Figure 16.18. Salt polygon system with thrust edges, Salar Pastos Grandes, Puna, Salta Province,
Argentina. Photo: C. Sancho.
a FORMATION OF SALT BLISTERS d THICKENING AND MARGINAL

TRIMMING -
~/ ,Rain+ +
Brine on plate
- - .
...,. ,,..
b FORMATION OF SALT POLYGONS e DEVELOPMENT OF THRUST

POLYGONS
Salt polygon Broken blister

+ +
- - . =,
...... - ~ Flow of - -
, ,
~------ ..=
.___ plastic wet mud
.
C LATERAL GROWTH OF POLYGONS f EXTRUSION OF MUD PINNACLES
Point of intense _..Dry

I coo,+nuo~176
,nO
salt-crystal growth
A
_,~.l~ ...... ...=~ .
. . . .
!
. . . . i
i: - 1 I
Figure 16.19. Stagesin the development of an irregular salt crust, Gran Kavir, Iran (Krinsley, 1970, in
Mabbutt (1977), Figure 58).
from brines rising along the mud cracks, particularly those on the upwind side of the
prevalent dry winds. The growth of the edge of the saline microplate produces small
thrusting surfaces over the adjacent plates to leeward, creating an illusion of
overthrusting (Reeves, 1968) (Figure 16.20). Concave-shaped plates fill up with brine
and consequent salt precipitation causes their thickening. Mud pinnacles are produced by
the extrusion of the underlying saline mud. This process occurs because of either
the thermal expansion of the black mud under the light-coloured saline crust or the
load effect of the thick salt crust (up to 30 cm) over the underlying plastic wet mud.
Figure 16.20. Thrust polygons developed on a halite bed. The halite crust is mostly covered by aeolian
sand, Sabkha Matti, United Arab Emirates.
The combination of saline ramparts and mud pinnacles makes the surface of most saline
playa lakes very irregular and hard to walk on (Figure 16.21).
In areas with saline sedimentation, patterned ground occurs, which is similar to those in
periglacial environments. This is a clear example of convergence of morphologies. Hunt
and Washburn (1966) identified them in saline playas close to alluvial fans in Death
Valley, California. They describe polygons, nets, steps, and bands. In the carbonate zone
of the playa lake, classified and nonclassified nets occur, with detrital sediments filling up
the cracks and causing a clastic classification. Polygons develop when a gypsum or salt
layer underlies a clastic layer; desiccation of the salt produces cracks that are filled up with
the overlying clastic material. The filled cracks provoke a lateral expansion of the
polygons when they are wet, which causes the uplifting of the central part of the cell.
5. Littoral sebkhas
Littoral sebkhas are supra-tidal plains where a number of different depositional processes
concur. During the last two decades, several sebkhas have been intensively studied
because of their economic interest, among them, the littoral sebkhas of the Persian Gulf,
particularly in Abu Dhabi (Ortf, 1989). Modem and recent depositional processes in these
littoral depressions are used as depositional models to interpret pre-Quaternary rock
formations, as with the Permo-Triassic in Europe.
Figure 16.22. Dome-shaped stromatolites, Sebkha el Melah, Zarzis, Tunisia.

Figure 16.23. Sandy-silt,organic-rich tidal plain, Sabkha Matti, United Arab Emirates.
The sediments accumulated in littoral sebkhas are composed of shell fragments, ooids,
algal mats, stromatolites (Figure 16.22), evaporites, and organic matter (Figure 16.23).
Some of the large littoral sebkhas show lateral changes in depositional environments. One
example is Sabkha Matti in the United Arab Emirates, which inland becomes an interior
sebkha dominated by alluvial deposition (Glennie et al., 1994). Deflation is also intense
and causes small nebkhas, microyardangs, and dunes. Water sources for littoral sebkhas
are large tides, storm waves, and sometimes, continental sources. Towards the sea,
intratidal plains, lagoons, barriers, and islands develop (Cooke et al., 1993). The evaporite
surface of the sebkhas present similar micromorphologies to those described for saline
lakes.
Chapter 17
Aeolian processes and erosive landforms
1. Introduction
Aeolian processes are those derived from wind activity, and they comprise erosion,
transport, and sedimentation. Small-sized particles are mobilized and, finally, deposited in
different environments, sometimes at a great distance from their source area. Desert winds
are able to build large dune fields, such as those of the Rub al Khali desert (Figure 17.1),
in the Arabian peninsula, that covers about 560,000 km 2 (Allison, 1997). Also, as a
consequence of aeolian processes, more than 300 m of loess were deposited in China
(Derbyshire, 1983a). Aeolian deposits are studied by different scientific disciplines such as
physics, geology, biology, and agronomy (Pye and Tsoar, 1990). This multidisciplinary
scientific research has provided significant advances in the knowledge of aeolian activity
and its effects. Recent research on aeolian geomorphology deals mainly with
quantification of processes, analyses of forms and internal structures of the aeolian
deposits, and also with the information that those supply in order to understand the
climates of the past.
From an historical perspective, during the 19th century, wind was considered as a
transporting agent of limited importance, in comparison with water. At the end of that
century, and at the beginning of the 20th century, the scientific trend reversed, and some
researchers (especially German ones such as Passarge and Walther) pointed out that
aeolian erosion was predominant in deserts, and inselbergs with their surrounding
plains were a consequence of wind activity (Goudie and Wilkinson, 1977). Research
activities focused on exploration of new territories and the description of landforms, and
were not particularly concerned with the connections between forms and processes.
Bagnold's (1941) work was a milestone in this subject because he quantified wind
characteristics and transport processes. Since then conventional aerial photographs have
become an important tool but the decisive boost started with research using satellite
images, clearly showing the geomorphic work of the wind (McKee, 1979a,b). With
these techniques it is possible to analyse yardang systems, that can extend up to
100,000 km 2 (Lybian desert, Tibesti massif, desert of Lut in Eastern Iran, and others).
Recent studies on orbital images from Mars and Venus show clearly the importance
of wind processes on these planets (Greeley and Iversen, 1985). The discovery of oil
and hydrocarbons in Mesozoic aeolian sandstones from Europe and USA was also a
stimulus to work on these subjects. Knowledge of the aeolian domain has increased with
the large number of absolute dates, carried out since 1980, using C-14 and luminescence
techniques.
Figure 17.1. Dune field. A1 Liwa, desert of Rub al Khali, United Arab Emirates.
2. Origin and characteristics of aeolian particles
Materials transported by the wind are commonly of diverse composition; mineral and rock
fragments, and organic particles, such as shell remains. The grain sizes of the mobilized
particles are usually silt and clay, but sometimes clay aggregates can be transported if
salts are present (Bowler, 1973; Wasson, 1983). Quartz is the most commonly occurring
material, because it is abundant in the Earth's surface and, moreover, it is very resistant to
chemical weathering (Bagnold, 1941). Feldspars and heavy minerals are also present, in
lower percentages, and, in addition, they undergo a progressive weathering.
The main physical characteristics relative to aeolian particles that influence erodibility
and wind transport capacity are the specific weight, size, and shape of the particles
(Watson, 1989b). Particles have different specific weights as a consequence of
mineralogical differences. Commonly, aeolian particles show bimodal or polymodal
grain-size distributions, showing the influence of the mineralogy in mobility (Willets,
1983). Aeolian sand accumulations are usually well sorted, although they have
significant variations, both in size and sorting, in different deserts, in the same erg,
and even in individual dunes (Watson, 1989). Grain shape also affects mobility.
Therefore, fiat-shaped particles are less prone to be mobilized than rounded grains
(Willets, 1983). Generally quartz sand grains from desert dunes present subrounded
shapes and, considering an aeolian accumulation as a whole, the biggest particles are
more rounded than the smaller ones. It is also necessary to point out that abrasion during
transport has an influence on the degree of roundness (Krinsley and Wellendorf, 1980).
Aeolian particles present a higher degree of roundness than those of glacial, fluvial, and
littoral environments.
Aeolian processes and erosive landforms 411
Figure 17.2. Aerial view of a group of coastal dune ridges in the Guadalquivir river mouth, Dofiana
National Park, Huelva Province, Spain.
The origin of particles from desert aeolian accumulations is very variable (Demangeot,
1981; Pye and Tsoar, 1990). Wind-transported particles can be weathered grains entrained
from the surface of a rock, but most of the particles come from loose materials. These are
localized in continental areas in the bottom of desert rivers and its terraces, pediment
covers and alluvial fans, lacustrine terraces, and in the bottom of sebkhas or playas. Marine
beaches are also an important source area and they originate coastal dune system ridges.
(Figure 17.2). During the periods of gradual sea-level lowering as a consequence of
glaciation, large submarine surfaces became subaerial regions, and some were affected by
aeolian activity (Glennie, 1972).
On the other hand, besides the transport of sand-sized particles, mobilization of aeolian
silts also occurs. Aeolian silts commonly travel as dust storms and can reach distances of
thousands of kilometres.
3. Particle mobilization by the wind
The entrainment of the aeolian sands is carried out in different ways, including complex
processes. Loose particles, subjected to wind action, are affected by opposing forces, like
gravity, friction, and cohesion (Iversen et al., 1987). Wind speed increases over any
ground protuberance, as pressure decreases, according to Bernouilli's equation. As a
consequence, a loose particle can undergo uplifting, and this force is much more intense in
the contact with the ground, where wind-speed is maximum (Cooke et al., 1993). The
lifting is much more important in dust turbulences (vortex and whirl-winds). Creeping
includes friction between the particle and the air, called surface creeping, that causes
rolling, and shape creeping, due to the difference of pressure between the stoss side and the
lee side of the particle (Livingstone and Warren, 1996). When grains are airborne,
bombardment against the substrate and collisions between grains begin. These processes
cause the possible creeping and lifting of other loose particles.
One of the fundamental parameters is the minimum speed required to mobilize particles
of a particular size. Entrainment is a function of the average size of the particles and the
shear-wind speed. Bagnold (1941) studied this relationship and obtained values for the
critical thresholds of shear speed (U*) considering different particle sizes. This
relationship is represented in Figure 17.3, where it is possible to observe that the bigger
particles need a higher speed to be mobilized, but the smallest particles (< 0.06 mm) also
need high wind speeds to be entrained. This is due to the electrostatic cohesive forces.
Sand size particles (0.04 to 0.40 mm) are the most easily mobilized. As lifting and
dragging energy increases, there is a critical value of wind-speed shear to start particle
60-
40-
l
.H
E l
20- \ /
/ /
///
4,
0 I I I I I
.Ol o.1 1.0 2.0 3.0
Grain diameter (mm)
Figure 17.3. Relationsbetween fluid and impact threshold shear velocity (U*) and particle size (Bagnold,
1941).
mobilization. This is the fluid threshold (or static). When entrainment of grains has begun,
ballistic impacts reduce required energy to keep the movement of particles. This is the
impact threshold (or dynamic), where slower speeds are required to move a stationary
grain by fluid pushing, and it is equivalent to 80% of the static threshold (Andersson and
Haft, 1988). These equations adjust correctly on flat, homogeneous, dry surfaces, with
isolated, loose clasts (Williams et al., 1994). The values of the movement thresholds,
however, are also affected by the variability of sediments, surface crusts, slope inclination,
humidity, and roughness.
4. Types of aeolian transport
When particles pass the mobilization threshold they can be transported in four different
ways; creep, reptation, saltation, and suspension (Figure 17.4). The transportation type is
mainly dependent on the grain size of available particles (Bagnold, 1941). The boundaries
between these different transport types are not sharp and the transition from one to another
is not very clear.
Creep involves translation or turning of coarse particles (0.5 to 2 mm) on the surface,
pushed by the wind force or by the ballistic impact of grains during saltation. Creep can
begin some moments before the mobilization of the particle by saltation (Nickling, 1983).
It is difficult to establish the amount of material that moves by creeping because is very
difficult to distinguish creep from reptation. We can estimate, however, that creep
represents one quarter of the mobilized bed load.
Recently, a transitional type between creep and saltation has been defined. It is called
reptation, in which particles move in small jumps caused by the impacts of high-energy
~I~~ ~long_term
TURBULENT suspension (<20 ILtm)
WIND FLOW A
short-term
suspension
/
creep
1 15m(>500 l~m) (70- 500 lam) ~k,
.~.-..!
9.~"'":"'~i
.?,
Figure 17.4. Types of particle transport by wind. Indicated particle size-ranges in different transport
modes. (Reproduced with permission of Pye (1987)).
Figure 17.5. Sandgrains mobilized by saltation. A1Liwa, desert of Rub al Khali, United Arab Emirates.
grains (Andersson and Haft, 1988). As a result of the impact, grains can loose 40% of their
speed, and the transmitted energy can mobilize other grains from the bed. The main
difference between reptation and creeping is that, in the former, particles change
continuously from reptation to saltation transport.
Saltation (Figure 17.5) consists of the uplifting of a particle, that covers a distance in a
ballistic trajectory (Bagnold, 1941) and falls down, impacting on the surface. The energy
of this impact is used in the uplifting of other grains. Therefore, saltation is a very
important transport mechanism with cascade effects (Nickling, 1988). Uplifting angles
fluctuate between 30 and 50 ~ and speeds are higher in the stoss side facing the wind. Most
particles, according with different models and experiments, descend with angles between
10 and 15~ (Anderson, 1989). Jump distances reach 12 to 15 times its height (Livingstone
and Warren, 1996), and the smaller particles, entrained at a higher speed, can reach more
height and so the distance of jump is longer (Anderson and Bunas, 1993). Height reached
during jumps can be 3 m, but the average is 0.2 m (Pye and Tsoar, 1990). Bagnold (1941)
recognized that jumps are higher over a rocky surface or a surface with pebbles than over a
bed of loose sand (Figure 17.6).
Pure saltation develops when turbulent vertical speed has no significant effect on the
particle trajectories. Between pure saltation and pure suspension the trajectory of the
particles is modified due to speed settlement and inertia. This type of transport has been
called "modified saltation" (Figure 17.7) (Nalpanis, 1985). This indicates that there is not a
sharp transition between saltation and suspension.
The smallest particles (< 0.06 mm) are transported by suspension within a turbulent
airflow (Figure 17.7) that can keep this fine sediment at a considerable height during many
days, and finally deposited it as dust or loess. This constitutes a deposit generally between
30 and 60 m, occupying large surface areas, and being able to travel along more than
I I
I i
I l
I i
Figure 17.6. Schematicrepresentation of saltation trajectories over (A) a loose sand and (B) a pebbly
surface (Bagnold, 1941).
300 km. Within the suspension transport we can distinguish the airborne grains for long
periods of time (long-term suspension) from those that fall quickly on the surface (short-
term suspension) (Tsoar and Pye, 1987) (Figure 17.4 and Figure 17.7).
5. Ripples
Ripples are the smallest aeolian accumulations, and they represent the initial response of
sand surfaces to wind transport (Figure 17.8). Although they are depositional forms, they
will be analysed in this chapter, because they are closely related to the types of transport
previously described. Ripples are oriented perpendicularly to the wind transport direction,
even though on slopes they flow obliquely, because they are affected by gravity. Most of
them have a short-term existence and they move much more quickly than dunes
(Livingstone and Warren, 1996). Usually tipples are asymmetrical in a transverse section.
Sand Sheets
- ,,,. /
ed load
(rolling and saltation)
i /
II
Dune Sands
0.1
(
l i,'/
.;';o,'~ /
E
E
v
z- ey" #
iioess
L_
E
"O
t'- / fff
tl:i
(.9 / / / .neg~\Og
0.01
tl:i
(O
dust storms
0.001 i 1 i i i 9 9 l l l 9 1
10 30 50 70 100 200
U* (cm s-1)
Figure 17. 7. The relationships between grain diameter, shear velocity, and mode of sediment transport.
Uf is the particle fall velocity (Tsoar and Pye, 1987; modified in Thomas (1997d), Figure 16.2).
The stoss side is convex with angles of 8 to 10 ~ and the lee side has a slope angle between
20 and 30 ~ (Sharp, 1963). Some tipples have a sharp crest; some others present a rounded
one. Wavelengths vary between a few centimeters up to tens of meters, and the heights
range from 1 cm to about 30 cm. Wavelength and height are the result of the particle sizes;
tipples of coarse sand present a longer distance between crests than those of fine sand
(Sepp~il~i and Lindt, 1978).
Megaripples are remarkable landforms, with a wavelength between 1 and 25 m. They
are originated by strong winds and are composed of coarse sand. In cross section they are
generally more symmetrical than the smaller tipples, and present a higher contrast in grain
size between crest and trough areas. According to Bagnold (1941) several centuries are
Figure 17.8. Ripples in A1 Liwa, desert of Rub al Khali, United Arab Emirates.
necessary to build megaripples, however, Sharp (1963) estimated that during a wind storm
a few hours are enough. Chiflones are wavy ribbons of coarse and fine sand, 100 to 150 m
long and 1 to 1.5 m wide (Simons and Eriksen, 1953). They form when coarse sands are
channelized between obstacles such as shrubs.
Several models have been proposed to explain tipple formation. The first one (Bagnold,
1941) established a relationship between tipple wavelength, observed in experimental
tunnels, and the length of sand paths during saltation. Irregularities in the surface act as
sites for erosion and sedimentation. The stoss side is bombarded by particles transported
by saltation that transfer energy when they collide, as a consequence new grains are
mobilized and the process continues. Trajectories have the same length as ripples if sand
size and wind speed do not change. This hypothesis is very attractive, because it explains a
large number of characteristics and variables related to tipple formation. Sharp (1963)
estimated that particles in tipples move essentially by reptation. Irregularities on the
bottom surface and interactions between moving grains form local elevations on the
former surface, originating incipient tipples of short wavelength and low height, which
progressively grow bigger. A third model, proposed by Anderson (1987), was developed
using experimental data and numerical simulations. The model establishes differences
between low energy impact reptations and long trajectory high-energy impact saltations. It
identifies different saltation paths, some of them longer than the wavelength of tipples.
This model disagrees with Bagnold's model, because saltation impact does not contribute
directly to tipple formation. According to Anderson's work, reptation processes are
dominant. Because reptation distances increase with wind shear strength, the tipple
wavelength also increases. This agrees with data obtained in experimental tunnels
(Sepp~il~i and Lindr, 1978).
In conclusion, aeolian ripples are closely related to sand saltation, although new
models suggest that reptation provides the initial conditions for tipple formation, and
they quickly evolve by saltation impacts and, as a consequence, by grains transported
by reptation (Lancaster, 1995).
6. Aeolian erosion and resulting landforms
When particles are lifted by the wind from the surface and impact on the rocks at
significant speeds, they produce aeolian erosion landforms. They are more widely spread
in areas of scarce or absent vegetation, especially in arid zones, but are also present in
periglacial and marine beaches environments.
Wind erodes by deflation, that is, the result of entrainment and transport of loose
material of the ground. Average deflation rate is about 1 mm/year (Cooke et al., 1993).
Abrasion is the mechanical wearing out of cohesive material. Ballistic impacts of grains
cause surface rock fracture, until the surface becomes stabilized. This suggests that
abrasion speed reaches a maximum and later decays. Most authors estimate that abrasion
is produced only by sand grains, however others think also that the aeolian dust originates
abrasion, although secondarily (Greeley and Iversen, 1985). Abrasion speed rates are very
variable, about several millimeters per year, and they depend on the type of rock, wind
speed and amount of grains provided by the wind. The height where maximum abrasion
occurs fluctuates between 0.1 and 0.4 m. These figures have been obtained from field
experiments during intense storms (Sakamoto-Arnold, 1981) and from theoretical models
(Anderson, 1986).
6.1. Ventifacts
Ventifacts are pebbles with smooth surfaces, polished and faceted by the impact of
particles carried by the wind (Figure 17.9). Pits, flutes, grooves, and helicoid patterns are
developed both in boulders and in rock outcrops. These micro-morphologies appear only
in weathering-resistant rocks.
Facets are oriented perpendicularly to wind direction. The number of facets is very
variable, and up to 20 have been described (Higgins, 1956). The most common, however,
is the presence of two or three facets that intersect forming edges. The German words
einkanter, zweikanter, and dreikanter are commonly used to indicate that they have one,
two, or three facets, respectively. Pits are small closed depressions that can be originated
by the action of wind over a nonhomogeneous rock or on pre-existent holes. They are
inclined 55 to 90 ~ from wind direction and, in consequence, they indicate the windward
side. Where the angle decreases considerably we find the transition toflutes (Sharp, 1949).
These have a spoon-shaped form, with dead ends and U-shaped cross sections. Commonly
they are open to the windward. Their length, width, and height vary greatly, from
millimetric sizes up to centimetric ones (Cooke et al., 1993). They appear in horizontal
surfaces or in gentle slopes, in very different kinds of rocks, and they are not related to the
structure (Breed et al., 1989; Laity, 1994). Grooves are opened in both extremes and are
longer than flutes. They indicate, as those described before, the predominant wind
Figure 17.9. Ventifacts (zweikanter-- two facets;

blow essentially from a single direction (Sharp, 1949). A well-accepted theory is based in
the rotation of pebbles around a vertical axis, with rare overturning caused by excavation
beneath due to deflation. The capsizing can be also caused by expansion and drying of
clays, and by animal activity (Livingstone and Warren, 1996). This explains that the
smaller the ventifacts, the larger the number of facets they present, and vice versa.
Moreover, overtuming justifies the existence of ventifacts with facets in the upper part and
in the bottom. Some authors have pointed out that these micro-morphologies (polished
surfaces, pits, flutes, grooves, and spiral forms) are probably originated by a vortex of
aeolian dust (wind rotating around an axis), more than by sand impact (Maxson, 1940;
Higgins, 1956). These micro-morphologies can be reproduced in wind-tunnel experiments
(Dietrich, 1977; Whitney, 1979). In any case, most of the shapes related to ventifacts are
explained by sand grain impact, and aeolian dust erosion (Breed et al., 1989, 1997).
Wind direction can be determined from ventifacts and their associated forms, and, as a
consequence, fossil ventifacts provide good records of wind circulation during past
geological periods. If, in addition, they are dated using desert-varnish techniques (Dora,
1995), it is possible to know the paleoenvironmental conditions and the intensity of
aeolian activity.
6.2. Yardangs
The word yardang has its origin from the eastern limit of the Taklimakan desert in China.
The shape of the yardang looks like a capsized ship hull (Figure 17.10) even though in
many cases yardangs are flat topped and lack a keel. The windward face is wider, whereas
the leeside becomes narrower and looses height. Yardangs show varied forms
(Figure 17.11). Halimov and Fezer (1989) described eight types of yardangs in the
depression of Quaidam in China, according to its morphology, and they suggest a temporal
succession. The ratio between longitude and width varies from 3:1 up to 10:1 (McCauley
et al., 1977). Yardangs appear usually grouped, and their long axes are parallel to the
direction of the more intense winds. They occur in every large desert all over the world,
except in the Australian desert. They have been also identified on Mars (Greeley and
Iversen, 1985). Some authors (Cooke et al., 1993) distinguish yardangs according to their
size: microyardangs (1 m) (Figure 17.12), mesoyardangs (10 to 100 m) and megayardangs
(1 km). The more useful classification by Livingstone and Warren (1996) of landforms up
to 100 m long, called yardangs, and those of longer size, that form ridges (Mabbutt, 1977,
Laity, 1994) or megayardangs (Figure 17.13). Cross sections of megayardangs present a U
shape, with the bottom as the trough corridor. They are commonly covered by sand,
pavements, and sometimes, by barchan dunes, parallel to the ridges. Megayardangs are
well developed in the southern part of the Tibesti Massif (Central-Eastern Sahara), where
the ridges can reach up to 200 m in height, up to 4 km in length and a separation between
0.5 and 2 km; they cover a surface area greater than 600,000 k m 2 (Hagedom, 1968;
Mainguet, 1972). Very large groups appear also in the hyperarid deserts of Egypt, Iran, and
Perf. Yardangs are carved generally in soft materials, such as aeolian sands and fluvial
sediments and Cenozoic lacustrine deposits, although many of the Saharan ridges are
excavated on Cambrian sandstones. They can be developed in a wide variety of
lithologies, from semi-consolidated sediments to granites or even harder rocks
Figure 17.10. Scheme of the general situation of yardang groups in the area of Borkou, Tibesti Massif
(Republic of Chad). 1: Cliff that separates the basin from a higher surface. 2: Lacustrine area, sometimes
subactual. 3: Small yardangs of the depression. 4: Scarp windwards over the lower surface. 5: Yardangs of
platform margin. 6: Sandstone crests and corridors (after Mainguet, 1972).
(Goudie, 1989b). Moreover, direction, dip and cleavage of the rocks affect the shape and
development of yardangs.
Yardangs are probably shaped by abrasion and deflation processes (McCauley et al.,
1979; Ward and Greeley, 1984), with slight modifications caused by water erosion (Laity,
1994). Abrasion contributes to polish and to basal excavation that can reach 1 or 2 m in
depth (Grolier et al., 1980). Deflation is important in the development of yardangs in
poorly consolidated lithologies. As happens with ventifacts, a vortex of dust particles
seems to be an important process in shaping the morphology of the yardangs (Whitney,
1985). Many yardangs have previously undergone fluvial erosion, and they appear more
elongated if the drainage network is parallel to the prevailing winds. Periods with fluvial-
lacustrine activity, and others with dominant aeolian action occur and deposition and
transport of sediment in the troughs alternate (Mainguet, 1972; Guti6rrez et al., 2002).
Finally, wind-tunnel experiments (McCauley et al., 1979; Ward and Greeley, 1984),
performed with synthetic sediments, show that it is possible to identify positive main
Figure 17.12. Microyardangs on a cemented dune. Sebkha Matti, United Arab Emirates.
Figure 17.13. Wind-sculpted sandstone landscape, originating a morphology of crests and corridors.
Borkou lowland, south of Tibesti Massif, Chad Republic (Hagedorn, 1968, in Mabbut (1977), Figure 40).
fluxes and secondary negative ones returning along the flanks. They conclude that abrasion
and deflation are the fundamental processes in yardang formation.
6.3. Deflation basins
In the previous chapter we dealt with the origin of closed depressions, and we analysed
some of the fundamental features that characterize the so-called erosive basins. In this
section we study some aspects more deeply that are essentially related to their geometry,
distribution and origin.
Deflation basins or blowouts are closed depressions that can reach 12% of large
surfaces (Cooke et al., 1993). They usually have a rounded, elliptical, or kidney shape,
with rounded margins. They are more common in the semi-arid zones of the world, and
have a great development in central and southern Africa, the Pampas and the Pantanal of
South America, southern and western Australia, and Manchuria and western Siberia
(Goudie, 1991; Goudie and Wells, 1995). Their size can vary considerably; with the
surface area of the basins in southwest Australia ranging between 0.004 and 100 km 2, with
an average size of 0.05 km 2 (Killigrew and Gilkes, 1974), whereas in the south of Africa
they fluctuate between 0.05 and 30 klTl2, with an average of 0.2 kin 2 (Goudie and Thomas,
1985). These authors calculate densities of 100 basins every 100 km 2.
Deflation basins are more commonly developed in poorly consolidated sediments.
Some lithologies, such as the sandstones from the south of Africa, are the most suitable for
the development of these closed depressions. Some rocks quickly disintegrate by saline
weathering, such as some claystones with a high salt content, making deflation easier. The
presence of swelling clays implicates that during dry periods clayey regolith cracks
deeply, and the product of this disintegration can be easily exported by the wind (Price,
1963).
During the Late Pleistocene period desiccation of the great pluvial lakes in the western
USA, a period of intense deflation occurred that caused excavation of the lake bottoms,
mobilization of dune fields and generation of lunettes (Goudie and Wells, 1995). Similar
effects are identified on other continents.
Ancient drainage systems developed in large areas of low relief are suitable places to
generate deflation basins. Excavation of some areas of the creeks and small rivers and
generation of small deflation basins may disconnect the drainage networks, although
paleochannel paths are still identifiable (Bowler, 1986). Dunes can also cut off the
channels, producing deflation basins.
In sandy deserts, there are large surface areas with interdune deflation basins. They are
formed as a consequence of the growth of dune crest fields derived from parabolic dunes.
The development of these dunes implies the formation of deflation basins of an elliptical
shape (Lancaster, 1978).
Also in littoral areas, with more humid conditions, is possible to recognize large fields
of deflation basins. In the Carolina Bays, between Maryland and north of Florida, they
cover a strip parallel to the coast of about 1100 km, and about 500,000 basins have been
identified (Prouty, 1952). They were apparently formed during the Wisconsin glaciation as
the elliptical depressions of giant parabolic dunes, built by northwestern winds that blew
along the sandy surfaces of the coastal plain, when the phreatic level was low. When the
water table rose, depressions were stabilized. After that, during dryer periods, many
lunettes were built (Price, 1968; Goudie and Wells, 1995).
In the leeward margins of the deflation basins sand-clay dunes with a half-moon shape
developed commonly. They are called lunettes. They often reach a height between 10 and
50 m, however in the Sebkha Kelbia, in the centre of Tunisia, their maximum height is
165 m, with a 17-km length and a 9-km width, being the highest lunette in the world. The
surface of these dunes is often colonized by vegetation, which facilitates accretion.
Particles from the bottom of the deflation basin become fragmented into micro-
polyhedrons because of desiccation, and they are mobilized by winds with a speed higher
than 3 m/s, being able to transport these particles and deposit them in the outer rim of the
closed depression (Perthuisot and Jauzein, 1975). The presence of salt, which blocks
vegetation development in the bottom of the depression, causes the aggregation of clay
particles that will be also exported out of the basin. Pellets of salt and clay that form the
lunette are lixiviated by rain, and salt returns in part to the sebkha bottom; so a cycle of salt
is established (Tricart, 1969).
In some cases in the leeward area several lunettes can occur (Figure 17.14). The most
external dune has a much higher content of quartz, and the rate of clay increases toward the
internal dunes, becoming clayey dunes (Bowler, 1973, 1986). This has been explained
"(6 oang!~t
's "d '(g66I) SllOM pue o!pno D u! '8L6I 'OODOlAIpue .~Ol~O~l) e!lealsnv tuo~seoqlnos 'soieAX qlnos aXON
jo :~soaxq~ou u[ sttuoj ounp pue so~oun I pme!oosse q~!~ 'sued ops.uo~oe.nzqo jo dnoag V "t,['Z [ a.:n~'.~j
. _.-._-.~-- _ _ . . ~ y . ~ .%_~' ~ -~ .
01. k~ 0
SOI!LU ..,,.~-~ ~ .. -.- \...-... ..-- .-

9
sounp O!loq~J~d ~. ~ -._-..-
-qns Jl~In6eJJl _ _- -----
sounp // ",uln, ~ " --~''-'' ~ = ~ ,'~
j~OU!l Jeln~Ol:l "~//'~/
o]~ounl ~
pu~ o~l'elf~fi]
. _ _ . . - - - __ :..
. I . I~-
...- ..
-~ ~ . ~
Jr" .... - - -" o
,.L. "," ~ ~ t. --'~ C
r ,....
~t C '
/ f
f ::::::::::::::::::::: L
r
*"
9 t"
t f"
t I0610~e ~ : ':":':i:~.*,:
''.'~''~'.'
E":~':'.~::::.:..:,;.:.~:.; r ~,
'~ r.~ ~., ~,"

~.
r r r
9 L ~ L
t , 'c c,% ~. o ".t t C
o ":':'Y,:,X:"y'.!::'..:':".:':':'7:~''" / / ~ $..,! r ~ . - ,.
9 ~ ,..~r ,,,,
.,-.9:.'::._. ' , r
t~ ' r ' s
"~.C,e it
~ s "~)r
"' "
r
,,
s (r
C
I ~.f'f. C , t.
'('( ~ Cr" I.. ,"

"~" ' , , ~ , r ', ,
g'LV ~ u a o f p u t ~ 1 aa~.~o.:a p u v ~as~ago.ld UV!lOaV

by deflation of fluvial sediments to build the external dune. On the other hand, the internal
lunettes are developed by deflation from playa deposits, accumulated after the initial
deflation stage (Goudie and Wells, 1995) (Figure 17.15).
(a)
Drainage channel
I Aeolian coversands
"'"""'"'" . . . . . . ...'........'...'.'.'.'::.:.":::'.".'.'.::'::.".:.'?'.i:i"('". . . . . . . . :""' '"
(b)
Lunette derived from deflated

channel alluvium and coversands
II
Lunette derived
i
from playa sediments
Figure 17.15. A model of lunette development caused by deflation of channel alluvium, aeolian sheets
and playa deposits (after Goudie and Wells (1995)).
Great deflation basins (Figure 16.9), such as the Quattara depression, commonly have a
polygenic origin, although deflation plays the fundamental role. Albritton et al. (1990)
show that this depression was initially a fluvial valley that was later split by karstic
processes during Upper Miocene, and later it was deepened by deflation. Moreover, mass
movements and fluvial processes modified it somewhat. In addition, saline weathering
prepared materials for subsequent deflation. We should also consider the role that
Quaternary climatic oscillations played, modifying the geomorphological processes and,
as a consequence the genesis of depressions. We can conclude that great deflation basins
have a long and complex history. In smaller deflation basins deflation is dominant,
although other processes are also present, as we previously indicated.
Chapter 18
Aeolian accumulations
1. Introduction
Aeolian accumulations are the result of sedimentation of particles carried by the wind,
producing deposits of sand and dust. We shall begin with the study of the aeolian sands,
analysing their distribution, processes and types of accumulations.
Dunes are, undoubtedly the most outstanding of desert forms (Figure 18.1). Research
on these landforms has experienced a boost during the last two decades with the use of
satellite images. Aeolian sands cover about 5% of the Earth's surface (Thomas, 1997d),
and 20% of the arid zones of the world are occupied by sand accumulations. The
proportion of this cover varies according different deserts, less than 1% in North and South
America (Lancaster, 1995) up to 50% in Australia (Mabbutt, 1977). The word erg or sand
sea means a large surface covered by sand, at least 125 km 2 (Fryberger and Ahlbrandt,
1979; Thomas, 1997d). The erg is called koum in Central Asia and nafud in Arabia. We
can see the global distribution of great ergs in Figure 18.2, where active ergs are
distinguished from marginal ergs, stabilized by vegetal colonization. Most ergs are active
with an average annual precipitation of 150 mm (Wilson, 1973). The largest number of
sand seas is concentrated in the Northern Hemisphere. Wilson (1973) identified 58 ergs
with a surface extent greater than 12,000 km 2 in the different deserts all over the world,
most of them in the Sahara desert, where the erg Chech in southern Argelia (312,000 km 2)
is the largest in this desert. The erg of Rub al Khali, which occupies one-third of the
Arabian Peninsula (Figure 18.3), has a surface area of 560,000 km 2, and the Great Sandy-
Gibson Desert, in western Australia, covers a surface area of 630,000 klTl 2.
In some areas it is difficult to distinguish between coastal and desertic dunes. Examples
dealing with this problem occur in the coastal zones of Namibia (Lancaster, 1982;
Lancaster and Ollier, 1983), Pert] (Finkel, 1959) and Oman (Goudie et al., 1987). In some
zones, continental dunes obtain their sand supply in the coastal regions, whereas in other
areas ergs have prograded towards the sea (Pye and Tsoar, 1990).
A hierarchic classification of ergs (Wilson, 1972; Lancaster, 1994) is based on
the identification of three classes with distinct spatial patterns: (1) aeolian ripples (0.1
to 1 m); (2) isolated individual dunes (50 to 500 m); and (3) compound and complex
dunes (> 500 m). Each element in this hierarchy responds to the dynamics of the wind
pattern of the region.
Other large bodies of aeolian sands do not have dune forms and are called sand sheets.
Their surface area ranges from a few square kilometers up to 100,000 K1TI2 in the eastern
Sahara (Breed et al., 1987). Their development is controlled by vegetation, grain size,
shallow phreatic levels, seasonal floods, and the occurrence of superficial crusts (Kocurek
and Nielson, 1986).
Figure 18.1. Dune field in Jaisalmer, Thar Desert, India.
Figure 18.2. The global distribution of aeolian deposits (Thomas, 1997).

Aeolian accumulations 431
Figure 18.3. Barchanoidforms in the southern erg of Rub al Khali, United Arab Emirates.
2. Factors controlling erg development
Three requirements are necessary in order to build ergs: (1) large supplies of sand; (2)
suitable climatic and topographic conditions; and (3) enough wind energy (Pye and Tsoar,
1990). About 90% of particles from active ergs are sand-sized because silt is
characteristically transported far away from deserts, and clays are mobilized by the
wind only as pellets. Most of the sand comes from fluvial, littoral, and lacustrine deposits.
In the Namib desert (Lancaster and Ollier, 1983) and in the Sinai desert, dunes have been
originated by aeolian transport in fluvial and littoral, environments. Sands can be
mobilized and flow by fixed paths, covering large distances, and reaching the distal zones
of the ergs (Wilson, 1973). Maximum speeds have been obtained in aeolian sheets from
Mauritania, fluctuating between 62.5 and 162.5 m3/m width/yr (Sarnthein and Walger,
1974), whereas the typical speed is about 3.49 m3/m width/yr in barchans. Where sand
supply speed is high, large dune fields can grow in only decades or centuries. Nevertheless,
the time required for the development of ergs is much higher. The average thickness of the
Sahara's ergs is about 21 to 43 m, although in the Simpson desert Australia it is 1 m
(Wilson, 1973). These values are very low by comparison with the thickness of several
hundred meters of Paleozoic and Mesozoic aeolian formations. This implies a subsidence
in the sedimentary basins, and possibly a greater aeolian activity during these periods,
especially before the development of many terrestrial plants (Kocurek, 1988).
Aeolian processes are usually more important when vegetal cover is scarce, with the
result that extensive ergs occur in zones with < 250 mm of rain, although in the south of
Africa the limits of active dunes vary between 100 and 150 mm (Lancaster, 1981). In some
ergs, distinctive accumulative phases and hiati can be identified as an evidence of past
climatic changes, as we can see in the Gulf of Oman (Cooke et al., 1993).
According to their mobility, ergs can be dynamic or static. First, they occur in flat areas
where the migration of the erg, in the direction of the predominant wind, does not
encounter significant obstacles (Pye and Tsoar, 1990). On the contrary, static ergs are
localized in topographical depressions, limited by steep slopes (Figure 18.4a) (Wilson,
1973). A clear example appears in the Great Eastern Erg, located to the south of the cliffs
of Tademait plateau in Algeria (Fryberger and Ahlbrandt, 1979). These authors proposed
different models of topographic influences in the development of ergs (Figure 18.4b). They
occur: (1) in the shadow of topographic barriers; (2) in shallow desert depressions; and (3)
TIBESTI
Erg Erg MASSIF Erg Erg Erg SW
NE Calanscio Rebiana Bilma Tenere TAMGAT Azouak
F '
Benghazi
Libya o 500
(a) L 9 9
km
9
" '
Reduction of total Reduction of resultant

sandflow energy sandflow energy
(b)
Figure 18.4. (a) The relationships of ergs in North Africa to topography (after Wilson, 1973). (b) Models
of topographic influences on ergs development (Fryberger and Ahlbrandt, 1979).
Figure 18.5. Sandflow across the Sahara showing major source and deposition areas (after Wilson, 1971;
Mainguet, 1978).
in zones with a reduction in sand-transporting energy. On the other hand resultant energy
can dissipate; (4) when surface winds are deflected leading to sites of favourable
accumulation; and (5) when winds blow from opposite directions.
When sand transport speeds are high, flow is higher than sedimentation, and aeolian
sheets and barchans can develop. On the other hand, if sand flow is saturated and
accumulation exceeds transport, ergs are generated (Wilson, 1971). This author used the
sand path flow obtained by weather stations to draw sand flow maps in the Sahara desert
(Livingstone and Warren, 1996) (Figure 18.5). In this way he obtained the source areas
and the sedimentation areas of Saharan ergs.
In tropical deserts, the environments with high aeolian energy are located in the Trade
Wind belts that border the anticyclonic areas, whereas the zones of low energy are situated
in the proximities of the high-pressure centres (Livingstone and Warren, 1996). The Sahel
region, in the south of the Sahara desert, is a low energy area where particles transported
by the Trade Winds from the northeastern areas of higher energy are accumulated.
3. Dominant processes in dunes
The onset, development, and achievement of a dune-equilibrium morphology reflects the

balance between erosion and sedimentation, both controlling the dune's shape (Lancaster,
1995). A decrease in sand transport results in an affluence that exceeds the evacuation,
with resulting sedimentation and increasing accumulation. On the other hand, ground
irregularities have considerable impact on transport speeds, because of the interaction with
the flow and by the generation of secondary flows (Lancaster, 1994).
The onset of dune formation is not a well-known process. The starting of a dune implies
a local sedimentation, due to the descent of the transport speed. Kocurek et al. (1992), in
their studies about the development of dunes in Padre Island (Texas), indicated that
accumulation begins in small irregularities of the terrain surface, or small vegetation
obstacles or boulders, originating shadow dunes. In this work they recognized five steps of
dune evolution, with progressive changes on the lee-side slope, due to the separation and
expansion of the flow (Figure 18.6): (1) small irregular patches of dry sand, a few
centimeters high; (2) protodunes of 0.1 to 0.35 m high, with tipples all over its surface; (3)
protodunes 0.25 to 0.40 m high with sand grains falling down on the lee side slope; (4)
grainfall protodunes 1 to 1.5 m; and (5) dunes with grainflow, 1 to 2 m high. The three first
stages can be considered as generating embryonic dunes.
Bagnold (1941) pointed out that an intense sand transport implies an instability
transverse to the flow, originating sand deposition in longitudinal strips 1 to 3 m wide. He
pointed out that these strips can be the core of further longitudinal dunes. In small sand
accumulations, erosion prevails with low and intermediate wind speeds, but when these
are higher, accretion processes dominate (Greeley and Iversen, 1985). Dunes can be also
built by strong wind gusts able to sweep the sand, depositing it when wind intensity
decreases; small aeolian accumulations can be the beginning of dune formations (Warren
and Knott, 1983).
Many hypotheses have been considered to explain the regular repetition of successive
dunes, with comparable shapes and similar separation. Models are based in dunes
developed under water, and their evolution is founded in the kinematic instability and in
turbulence physics (see Cooke et al., 1993; Livingstone and Warren, 1996). On the other
hand, as dunes grow there is a streamline convergence of wind towards the crest, and as a
Sand patch
(a) ~__. . . . . . . . . . . . .
W i n d - ripple protodune- flow expansion
(b) _ .,.-_- - _ ._
Grainfall protodune- greater flow expansion
(c)
Grainfall protodune- flow separation
. . .... . .... ..... . .... .... ....

....... .;.:. _, ,_,.
Dune with grainflow-flow separation
(e) - ...... ,--,.
Figure 18.6. Evolutionof lee-face airflow pattems for stages of protodunes. Internal structure indicated
by wavy dashes for wind-ripple laminae, dotted lines for grainfall laminae, and solid lines for grainflow
cross strata (after Kocurek et al., 1992).
m12.~
10
8 "- 1 . 2 0
Z 6 ~ 1.00
~ - ' " " ~ 0.80
4 o02
2
.20
0
-1.0 -0.5 0.0 0.5 1.0 m
X/L
Figure 18.7. Wind flow pattern over an isolated barchan, in Salton Sea, California. A convergence of
stream lines, with compression, is observed on the windward slope, and an expansion in the lee-side slope
(after Lancaster, 1995).
consequence, wind speed increases, but conversely an expansion and divergence on the lee
side slope, originating varied secondary flow systems (Lancaster, 1995) (Figure 18.7)
Windward slopes of dunes are zones of important changes, becoming apparent by
modifications of slope inclination, roughness, shear speed and sand discharge (Livingstone
and Warren, 1996). The entire slope is a zone of maximum erosion (Bagnold, 1941). Wind
speed increases towards the highest point. Differences of wind speed between slope and
crest, measured on barchans, can reach a ratio of 1:42 or even more. These conditions
occur on straight slopes with an inclination of 5 to 10~ High wind speeds, however, have
been pointed out on very steep slopes (Lancaster, 1985), but other authors consider that in
zones of strong winds the inclination of slopes is low (Gaylord and Dawson, 1987). A clear
disagreement exists, indicating that further research is needed.
Crests can have a straight form or occupy the highest point, separated from the brink (at
the top of the leeward face) (Figure 18.8). These variations are explained considering first
Crest and brink

/
(a)
Crest
Brink
/
(b) ..
Figure 18.8. Dune crest (the highest point) and its brink (the top of the slip face). (a) Crest and brink are
together. (b) They are separated (Livingstone and Warren, 1996).
Figure 18.9. Bottleneck-shapedgrain flows in the lee face of a dune, Ythan Estuary and Sands of Forvie,
Aberdeenshire, Scotland. Photo: F. Guti6rrez.
small dome-shaped dunes, with a clear separation between crest and brink, becoming
bigger dunes where the separation is much smaller, evolving finally to a straight slope with
a crest (Lancaster, 1987). Inverse evolution has also been suggested (Capot-Rey, 1963).
Another interpretation, indicated by Bagnold (1941) points out that the separation between
crest and brink is due to variations in shear speed and sand transport speed. Finally, this
author indicates that in dome-shaped dunes the crest can be mobile whereas the bottom
remains static. These variations have been documented by Lancaster (1985), who pointed
out speed relations between the bottom speed of the dune with regard to the crest of 158:1
and 13:1, for strong and gentle winds, respectively. These observations indicate that sand
transport occurs only in the crest areas of the dunes during low wind speed periods,
whereas during strong winds mobilization is present all over the dune. Crest height
decreases during periods of low speed winds, and increases with high speed winds.
In lee-side slopes, wind speed and sand transport decrease quickly because of the flow
expansion between crest and brink and their separation in the lee-side face. Flow is very
irregular, gusty and with flow inversions, which can be detected with smoke. Wind speeds
in the bottom of this slope are low and variable; in crescentic dunes they fluctuate between
Figure 18.10. Slid sand plates on the lee side slope, Beach of Trabucador, Ebro delta, Spain. Photo:
F. Guti6rrez.
0.4 and 0.8 respective to the crest values in the desert of Namib, and up to 0.04 in the dunes
of Padre Island, Texas (Sweet and Kocurek, 1990). In longitudinal dunes and in star dunes
flow is complex, especially if the crest line is sinuous. Flow separation takes place on the
crest, but when it is oblique to the crest, whirls appear on the leeward slope and sand is
transported along the dune (Tsoar, 1983a). Sand grains can fall down from the crest or
cross it by saltation. They are also affected by mass movements, from fracture places
with scarps of 5 to 10 mm, flowing as avalanches, causing bottleneck morphologies
(Figure 18.9). These avalanche tongues are a few centimetres deep, about 0.5 m wide, and
move at speeds around 0.2 m/s (Hunter, 1985). With morning dew or in saline
environments, increased superficial cohesion may allow mobilization of sand plates
(Figure 18.10) (McKee, 1979a,b).
4. Classification of dunes
A large number of dune types have been differentiated, even though the main processes,
previously described, are always present. Besides this, a large number of terms have been
applied to the same form (Breed and Grow, 1979), giving a certain degree of confusion to
the attempts for classifying dunes and their associations.
Existent classifications belong to one of the following two groups (Lancaster, 1995).
First are the morphodynamic classifications that relate the dune type to the generating
winds or the sediment supply. The second group, morphological classifications, only
consider the external shape of the dune. A large number of morphodynamic classifications
have been proposed (Aufr~re, 1928; Clos-Arceduc, 1969; Hunter et al., 1983). These latter
Table 18.1. A classification of aeolian dunes (Livingstone and Warren, 1996).
DUNES
FREE ANCHORED
Transverse Linear Star (Sheets) Vegetation Topography
Transverse Sandridge Star Zibar Nebkha Echo

Barchan Seif Network Streaks Parabolic Climbing
Dome Coastal Cliff-top
Reversing Blowout Falling
Lee
Lunette
authors, relating dune courses and wind directions, distinguished between transverse,
oblique, and longitudinal dunes. Research about dune dynamics demonstrates, however,
that different parts of the same dune can be at once transverse, oblique, and longitudinal
(Tsoar, 1983a; Lancaster, 1989b). Mainguet (1983, 1984) differentiated between erosive
forms (parabolic dunes and sand ridges) and depositional morphologies (barchans,
transverse chains, longitudinal dunes, and pyramidal dunes). The morphological
classification elaborated by McKee and Bigarella (1979) is based upon the shape and
number of faces. Taking into account this classification, Lancaster (1995) and Thomas
(1997d) proposed other classifications of dunes. Moreover, yet another type of dune
classification is based on practical finalities, as to their limitations for traffic.
We shall adopt, because of its simplicity, the classification proposed by Livingstone
and Warren (1996), resulting from the modification of a previous one (Cooke et al., 1993),
which is based on the form and the movement characteristics of the dunes, related in some
cases to the presence of vegetation and topographic obstacles (Table 18.1).
4.1. Transverse dunes
The main flow direction is normal to the crest, during almost all the year. These dunes
grow and keep an equilibrium in shape and size. They are longer in the direction
perpendicular to the predominant wind and the windward slopes present less inclination
than those on the the lee side.
The word barchan has a Turkish origin, and is used for isolated, half-moon-shaped
dunes, developed over a coherent substratum, like a pediment, a desert pavement, or a
sebkha. They commonly occur on the margins of ergs and in sand transport corridors
associated with depositional zones. As a general rule, they are small sized, and their height
is one tenth of their width. The smaller ones are quickly mobilized, whereas those that
reach several decameters high migrate very slowly. The stoss-side slope is clearly convex,
Figure 18.11. Smallbarchan in the foreground, and nebkhas in the background, Jaisalmer, Thar Desert,
India.
with an average angle of 12~ whereas the lee-side slope presents an inclination of 33 to 34 ~
(Pye and Tsoar, 1990). The horns point out in wind direction (Figure 18.11). Barchan
advance speed is directly related to sand transport speed over the crest, and presents an
inverse relation with the height of the crest (Bagnold, 1941). Speeds are very variable, and
can reach up to 63 m/yr (see Thomas, 1997d). As sand supply increases, barchans join
together laterally to form barchanoid ridges (Kocurek et al., 1992), also called
megabarchans (Cooke et al., 1993; Lancaster, 1995; Thomas, 1997d). If the amount of
sand increases even more barchanoid complex ridges are produced (Figure 18.3) (Thomas,
1997d), also called compound crescent dunes (Breed and Grow, 1979; Lancaster, 1989a).
These dunes characteristically present a principal ridge, from 20 m up to 80 m over the
interdune area, with a large number of barchanoids on the stoss-side slopes and in the crest
zones. The separation between the bigger ridges ranges from 700 m up to 2000 m. Single
barchans, and barchan associations occupy about 40% of sand deserts (Fryberger and
Goudie, 1981).
The main characteristics of erosion and sedimentation in barchans are the presence of
erosion in the stoss-side slope and sedimentation in the lee-side slope. Howard et al. (1978)
measured, during a period of 2 weeks, the changes of height of the barchan surface. They
essentially found maximum erosion in the centre of the stoss side, and sedimentation in
almost all the lee-side areas (Figure 18.12).
Dome dunes are included in the group of transverse dunes, because they have a similar
orientation and the same type of sand transport. Usually they are low, with a flat crest,
without a slope more abrupt than the others, and with a circular or elliptical form. They are
not common in most deserts, but they are very abundant in the Chinese desert of Taklimakan
(Zhende, 1984). In contrast to other sandy areas where their size is very small, here they
-0.s-....._> /
..; .22,..=-:: z -
Figure 18.12. Changesin surface elevation of a barchan dune during a period of 2 weeks (after Howard
et al., 1978).
reach 40 to 60 m high and 500 to 1000 m in diameter. Dome-shaped dunes are common in
the margins of dune fields. Glennie (1970) pointed out that if wind speeds are low, oval hills
appear and with high wind speeds longitudinal forms are built. Other authors indicate that
the action of strong unidirectional winds is the cause of the origin of dome-shaped dunes, as
it impedes crest forming (Breed and Grow, 1979; McKee and Bigarella, 1979).
Reversing dunes are included in the group of transverse dunes because sand transport is
perpendicular to the crests. They are originated by totally opposite winds, corresponding
to two different seasons. So, the reversing dunes of Oman (Cooke et al., 1993), are formed
by the strong summer monsoons (the bigger ones), and by the opposite weaker winter
winds (the smaller ones).
4.2. Linear dunes
These are the most common dunes and can have a great length, up to more than 200 km.
They are quite straight-edged, parallel, with a regular separation between ridges. The
transverse section is more symmetrical than in the transverse dunes. A large number
of crests and lows can be observed in the longitudinal profile. These are typical for the
great dunes and the middle-sized ones. The smaller ones, however, are quite straight.
These dunes usually move in the direction of the dominant wind, but in some cases they
Figure 18.13. Seif "The Dragon," on the coast of Iquique, north of Chile. Photo: J. Rodr/guez Vidal.
are mobilized in an oblique direction. Linear dunes are also called seif, sif, silk, etc., and
the corridors between sand ridges are known as gassi, goud, etc. (see Cooke et al., 1993).
Among linear dunes it is possible to differentiate simple, composite, and complex
varieties (Pye and Tsoar, 1990). The simple ones are narrow ridges (Figure 18.13) that can
be longer than 200 km with a straight or wavy crest. The transverse profile is straight or
rounded, with a regular interdune spacing and a high dune to interdune area ratio. They are
widely developed in the Kalahari and Simpson deserts (Lancaster, 1995). In these regions
they are from 2 m up to 35 m high, with widths of 150 to 250 m, an interdune spacing of 200
to 450 m, and they can reach up to 200 km long. Sometimes these dunes appear partly
covered by vegetation, as in Nizzana, on the border between Israel and Egypt (Figure 18.14).
A characteristic of these dune fields is the coalescence of crests forming a "Y" or in a diapason
shape (Figure 18.16), which are common in areas where dunes are very close. Interdune zones
can be covered by sand or can be desert pavements, partly covered by vegetation.
Composite linear dunes are formed by two or more dune ridges, very close or
superimposed on the crest. They have been well studied in the Namib desert (Lancaster,
1983), where ridges are asymmetrical, from 25 m up to 40 m high, with a spacing of 1200
to 2000 m, and the interdune areas are covered by sand. Respective to the linear complex
dunes, the big dunes of the Namib, 50 to 150 m high and with a spacing of 1 to 2 km, have
a wavy main crest, and crescentic dunes on the flanks. The interdune areas are covered by
sand with tipples subparallel or slightly oblique to the main crest.
Wind patterns in longitudinal dunes are variable. The most common are unimodal or
bi-directional but, sometimes, several directions have been identified. They are often
essentially associated with permanent winds from a single direction, with superimposed
oblique winds.
Figure 18.14. Simple linear dunes partly covered by vegetation, Experimental station ofNizzana, Negev
Desert, Israel.
I III
~ _ _ A /
g - ',,....'_ . . - " g__.__ D
_ _ _
Figure 18.15. A model of the development of linear dunes (seifs) from barchans; "s" refers to strong
winds and "g" to gentle winds (Bagnold, 1941).
double helix
'Y' junction
/ /
.J /
j ,
linear dune
Figure 18.16. Bagnold's hypothesis for linear dune formation by induced roll-vortices (interpreted by
Livingstone and Warren, 1996).
The origin and dynamics of linear dunes have been the aim of a large number of studies,
reflected in the syntheses of Pye and Tsoar (1990), Cooke et al. (1993), and Livingstone
and Warren (1996). Bagnold (1941) proposed that seifs are derived from barchans by the
lengthening of one of the horns in an area with winds from two directions of different
intensities (Figure 18.15). Others consider that lengthening is produced in the front of a
parabolic dune (Verstappen, 1968). The hypothesis of the origin of linear dunes from
helicoidal flows (Bagnold, 1953) is still considered at present (Figure 18.16), supported by
field and wind-tunnel research (Tseo, 1993). Dunes develop in convergence zones and
corridors occur between them. This model seems to explain also the spatial regular pattern.
4.3. Pyramidal and network dunes
The common characteristic to both types of pyramidal and network dunes is the
occurrence of a certain number of slip faces developed as a consequence of the coexistence
of different wind directions. Pyramidal dunes are also called star dunes, cones, sand
mountains, ghourds, and other local names (Pye and Tsoar, 1990; Cooke et al., 1993).
They are the biggest dunes present in ergs, and in Algeria they can be more than 400 m
high (Wilson, 1973). As a consequence, they support a large amount of sand and they are
placed in zones with a high sedimentation rate. Besides their large size and pyramidal
morphology, they also present three or four radial arms (Figure 18.17). About 11 per cent
of all dunes belong to the pyramidal type (Fryberger and Goudie, 1981) and in the Great
Eastern Erg of Algeria they occupy 40% of the region (Breed and Grow, 1979). The arms
coalesce on the top of the dune, where the slope inclination ranges between 15 and 30 ~,
whereas near the bottom of the dune, slope inclinations are lower (5 to 10~ and it is
possible to see small crescentic and reversing dunes there. Spacing between dunes varies,
according to Lancaster's compilation (1995) between 150 and 6700 m, and dune widths
Figure 18.17. Pyramidaldune, Erg Chebbi, Morocco. Photo: J.L. Pefia.
between 180 and 6000 m. Generally there is a clear direct relationship between height of
the dune and spacing in pyramidal dunes (Lancaster, 1989c). These dunes can grow
several centimeters every year, and it seems that they migrate slowly, instead of keeping a
compensatory balance (Nielson and Kocurek, 1987).
The wind studies done over pyramidal dunes in different deserts indicate that winds are
multidirectional with variable energy (Lancaster, 1995), especially during the months
when most of the sand transport takes place. As a consequence, net transport is usually
low. The interactions between the seasonal wind changes concentrate sand sedimentation
in the central parts of the dune. The main arms appear approximately transverse or slightly
oblique to the directions of the local prevalent wind. Minor arms of the pyramidal dune are
placed parallel and transverse to the beating strike of secondary winds (Figure 18.18)
(Lancaster, 1989c). Besides this, topographic barriers have exerted a clear influence in the
existence of pyramidal dunes (Breed and Grow, 1979) because they modify the regional
winds, increasing their variability or creating traps affecting sand transport.
Network dunes are developed under a continuous cover of sand, and are formed by dunes
a few meters high. They are interlaced, alveolar, or rhombic-shaped networks of several
hundred meters. They have been interpreted as groups of transverse dunes overlapped as a
consequence of the different directions of the seasonal winds (Aufrbre, 1935). Their
complexity is variable, depending on the intensity and persistence of the seasonal winds.
4.4. Sand sheets and zibars
Sand sheets are accumulations of small dunes without slip faces that are developed on a
flat, smooth terrain (Figure 18.19). Sand strips and sand ribbons are elongated form
Transverse dune
Seasonal reversal
of crest line
Reversing dune
with incipient
star dune arm
Development of arm
by secondary flow
Accentuation of
arms by third wind
direction and secondary
flow
Figure 18.18. Modelfor the formation of pyramidal dunes (star dunes) by the development of secondary
wind flows, while the dune is mobilized in a multidirectional wind regime Modified from Lancaster,
1989b.
varieties of the sand sheets (Breed and Grow, 1979). The surfaces can be irregular,
undulated with or without ripples, and so forth. Many sand sheets are the basis or support
of other mobile dunes. Globally, the extent of the sand sheets is more than one and a half
million k m 2 and, at the beginning of this chapter we have pointed out that the maximum
surface known is ---100,000 km 2, in the borders between Egypt, Sudan, and Libya.
Figure 18.19. Wavysand sheet, with some ripples. Matagusanos valley, Andean Precordillera, San Juan
Province, Argentina.
The thickness of the deposits in sand sheets varies from a few centimetres up to 10 m
(Breed et al., 1987). Grain size fluctuates between fine sand, sometimes with a significant
content of silt, and poorly classified coarse sand (Pye and Tsoar, 1990). In Egypt we deal
with sandy plains with residual deposits of fine gravel and pebbles that form the surface
levels. Generally, with rare exceptions, sand sheets are poorly sorted deposits. Underlying
levels are usually formed by subhorizontal layered sand beds (Lancaster, 1995).
Sand sheets are affected by several factors conditioning their formation (Kocurek and
Nielson, 1986). Vegetation reduces the movement and dune growth, and consequently,
accretion of horizontal sand laminas occurs. These circumstances of sparse vegetation
are common on the margins of deserts. Where a coarse sand layer occurs, it is likely that
the fine fraction has been exported by deflation. If there is a shallow phreatic level or
periodic or seasonal floods occur, formation of dunes is made difficult or impeded.
Finally, the presence of surface crusts or algal mats makes impossible the transport of
sand in order to build dunes.
As happens in sand sheets, zibars are flattened areas with dunes without slip faces, with
thin deposits. They form tipples and megaripples, sediment is coarse sand, and they are
settled on a hard substrate. This explains that the term zibar, derived from the Arabic word
zibara, means a hard surface passable by vehicle. They are common in areas where fine
materials have been deflated. They can be found in different environments, from the centre
of the Sahara Desert to the zones with scarce vegetation in southern California (Cooke
et al., 1993).
4.5. Dunes anchored by vegetation
The areas that present dunes with low mobility are suitable places for plant growth,
because water retention by sand occurs (Tsoar and Moiler, 1986). This vegetation cover
modifies the wind flow and favours aeolian sand trapping.
Dunes anchored by vegetation, also called phytogenetic dunes (Cooke et al., 1993), are
originated by the stabilization of aeolian sand around the plants (nebkhas and coastal
dunes) and by the erosion of surfaces covered by vegetation (blowouts and parabolic
dunes). Even though they are almost fixed dunes, many of them can undergo changes in
their shape and size, during seasonal or even longer cycles.
The most common type of dune anchored by vegetation are the nebkhas, also called
shrub dunes, shadow dunes, hummock dunes, rebdou, and other terms (Pye and Tsoar,
1990; Cooke et al., 1993). They occupy enormous areas in the lowlands of the semiarid
areas. They look like small ridges trapped by the plant, elongated in the sense of the sand
transport, and located on the lee side (Figure 18.20 and Figure 18.11). The particles
forming the nebkhas are sand, silt, and clay aggregates. The shape of a nebkha is a
consequence of its size, density, and the growth of plants (Capot-Rey, 1957). Plants need
to be at least 10 to 15 cm high for an effective trapping of sand. The size of the nebkhas
varies considerably; they can reach up to 10 m high and 1 km long when they get trapped
by a group of trees (Cooke et al., 1993). The dune reaches its maximum height near to the
plant in the lee-side zone, because here the inverse flows are higher. Plant longevity is also
important, because only long-living species can retain a large quantity of sand. On the
contrary, many nebkhas are ephemeral, because they are placed on annual plants. In some
cases perennial plants suffer changes in the phreatic level, and rain and sediment supply
Figure 18.20. Groupsof nebkhas in the margins of the Thar Desert, Jaisalmer, India.
produce sudden modifications in the aeolian dynamics (Gile, 1975). Some nebkhas are
affected by processes other than the aeolian ones, such as impact and splash by raindrops,
runoff and animal activity. Therefore, there is a sum of processes, and it is difficult to
evaluate the input of each one (Cooke et al., 1993).
Blowouts or deflation basins are morphologies produced by deflation, as closed
depressions on unconsolidated aeolian sands, up to the limit of plant roots (Livingstone
and Warren, 1996). Some authors use the term blowout in a restricted way, only for those
generated in vegetated areas, whereas others such as Lancaster (1989c) extend
the terminology to those shaped on bare dunes (Figure 18.21). Blowouts occur in
continental dunes, essentially in the desert margins, and in vegetated coastal dunes. One of
the most studied groups of blowouts is the Meijendel one, in the coastal dunes of Holland
(Figure 18.22) (Jungerius, 1984). Blowouts elongate in the direction of the wind, and the
major axis is 10 to 30 m long, exceeding 100 m in some cases. Vegetal cover is very
important in the development of these forms. Therefore, loss of vegetation reduces the
roughness, increasing, in consequence the scour surface and the export of material. This
loss can be produced by overgrazing, rodent activity, and fire. Also the plants of the upper
parts of the dunes are the first to become dry and so aeolian erosion takes place
preferentially in these areas. Sometimes vegetation can be pulled out by hurricanes and
unusual storms. Deflation deepens in the area without vegetation. Winds reach their
maximum speed in the bottom of the depression, originating vertical vortices that spill the
sand towards the edges of the hollow. In areas where rain precipitation is considerable,
water erosion helps to widen the blowout (De Ploey, 1980). Exported sand is settled on the
plants at the edges of the closed depression, and the aeolian erosion becomes maximum
during the dry season, where the depression widens and it is able to migrate. On the other
Figure 18.21. Blowout developed on the dune chains of A1Liwa, Desert of Rub al Khali, United Arab
Emirates.
Figure 18.22. Blowoutwith sandy bottom, and vegetation of shrubs and trees. Small blowouts can be
seen in the foreground, Meijendel, Coastal dunes of Holland.
hand, most blowouts have a short existence because they get quickly covered by
vegetation, and so, half of the blowouts of Holland disappeared in about 9 years (Jungerius
and Van der Meulen, 1989). Entrained sand takes the form of a fine layer with a plume
shape and, when erosion is intense, a dune appears in the leeward zone, which advances
over the existing vegetation. As a consequence, the area with floating sand is enlarged and
a parabolic dune can be generated.
Parabolic dunes are U- or V-shaped, with the two arms partly covered by vegetation in
the stoss-side area. They are localized in cold climates, coastal areas, and desert margins.
The largest zone of parabolic dunes is found in the Thar desert of India, covering an area of
about 100,000 km 2 (Verstappen, 1970) (Figure 18.23). Arms are 1 to 2 km long, and the
front is 10 to 70 m high: they essentially advance by means of avalanches. Many of these
dunes coalesce, because they have different movement speeds, giving rise to imbricated or
multilobated parabolic dunes, with multiple crests and slip faces (Wasson et al., 1983). Few
data are known about advance speeds of parabolic dunes. They fluctuate between 0.05 m/yr
in zones of dense vegetation and 13 m/yr in some coastal regions. These dunes seem
basically associated with the existence of a vegetal cover with a moderate development and
with a unidirectional wind regime. The role of vegetation in the genesis of the parabolic
dune is fundamental, because it protects the less mobile arms against aeolian action,
allowing the advance of the central part. Parabolic dunes seem to derive from blowouts and
they grow using the sand supplied from the underlying sediments. The sand supply can stop
if a cohesive bedrock crops out or if the capillarity front, where sand is moist and unable to
be entrained, is reached. In other cases the phreatic level can appear, originating swamps and
Figure 18.23. Parabolicdune front slightly covered by vegetation, Tena-Jaisalmer, Thar Desert, India.
lakes. After this, the arms of the parabolic dune converge, decreasing their height and finally
a sand tongue passes over the apex (Figure 18.24) (Pye, 1982).
Coastal dunes are the result of a combination of aeolian processes interacting with
lithological, marine, fluvial, slope, edaphic, and cultural processes (Livingstone and
Warren, 1996). The type and extent of these dunes depend on the quantity of available
sand and the supply rate. This is a function of the existence of a sediment source and the
strong littoral drift towards land. Their origin is related to rivers, beaches, coastal erosion,
and estuarine sands.
Coastal dunes can be classified, in a simplified way into primary dunes, with their
development directly related to coastal processes, and secondary dunes, where the sand
has a littoral origin, but other processes also take place. These last dunes include aeolian
sheets, blowouts, and parabolic dunes.
If sand supply to coastal dunes is poor, the sand spreads inland forming a thin layer.
On the other hand, if sand supply is abundant, ridges parallel to the coast line appear
(Figure 17.2), that move slowly towards inland, and they can be more than 100 m high,
keeping their slip faces straight (Goldsmith, 1985). Where sand supply is very abundant,
vegetation can be buffed, and dunes very similar to the desert ones appear. These dunes
are known as transgressive, and are usually of a transverse type. If vegetation is very
high, a continuous ridge is developed or, sometimes, a system of parallel ridges in a
prograding coast appears. These ridges can grow at high speeds. Where vegetation is
sparse as in the arid coasts, ridges are discontinuous and are formed by sand mounds
lined up with vegetation.
The evolution of coastal dune ridges can originate very complex morphologies, with
blowouts, transverse dunes and parabolic dunes, by successive building of parallel ridges,
which development can last for hundreds of years.
Figure 18.24. Stages in the growth and eventual dissipation of elongate parabolic dunes in the Cape
Flattery area, Australia (Pye, 1982).
4.6. Topographically anchored dunes
When moving aeolian sands encounter any topographical obstruction, like hills, cliffs,
blocks, or buildings, accelerations and expansions of the wind flow around the obstacle
occur. These circumstances provoke erosion or sand accumulation, and static dunes
occur.
Experiments carried out in wind tunnels (Tsoar, 1983b) have provided a large number
of data about the relationship between topographic obstacles and the generation of dunes.
If the wind blows perpendicularly into a slope presenting more than 60 ~ of inclination,
echo dunes appear, separated from the obstacle a distance equivalent to three times the
height of the obstacle. At the bottom of the cliff an inverse flow appears, sweeping the sand
and generating a corridor between the obstacle and the sand ridge. The small initial echo
dune increases its height by this mechanism, until it reaches an equilibrium condition,
corresponding to 0.3 to 0.4 times the height of the obstacle. If the cliff is sinuous, the dune
shape changes along its length.
Where the slope of the obstacle presents an inclination of less than 30 ~, sand is
transported over it, but with higher inclinations sand is deposited and climbing dunes are
Figure 18.25. Climbing siefs, Aswan, Egypt. Photo: J.L. Pefia.
formed (Figure 18.25). They are very common on desert hills and cliffs. The sands that
reach the crest of the cliff settle down because they enter a zone of reduced speed. These
aeolian accumulations are called cliff-top dunes. Leeward of wide obstacles the wind
calms down and falling dunes are deposited. When the obstacle is narrow, lee dunes
appear, sand ridges are commonly straight, which because they are sheltered by the
obstacle, can extend for long distances. In the Draa of Malichigdane in Mauritania they get
up to more than 100 km long (Breed and Grow, 1979). Several dunes present around
rounded craters of Mars have been interpreted as lee dunes (Greeley and Iversen, 1985).
Lunettes, half-moon-shaped dunes, have been described in the previous chapter.
5. Desert dust
During the last two decades of the 20th century the study of aeolian dust deposits as
paleoenvironmental records has steadily increased. They include terrestrial sequences,
deep-sea sediments, and cores from the polar ice caps. Research provides information
about changes on the extent and intensity of aridity, glacial periods and localization of
glaciers, and changes in the global atmospheric circulation (Pye, 1995).
The entrainment, transport, and sedimentation of aeolian dust are very important for the
geomorphologist, but also present a relevant interest for meteorologists, climatologists,
ecologists, and environmental scientists. Besides the geomorphic implications that open a
whole range of research possibilities (loess, varnish, crusts, erosion, and so forth), there is
also the interest of applied studies, which are more and more important because of the
increasing number of inhabitants in arid zones (P~w~, 1981; Middleton, 1997).
5.1. General characteristics
Desert dust consists of small particles carried away by the wind of arid environments in
suspension and deposited afterwards. Atmospheric dust can come from different
environments than those of arid zones (cosmic dust, volcanic dust, marine salts, etc.)
and the source areas can be distinguished by their composition and size.
In western Europe desert dust is a familiar feature, because almost every 7 years or so
occur the so called "red rains" of aeolian dust originating from the Sahara desert, that can
be observed as a thin film on top of cars. This dust settlement rate is on average
0.25 kg/m2/yr (Goosens and Offer, 1990). Globally it is possible to estimate that dust
carried out from deserts is about 1800 to 2000 million tons per year (D'Almeida, 1989).
A dust storm is recognized when visibility is less than 1000 m. In Egypt some ten storms
are observed every year, about 30 in China, and in the Mexican capital and in Kazakhstan
an average of 60 storms occur every year (Goudie, 1978).
According to grain size, there are two main types of desert dust that reflect the distance
they have travelled. If dust has been transported less than 100 km, sizes are between 0.005
and 0.05 mm; if transport distances are longer, the diameter is smaller than 0.002 mm
(P6w6, 1981). The largest particles are transported by wind whirls and dust storms and
produce extensive loess deposits, whereas the smaller ones move like an aerosol in the
troposphere, and remain suspended until they are entrained by the rain and deposited.
Desert dust is composed of inorganic and organic materials. The most abundant mineral
is quartz, accompanied by feldspars, calcite, dolomite, mica, clay minerals, oxi-
hydroxides, heavy minerals, gypsum, halite, and opal. The organic part is composed of
diatoms, phytoliths, spores, and pollen. As we could expect, composition changes
according to the source area and the distance.
5.2. Source area and generating processes
The main currently active areas providing aeolian dust are the subtropical deserts, which
form a wide belt from western Africa to Central Asia, and the arid and subhumid regions
where dry soils are exposed to strong winds during several periods through the year (P6w6,
1981; Middleton et al., 1986)(Figure 18.26).
The areas covered with loose sediments that contain significant quantities of sand and
silt, but are poor in clay, are the most favourable for generating aeolian dust. One of these
environments is the poorly-sorted and unvegetated sediments of the ouads. They are
usually braided channels with a great lateral migration.
The playas, chotts, and littoral sebkhas, although they vary enormously in their
composition, contain zones characterized by the presence of fine grained material, which
are another source of aeolian dust. Alluvial fans are supply zones, especially their middle
and distal areas, where the percentage of fine particles is important. Ancient dry lakes also
can supply diatom dust, salt, gypsum, and clay minerals. The large regions covered
with loessic sediments are important dust source areas, particularly where vegetation is
nonexistent. An important source area of dust comes from volcanoes, as with the
emissions from Mount St. Helens in USA in 1980, and Pinatubo in the Philippines in 1991,
where large quantities of dust were emitted into the atmosphere, and transported
afterwards (Pye, 1987; Cooke et al., 1993; Livingstone and Warren, 1996). Mechanisms
Figure 18.26. Majordeserts, directions, and distances of dust transport (after P6w6, 1981).
that form fine particles are quite varied. Clay-sized particles are produced essentially by
chemical weathering, but the origin of silt-size particles, that basically form the loess, is
controversial. Weathering by frost can play a fundamental role in the generation of fine
particles in cold deserts. Processes of thermoclastism, wetting and drying, besides the
haloclastism or salt weathering, are the basic mechanisms of fine-grained material
production in deserts. Besides these, we can add aeolian abrasion. In beach environments
agglomerates are produced from silt and clay particles that are exported by the wind.
Aeolian dust can incorporate particles from chemically-weathered profiles, developed in
tropical humid environments. Finally, another source is of organic origin; phytoliths,
diatoms, radiolarians, spicules of echinoderms, pollen, spores, and so forth.
5.3. Entrainment, transport, and sedimentation
The entrainment of dust particles from the soil surface is a function of wind speed, and it is
controlled by the type of wind, sediment characteristics, and the presence of obstacles
(vegetation cover, topography, and so forth) (Middleton, 1997). Strong winds able to
mobilise and transport dust particles occur under very different conditions, and at different
scales (Pye, 1987). These dust masses are identified and studied using satellite images and
the classical meteorological observations.
Wind types transporting dust vary from simple gusts to dust storms with advance fronts
up to 1000 km long (Figure 18.27). Dust devils are small convective vortices that appear
when an intense warm up of the ground surface has taken place, causing the movement of
air and dust upwards. These vortices are visible from a long distance and present an
inverted cone shape. They commonly reach heights between 3 and 100 m, although they
can exceed 1000 m (Idso, 1974). Diameter on the ground is 0.5 to 3 m. They move
laterally and can cause significant damage.
Figure 18.27. Schematicrepresentation of modes of aeolian dust transport and deposition (Pye, 1995).
Dust storms, also called haboob (an Arabic word that means violent wind) are probably
the result of descending air currents from large cumulonimbus clouds. They look like a
dust wall of 300 m up to 3000 m high, which can advance at speeds up to 200 m/s. As we
can see on satellite images they measure 500 to 600 km wide and extend up to 2500 km
with a plume shape (P6w6, 1981).
Other winds that also transport very fine particles are the dust plumes that move parallel
to the soil surface as a horizontal vortex. The shamal is a wind that blows towards the north
during summer in Iraq and the Iranian Gulf. Most of these winds loaded with dust do not
reach 1 km in height and their ground speed is less than 40 km/h. Another transport of
aeolian dust is the katabatic wind, that sweeps the sediments from the foothills of large
mountain ranges, such as with the zonda wind in Argentina and the Santa Ana, in
California. Similar winds occur in the Andean altiplano and on the plateau of Tibet (Pye,
1987; Livingstone and Warren, 1996; Middleton, 1997).
In semi-arid zones of higher latitudes, the passing of cold fronts is probably one of the
most important causes in the mobilization of aeolian dust. Huge walls of dust appear, as in
the haboobs, up to 3 km high, and move at a maximum speed of 40 m/s. They have been
described in the north of China, Israel, Egypt, in the High Plains from Texas and New
Mexico, and in the south of Australia (Pye, 1987).
Sometimes dust reaches considerable heights, and it is transported at a great speed by
the jet streams, coveting distances of thousands of kilometers (Figure 18.26). Thus
Saharan dust travels to central and northern Europe, and crosses the Atlantic Ocean,
reaching the Caribbean Islands (Prospero et al., 1970).
Sedimentation of aeolian dust takes place in different ways: (1) wind speed and
turbulence reduction; (2) aggregation of particles with electrostatic charges; (3) dragging
of dust suspended in the atmosphere by rain (Pye, 1987). Although sedimentation also
takes place in oceans and wet areas, we shall deal essentially with dust deposited in arid
zones. The quantities deposited are variable, although the average is about 100 t/kmZ/yr
near the source areas (Goudie, 1995). This amount decreases quickly with the distance
from the area of origin. Desert dust with a coarser size travels within the lower parts of the
atmosphere, and often is deposited when it encounters obstacles perpendicular to the wind
flow. Nevertheless the finest particles that travel in the higher parts of the atmosphere are
not affected by the roughness of the terrain, and can remain suspended for long periods of
time (Coudr-Gaussen and Rognon, 1988; Coudr-Gaussen, 1991).
In hyperarid zones only small quantities of dust occur, because precipitation is rare and,
generally, obstacles are not common. On the margin of deserts, however, corresponding to
semiarid zones, there are great quantities of aeolian dust deposited. These particles are easily
trapped by the disperse vegetation. In the north of Nigeria, up to 3 m of desert dust have been
deposited during the last 40,000 years (McTainsh, 1987) and, at present, sedimentation rates
vary between 0.1 and 0.7 mm/yr in different zones of Israel and the Sahel.
5.4. Geomorphological implications
Desert dust has a considerable importance to the better understanding of certain

geomorphic processes that have strong implications for human and economic activities.
Some superficial formations appear because dust deposited in the vicinity contributes to
their genesis in semi-arid and arid areas. So, the supply of calcium carbonate to the soils
produces caliches in areas without calcareous rocks, as we can interpret from the studies
on calcretes in Namibia and southeast of Spain (Bltimel, 1982). Also dust rich in silica can
be important in the formation of silcretes in deserts (Summerfield, 1983). Likewise,
sedimentation of gypsiferous dust is the main cause of the formation of gypsum crusts in
southern Tunisia (Coque, 1955). Desert varnish, that forms a thin film rich in manganese
and iron, covers rocks in arid zones. As was pointed out in a previous chapter, particles that
form the varnish are provided by the sedimentation of aeolian dust. Laminations that
constitute the varnish are analysed carefully and paleoclimatical interpretations are carried
out, according to the fluctuations in manganese and iron contents (Oberlander, 1994).
Desert dust can also affect the weathering developed on buildings. In dune zones,
sedimentation of dust and salts can form a crust that is enough to stabilise the dunes,
allowing the establishment of biocrusts and plants (Pye and Tsoar, 1987). Aeolian dust is
also collected in playas, sebkhas, alluvial fans, and ouads causing modifications of their
sediment composition. These sediments are quickly mobilized by water within their
depositional environments. Many soils from desertic and peridesertic regions contain
significant quantities of aeolian dust, and some soils are virtually formed by these
allochtonous supplies. According to Rapp (1984), the terra rossa from Spain, Italy, and
other Mediterranean countries are formed essentially by reworked dust coming from the
north of Africa. This desert dust fertilizes many soils, affecting even some of those
developed in the Caribbean Islands.
Desert dust deposited in oceans and seas constitutes a significant percentage of oceanic
sediments. The study of these accumulations provides very useful information on
modifications of wind systems, and about environmental changes that have taken place on
the continents during the Quaternary period (Middleton, 1997). Therefore, this research
provides information on the arid phases and their intensity in terrestrial environments. Ice
accumulations, especially in the polar ice caps, preserve a record of aeolian dust that is
known because test drillings have been done. In Greenland, during the Last Glacial
Maximum (18,000 years ago), dust concentration was 40 times greater than now. Phases
with high dust content are followed by others with small or nonexistent quantities of dust,
indicating the alternation of periods with high and low aridity.
5.5. Loess
Loess, a term derived from the German lgss, is a clastic terrestrial sediment, composed
essentially of particles of a silt size and formed by the accumulation of aeolian dust (Pye,
1995). It presents a variety of colours; gray, white, yellow, brown, and red. It is possible to
differentiate between primary loess, with an aeolian origin, and secondary loess,
redeposited or originated by nonaeolian processes. In loess outcrops, a certain degree of
chemical and biological weathering occurs, and also edaphogenic processes.
Loess accumulations (Figure 18.28) cover about 10% of the Earth's surface above sea
level (Prcsi, 1968), being mainly developed in Eurasia, USA, and Argentina (Figure 18.2).
In China loesses occupy more than 1 million km 2. The layer of loess has an irregular
thickness and covers the irregularities of the terrain. Thickness is usually less than 30 m,
although in Lanzhou, China, it exceeds 300 m (Derbyshire, 1983b).
Figure 18.28. Loess accumulations in the region between Jaipur and Agra, India.
Loess mineralogy is very variable, reflecting the composition and evolution of the
source area. Usually quartz predominates (45 to 55%) accompanied by feldspars,
carbonates, heavy minerals, volcanic flint, and clay minerals. Loess has more than 50% of
silt-sized particles, but the contents of clay and sand are very variable. Where loess
contains more than 20% of sand, it is called sandy loess, and if the clay content exceeds
20%, clayey loess (Pye, 1987).
Loess has a mechanical behaviour with a tendency to collapse. It supports vertical loads
when it is dry, but its shear strength decreases considerably when wet, and the underlying
material flows and slides (Middleton, 1997). Where loess is dry it presents vertical fracture
systems in scarped outcrops.
The formation of a loess deposit can be simple or quite complex. Loess can be
generated by weathering and further deflation. Usually there is an intermediate phase of
fluvial transport, previous to the aeolian action. Initial particles can also be the result of
glacial weathering and/or erosion, or can be transported and deposited by fluvioglacial
action, and finally mobilized by deflation (Smalley, 1972).
There are many processes producing particles of a silt size; particles from bed rock;
glacial crushing; gelivation; fluvial abrasion; aeolian abrasion; haloclastism; chemical
weathering; clay-pellet aggregation, and biological processes (see Pye, 1995). According
to the generating processes some authors denote the extensive loess deposits related to
Quaternary continental glaciations, periglacial, or glacial loess. Others refer to "peri-
desert" loess in connection to those on the margins of arid zones.
In the deserts of central Asia particles are weathered in high mountain environments
and later are transported and deposited by the rivers in lower zones, where they are
entrained by deflation. In those areas with active tectonics, large quantities of sediments,
including silts, were produced during the Quaternary era (Smalley, 1990). Due to the
environment where they were generated, they are called "peri-mountain" loess. In deserts
located on shields and platforms, extensive deposits of loess do not occur, because those
areas have a low relief, are tectonically stable and, as a consequence, have low sediment
production, in contrast to the deserts of central Asia (Smalley, 1995).
An accumulation of thick and extensive loess deposits needs an important supply of
dust during tens or even hundreds of thousands of years in order to be produced, and the
existence of topographical and vegetation barriers to stop the dust. In the deserts of low
relief the production of sediments is poor (peri-desert loess) whereas in deserts of
mountains with neotectonic activity sediments are generated at a very high rate (peri-
mountain loess).
Chapter 19
Applied geomorphology and arid regions
1. Introduction
The applied geomorphology of desert areas is not very different from that of other
morphoclimatic areas, although it does have some peculiarities. The lack of water, scant or
zero plant cover and much eroded soils leave the regolith with no efficient protection
against the action of atmospheric agents (Tricart, 1969). The most important processes in
arid areas are wetting and drying, salt weathering, water erosion, channel changes,
flooding, and wind erosion.
Practical studies on arid regions are important because of the extraordinary extension of
these environments; the world's two largest morphogenetic systems are those made up by
the arid and cold regions. It is difficult, however, to generalize about the effects of
geomorphology in these areas because of their great climatic, geological and tectonic
differences. Further, these regions are usually underdeveloped and scantily populated, and
data on even their basic features, such as their climatic and hydrological characteristics,
are few and geographically disperse. In most cases, the only data available come from a
few rare and rather specific studies (Guti6rrez and Sancho, 1993).
In 1979, the population of the world's arid areas - 651 million - was 15 per cent
of the planetary total. Between 1960 and 1974, however, an increase of 63.5 per cent
was experienced. Some 1200 million people inhabited these regions by the year 2000.
This population is distributed mainly as large concentrations; there are 355 cities
over the world's arid regions with more than 100,000 inhabitants, although most of
them are in Asia. The geomorphological and environmental problems caused by the
expansion of these cities are numerous (Cooke et al., 1982; Goudie, 1990a,b;
O'Hara, 1997).
Many of the landforms of these dry regions were generated under climatic conditions
different to those of the present day. A detailed study is therefore needed to differentiate
between those shaped by modem processes and those that are relicts of earlier times
(Guti~rrez, 1986). It is also important to decide on the basis and goals of any practical
geomorphological study to be performed in a desert region. This has been clearly stated by
Cooke (1977):
Applied geomorphology in deserts is primarily concerned with the survey and evaluation of
landforms, superficial materials and processes, and with managing, monitoring and predicting
landform and process changes. It relates to practical problems in which geomorphological
information normally only forms a part of the required body of knowledge, and its
philosophical basis (in so far as it exists independently of "pure" geomorphology) is
inevitably dictated by the demands of specific problems. Axiomatically, the applied desert
geomorphologist must be able to talk to and work with those in other cognate disciplines, and
he (or she) must be able to communicate ideas comprehensibly to others who may have no
geomorphological knowledge.
2. Weathering
The weathering of desert areas is mainly driven by strong variations in temperature and
humidity. Fluctuations in temperature give rise to insolation weathering, and at high
altitude to freeze-thaw phenomena. Variations in humidity encourage wetting and drying
processes as well as haloclasty. The lack of water impedes chemical modification, and
biotic weathering is mainly limited to the physical and chemical degradation caused by
algae and lichens. Weathering in these areas is therefore rather peculiar and gives rise to a
set of micro-landforms such as honeycomb, tafoni, caverns and gnammas, which can be
very abundant in medium-grained rocks (Mabbutt, 1977).
2.1. Weathering by wetting and drying cycles and salt weathering
Weathering by wetting and drying cycles leads to the disintegration of rocks by flaking and
cracking (Oilier, 1984). It is an important agent of erosion in desert areas, causing
considerable breakage of rocks (Oilier, 1977). The water circulating in the pores of rocks
carries very soluble salts, especially chlorides and sulphates, and when these precipitate
because of evaporation they can have an even greater disruptive influence than freeze-
thaw. This process, known as haloclasty, acts alongside wetting and drying. Of all the
weathering processes at work in desert areas, haloclasty, which involves the growth of
crystals, volumetric expansion, and salt hydration (Cooke, 1986; Cooke et al., 1993), is the
most important (Doornkamp and Ibrahim, 1990). It can disintegrate rocks into particles the
size of silt. Evidence of this is mainly seen along the edges of salt lakes, in ephemeral
channels, and where the area of capillarity intersects the soil surface. The builders of
ancient civilizations knew the disruptive effects of salts on rock very well (Oilier, 1977).
The Great Pyramid of Egypt (Fig. 19.1), for example, was originally covered in a
protective layer of resistant limestone slabs. Unfortunately, about 1000 years ago part of
this covering was destroyed and the weathering of the pyramid's core began (Emery,
1960). Although natural materials are still used in construction today, many structures are
made of concrete, but the weathering behaviour of this material can be studied in the same
way.
The effectiveness of haloclasty (Cooke et al., 1982) depends on environmental
conditions such as those produced by daily fluctuations in temperature and humidity
(which give rise to the crystallization or hydration of salts), and on fluctuations of the
capillary front. Equally important are the properties of the materials affected, especially
those related to porosity and pore distribution. Finally, different salts have different effects:
the most aggressive are sodium and magnesium sulphate (Fig. 11.7) (Goudie et al., 1970;
Goudie, 1974 and Cooke, 1979).
The most common and most serious problems related to engineering structures (Cooke,
1986; Goudie, 1994a; Goudie and Viles, 1997; Goudie, 1998) concern fluctuations in the
capillary front. These can affect the foundations of buildings, roads, canals, oil pipelines
Figure 19.1. The upper part of the Great Pyramid of Egypt is covered by unweathered limestone slabs.
The underlying stone, however, shows marked weathering. Photo: J.L. Pefia.
and other structures. Strict control over the materials used is therefore vital, a good idea
being to use only those with low salt content and which are highly resistant to salt
weathering. Laboratory analysis of their durability is recommended. Some buildings
clearly show the height reached by the capillary front. In the arid areas of Bahrain, Dubai
and Egypt, this is never higher than 4 m, and usually between 2 and 3 m (Cooke et al.,
1982). In the semi-arid area of Alcafiiz, in the Ebro Basin of Spain, however, heights of
5 m have been reached (Fig. 19.2). The chlorides and sulphates of sodium, calcium and
magnesium are the salts that most affect concrete, in which they can cause expansion and
disintegration. The use of sulphate-resistant cements is recommended (Mehta, 1983).
In the construction of roads, areas made wet by capillary rise (Fig. 19.3) should be
avoided. In such places, salt weathering causes cracking, hollowing, crumbling and
disintegration (Cooke et al., 1982). A way of assessing the risk of haloclasty is to measure
the electrical conductivity of water in the vadose zone and to make salinity maps
(Fig. 19.4).
2.2. Dissolution
Sweeting (1972) indicated that karst landforms in carbonate formations do not develop
when rainfall is below 250-300 mm. Therefore, the karstification seen in the calcareous
rocks of arid areas can only have been produced during times that were wetter than the
present. Chlorated and sulphated materials, however, are more soluble than limestone and
are susceptible to karstification. According to Jakcus (1977), gypsum and halite are,
respectively, 183 and 25,000 times more soluble than calcite in distilled water at 20~
These figures show the intensity and velocity of karst development in salt and gypsum.
Figure 19.2. An undulating capillary front in the sandstone slabs of the Calatravas Castle (18th century),
reaching 5 m above ground level. The most important weathering is seen in the lower slabs. Alcafiiz,
Province of Teruel, Ebro Basin, Spain.
Only minor salt enters road

by vapour transference or
migration along grain surfaces
Salt transferred into road by I
'Dry' salt efflorescence and capillary moisture movement I
precipitation on voids I /
Moisture zone E ~ / ~ t
, _.. ~ J ~ / Maximum salt crust growth ' "~- / N,
YIIIl~lllll)Clllllb/u/,,.~ I I . \ /////H/1/1/1/1///////,////////// ,
~ . . . . ! ~ . . ] . i . Z , , ~ ' ~ r / / ~ /z// Damp salt
Damp - crusl
salt-crusl ~ j~l l////~//////I/t/ll c//a~)~(/a~/~/ri/n/g/el ~zone'D~:.....~......~,~
~_
. " ~::i iiiiiii:ii.~i~i:.:.~i::.ii~i~i:.~i~i:::
:::::: i
c,Water tab,; =========================

::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::~~i:.
~:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
:i:!:i:i:i:i:!:i:i:i:i:~:~i::::::::.'::.::.':-~'."::':.:::!:.. ":..., ,:::::,. .:::::.. :.:.~.. ..:~i. ~i.': ~w
~,.~~~~';:".'~::?.i
[ ::::::::::::::::::::::::::::::::::::::::::::::::::::::::
~ i . ...............
9 .....9 .,.,,, ........ :.,
~.~:~.~..~..~..~.~..~..-~.~.~...:~.~..~..:~
~ ~ i i . . ~......~.~.~:~.
i i i . ~ . ~ i :..::~.~:?~,...:.,,.~!~,-., ,:,..~~1,...~'~.~,.~'.
............ ..~,...........
.,,,,,,,,:,,,,,,,,,,.,,,,,,,,,,,,,,,,,,.,,.,,,:,:,-::,~,,-~,,,.,,~,:~,,,,,,,,,, ,, ~.,,,,
~~e,..,.~.,,,~,,,,.":~.
:~.,,~..~~.'
,,.., ,,~,,,~ , ~ !
.o ~ .,..:..,.,..<:.
~ , ~,.~.
.
~,.,,,~,,,," ,,-, ~ ~ ,','
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
::::::::::::::::::::::::::::::::::::::::::::::::::::::
::.:~.'::::'.......':i~:::':::':'.........:...:... ~ :~.~x.,. . . . . . . . ~'~..~4.::
::::::::::::::::::::::: ................. ..~:.:,.,:.:.:.:.:.:...:.:.:.:.:.:.:~:.:'..:.:.:~:~:.:.x.~.~:~:~:~*~:.:.:~:i:!:i:::~.~::;~::~:.::~i:!.~ i:~ ..
Salt transferred into road

by evaporation
Figure 19.3. Schematic section to show the relevance of soil moisture zones to road construction (after
Fookes and French, 1977).
44 46 48 50 52 54 56 58 60 62 64
1 r [ ~ I "I !. r
t
- ] 1 1 1 1 I t I 1 I I T
9 Sample point CONDUCTIVITY 08

08 _ __ n Capillary fringe boundary 3.80
re,o-- 10 metre contour
~ Playabasin ~ 7.60
~ 111"40 06
06 0 , Miles , 4
5.20
() ' kms ' J,
06 04
Electrical conductivity
in pmhos/cm (x104)
AL MUHARRAO
02
~176176 AL MAN~,MAH
- 00
98 - 98
96 - 96
94 - 94
92 SITRAH 92
AR' RIF,~
90
ASH SHARO]-
AR RIF,g,'_~ i
AL GHARBI
I- 88
_
- 86
- 84
-~ 82
44 46 48 50 52 54 56 58 60 62 64 66
GRID DATA:- Universal Transverse Mercator. Grid Zone 39. Origin. Long 51~ Lat. Equator. Numbered at 2,000 metre intervals.
Figure 19.4. Aggressiveground conditions in northern Bahrain: the capillary fringe limit, and depth to
groundwater beneath the surface (after Power and Water, Unpub. data; in Cooke et al., 1982).
The outcrops of these evaporites around the world occupy a surface area of 60 million klTl2
(Ford and Williams, 1989); although karst is found mainly in the northern hemisphere, it
appears in all kinds of climates (Klimchouk et al., 1996).
The high solubility of the gypsum in the Tertiary formations of the Ebro Basin, where
conditions are basically semi-arid, has given rise to extensive karstification and a series of
associated dangers (Guti&rez and Guti&rez, 1998). Dissolution is much more marked
where the gypsum is covered by alluvial deposits (glacis, terraces, infilled valleys).
This gives rise to the risk of subsidence, which can have important socio-economic
consequences (Guti6rrez and Pefia, 1994). The alluvial deposits overlaying the evaporitic
formations, which can be over 100 m thick, have been explained by sedimentary
karstic subsidence (see Guti6rrez and Guti6rrez, 1998). The alluvial cover subsides
because of ductile flexure or fragile collapse.
Around Zaragoza, Spain, subsidence affects roads (Figs 19.5 and 19.6) and railways
(Benito and Guti6rrez, 1988; P6rez del Campo, 1989; Benito and P6rez del Campo, 1991;
Benito et al., 1995), houses, warehouses and even the water supply to the west of the city
(Casetas-Utebo) (Zuidam, 1976a,b; Guti6rrez et al., 1985; Soriano et al., 1994; Soriano
and Sim6n, 1995; Guti6rrez and Guti6rrez, 1998). The most spectacular case, however, is
that of the new town of Puilatos (about 30 km north of Zaragoza), whose construction had
to be abandoned. What had been built had to be destroyed because of the numerous cracks
caused by subsidence through karstification (Benito and Guti6rrez, 1988).
The area's lined irrigation network has also suffered the effects of subsidence. In 1954,
the La Violada concrete canal gave way because of a collapsed doline 25 m in diameter
(Llamas, 1962; Riba and Llamas, 1962). Non-lined canals, such as the Canal Imperial de
Arag6n, supplied water to the underlying gypsum, starting a karstification process and
the generation of dolines. A relationship exists between the density of dolines and their
proximity to irrigation ditches in the area of Villamayor in Zaragoza. The highest
concentration of dolines is close to the irrigation ditches, especially at their confluence
(Fig. 19.7) (Benito and Guti6rrez, 1988), showing that water leakage from the ditches
is the main cause of dissolution in the area and therefore of the ensuing subsidence.
Figure 19.5. Flexurebetween the Pikolin factory and the Zaragoza-Arag6n highway in Spain caused by
karstic subsidence of the underlying gypsum (June, 1996). Photo: F. Guti6rrez.
// /,~/ Relative height of

9 contours in meters
., .- -" Cracks
I' . / Motorway border
Factory border
0 15 30 45 FEET
; J
0 5 10 1,5 METERS
Figure 19.6. Karsticsubsidence affecting the Zaragoza-Alag6n highway and a factory. The area has been
filled with 3 m of asphalt agglomerate over the highway's 25-year existence. Current subsidence is about
25 cm/yr and is affected by traffic and the interannual variability in rainfall (according to Gonzfilez, Unpub.
data; in Benito and Gutirrrez, 1988).
The irrigation of crops can play a role in doline formation through the pumping out of
water (Benito, 1987). Finally, some dolines are used as rubbish dumps, which
contaminates the groundwater.
The semi-arid area around Calatayud in the Province of Zaragoza, which lies in a
tectonic graben filled with Neogene and Quaternary sediments, experiences very similar
problems to those of the area around Zaragoza. The Quaternary alluvial vary greatly in
their thickness (they can be 100 m thick, e.g. at the confluence of the Jalrn and Jiloca
rivers) because of karstic subsidence. This process is very active on the alluvial plains and
affects drainage canals, communications, irrigation ditches and the buildings of Calatayud.
The conservation of buildings of cultural interest in the city is faced with serious problems
(Fig. 19.8)(Gutirrrez, 1995, 1996).
Other environmental problems of gypsum terrains include the poor quality of water for
consumption, which can be polluted by fertilizer, oil products, and even radioactive
materials (such as those deposited by the Chemobyl disaster). Gypsum mining can also
cause collapses and flexures, and the removal of groundwater can intensify karstification
(Klimchouk and Andrejchuk, 1996).
3. Changes in volume
Soil volume changes can be caused by high sodium concentrations in the cationic
exchange complex of clays, and by the presence of shrink-swell clays. High sodium
concentrations cause clay dispersion and tensional stress to manifest themselves at the
surface as a system of cracks with domal blocks. In the Miocene shales of the Ebro Basin
N ,.~"'" '":-'.: .. ~ Density of dolines

I ,,~~. ,. ,. "'::-',
,;...~,.,, ,"-1" ~ Dolines
"~,-' 2,'
'" ~'": "~2 ,'~ ~1---; ~ ] Irrigation ditches
'-/,
"
"i ~ Villamayor
""x''~"
,1-'> ~.'...-.,
0 lo0' \'.....
"~""' " "-,,~-:,
"
(a) 0 3000FEET
10
8
lib
6
N/S
4 o
woo 9 9 9
omwoom 9 9 so 9 mm o 9
i i ,i ,i ..t i t
0 400 800 1200 1600 METERS
,'--- " : ~ ! I I t I I I
0 1200 2400 3600 4800 FEET

(b) Y~US
Figure 19.7. (A) Relationship between doline density and the network of non-lined irrigation ditches.
(B): Relationship between doline density and distance from the nearest irrigation ditch. N = number of
dolines. S = unit of surface area, ,~L -- sum of the distances of the dolines per unit area with respect to the
nearest irrigation ditch (Benito and Guti6rrez, 1988).
in Spain, where shrink-swell clays are absent, swell values of around 12 per cent have
been recorded (Guti4rrez et al., 1995).
More important is the expansion caused by wetting when the soil contains significant
percentages of shrink-swell or expansive clays (smectites, vermiculite and halloysite). In
semi-arid areas, this process is much more common in the lowest lying plains (Cooke et al.,
1993). Variations in the water content of the soil cause its expansion and contraction, and
pressures of up to 10 kp/cm 2 can be generated (Salinas, 1988). This swelling and shrinking
leads to gilgai soils that are characterized by a gently rolling landscape. In this micro-
relief, only a small percentage of the land is suitable for cultivation. Most fertile land has
to be used for grazing because, if machinery is used to level the land, the characteristic
micro-relief returns within 3 years. Even stakes driven into this type of soil are pushed out
by movements generated by vertical forces (Harris, 1968).
In the USA, more than 250,000 houses are built every year on expansive soils - and
about 10 per cent of them suffer significant damage (Jones and Holtz, 1973). In Spain,
Figure 19.8. Crackscaused by karstic subsidence in the Colegiata de Santa Maria la Mayor, a mudejar
church of the 12th century. Calatayud, Province of Zaragoza, Spain.
about 32 per cent of all geological formations contain expansive clays, and about 67 per
cent of the country experiences the type of climate which can lead to swelling (by about
10%) and shrinkage (Salinas, 1988). An increase in the soil volume of about 3 per cent can
cause damage, so special foundations are required. If these are to be made properly, it is
important to know the depth to which such expansive behaviour occurs. This zone given to
variation in volume is known to geotechnologists as the "active layer." According to
Salinas (1988), volume changes cause the deterioration of earth cuttings (slopes and road
cuttings), the breakage of pipes, the destruction of walls, the deformation of paved
surfaces, the breakage of ditch bottoms and pavements, the distortion of foundations, and
cracks in buildings (Fig. 19.9). Some damage can be made worse by the breakage of
drains.
Another manifestation of this activity is the production of giant desiccation fissures
resulting from the shrinkage of playa materials belonging to ancient desert lakes. These
7-
~ Z
uJ l.iJ co
z- co z
I- n w I- I.I_
< w z ~ on'Z < rr
oc: uJ _ ,,, ~
w oc < > ~s
w ~ cI:
Figure 19.9. Problems caused by the presence of expansive clay. The black arrows mark the moisture
transfer lines. The open arrows indicate possible damage (Salinas, 1988).
are dangerous and can form suddenly, causing serious structural damage to linear
constructions and introducing trafficability problems (Cooke et al., 1993).
4. Fluvial systems
Despite the low rainfall of arid regions, they are significantly shaped by drainage processes
(Tricart, 1969). Rainfall is often local and short-lasting, as are the ensuing runoff and
drainage processes. Areas with scant vegetation and gentle topography can cause
unconfined flow. During times of flood, sediment transport is discontinuous and channel
morphology can change rapidly from the meandering to braided type.
Within a drainage system, surface runoff and sediment production are related to rainfall
and temperature. These factors also influence the development of plant cover, and a set of
complex relationships between these variables can be established. For example, when
rainfall increases, so too does runoff. If the temperature goes up and there is more constant
rainfall, runoff is reduced because of evaporation. These two variables influence the
production of sediment in drainage basins.
For a given amount of rainfall, the sediment concentration increases with temperature;
and for a given mean annual temperature, the concentration falls with an increase in annual
precipitation. This means that in arid regions, more sediment is moved per unit of water
than in other regions, but because runoff is not very common, the total amount of sediment
transported per unit time is less.
Temperature and precipitation have a clear relationship with plant cover. Plant cover
increases with rainfall, and this leads to a considerable reduction in erosion. Knox (1984)
established a critical plant cover threshold of around 70 per cent (corresponding to a mean
annual rainfall of 400-500 mm) for the prevention of serious erosion. Where plant cover
is below 70 per cent, the amount of soil vulnerable to erosion increases rapidly. Francis
and Thornes (1990) lowered this threshold for the conditions of the semi-arid
Mediterranean, although they did differentiate between low and high density rainfall.
4.1. Water erosion
The erosion of slopes by water involves the detachment of particles from the soil and their
consequent transport. The energy required for this is supplied by the impact and the
splashing of raindrops, by surface flow, and by interactions between these. The size
distribution of soil particles, structural stability and crusting are the main properties of the
soil, determining its vulnerability to water erosion in arid environments. Land use can also
influence this (Lal, 1990). About nine per cent of the world's dry regions are affected by
water erosion induced by human activity (Thomas and Middleton, 1994). The countries of
the Mediterranean have some of the highest human activity-induced erosion rates in the
world (Woodward, 1995).
The techniques for calculating water erosion rates are numerous and include the use of
experimental stations, measurement of the sediment load in rivers, the use of
radioisotopes, satellite imagery and map analysis, and the revised universal soil loss
equation (RUSLE), and so forth. The results obtained with these techniques can be used in
soil conservation, the prediction of reservoir longevity, and to decide on the best places to
store waste, and other factors.
The influence of human activity on water erosion was studied by Iverson (1980) in the
Mojave Desert. The intense use of four-wheel drive vehicles has caused much surface
disturbance (Fig. 19.10) that has led to greatly increased runoff and erosion. The responses
to this activation were studied with a rainfall simulator, and it was found that damaged
areas had a reduced infiltration capacity, that there was more production of sediment (Fig.
19.11), and that runoff became channelled into wheel ruts.
Evenari et al. (197 l a,b), who worked in the Negev, reported another case related to the
water erosion of slopes. Some 2500 years ago, water was collected on slopes in this desert
and channelled towards the valleys for irrigation purposes. The runoff increased
considerably during this time, stones accumulated on the surface, and the underlying
loess became encrusted through the action of the rain, reducing its infiltration capacity.
Human activity was also involved in the rapid reduction of the suspended load carried
by the Colorado River (in its Grand Canyon stretch) between 1940 and 1946. This
reduction (of around 50-100 • 106 metric t/yr), was attributed to conservation practices,
flood control, a reduction in stock raising, and drought (Schumm, 1977).
Badlands are areas of intense water erosion that can affect man-made structures. Rilling
and gullying destroyed a large part of the town of Pastici in the Province of Materna, Italy,
in 1688, and a new town had to be built (Zachar, 1982). On the Yellow River in China,
intensive gully erosion produced a rapid gully head retreat affecting houses in its basin
(Zachar, 1982). In a 5-year study, Crouch (1983, 1990) used a microtopographic profiler to
obtain gully head retreat values of 0.5-1 m/yr and of 2.5 m/yr in areas with intense
subsurface flow.
Figure 19.10. Erosion of slope by vehicles. Kalia, Judean Desert, Israel.
9 Plot used by vehicles 9

J
1400 o Plot not used by vehicles 9
1200
E 1000
v
~,
II)
800 " /Z "
,,i--,
r
E
.m
-o
II) 600
Y.
(/)
400
200
/o
Or
0 0.01 0.02 0.03 0.04 0.05 0.06
Average runoff power (watts m-2)
Figure 19.11. Computed regression lines for runoff power versus sediment yield from meter-square
erosion plots used and unused by off-road vehicles. Mojave Desert (Iverson, 1980).
Formations with expansive clays or high exchangeable sodium percentages (ESP) are
potentially capable of causing piping (i.e. of making subsurface conduits). In detrital rocks
such as shales, the effects of piping look somewhat similar to those seen in soluble rocks,
with the production of tunnels, caves, dolines, natural bridges, and subterranean drainage
systems similar to karst systems (Parker et al., 1990). Piping is a serious problem for roads,
and can cause the undermining of bridge abutments, the breakage of lateral concrete
drainage systems, the undermining of concrete drains, and even the breakage of the road
itself (Fig. 19.12) (Parker and Jenne, 1967).
Many earthen dams are subject to piping, a process that has caused more than a few to
break. In Australia, fracturing in dams is due to the presence of montmorillonite group
clays and the high percentage of exchangeable sodium (Ingles and Aitchinson, 1970).
Studies performed in the USA and other countries report the same conclusions (Sherard
et al., 1977; Sherard and Decker, 1977).
In some cases, pipes can cause landslides and subsidence and vice versa. Pipes can
make slopes susceptible to these processes, which end up causing sheet and rotational
slides. The undercutting of piping helps to generate mass movements on slopes (see Jones,
1981).
Piping has important effects on agriculture (Fig. 19.13). In some parts of Australia, deep
ploughing is used to temporarily destroy existing pipes (Hudson, 1981). Generally, intense
piping occurs in gullied terrain of little agricultural value. Piping often occurs because of
inadequate deforestation practices or overgrazing. The soil conservation agencies of
several countries, therefore, try to ensure the rational use of agricultural and forestry
resources.
The prevention and control of piping is an important civil engineering and forestry
problem. The solution used in earthen dams is to install sand filters to seal the escape of
dispersive clays (Sherard et al., 1977), or to intensely compact the material to reduce
hydraulic conductivity. The remedies to alleviate piping in linear constructions are based
on surface drainage (Parker and Jenne, 1967). Some authors suggest the reforestation of
the affected areas as a long-term solution. Reforestation would diminish the rapid drying
that occurs during summer and prevent subsequent cracking. It would also increase the
organic matter of the soil (Newman and Philips, 1957). Unfortunately, it is difficult to get
vegetation to take root properly in arid areas. Another way to alleviate the effects of piping
is to add gypsum to the soil to reduce the ESP, and therefore reduce clay dispersivity
(Stocking, 1976).
4.2. Fluvial channels
The types of channel that exist in desert regions are straight, meandering and braided, but
their topography can change rapidly. Research into hydraulic geometry and the behaviour
of drainage channels in these areas has led to important advances in fluvial
geomorphology. Though the types of channel are the same as those occurring in wet
regions, some aspects of fluvial theory, such as plain flooding, cannot be extrapolated to
arid areas because the feed mechanisms are different (Graf, 1988). The instability of these
channels is quite obvious, and human activity commonly causes fluvial changes that entail
considerable risk.
4~
.,..j
9 " " " ' ~ - " " ." . ~ . . .9 . % : . ' : : x " ; . . . . '"'"
9 - -~. 9 . < . . . -:...:~s.

~ ._,,"."-".
...-~.-~-..-_-..' US Highway140 4~
- ~. ~ 9 . . .,. . , ~o--- ~'-....s ..." :....:-:"~163
, ;
9 - . 9 . ~ . .
9 - 9 ~--~Z.~ 9. ~ '-~,
9 . .. -~-..., ._,.:.j.. ~.~:..::
9 . o
. " . " ." . ' ' - . . - s
9 . . . 9 : .----..:.:....,
9 . . - - , . . ,,a=..&..i.~;.
~ ' . . . '. . ~ - ".. . . . -:...
.
9
.
.
.
~
.
9
9 .~.
...'~ ...
%
.''....'-.
.
9 .
9
9 , "" ~
:!:.',.::..".,,,
|
. .. v
9 ' ~. " 9~"2""-.~'-:'.":
" "7. {.'-'"
. i' ~ . ~ . " ~ ' ' ' ? " '
"~J~'"'~l~'" )-._.~ . ~ ~ - -
i::ii ~ . . . . .
.'7"-- -'.--~.'. "~.." :: :":

d~"
9 9 .--... .... ".'.' . 9..~.
9~--, '. :..::::!~.:.~:..~'.z,". " =.~. ---~ !.

9<_ ~.:.~.. -.,: .: -...,~". ..: , ~ : . ~ ~ 9 . .
_ _
Explanation
Q
Q Shaleand sandstoneof Cretaceous G Pipedgully
Mancos Shale G Naturalbridge
Q Tan silt and clay, sandyin places,
Quaternary Q Cave
~ Floodplan of AztecWash O Culvert
Aztec Wash low-flowchannel Q Gravitycrack
Pipe system system, piped c~
C~
Figure 19.12. Idealized block diagram of Aztec Wash, southwestern Colorado, showing damage to US Highway 140 where it traverses a dissected and
extensively piped valley fill. 1. Cretaceous shale and sandstone. 2. Quaternary silt; clay and sand. 3. Floodplain. 4. Low-flow channel. 5. Pipe system. 6. Piped
gully. 7. Natural bridge. 8. Cave. 9. Culvert. 10. Gravity crack system, piped (from Parker and Jenne (1967), in Costa and Baker (1981), Fig. 10.15).
,,<
Figure 19.13. Pipe in a field of cereals. Azlor, Province of Huesca, Ebro Basin, Spain.
Changes in channel type are relatively frequent, and have been recorded on several
rivers of the western USA (Table 19.1), where meandering channels have become braided
(Graft, 1988).
One of the most surprising examples is the metamorphosis of the Cimarron River in
Kansas, which was analysed by Schumm and Lichty (1963). This fiver was a deeply cut
Table 19.1. Changes in channel width in rivers of the western USA (Graf, 1988).
River Change Period
Canadian River, Oklahoma 0.8- 3.2 km 1906 flood

Rfo Salado, New Mexico 15-168 m 1882-1918
Red River, Texas-Oklahoma No change 1874-1937
Red River, Texas-Oklahoma 1.2-0.8 km 1937-1953
Cimarron River, Kansas 15- 366 m 1874-1942
Cimarron River, Kansas 366-168 m 1942-1954
Platte River, Nebraska 1161-111 m 1860-1979
S. Platte River, Colorado 790-60 m 1897-1959
N. Platte River, Wyoming 1200-60 m 1890-1977
Gila River, Arizona 45-90 m 1875-1903
90-610 m 1903-1917
610-61 m 1917-1964
Salt River, Arizona No change 1868-1980
Fremont River, Utah 30-400 m 1896 flood
meandering channel about 15 m wide which, after the floods of 1914, became a straight
bedload channel about 336 m wide. This dramatic change was interpreted as being due to
climatic fluctuations, agricultural activities, and the destruction of the natural vegetation.
Another no less surprising example is that of the Gila River that runs through Arizona
and New Mexico (Graf, 1988). Before 1890, the channel was narrow and meandering, but
in 1905 flooding modified it to become braided, and to have a width of 1 km in some
places. Later, the growth of vegetation and sedimentation narrowed the channel, and by
1980 - almost a century later - it had returned to its old meandering nature. These results
show that the change from braided to meandering channels occurs much more slowly than
in the opposite direction.
The effects of human activity have been very clear in changes to the South Platte River
of Colorado, which now has a braided flow. This fiver went from being 800 m wide in
1897 to only 60 m in 1959, becoming a narrow channel with a tendency to form meanders
(Fig. 19.14). This profound modification was related to the regulation of the river and the
loss of depth due to the use of water for irrigation. This caused an invasion of vegetation on
the floodplain and sandbars (Nadler and Schumm, 1981).
The erosion of the concave banks of meanders and sedimentation on their convex parts
creates lateral instability. It also causes the migration of meanders downstream that can
interfere with man-made constructions. Such is the case of the Cimarron River near
. ",',,"/ ,"r " "

(a) ,. . r " < ;,:'..i, ."-' :, ..... , - ' . < ' (b) ., . . .. . . .. . .., ~ . ~ j ," .~.. 9 ~.,., t .$. '';.,
.~ i ~ l ) , . . ...,t . . . a . "~A~ 9 --' 9s- ;,'
,-,.r.l,,. , .. :. . . ~ - ~ ' . ~ ." ":' i ",::~
c. I s l a n d s
450m
..' ~'J.,' ~ : 5Y J, J-'.~tJS,'l ' ~ " ?~t"7./J~:.,2"~

9j ga jl'r..~a~s,'i ja,a.~ #, .~j,ij-
I (C l (d) _ .., .... , ..... ,,,.~ , : ..... , ..

.a # .
~,~,< 9v,., ,,,, ,.~-,. r,,<i;,'r,J3;"

;"~'" '" ~ ' - ' ~ "
: - ' ~ 9,, , - ~ ' . .". ' T Z ~ . ~ ' f ,.'t~?
$t:4"~,' * ~ " ~ i : ? 9 r~;7,-: r.?~'l,.~'~%.:,;L-./
0 ,..?.
~ ~ ~ . 9 , "..:,.rl,~ "
1".~ ,.. ,. ;., . / ,., ,..s~'~ ~''

..,.7.
9 ~. * -- ,"a,.4.'.~ ..,,y..,.~.,
Figure 19.14. Model of South Platte River metamorphosis: (a) early 1800's, discharge is intermittent,
bars are transient; (b) late 1800s, discharge is perennial, vegetation is thicker on floodplain and islands; (c)
early 1900s, drought allows vegetation to establish itself below mean annual high-water level, bars become
islands, single thalweg is dominant; (d) modern channel, islands attached to floodplain, braided patterns on
floodplain are vestiges of historic channels (from Nadler and Schumm, 1981).
Perkins, Oklahoma. When it was built, the fiver was straight and braided in the area of the
bridge, but there was a large meander one mile upstream. The position of this meander
drifted, and in 1968 it affected the structure of the bridge (Chorley et al., 1984). Gravel
extraction from fiver beds can also cause headcut erosion and the undermining of bridge
pillars, as occurred on the Salt River in Phoenix, Arizona (Chorley et al., 1984).
Dams are perhaps the clearest form of human interference in fiver systems. They cause
the deposition of sediment in the reservoir while downstream channels become degraded
and rapids can form. This occurred after the construction of the dams on the Colorado
River (Graf, 1980).
The effects of the Aswan dam on the Nile have been truly important. The downstream
delivery of silt has been reduced and the delta has retreated because of a lack of water in its
channels. It is estimated that this retreat might be as much as 30 km in some places by
2100: this would have dire consequences for the area's 1,000,000 plus inhabitants. There
has also been a generalized fall in the water table and an increase in the salinity of nearly
all the delta lands (Mainguet, 1991; Beaumont, 1993).
The River Ebro in northeast Spain also suffered an important fall in its bedload during
the 20th century due to the construction of numerous dams. During the first third of that
century, the sediment load was estimated at 1,000,000 metric t/yr. During the period
1950-1970, however, it had dropped to some 730,000 metric t/yr, and between 1970 and
1990 it was down to 300,000 metric t/yr. For 1990-2000, the mean was below
100,000 metric t/yr. When these results are compared with the transport capacity of the
coastal part of the Ebro Delta (320,000 metric t/yr), a negative sedimentation balance can
be appreciated which, since about 1970, has caused the acceleration of the delta's erosion
(Guill~n et al., 1992). Further, the horizontal and longitudinal instability of its channels -
many of which are found in arid regions - are frequently accompanied by quite serious
vertical instability. This is shown in successive stages of accumulation and incision, which
are reflected in the morphology of the terraces. Human activity in semi-arid areas has also
caused the formation of gullies. Several examples are analysed in the chapter on climatic
change in arid areas.
As a consequence of the spectacular increase in the settlement of arid areas, fiver
systems have suffered important changes. During construction, natural vegetation is
removed and the surface is strongly modified by earth-moving machinery. This increases
soil erosion. When construction is over, soils are impermeable and consequently there is
an increase in runoff during times of rainfall and an increase in the depth of the water
carried in the channels (Leopold, 1968, 1994). Flash floods can also occur.
4.3. Alluvial fans, debris flows and pediments
Alluvial fans, characteristic features of piedmonts in arid and semi-arid areas, are formed
by the discharge of sediments transported by rivers from mountainous areas with gentle
slopes. Their morphology is triangular and coalescence is very frequent.
Commonly, alluvial fans are controlled by tectonic, climatic or vegetation changes that
lead to modifications in the production of sediments and runoff (Harvey, 1997). Deposition
occurs locally when the supply of sediments is greater than the transport capacity, i.e.
when the transport potential of the fan is below that necessary to move the sediment
through the system. This is a critical energy threshold of the stream flow (Bull, 1979).
Temporal variations can cause modifications to the sedimentation velocity and even the
aggradation or dissection of the fan surface.
Alluvial fans present serious difficulties to human activity (Beaty, 1974; French, 1987).
Flows are rare and unpredictable, and there can be wide variation within and between
successive flows. Further, they can be dominated by water or by detritus which can even
block channels. They may also undergo important migration and avulsions are common
(Cooke, 1984).
The risk of flooding is evident, but not all parts of fans are subject to the same degree.
The risk is extreme along the active channel, moderate at the head and distal areas of the
fan, and low in the mid sector (Kesseli and Beaty, 1959). A common planning and
engineering solution is to design and maintain a single channel for the flow (Graf, 1988).
But the best defence against flooding is to identify the most dangerous areas and make
flood risk maps such as those produced for the city of Suez and its environs (Cooke et al.,
1982). These maps help to determine construction densities and to situate projects away
from the channels (Fig. 19.15). Sediment deposition can also be fought using check dams
and by dredging the channels in the catchment area. Flood alert systems can also be
installed (Porath and Shick, 1974).
The channels running across fans are unstable and frequently overflow because of
boulder sized sediment on their bottoms. This creates new channels which on occasion
form part of abandoned channels. Important floods occurred in the south of Jordan in 1966
that partially destroyed the town of Maan, killing 70 people. The cause was mainly the re-
use of an abandoned channel by the rising water (Graf, 1988).
Debris flow is one of the most common types of mass movement in semi-arid areas. It
can be caused by intense rainfall moving abundant colluvial material through channels and
over alluvial fans (Johnson and Rodine, 1984) and depositing material in areas with gentle
slopes - commonly at the head of the fan. The distinction between the deposits made by
debris flows and fluvial processes is not always easy to make because there is a continuous
gradient between them (Hooke, 1987). Both are generated by concentrated flows of large
transport and erosion capacities, but with different viscosities.
In their movement, debris flows can uproot vegetation, cover roads, channels and railways,
drag along cars, destroy constructions and cover fields (Hooke, 1987). In semi-arid areas with
clay soils, dangerous mud flows can occur that become channelled down fills and slopes.
Debris flows can be started by landslides, along with materials supplied from other
sources and by a sudden saturation in detritus (the fire-hose effect) (Johnson and Rodine,
1984). Debris flows are very common in the semi-arid regions of the world and have been
frequently studied, especially in California. In the La Canada Valley, Los Angeles County,
debris flows occurred in 1934, 1938 and again between 1962 and 1971, causing the deaths
of 63 people and important material losses (see Johnson and Rodine, 1984).
The area around Gissar in Tajikistan, close to the Afghanistan border, suffered an
earthquake in 1988 that caused landslides on loess hills. The deposits were saturated by
irrigation water and in the initial moments reached a liquid state producing a debris flow
that moved some 20 million m 3 of material a distance of 2 km. Houses on the outskirts of
the town were buried (Ishihara, 1989).
In some areas, urban development reaches up to the apex of alluvial fans, allowing
detritus and mud flows to extend beyond their source area. Urbanized substrates facilitate
Figure 19.15. The flood hazard map of Suez, Egypt (Halcrow and Partners, Unpub. data; in Cooke
et al., 1982).
the propagation of the flow and, because they are impermeable, prevent the loss of water
(Rantz, 1970).
Perhaps the most important problem in determining the risks associated with these fans
are that they remain dry for long periods of time. In southern California, for example, the
flood recurrence interval is 70 years. Disasters only occur, therefore, from time to time,
and only the really important ones become the object of public note (Burton et al., 1978).
Pediments have high drainage densities, but in contrast to alluvial fans they have no
radial network. Pediments can be very problematical areas for human activity because
they are at risk of flooding, channel instability and sedimentation (Rhoads, 1986). Because
of their many small channels, linear constructions are often flooded and eroded. To solve
this problem, dams are constructed on the upper ramp of the pediment to deviate water for
flood control. The sediments transported also influence human activity by blocking drains,
sewers - even narrow bridges - and thus hold water back and increase the flow. The small
channels can overflow into one another. All this makes it more difficult for engineers to
stabilize the channels. Where there are pediments, the best solution is to stabilize the
largest channels and to be prepared for rising water (Graf, 1988).
4.4. Floods
On alluvial fans, flood waters are mainly sheet flows running across the fan surface.
Floodwater poses important risks to human activity, and on occasion can be the cause
of catastrophe. Buildings can be damaged or destroyed, and railways (Fig. 19.16), roads
(Fig. 19.17), canals, bridges (Fig. 19.18) and agricultural land damaged by erosion or the
accretion of the solids carried by the water.
Figure 19.16. The Arroyo del Medio fan, over which a railway passes, has been cut by the Rfo Grande.
The tracks have been disturbed and left hanging in the air. Cordillera de los Andes, Province of Jujuy,
Argentina. Photo: J.L. Pefia.
Figure 19.18. Bridge broken by flood water. Be'er Sheva, The Negev, Israel.
The inherent risk of flooding is increased by constructing close to rivers and by the
agricultural use of alluvial soils; this increases the population close to rivers and their
floodplains.
One of the most catastrophic flood events of recent history occurred in the spring of
1887 when the Yellow River or Huang Ho (China) broke its banks, killing between 1.5 and
7 million people (Davis, 1992). It affected 2000 towns and villages and material losses
were huge. Similarly, in September of 1911, the Yangtze flooded, affecting four Chinese
provinces, killing over 200,000 people and leaving half a million homeless. On the 17th
September 1954, in Farahzad, Iran, 2000 pilgrims who were camped in a narrow gorge
were killed when a flash flood hit them. In semi-arid Spain, the worst ever flooding
occurred in Valencia in November of 1957 with 86 deaths, and in Tarrasa-Sabadell, in
September 1962, where 267 people lost their lives (Martfnez et al., 1987). L6pez-
Bermfidez et al. (2002) developed an historical analysis of the floods which occurred in the
Spanish Mediterranean zone.
The fight against flooding has gone on throughout history in arid areas. The town of
Calatayud in the Province of Zaragoza, Spain, was founded by the Arabs at the
beginning of the eighth century, and it partially occupied an alluvial fan exposed to
frequent flooding during storms (Guti6rrez, 1998). With the aim of reducing this risk, a
weir and channel were built that deviated the water from the Barranco de la Rfa and the
Rambla de Ribota (Fig. 19.19). In the Barranco de la R6a basin, the runoff over a
surface area of 6.3 klTl 2 (total basin area 8.2 km 2) was diverted by Arab engineers. As
the population grew and settled closer to the River Jal6n, mainly in the 13th and 14th
centuries, the risk of flooding increased. During this time, or perhaps a little later, the
Balsa de Valparaiso reservoir was constructed close to the fan apex to alleviate the
problem. After the 16th century, a tunnel nearly 1 km long was excavated that drained
the water towards the Barranco de Longia, and in so doing some 98 per cent of the
runoff of the Barranco de la Rtaa was diverted from its drainage basin, avoiding the risk
of flooding. The Barranco de las Pozas, with its smaller drainage area, still causes
problems for Calatayud (Guti6rrez, 1998). In the neighbouring town of Daroca in the
Province of Zaragoza, the Mina tunnel, also about 1 km long, was excavated in 1555, to
divert water from the main throughway, which passed through what is now the Calle
Mayor (Main Street) where it caused serious flooding.
To avoid or at least reduce the chance of flood disaster, many and costly projects
are undertaken in the world's arid regions, although recent studies suggest that the
engineering solutions adopted are not always very adequate (Cooke et al., 1982).
Dikes, dams and ditches are all part of these projects. Small dams at the headwaters of
channels reduce erosion and control the runoff over small areas, whereas large dams
downstream allow the attenuation of large rises in water levels (Costa and Baker,
1981). Dams, however, do not provide control over flooding, but rather "partial or
specific protection" (Leopold and Maddock, 1954). It is also very important to control
human use of floodplains with adequate legislation. Each local government should have
detailed information concerning areas at risk of flooding. Flood alert systems have
been improved by the installation of automatic hydrological information systems
(AHIS), the aim being to establish emergency alerts that minimise flood damage
(Ortiz, 1993).
OO
"(866I 'zz~P!mD) uFds 'ezo~eaeZ jo gOU~.AOId 'pn~mele D u! ~u!poog lsu!e~e lq~y oq,L "6I'6I d.mSJ.2
w~4L 0
I
.
AIInO
os!eJedleA -OOZ
.,2
~o
008
-z= /
leuueqo le!o!J!lJe / /
/ /
"t=
/x~co ,,me(] opeJB ,, 006
~b
~,L%~896
~b
~o SgZNVI41UV
"t=
5. Lake systems
Closed depressions, or undrained lakes (Langbein, 1961), are relatively common in arid
areas, and are often found grouped together. They are not usually very deep and are
ephemeral in nature. Because evaporation commonly exceeds water input, they are often
dry and saline, and the water level fluctuates because of the high variability in available
water.
These lakes can, however, overflow as a consequence of excess inflow. In September
and October 1968, it rained for 38 days over the greater part of Tunisia. Rivers burst their
banks and lakes overflowed, and the excess water ran down overflow channels giving rise
to flooding. Some 542 people were drowned and hundreds of thousands made homeless.
Damage to construction and agriculture was severe (Davis, 1992).
The salinity of these desert lakes can be related to the groundwater flow toward the
lake, as occurs in the closed Bujalaroz basin in the Province of Zaragoza, Spain. The
continuous supply of water provides salts to the system - and the only way to get rid of
them is the wind. This makes the use of such lands for irrigated agriculture inadvisable; the
water supplied to the irrigation system increases the level of these lakes, perhaps making
them permanent. This can affect the flora and fauna of the region. Additional irrigation
water can cause changes that may lead to serious environmental problems (S~inchez et al.,
1993).
Human activity associated with the lakes of arid regions can be very varied, but often
they lead to their slow eutrophication or degradation (Goudie, 1986). Sometimes, in semi-
arid environments, lakes are partially dried to make way for crops. This is quite a normal
fate for small lakes. In the Caspian Sea, the biggest lake in the world, man-made activities
have caused a fall in the water level of over 3 m since 1929 (Goudie, 1992). In Israel, the
largest hydrological project ever undertaken was the National Water Carrier, which
transports water from the Sea of Galilee to the whole of the country' s Mediterranean coast
where it is required for irrigation (Beaumont, 1993). Consequently, the salinity of the
Dead Sea has increased considerably as the quantity of fresh water flowing into it from the
River Jordan, rising just north of the Sea of Galilee, has decreased. This increase in salinity
has killed off the Dead Sea's animal and plant life, although it has favoured the salt
extraction industry at the lake's southern tip.
Over the last few decades, the Aral Sea, the fourth largest lake in the world, has been
the focus of one of the greatest ecological disasters ever recorded. This lake lies in a classic
arid region and is fed mainly by the Rivers Amu Daria and Syr Daria, which arise in the
mountains of Pamir and Tian Shan, respectively. Since 1960 this lake has suffered a
serious reduction of its perimeter because of huge irrigation projects (L6tolle and
Mainguet, 1993) (Fig. 19.20). It once had a mean depth of 16 m and a maximum of 68 m
close to the eastern shore. Table 19.2 shows the main hydrological data for the period
1960-1990.
The removal of water has left ports high and dry, fishing has fallen from
48,000 metric t/yr in 1957 to zero, fishing villages have been reclaimed by the desert,
the water is polluted by pesticides and its salinity has increased considerably, the water
level in the area's wells has dropped by 10 m and is polluted, and the fauna and flora are
degraded, etc. (L6tolle and Mainguet, 1993). Moreover, many of the irrigated soils have
become saline and have been abandoned (Beaumont, 1993).
1960 1971
1976 1987
1990 2000 ?
Figure 19.20. Evolution of the Aral Sea during the period 1960-2000 (L6tolle and Mainguet, 1993).
Table 19.2. Hydrology of the Aral Sea from 1960 to 1990 (L~tolle and Mainguet, 1993).
Year Discharge of rivers Water level Area Volume Mineralization Fishing

(km3) (m) (km2) (km3) (g/l) (t)
1960 40 53.5 67,900 1090 10 43,430

1965 31 52.5 63,900 1030 10.5 31,040
1970 33 51.6 60,400 970 11.1 17,460
1975 11 49.4 57,200 840 13.7 2940
1980 0.5 46.2 52,500 670 16.5 0
1985 0 42 44,200 470 23.5 0
1990 0 39 41,000 330 26.5 0
6. Subsidence caused by human activity
Human-induced subsidence has different causes including the extraction of fluids

(water, oil, gas and geothermic fluids), hydrocompaction, compaction and later
expansion due to the reduction of the load, the drainage of peat bogs, the drying of
clay, and underground mining (Costa and Baker, 1981; Waltham, 1989). In arid areas,
the most common type of human-induced subsidence is that caused by the exploitation
of groundwater. The lack of water in these regions obliges people to extract it from the
aquifer so that agriculture can be practiced. Abusive extraction can induce subsidence
when the hydrostatic pressure in the storage rock falls. This subsidence is more
frequent in non-consolidated Tertiary and Quaternary detrital deposits, formations in
which sand and clay alternate. With the extraction of water from the sands, the pressure
in the pores falls and the grains of sand have to support the overlying pressure. Sands
do not compact much, but the compressibility of clays is high, and this allows surface
subsidence (Marsden and Davis, 1967). This kind of subsidence generally affects large
areas, producing shallow depressions that cause damage to buildings and other features.
These endorheic areas of anthropic origin are sensitive to flooding, and if such
subsidence occurs on the coast the sea may invade the area. This happened in
Galveston Bay in Texas (Fig. 19.21), where residential areas were flooded (Coates,
1983, 1987).
In this same area, the extraction of groundwater has caused a wide sink hole and more
than 160 faults with a total length of some 500 km (Fig. 19.21) (Holzer, 1984). Of these, 86
have been historically active, and are probably the result of the reactivation of Tertiary
faults in the overlying Quaternary sediments. The groundwater is extracted mainly from
sand and clay beds of the Chicot Aquifer that form part of a thick sequence of non-
consolidated river and marine sediments. Pumping began at the end of the 19th century but
increased during the 1930's. The result was the generation of an area of subcircular
subsidence of about 12,200 km2; maximum sinkage was 2.5 m. Although the Goose Creek
oil field lies in the same area, its contribution to this subsidence appears to have been very
small.
The subsidence of this area is accompanied by faults at the centre and the flanks of the
depression, the density of faults being variable. Historically these faults have become
95045 ' 95035 ' 95~ 95o00 ' 94045 '

I J I I
30000 '
~ 7
-1.,
- \J sf i ~"-" Io ~' /
29o45 '
U
\ _
1
'~
,,VVe.,Io ...'_t__J
LJ 6 5 - 2 2 - 3 1 7 f (x' -~ . ,
i
~ L I(
BAY
29o30 '
EXPLANATION
-r- Fault
-- Line of equal
subsidence, in meters
0 10 K I L O M E T E R S
29~
Figure 19.21. Land subsidence from 1906 to 1978 and subsidence faults in the Houston-Galveston,
Texas, region (in Holzer (1984) with data from several authors).
reactivated and escarpments of 1.12 m have been observed. These faults are easily
identified where they cross roads, airstrips, and other features.
In terms of the magnitude and variety of subsidence caused by water extraction,
however, the record goes to Central Valley, California. The San Joaquin Valley is the most
important area with 13,500 klTl 2 affected by settling and 9 m of maximum subsidence. This
began back in the 1920' s, and gradually increased until the mid 1950's when the sinking
rate reached 0.55 m/yr (Poland and Davis, 1969). It then fell somewhat to 0.33 m/yr
between 1963 and 1966, and in 1973 was imperceptible, all due to the bringing in of water
from northern California. In 1976 and 1977 the subsidence again accelerated due to heavy
drought that provoked the need for groundwater.
In central-southern Arizona, groundwater was heavily exploited in the mid 1940' s, and
this again caused subsidence. More than 50 areas were identified, with a total area of
8000 k m 2 affected that became subject to subsidence, cracking and faults (Holzer et al.,
1979). The cracks are usually about 1 m wide and several meters deep. The lengths of the
faults vary from 1 - 16.7 km, and their escarpments are 1 m high. Serious damage has been
caused to roads and irrigation systems, and the trajectories of some aqueducts have had to
be modified (Coates, 1983).
In the San Joaqufn Valley and central-southern Arizona, a slow uprising has followed
the pumping out of groundwater. According to the US Geological Survey, this reached
6 cm between 1948 and 1967 in the area close to Casa Grande, Arizona. It is thought this
was caused by isostatic rebound following the extraction of the water, which eventually
reduced the load on the underlying materials (Costa and Baker, 1981).
7. Wind action: problems and control
The mobilization of sand and dust by the wind occurs in many types of climate, but in
arid areas it is at its greatest. Given the scant or absent vegetation in these areas, much
of the soil is subject to wind action. Further, the limited development of soils leads to
high erosion vulnerability with some 39 per cent of the world's arid land affected
(UNEP, 1992). Locally, tornadoes can have intensely destructive effects on construction
and agriculture, as occurred on the 8th August 1992 at Ejea de los Caballeros, Province
of Zaragoza, Spain (Fig. 19.22). The movement of wind-driven particles has a
considerable effect on human settlements, and can cause enormous damage to
buildings, crops, and transport and communications networks. The accumulation of
wind deposits can destroy vegetation and increase the problem of desertification
(UNESCO-FAO, 1977).
Figure 19.22. Destruction of grain silos by a tornado on August 8th 1992. Ejea de los Caballeros,
Province of Zaragoza, Spain. Photo: C. Sancho.
Problems arise because of deflation, transport and sedimentation. Deflation can reduce
the quantity of fine particles and nutrients in the soil, causing it to lose its fertility
(Middleton, 1990). It can also undermine structures (e.g. lampposts or telephone posts),
causing them to collapse. During transport, the particles cause abrasion that produces
hollows and furrows. Dust in the wind can also reduce visibility at airports and choke
people and animals (Pew6, 1981; Pye, 1987; Middleton, 1997). For these reasons, a dust
storm alert system has been set up in Arizona, with electronic signboards on roads and
announcements on public radio warning of any threats. Dust can also enter houses, causing
health problems and contaminating food and drinking water. If the dust is saline, it could
affect buildings. The positive effect of wind-blown dust is that it can provide soil nutrients
to new areas.
The quantities of dust mobilized are of the same order of magnitude as those
transported by rivers (Livingstone and Warren, 1996). A significant part of windblown
dust comes from human activities. Ploughing activities break up soil aggregates, and in dry
periods plumes of dust can be seen behind tractors (Lee et al., 1993). Another source is
traffic driving over unsurfaced roads. The removal of vegetation by grazing or fire can also
increase wind erosion (O'Hara, 1997). The drying of the Aral Sea has increased the
amount of windblown dust (Goudie, 1994b). Wind sedimentation, on the other hand, can
bury houses, crops, canals (Fig. 19.23) and roads (Fig. 19.24). Digging them out can be
very costly if large areas are affected (Livingstone and Warren, 1996).
Control measures should be aimed at reducing the wind's transport capacity, reducing
the supply of sand and, if required, deviating windblown sand (Watson, 1990; Pye and
Figure 19.23. An irrigation ditch, partially buried by sand, that was part of the infrastructure for irrigating
an area of the That Desert using water from the Indira Gandhi Grand Canal. Jaisalmer, Province of
Rajasthan, India.
Figure 19.24. Dunes advancing across a road. Maharaga, United Arab Emirates.
Tsoar, 1983a; Mainguet, 1991). The techniques employed (Cooke et al., 1982; Watson,
1990) involve stabilizing areas by recovering the natural vegetation or the planting of
other species, depending upon how arid the area is. In hyper-arid areas, the extreme lack of
water may render this impossible. It is important to plant fast-growing species (e.g.
tamarix sp., acacias and eucalyptus) (Garcfa Salmer6n, 1967; Pye and Tsoar, 1990).
Around the railway from Bagoton to Lanchou in China, vegetation corridors 300-500 m
wide have been planted to slow down erosion (Watson, 1990).
Another control method lies in the stabilization of the surface. This could involve
covering it with stones, oiling it, or covering it with asphalt, synthetic latex, polyvinyl,
gelatin or cellulose fibres. These can be effective but for periods of less than 5 years.
Chemical sprays have been widely used to produce a chemical crust a few millimetres
thick which, combined with the reestablishment of vegetation, has provided some good
results (Garcfa Salmer6n, 1967; Cooke et al., 1982; Livingstone and Warren, 1996).
Windbreaks are commonly used in desert areas to slow down the wind and trap the sand
it carries (Fig. 19.25). They are commonly made from palm leaves, but also from low-cost
wood produced by local vegetation. Such structures reduce the transport capacity of the
wind both in front of them and behind them. The volume of sand that accumulates is
proportional to the height of the windbreaks used. These should be installed some distance
from the object to be protected and perpendicular to the prevailing wind. It is a good idea
to set up a series of windbreaks at different distances from the object. The furthest away
will become covered rapidly, but it can increase the lifespan of those in the intermediate
distance by a factor of four, and of the closest by a factor of nine (Fig. 19.26) (Kerr and
Nigra, 1952). Some windbreaks are set up in a zigzag fashion to restrain winds from
several directions (Fig. 19.27) (Watson, 1990).
Figure 19.25. Windbreak system for protecting a palm grove. Nefta, Tunisia. Photo: J.L. Pefia.
Dune stabilization is the objective of several techniques, but their effectiveness is very
variable. Sometimes, however, it is only necessary to deviate the direction of a dune. If
dunes are small, it is sometimes decided to simply move the sand somewhere else. This
procedure is cheaper if it can be used in construction. The destruction of a dune by digging
ditches parallel to its axis is an expensive solution - and only temporary. Stabilization
with vegetation can be very expensive and the use of windbreaks or covering dunes with
Figure 19.26. Three fence system to protect extensive areas, such as villages, shops, yards, industrial
plants, and other features (after Kerr and Nigra, 1952).
Figure 19.27. Windbreaks arranged in a zigzag for trapping sand blown by winds coming from different
directions. Erfoud, Morocco. Photo: J.L. Pefia.
stones is more useful. The surface treatment of the arms of crescent-shaped dunes with oils
allows the deflation of the central part, and an accumulation of sand (very much greater
than that associated with the original dune) occurs to windward (Fig. 19.28) (Kerr and
Nigra, 1952).
8. Degradation of arid areas. Desertification
The degradation of soil results in a reduction of its quality and, therefore, of its potential
productiveness. Such degradation can be physical - manifested by the movement of the
smallest particles, biological - related to the loss of organic material, or chemical - linked
to the concentration of highly soluble salts. This degradation of the soil is the main factor
causing desertification (Mainguet, 1991). Desertification affects some 65 million hectares
that were previously productive and on which some 800 million people live (Fig. 19.29).
Desertification and salinization are two of the nine main problems cited by the World Bank
(Goudie, 1990a,b).
In Spain, some 27 million hectares, 53.4 per cent of the country, are affected by
important soil degradation. Annual soil losses are about 1,000,000 t (MOPU, 1987). In
south-eastern Spain, desertification is a severe problem with erosion rates of up to
300 metric t/ha/yr (L6pez Bermfidez, 1986). Data on other processes such as salinization,
wind erosion and the physical and biological degradation of the soil are not available for
the country as a whole (L6pez Bermfidez, 1988). Recently, the study of water erosion in
Spain has intensified with the Lucdeme Project. The Network of Experimental Stations for
Figure 19.28. Dune stabilization by oiling (after Kerr and Nigra, 1952).
Monitoring and Evaluating Erosion and Desertification (Red de Estaciones Experimen-

tales de Seguimiento y Evaluaci6n de la Erosi6n y Desertificaci6n - RESEL), which
covers a large part of the country, is now installed (Rojo and S~inchez, 1996).
At the 1977 World Conference on Desertification (UNWCD) in Nairobi, the United
Nations proposed a plan of action for combating desertification. Desertification was
defined as the reduction or destruction of the biological potential of land that could finally
lead to conditions similar to those of deserts (UNESCO-FAO, 1977). There are many
other definitions that emphasize the role of human activity or the climate (Baumer, 1990;
Mainguet, 1991; Thomas and Middleton, 1994). The risk of desertification comes from the
interaction between natural physical and biological systems and human activity. Such
interaction has become more intense because of the explosion of the human population in
the 20th century (Goudie, 1986, 1990a).
Physical processes such as water and wind erosion (mentioned above) intervene in the
degradation of deserts. But the most important physicochemical processes in arid areas are
~]]]]~ Moderate
Figure 19.29. The United Nations Conference on Desertification (1977) map of areas at risk of
desertification held in Nairobi, Kenya (Goudie, 1990a,b).
those of soil salinization and sodification (sodium enrichment) which result from the
accumulation of soluble salts (chlorides, sulphates and carbonates of sodium, magnesium
and calcium) (Rhoades, 1990). According to the FAO, a soil is saline when its content in
salts exceeds 1 - 2 per cent in the upper 20 cm. In such soils, accumulated salt rises because
of capillary action to give salty coverings and crusts.
Human-induced salinization is a consequence of the irrigation of arid and semi-arid
areas (Fig. 19.30) and it affects about one third of the world's 200 million irrigated
hectares (Table 19.3) (Goudie, 1990a,b). Its effects include modification of the soil
structure, reductions in permeability and porosity, and the reduction or even complete loss
of natural vegetation (Fig. 19.31). Salinization can also be linked to the over-exploitation
of aquifers in coastal areas, which can allow seawater to gain access to them. The failure of
fresh water to reach delta areas because of dams can also be a cause of salinization, as
occurs in the Nile Delta (Mainguet, 1991).
The salinization of the Ebro Basin in Spain (Fig. 19.32) is due to the area's dry climate,
the salt content of the bare rock, the redistribution of salts, geomorphological conditions,
intrinsic features of the soil, and the agricultural practices of the area (Alberto, 1989;
Herrero et al., 1993). Among the latter, the levelling and irrigation of alkaline soils are
among the most important. Aspersion irrigation or drip-feed systems are the best for
preventing salinization. The problem of irrigation-associated salinization is economically
very important (Zekri et al., 1990). In the central region of the Ebro Valley there are
5360 km 2 of irrigated land, of which 1000 km 2 are severely affected by salt and another
1500 km 2 require some kind of control (Herrero and Aragties, 1988; Herrero et al., 1993).
Similarly, in the San Joaqufn Valley of California, 160,000 ha are affected by salts; by the
year 2080, 445,000 ha will be unproductive (Sheridan, 1981).
Rainfall varies enormously in desert areas, and the discharge of rivers can fluctuate
widely. This leads to important variations in salinity, as reflected by the River Casamance
Figure 19.30. Channelled water from the San Juan River originally from the Andes. San Emiliano Dyke,
San Juan, Argentina. This water is put to several uses in this very dry area.
of Senegal (Fig. 19.33), which affect the availability of drinking and irrigation water. In
October 1968, salinity was low, coinciding with the rainy season. But in May 1984, the
whole river had a salt content higher than that of sea water (35%o). Close to its mouth, the
salinity of the fiver tripled this value (Farmer and Wigley, Unpub. data; in Mainguet,
1991).
As mentioned above, inadequate agricultural practices and over-irrigation cause
salinization, but they can also cause lixiviation. As stated by Mainguet (1991): "It is
absolutely vital that irrigation in arid or semi-arid ecosystems be considered a last resort,
especially for large projects."
Table 19.3. Approximate percentages of irrigated land affected by salinization in different countries
(Goudie, 1990a,b).
Algeria 10-15 India 27

Egypt 30-40 Iran < 30
Senegal 10-15 Iraq 50
Sudan < 20 Israel 13
USA 20-25 Jordan 16
Colombia 20 Pakistan < 40
Peru 12 Sri Lanka 13
China 15 Syria 30-35
Cyprus 25 Australia 15 - 20
Figure 19.31. Cereals affected by salinization in the area of Los Monegros in the Ebro Basin. E1
Tormillo, Province of Huesca, Spain.
0 so 100k~ ' ~ , ~ "~-,~rMJ
Figure 19.32. Distribution of saline areas in the Ebro Basin, Spain (Alberto, 1989).
Salinity
ppt
100-
50-
___ - - - - - - ~_ =.....~ _ \
. \
e'ee* ~ ee ~ e e~
0 ' | !
100 200 km
..... 1 .... 2 3 ------4
Figure 19.33. Casamance River salinities as a function of distance (in km) upstream from the mouth
(Pages, CRODT, Dakar, Senegal). The ordinate is in parts per thousand (ppt). Seawater salinity (35 ppt) is
shown as a horizontal line. 1 October 1968; 2 July 1969; 3 May 1984; 4 September 1984. (After Farmer
and Wigley, 1985); in Mainguet, 1991, Fig. 44).
Figure 19.34. Experimental plot for the study of water erosion. Four years after its set-up, more
vegetation was found inside than outside the plot, which became affected by overgrazing by sheep and
goats. Mishor Adumin, Judean Desert, Israel.
,, . . ,,z. . .'..':.~' % . " -~," """

"- % " - . ,, /
, ~.'.'.'.
*...,'.. ~r o . . J
,~.:. * "zi::
**. 9S J'..;:,, 9
, " .~
!
". . . . . 9... . . . . ~!.1
* -*/ 9 *':. " 9
,- I! /-.,~ "'-" '.'',%'.'.j,:'.. 9149 '. I ., "" /

" "-,. ,," ": :. ~ .'7 ~'~:" "."::.'11.',.'. :':'.'.'." 9 . e "" ..!
",, t .-.::. ,,. 9149 ".'.'.-..'.. I .'.'"1
~'~-,, 9 " . 9 ~'~ " . " 9 " : ; 4 : : . ~ ' . ' : 9" 9 . . . . "..I% : ' : . " "1
",".'.: .... .'. . : 9." . .~. " 9 x .. " 9 . "
,/' " ",..-.".' 9 ..... : . . . . . ~.~ ".-'.'." ". 9 ,t~. "-:'.;"1
., .:"<'.',.'-'~"::'";'-..': : . : ".'. ",. :: "a 9 -_ "',,-.:!.1
;:~'.
s ...: .
"~ | "<:'.>'.'.'"
:..'.. *-... 9 *
"'.'
". :.
-:"
. ."
" . .. ' .9' .,', . , ~ ;- " ' " , . - " ~.
"'1
9 .
9"
,,.;.
:
9
" "" " ""'~"t."" ........ ~" ;"
,.,.'<-. ,,-":'/
"1" ~'-" '''|
".. :z-'" I "-:" "~: " . " . ' . ' ~ ' 9 9 : ".:,;~" " , .
'~** II ,.4...
-- . . 9L I I... ~* "%
,~ : .*:' .":~/ ",~. ..l:e. ..7, ~ . . . : 9
. .~,' , . ~4
.- /
.. . ,,,,,'.:: .:. :.a. " - - ,. - L . / " : . ' / . ' : 9 ~"
9: ' . . /I .:.:., :." i '", t ,',." "~,'.:~, " 9. . . . . "4.

I " " 9 9 9 "."
:.. : ' ; s , ' l ,4. "::
9 ,/~.:
'. " "."
9162 i- .'.'. ,, 9 ":.:l. ","9 :"
9 , 9. . . . . t 9 4' "(~ I I 9' . ' :'" 9 ~ % ",0
9. . . ..... ." ,, I 9 4 9 4- %% . ; t ( ' " 9 .'..- %
I :: "::'.', . - " .} "-': "14". " "~', .,' ':.': 9 : ~ ..
L-:, ;?,Y.'::. ~.';,.~: ~, , 4 "~. .,..v. "'. :;:',':.:"" "
0 35 km
l" ,, [i ~ 13 I'.'--4 L
Figure 19.35. Degradationcaused by overgrazing. 1. Barren alluvial surface; 2. Grazing land; 3. Live-
stock sheds and wells" 4. Road and path. (Mainguet, 1991).
Another human-induced cause of the degradation of desert areas is the abusive

agriculture sometimes practiced. Sometimes, degradation is the consequence of
introducing inappropriate agricultural practices (Thomas and Middleton, 1994), and
may lead to the complete loss of productivity. The land is then abandoned and becomes
subject to wind and water erosion. Many examples of this can be seen in China, the Sahel,
Rajasthan, and elsewhere. Deforestation is also a cause of degradation. Deforestation for
fuel is severe in some places where electricity or gas is too expensive for the majority of
people. Acacias have disappeared from Khartoum in Sudan for a radius of 100 km and
wood now has to be brought in from over 500 km away (Goudie, 1990a,b).
The number of cattle in the world increased by 38 per cent between 1955 and 1976, and
the number of sheep and goats by 21 per cent (Goudie, 1990a,b). In many areas, the
number of livestock animals has exceeded the carrying capacity of the land and
overgrazing has occurred (Fig. 19.34). Land has also been brought under cultivation which
has further reduced grazing areas. The loss of plant cover has brought with it both wind
and water erosion. In northern China, steppe vegetation has been degraded by overgrazing
and a series of rings can be seen around wells (Fig. 19.35). These are arranged as a central
area with a radius of 300-500 m where there is little vegetation (often just bare ground), a
second ring with a radius of about 500-1000 m where some resistant grasses and bushes
remain, and then the vegetation returns to normal some 1000-3000 m from the well
(Mainguet, 1991).
Desertification is a serious problem. Technically it can be stopped and many degraded
areas can be restored to production, but the solutions are costly. Preventive measures for
saving plant cover are based on introducing new species more resistant to the lack of water
and the presence of salts. Water usage must also be improved, and wind and water erosion
combated. Ignorance must also be fought, perhaps with social awareness/training
programs aimed at all levels of society.
Chapter 20
The humid tropics: weathering and laterites
1. Introduction
The tropics are the regions located between the Tropic of Cancer (23.5~ and the Tropic
of Capricorn (23.5~ These limits involve areas where the sun can reach the zenith. As a
result, these zones of low latitudes receive a great amount of solar radiation and continuous
heat, with important biological consequences such as the development of a dense and
widespread vegetal cover, which has a strong influence upon the morphogenesis of these
regions. In the intertropical zones there exist both large areas warm deserts and important
zones with high rainfall. A separation between different environments is needed, which
can be based upon several criteria such as climate, biology or geomorphology, among
others. The subdivision is always difficult, especially in continental areas (Reading
et al., 1995).
Some authors discriminate the humid tropics by using vegetation criteria, concluding
that these regions are characterized by the scarcity or absence of xerophytes and the
development of rainforests, deciduous forests and savannas (Fosberg et al., 1961). Other
authors use climatic criteria and establish a limit on an average temperature of 18~ for
the coldest month (Krppen, 1936). The mean annual precipitation has also been proposed,
but great discrepancies exist among researchers. Krppen (1936) used the range of
450-600 mm and Tricart (1974b) between 750 and 800 mm.
About 2400 million inhabitants live in the tropical zones, which means nearly 45 per
cent of the world' s population. Nearly all these people live in the humid tropics, and about
60% in western and southern Asia. The most important economic sector is agriculture,
which obviously depends a great deal on the climatic fluctuations and also on catastrophic
events (McGregor and Nieuwolt, 1998).
Most parts of the surface occupied by the non-arid tropics are located on stable cratonic
zones, which are basically constituted by Precambrian plutonic and metamorphic rocks,
mainly in central Africa and the Brazilian shield. These materials underwent long-lasting
denudational processes, which favoured the development of extensive planation surfaces.
Due to the low topographic gradient, the erosion velocity is slow and weathering
progresses with no difficulty. For this reason, landform evolution over long periods of time
can be studied, under conditions of crustal stability (Thomas, 1974; Wirthmann, 2000).
In the opposite situation, regions affected by alpine tectonics and with high relief,
mainly in Indonesia, SE Asia, Central America and the Antilles, suffer high denudation
rates and weathering cannot deepen. In addition, prevailing rocks are different from the
stable regions, like the great calcareous outcrops where singular karst landforms develop.
As previously indicated in preceding chapters, in the arid regions, and especially in
the semi-arid zones, typical humid tropical weathering profiles are quite common.
This indicates that in the past these dry zones underwent conditions much more humid
than at present. Hence, the recognition of processes and deposits in the humid tropics helps
in the interpretation of climatic variations in past times.
2. Climatic characteristics, vegetation and morphoclimatic domains
Climatic parameters play an essential role in the dynamics and intensity of geomorpho-
logical processes, as well as in the nature of the resulting landforms (Faniran and Jeje,
1983). For this reason, it is convenient to know the magnitude of these parameters and
their variability.
Temperatures in the humid tropics are moderate and do not exceed 34~ very different
from the tropical deserts, with values higher than 50~ Mean annual temperatures
fluctuate between 24 and 30~ measured at sea level. Seasonal and annual variations are
lower than 2~ This thermal uniformity is broken with increased altitude, which is
accompanied by cloud intensification.
In the humid tropics humidity is always high. Relative humidity often exceeds 80%,
and normal values throughout the year do not move far from this number. Maximum
potential evapotranspiration is recorded in the equatorial zones and varies between 1000
and 1500 mm/yr.
With the exception of very elevated areas, precipitation is in the form of rain. As
temperature is quite uniform, seasonality is marked by precipitation variability.
Thunderstorms are very common, and contribute some 90% to the total rainfall in
savanna zones. Most thunderstorms start in the afternoon and have short duration. The
strongest one ever recorded produced an 1870 mm rainfall in a single day, 16 March 1952,
in Cilaos on Reunion Island. In the humid tropics 40% of the precipitation has an intensity
higher than 25 mm/h and several records are known with intensities of up to 340 mm/h
for some minutes (Reading et al., 1995). Precipitation regimes are as follows (Birot, 1973):
(1) Tropical areas of equatorial regions influenced by the Intertropical Convergence
Zone, rainfall through all the year and one single maximum. Annual average precipitation
is about 2000 mm or much higher. (2) Areas within 10~ latitude north and south of the
Equator, with two dry and two rainy seasons and precipitation ranging between 1000 and
2000 mm. In this humid and dry tropical regime rainy seasons can be shorter and
precipitation can reach 650 to 1000 mm. (3) Dry regime, with precipitation between 250
and 650 mm, located on the desert margins.
Vegetation in the humid tropics is a direct response to the existing climatic regime.
An equatorial rain forest (pluvisilva) can be differentiated, which is constituted by very
close trees of up to 40 m high with large perennial leaves and thousands of tree species
(Figure 20.1). The high vegetation density generates a continuous shadow on the ground.
Coastal vegetation is represented by mangroves. The rainforest mainly develops in the
Amazon and Congo basins, Central America and SE Asia and Indonesia. The savanna is a
consequence of a humid-dry tropical climatic regime. It is formed by very isolated trees,
sometimes grouped in corridors following valley bottoms, that arise upon a grassland
of tall graminoids, commonly higher than a person (Figure 20.2). Summer fires in these
yellow dry graminoids are very frequent and are responsible for the grassland predomi-
nance over the forest. The savanna is mainly located in central and southern Africa,
The humid tropics: weathering and laterites 503
Figure 20.2. Savanna with graminoid grassland and dispersed trees in Nairobi National Park, Kenya.
Because temperatures are very homogeneous, climatic subdivisions are made as a

function of precipitation. For the humid tropics and transitional areas a limiting value is
established with annual precipitation higher than 750 mm (Thomas, 1994). It includes rain
forest (Af and Am) and savanna (Aw and As) climates (K6ppen, 1936).
Morphoclimatic differentiation is quite similar, because precipitation and the resulting
vegetation cover play a determinant role in humid tropic regions. The numerous geo-
morphic inheritances from other climates existing both in equatorial and in humid and dry
climates, however, must be taken into account, although more obvious in the latter, near
zones of dry climate. Dominant geomorphological processes in the humid tropics are
chemical weathering, mass movements, and fluvial erosion, the latter being much more
important in regions of humid and dry climate (Wilson, 1968, 1969).
Without any doubt, the intense deep weathering and its profile evolution constitute one
of the most characteristic features of this morphoclimatic zone. Some authors, however,
doubt the existence of a tropical climatic geomorphology (Stoddart, 1969c), and others the
differentiation of the tropical zone within the global morphoclimatic system (Douglas,
1978; Gupta, 1993). This last author indicated that processes and forms in the tropics
depend only partially on climate, and that there exist other important factors such as
neotectonics, lithology, episodic and catastrophic events (cyclones), and anthropic acti-
vities. Although this statement is true, it must be kept in mind that all these non-climatic
processes and factors act throughout the world, and they operate at the same time as
climatic causes. This vision would correspond to very populated regions undergoing
important neotectonics. This is a quite different situation from the one existing in African
and Brazilian regions, where old planation surfaces have favoured the development of
thick regoliths, while environmental changes have produced lateritic landforms, etch
surfaces, and other features typical of humid tropical zones.
Following Btidel (1977, 1982), humid tropical zones are completely different from
other morphoclimatic zones. They do not develop fluvial terraces, gravel plains, or flood
plains such as in the ectropics (temperate regions), nor loess mantles. The soil is much
deeper than in the ectropics, and contains plenty of duricrusts and hardpans. Inselbergs of
variable size are common in the tropics, standing out on extensive erosion surfaces, which
are weathered at the same time as they form. Mountain zones are characterized by deep
incisions. This author assumes that morphodynamics in tropical lowlands are completely
different, where landforms are controlled by the generation of etch surfaces.
Tricart and Cailleux (1965) differentiated an intertropical zone with warm temperatures
and constant fluvial flows. Depending the rainfall distribution, its annual amount, and
the vegetation cover density, they divide this zone into a savanna domain, with less
precipitation and less dense vegetation cover. Sheet wash is considerable and chemical
weathering is strong and discontinuous in time. The rainforest domain is characterized by
dense vegetation cover and higher precipitation. In this domain chemical and biochemical
processes reach a maximum intensity.
3. Weathering in the humid tropics
Weathering is the response of materials that were initially in equilibrium inside the
lithosphere to conditions at or near their contact with the atmosphere, hydrosphere and,
perhaps more important, the biosphere (Reiche, 1950). The importance of the biosphere is
perhaps overemphasized. A great part of the rocks have been submitted to high pressures
and temperatures, in the absence of air and water. After outcropping from the surface, they
change towards an equilibrium with pressures of about 1 atm, much lower temperatures
and the presence of air and water, which are the conditions existing in the contact
lithosphere-atmosphere (Carroll, 1970).
Climate constitutes one of the most significant factors in weathering development.
Water from precipitation is the most important reactive agent in the weathering process
and its quantity is a prime factor in the intensity and type of alteration. In humid tropical
regions water attacks crystalline structures and evacuates bases (Na +, K +, Ca 2+, and
Mg 2+) as well as silica to some extent. It also affects the velocity of biochemical reactions
in plants. Temperature has direct and indirect effects on weathering processes. High
temperatures accelerate chemical reactions. Following the Van't Hoff Law, every 10~
increase in temperature duplicates or triplicates the rates of chemical reactions (Tricart
and Cailleux, 1965). Moreover, heat increases biological activity. As a result of these
temperatures, bacteria consume great quantities of organic matter in decomposition, which
brings about the scarcity of humus in these humid tropical environments.
The result of the actuation of weathering processes, especially those related to
chemical alteration, gives rise to an intensively weathered rock (Taylor and Eggleton,
2001). This rock decomposition is a consequence of intense precipitation and temperature
and also of the vegetation response to these conditions. In these areas natural waters
reach pH values ranging between 4 and 6, and very possibly these acid conditions may be
due to silicate hydrolysis (Curtis, 1976). The important release of biogenic carbon dioxide
supplies great quantities of bicarbonates to the waters. Humification processes produce
humic and fulvic acids, which export cations as organometallic compounds (Thomas,
1994). The activity of all these processes produces thick alteration profiles that will be
analysed later in this chapter.
3.1. Physical and biological weathering
Weathering processes can be divided in three types: physical or mechanical, biological,

and chemical, and all of them can act simultaneously. Physical weathering breaks the rock
by different mechanical mechanisms. Most of the forces applied to the geological
materials is external, although they can also accumulate inside the rock.
Many rocks were originated or remained for a long time at considerable depths in
the lithosphere and hence, were submitted to high confining pressures. When uplifted to
the surface, these rocks suffered a pressure-release by unloading that generates joint-like
fractures parallel to the topography; a process known as sheeting (Figure 20.3) (Gilbert,
1904). Joint spacing increases with depth and is better recognizable in massive rocks such
as granites, conglomerates, sandstones, and so forth. The difficulty of recognizing this type
ofjointing in non-massive rocks is probably due to the existence of old bedding planes that
would distribute expansion stress (Ollier, 1965). Other authors suggest that sheeting is
a consequence of lateral compression (Twidale, 1973, 1982a; Vidal Romanf et al., 1995;
Vidal Romanf and Twidale, 1998). As will be discussed later, this jointing has great
importance in the landform development of tropical cratonic areas (Birot, 1968a,b).
Figure 20.4. Termite-mounds in the Masai Mara savanna of Kenya.

intensity in the high parts of stratovolcanoes, which interrupt the extended plains of
western central Africa, like mounts Kenya and Kilimanjaro. It also acts on the high
Andean peaks of Peru, Equador and the Nevados of Columbia.
Rock disintegration is controlled by the rupture of materials, due to the pressure exerted
by growing roots. The carbon dioxide supply to the solutions is also very important, which
results from the gas exchange or respiration of the flora and fauna existing in the soil and
from decomposition of organic matter. The fauna plays an important role in transferring
and mixing materials, moving the regolith to other areas. In the humid tropics this last
work is carried out by termites and worms and can be responsible for the sorting of the
upper part of the soil. There exist several thousand socially organized termite species.
They carry clay size particles to build termitaria, which they construct with segregated
saliva. The number of termites reaches values of 9 million~a in the Ivory Coast and their
amount decreases with latitude and altitude (Goudie, 1988). Termitaria appear as small
mounds (Figure 20.4) and towers (cathedral-termitaria of Tricart, 1974b) (Figure 20.5),
Figure 20.5. Inactivetermitaria tower of about 6 m high in the forested savanna of Eastern Bouak6, Ivory
Coast.
and they can reach up to 9 m high and a density of 1000/ha. Large termitaria are built by
only a few species, while smaller termite-mounds are associated with numerous species.
Growing rates are very variable, from 25 to 500 mm/yr (Goudie, 1995). When termites
leave the termitaria, it is then exclusively submitted to erosive processes, leading to its
destruction. Although the termitaria material is much stronger than the underlying soil it
is finally eroded, normally in less than 10 years (Goudie, 1988). Termites also build small
filiform corridors upon the soil and vegetation, made up of cemented soil particles. These
accumulations give an idea of the amount of material involved in the removal process.
Following calculations made by several authors in Nigeria, Senegal, and Kenya, the
accumulation rate ranges between 300 and 1050 kg/ha/yr (Goudie, 1988).
Terrestrial worms also build constructions like termites although of lesser size, hardly
reaching 30 cm in height. Earthworms excavate burrows of up to 70 cm depth, which
indicates the importance of these organisms in the removal and mixing of soil components.
The amount of removed material is estimated as 5 to 25 mm/yr, greater than the quantity
moved by termites (Goudie, 1988).
All these weathering processes have a certain importance, but chemical weathering
plays the most relevant role in the disintegration of minerals and rocks in humid tropical
zones.
3.2. Chemical weathering
Chemical weathering results from a change in the chemical environment (Loughnan,

1969), because rocks formed in high pressure and temperature conditions and in the
absence of air become unstable when put in contact with the atmosphere. Water, oxygen,
and carbon dioxide attack the rock minerals through exothermic reactions. Most soluble
products are eliminated by dissolution, leaving a residue rich in less soluble components.
This residue is structured to produce neoformed compounds, mainly clay minerals and
iron and aluminium oxihydroxides or sesquioxides, which are in a more stable equilibrium
with the environmental conditions.
Jenny (1950) proposed a disintegration mechanism for the mineral structures. Following
the Pauling Rules, the sum of the positive charges in a crystal must be equal to the sum
of negative charges, but on the crystal surface valences are not compensated, resulting in
an unstable equilibrium. When crystals are in contact with water, the OH- ions join to
the existing cations and the H + ions with oxygen and other negative ions. The H + ion has
a small ionic radius and can occupy cation places in the crystal surface. Cations are then
released as hydroxides. In addition, the loss of H + ions modifies the pH of the solution.
3.2.1. Weathering of silicates
Silicates are the most common minerals in rocks. Approximate proportions of mineral
species in the Earth's surface are: feldspars (30 per cent), quartz (28 per cent), clay
minerals and micas (18 per cent), calcite and dolomite (9 per cent), iron oxides (4 per cent),
pyroxenes and amphiboles (1 per cent) and other minerals (10 per cent) (Leopold et al.,
1964). Cations released in the weathering reactions can be easily washed away or remain
in the residue. This is due to the physico-chemical variations of the solution. In this section
a brief discussion is presented on these factors and their incidence on the mobility and
precipitation of the most frequent ions in silicates.
Following geochemical studies, fresh water pH fluctuates between 4 and 9 (Mason,
1958). Figure 20.6 presents the solubility of some of the most common compounds as a
function of pH (Loughnan, 1969). Sodium and potassium are soluble under any pH value
and calcium and magnesium hydroxides are also soluble under the pH of natural waters.
Instead, titanium and aluminium oxides and ferric hydroxide are insoluble in the marked
intervals. Only ferrous hydroxides, titanium, and calcium carbonates are influenced by the
pH of natural waters. The solubility of silica is low, although under a clear alkaline pH it is
washed away with a certain facility. Its solubility curve refers to amorphous silica and not
to quartz, which has a solubility 20 times lower than the amorphous silica (Carroll, 1970).
Nevertheless, several environments exist where acidity and alkalinity reach very high
values. For example, aluminium can be leached in the aluminates form in very basic
playa-lake environments. In soils with great quantities of organic matter pH can reach
values lower than 4 under which aluminium can be removed. Sulphur oxidation also
makes waters reach very high acidities.
The oxidation-reduction potential (Eh) has a clear incidence on the solubility of
polyvalent cations, such as Fe and Ti. Chemical elements under an oxidation state are in
equilibrium with the environment. Because solubility varies as a function of the different
valence states, the redox potential affects the mobility of the chemical element. Figure 20.6
shows the solubility of trivalent and divalent iron. Ferric hydroxide is only soluble with
10 . , ,!
TIO2
Fe(OH) 2
Fe(OH)3 Ca
J Mg(OH)~
~" 7 AI20 3
00
IJJ
._J
o 6
._1
._.1
~
>.-
5 /
I--
._1
ca 4
23
._1
o
oo 3
2 ......
f
1 -
0 -- 1 1 l 1 . . . . . i ~ .1 . 1
1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH
Figure 20.6. Solubility in relation to pH for some components released by chemical weathering
(compiled from several authors by Loughnan (1969)).
a pH lower than 3, whereas ferrous hydroxide is soluble with a pH of up to 8. Hence, the

value of the redox potential is essential to the mobility of iron. The main factors affecting
Eh in weathering environments are the availability of atmospheric oxygen and the presence
or absence of organic matter (Loughnan, 1969). Oxidation is an exothermic reaction that
takes place above the saturation zone. Below the water table a reducing environment
without oxygen prevails. For this reason, in these environments iron appears in its ferrous
state and is related to the gley horizon, generated in humid areas without any drainage. The
organic matter is a powerful reducing agent due to its rapid oxidation which produces
carbon dioxide. Potassium is a very soluble cation, although it is usually retained in the
structure of clay minerals of the illite type. Nevertheless, under conditions of high
precipitation it is dissolved, without leaving any remains in the system.
Chelation consists in the joining of a cation and a complexing agent (chelate), such as
the EDTA. As a result an organometallic compound is generated, bearing a ring-shaped
structure with the cation inside (Bland and Rolls, 1998). The cations fixed by these
chelates can be divalent or trivalent and the organometallic complexes are quite soluble. In
nature they are produced by the generation of humic and fulvic acids associated by the
decomposition of plants and root secretion (Keller, 1957).
Ion exchange is the most important process in chemical weathering (Carroll, 1970).
It consists of the reaction between the ions of a solution and the ones existing in minerals.
Clay minerals are the main agent responsible for the exchange capacity in rock weathering.
If the mineral is not electrically neutral, a cation adheres to the clay surface, neutralizing
the negative charge. A frequent replacement is Si 4+ by A13+. Each clay mineral has a
different ion exchange capacity and minerals of the montmorillonite group reach the
highest values.
Hydration is the addition of water molecules to a mineral structure. When hydrated,
iron oxides transform into iron oxihydroxides. Water is easily introduced into the
crystalline structure of expansive clays. The transformation of anhydrite (CaSO4) into
gypsum (CaSO4.2H20) also involves a considerable increase in volume.
The most important factor controlling the disintegration speed of crystalline structures
and resulting products is the quantity of solution water. This water removes the soluble
cations resulting from hydrolysis processes. The quantity of dissolution depends on the
water volume in contact with the mineral surface and on its solubility. When preci-
pitation is important and takes place over long periods, the most resistant minerals,
such as quartz, can disappear by dissolution. The Weipa bauxites of North Queensland,
Australia derive from the leaching of a substratum constituted by kaolinitic sandstones
with 90% quartz and 10% kaolinite. Monsoonal precipitation give rise to bauxites with
5% quartz and small quantities of kaolinite (Loughnan and Bayliss, 1961).
Several investigations have been made for determining the order by which the most
common cations are lost. Polynov (1937) compared the composition percentage of the
solved matter in rivers, for drainage basins on igneous rocks with the composition of
the latter, obtaining the following losing order: Ca 2+ > Na + > Mg 2+ > K + > SiO2 >
Fe203 > A1203. Other studies indicate that this order can vary depending on the rock type
(Reiche, 1950; Carroll, 1970).
Crystalline structure and atomic bond strengths play an important role in mineral
weathering. During magmatic crystallization the minerals that first form are the ones with
the highest melting point and the largest number of weaker ionic bonds. The sequence is
BOWEN REACTION SERIES GOLDICH WEATHERING SERIES
(Early crystals, largely removed as formed)

Olivine Olivine
Pyroxine~ Anorthite-rich)agioclase Calcic plagioclase
Augite
Calcic - alkalic plagioclase
Amphibole Intermediate plagioclase
/ Homblende Alkali - calcic plagioclase
Alkalic plagioclase
Albite-rich plagioclase
Biotite
\
Potassium feldspar
/ Biotite
Potash feldspar
;r
1
Quartz
Muscovite
Quartz
~1
(Late crystals formed from residual magma)
Figure 20.7. The reaction series of Bowen and the weathering sequence of Goldich (adapted by Oilier
and Pain (1996), Fig. 4.2).
represented by the crystallization series of Bowen (1928). Goldich (1938) obtained

the weathering sequence for the most common silicates from a detailed study on the
weathering of several igneous and metamorphic rocks (Figure 20.7). The Goldich
sequence is the inverse of the Bowen series, because the most easily weathered minerals
are the first that crystallise in a silicate melt. The most stable minerals crystallize later in
the cooling sequence and have a greater number of the more stable covalent bonds.
3.2.2. Dissolution of carbonates
Carbonates occupy important outcrops in alpine belts affected by humid tropical climates.
Their weathering is due to solution processes, which have been analysed by many authors
(Jennings, 1971; B6gli, 1980; Jennings, 1985; Trudgill, 1985; Ford and Williams, 1989,
among others).
Calcite is the dominant mineral in carbonates, although it can also be accompanied
by dolomite, whose dissolution is five times lower than that of the calcite. This last
mineral is slightly soluble in distilled water, although its solubility increases with the
dissolution of carbon dioxide in natural waters. Temperature affects dissolution of CO2
and is maximum at 0~ This relation is summarized in Henry's Law in which the total
amount of limestone that can be dissolved under saturation equilibrium per unit of water
volume is a direct function of the CO2 partial pressure of the air in contact with water,
and an inverse function of water temperature (Figure 20.8). The solution at the point
located on the saturation curve of 30~ can cool (towards C) and hence more limestone
will be dissolved. If it is heated (towards W) some calcium carbonate will precipitate.
If it is put in contact with air containing less CO2 quantity (towards L) it will lose
some CO2 and will precipitate calcium carbonate and, finally, if the solution finds an air
with a higher proportion of CO2 (towards H) it will take it and more limestone will
be dissolved.
450
SUPERSATURATED
400
350
300
._1
250
v
E
o'3
O
0 200
0
150
UNDERSATURATED
(Aggressive)
100
50 I
I
I
0 2'5 s'o 7~5 1()0 1~'5 1,50 1-75 200
Equilibrium CO 2 (mg/L)
Figure 20.8. Saturationequilibrium curves for solution of calcium carbonate at different temperatures as
a function of equilibrium carbon dioxide in solution (after Trombe (1952)).
The effect of temperature on carbon dioxide dissolution is very small, because at about
30~ it dissolves three times less than at 0~ Air in the soil, however, can reach CO2
values of up to 30 to 60 times the content of a normal atmosphere (0.03% in volume).
In only weakly aerated tropical soils values of 20 to 25% CO2 can be obtained. This seems
to be due to the decomposition of organic matter and to root respiration. Therefore, the
biogenic carbon dioxide is the primary factor in the dissolution of carbonates and its
content is a function of temperature and precipitation, factors that activate the vegetation
development. As a consequence, in humid tropical zones chemical corrosion of limestones
reaches very high values.
3.3. Weathering degrees and products
Weathering degrees of the alteration products are variable and several indexes are used
to differentiate them (Taylor and Eggleton, 2001). These are based on qualitative scales
of friability (Ollier, 1965, 1984), on the mechanical properties of the regolith (Lumb, 1962,
1983), and on the chemistry and mineralogy of the weathering profiles (Grant, 1969;
Sueoka, 1988; Geological Society, 1990). As an example, Grant (1969) correlated
abrasion pH to the relation between the most mobile elements (Na, K, Ca, and Mg) and
the residual clay minerals. The lower the index, the more intense the weathering. Sueoka
(1988) proposed the chemical weathering index, which consists of dividing the sum of the
molecular weight percentages of A1, Fe, and Ti oxides plus H20, by the rest of elements
(alkali, alkaline earth cations, and silica). Thomas (1994) included other indexes related to
the weathering degree.
As a consequence of the destruction of crystalline structures by weathering processes,
several cations are released. Those with high solubility are leached by water, while others
remaining in the residue are restructured, giving rise to neoformed minerals (Bland and
Rolls, 1998). The most common ones are clay minerals, quartz, and oxihydroxides of iron
and aluminium, which form lateritic hardpans.
Clay minerals are hydrated silicates of aluminium, iron, and magnesium ordered into
sheets, the reason for which they are named "phyllosilicates." The tetrahedral sheets are
formed by Si surrounded by four O ions, while the octahedral layers consist of six O or OH
ions located around A1, Mg, or Fe cations that are placed in a central position. These two
basic sheets can be grouped in three different ways. First, a tetrahedral sheet joints an
octahedral one to form structures of the type 1:1. Clay minerals of this type are those of
the kaolinite-halloysite group. Secondly, an octahedral sheet can be sandwiched between
two tetrahedral sheets. This type of arrangement gives rise to the 2:1 phyllosilicates, which
are represented by the smectite group (montmorillonite, beidellite, and nontronite) and the
illites. The third ordering case results from the joining of an octahedral sheet between
adjacent 2:1 sheets, and this type is known as 2:1:1 clay minerals, among which chlorite is
the most common.
The genesis of the clay minerals is complex due to the wide variety of weathering
environments. They can also originate as a result of weathering of non-phyllosilicate
minerals. Cations, silicon and alumina are released and they reorganise to from clay
minerals. When the altered mineral is a phyllosilicate, weathering is mainly produced in a
solid state (Birkeland, 1984).
Studies by Sherman (1952) on basalt weathering in Hawaii indicated an increase in
the clay percentage with precipitation, as well as changes in the mineralogy of the rock.
Figure 20.9 shows how montmorillonite is a typical mineral of low precipitation regimes
that transforms into kaolinite under conditions of higher leaching. When precipitation is
high the soil desilicifies and iron and aluminium oxihydroxides are formed. The figure also
indicates that if a dry period occurs, the alteration residue enriches in iron and titanium
oxides. When precipitation takes place throughout the whole year the main constituents
are aluminium oxihydroxides.
Taking into account the dominant weathering products, several weathering zones can
be distinguished in the humid tropics (Strakhov, 1967) (Figure 1.8). In this sense, Pedro
(1966, 1968) differentiated:
1. Bisiallitization zone with the formation of clay minerals of the 2:1 type
(montmorillonite, 4SiO2.A1203.H20), in areas with a precipitation lower than 500 mm.
2. Monosiallitization zone with kaolinite (2SiO2.A1203.2H20) as the predominant
mineral in the regolith, in zones with a precipitation of 500 to 1200/1500 mm.
100
80
60
Cb
o)
40 t ./" Gibbsite ~,
Diaspo.re ~Bauxite
~~. Boehmite"
2O
" - I ,I ~ | I I
76 152 228 304 380

Mean annual precipitation (cm)
Figure 20.9. Progressivedevelopment of clay types in soils of Hawaii, under a continuously wet climate
(after Sherman (1952), in Birkeland (1984), Fig. 11.12).
3. Allitization zone in which gibbsite (AI(OH)3) appears with kaolinite in regions with a
precipitation higher than 1500 mm.
Sanches Furtado (1968) indicated that kaolinite is the most abundant mineral in the
humid tropics. He also established a geographical differentiation in several regions as
a function of prevailing minerals and precipitation. Jenny (1941) analysed the role of
temperature in the generation of clay minerals. He indicated that the clay content in
regoliths of basic rocks is a direct function of temperature. At 10~ average annual
temperature we obtain 15% of clays in the soil, while at 16~ it rises to 50%.
Duchaufour (1977) differentiated among three main types of alteration in the tropics,
which broadly characterize different climatic zones. According to this author, fersiallitic
soils predominate in zones with precipitation between 500 and more than 1000 mm and
temperatures between 13 and 20~ These soils also have smectites and the beginning
of desilicification to generate kaolinite, with the iron appearing as oxihydroxides with
nontronite, which is an iron variety of smectite. Ferruginous soils generate in regions
with higher temperature and precipitation, and most parts of the neoformed clays belong
to the kaolinite group. Finally, the ferrallitic soils correspond to the terminal phase of
the evolution and alteration of soils, when primary minerals, except quartz, have been
completely hydrolysed and have released Fe and A1 oxihydroxides. Climatically they
develop with an average annual temperature higher than 25~ and precipitation above
1500 mm.
In general, the type and amount of clay minerals vary with depth, although some
profiles do not present substantial changes. These variations must be related to the intensity
of leaching, which removes many ions and silicon down to important depths. This is a quite
common modification. In this way, surficial gibbsite passes with depth to kaolinite or,
if this clay mineral is already at the surface, the transition is to montmorillonite (Loughnan,
1969). These sequences can be in equilibrium with environmental conditions,
although they can change. If the soil profile is eroded, the montmorillonite forms part of
the surficial outcrops, but if there is an aggressive climate the montmorillonite transforms
into kaolinite at the surface (Figure 20.10). If the surface lowering due to denudation is
Figure 20.10. Verticaldistribution of clay minerals as a function of leaching conditions within the soil.
(A) Leaching conditions favour the formation of kaolinite and montmorillonite at different levels in the
soil profile. (B) Rapid lowering of the surface by erosion results in the montmorillonite outcropping in a
soil environment of high leaching. (C) With time, the montmorillonite in the surface becomes desilicified
and alters to kaolinite. In the figure, the wider the arrows, the greater the leaching (after Birkeland (1984)).
slow and is balanced with the speed of kaolinite generation, the profile modifies its
mineralogical distribution (Birkeland, 1984).
Formation and transformation of clay minerals in the weathering profile are slow
processes. Hence, the study of clay minerals can be useful for making paleoclimatic
reconstructions (Birkeland, 1974). As previously indicated, the montmorillonite forms in
zones of low precipitation, whereas kaolinite and halloysite generate in regions of high
leaching. If we find kaolinite in soils under an arid climate, we can deduce that the area
underwent a humid tropical climate in the past. On the other hand, when an arid zone
bearing montmorillonite in the soil profile undergoes a climate change towards high
leaching conditions, the montmorillonite generated under that climate transforms into
kaolinite. In consequence, we can recognize humid to arid climatic changes, but not the
opposite (Pedro et al., 1969; Singer, 1979/1980). Because mineral transformations
are slow, for these changes to occur, the climatic change must have a long duration. If the
climatic change is brief then no detectable modifications occur. In volcanoes located on
humid tropical zones, pedogenesis is relatively fast, although it is frequently interrupted by
new volcanic emissions that bury and fossilize the soils. Under these circumstances the
endogenous exhalative activity is much more rapid than pedogenesis and if some climatic
change occurred, it would not be recorded in the soil (Figure 20.11).
Figure 20.11. Reddishpaleosoils buried by volcanic ashes in the GregoryRift Valley, Kabernet, Kenya.
3.4. The weathering profile
Many landforms exhibit a thin layer of weathered material, whereas others are covered by
an important alteration mantle, whose physical and chemical characteristics are of great
importance in the landform evolution (Gerrard, 1988). In situ altered rock is named
saprolite, although the more general term regolith is commonly used, which also includes
weathered rocks that might have suffered thin mass movements (creep) or the activity of
animals and plants (Ollier, 1984). French researchers call it alterite. The sandy material
resulting from granite alteration is known as grus or lem.
The limit between the unweathered rock and the altered material can be sharp as in
some granites, or diffuse as in some limestones and slates. Ruxton and Berry (1957) denote
this limit as the weathering basal surface, although this term has been criticized due to its
static implications. Because alteration deepens, this is a dynamic limit, and therefore it
seems preferable to name it the weathering front (Mabbutt, 1961a).
3.5. Weathering depth
Regolith development and deepening is a function of factors such as climate (temperature

and precipitation), biology (vegetation), geomorphology (surficial stability), hydrology,
geology (rock type and joint density) and chronology (climatic and tectonic changes)
(Faniran and Jeje, 1983; Thomas, 1994). The weathering depth results from a balance
between weathering intensity and surficial erosion, which is closely related to topography.
In the tropics overland flow erosion is low (Fournier, 1960; Douglas, 1969), which favours
weathering. Weathered material is easily eroded on slopes, but where these are gentle, or
where it develops on plains, erosion is very small. As a consequence, the maximum
weathering depth will be reached in plane zones with a climate suitable for rapid
weathering, smooth relief, and vegetation that hinder run-off. In these zones erosion is
minimum (Ollier, 1984).
Inside the profile, weathering reaches its maximum intensity in the aeration zone. In the
saturation zone processes like reduction, hydrolysis and chemical substitution also take
place, although with a lesser reaction velocity (Gerrard, 1988). These circumstances led
Lelong (1966) to differentiate between an upper alteration zone, where water flows easily
and increases hydrolysis processes, and a lower alteration zone in which percolating water
slowly flows through joints. All these processes help to explain the great thickness of
altered rocks.
Geological research in humid tropical zones focuses on outcrops of unaltered rock. On
the contrary, when the regolith is studied with an applied finality (civil engineering,
mining, agronomy, and others) it is necessary to know the surficial extent of rock outcrops
and the regolith depth. This is obtained by mechanical drilling supported with geophysical
methods. Quarry activities (Figure 20.12), open-pit mining works, terrain excavation for
road construction, and so forth, are excellent places for studying weathering depth. Ollier
(1965) recorded depths of 150 m during some tunnel drillings for a hydroelectric project in
Victoria, Australia. Thomas (1965) indicated a depth of 100 m in Nigeria and Ollier
(1960) found the same in Uganda. In South America, Nagel (1962) on metasediment
ridges in Brazil and Feininger (1971) on quartz diorites in Colombia, determined depths
more than 100 m.
Figure 20.12. Weatheringprofile observable on a granite quarry. Note the ferrallitic zone in the upper
part, thinning towards the valley, Aparecida, Sao Paulo, Brazil.
In the arid and semi-arid zones of Africa (Thomas, 1974) and Australia (Mabbutt,
1965b), in temperate regions of the USA (Kaye, 1967), and Europe (Demek, 1965;
Guti6rrez and Rodrfguez, 1978; Molina and Blanco, 1980) deep weathering profiles are
relatively common. It is obvious that the origin of such thickness is not possible in the
climatic conditions under which they remain. In the opinion of numerous authors, they
must be interpreted as generated under humid tropical climates, with a greater chemical
activity. Therefore, the existence of thick weathering profiles located in zones far from the
humid tropics is a good indicator of climatic change.
Thomas (1974) has analysed the spatial distribution of regoliths. The influence of rock
structure is apparent because it affects weathering depth. In this sense, the density of
splitting planes existing on the rock massif controls weathering thickness. The maximum
depth is related to the maximum density (Thomas, 1966) (Figure 20.13), which some-
times corresponds to intersection zones between fracturing systems. For this reason the
weathering front is very irregular and bears no relation to the topographical profile.
On the contrary, the deepest weathering develops on the fluvial divides. This can be due to
the fluvial erosion of a previously thick weathering profile (Figure 20.14) or to a faster
weathering on the divides (Thomas, 1994).
3.6. Weathering zones
The alteration profile is not an unstructured regolith. Instead, several differentiations can
be made within the profile, especially apparent on granitic rocks. The works by Ruxton
and Berry (1957, 1961) and Berry and Ruxton (1959) on granite rocks in Hong Kong were
the first key studies in establishing a zonation of weathering profiles. Wilhelmy (1958)
studied similar topics in eastern Asia, Africa, Central and South America, the Iberian
Peninsula, and Corsica. Mabbutt (1961b) analysed them in western Australia. Inside the
profile the number of flesh rock blocks decreases towards the higher parts whereas their
angularity increases with depth. In some cases joint blocks present a concentric structure
formed by shells of residual materials, with a corestone that can remain totally flesh
and coherent. This disposition is known as spheroidal weathering (Figure 20.15 and
Figure 20.16). Its origin is thought to be due to more rapid chemical attack on comers and
edges of joint blocks, compared to joint-block faces, as well as the migrations of elements
within the rock (Ollier, 1984). The upper parts of the profile are easily eroded down to the
zone where blocks predominate. The differentiation into zones is based on the alteration
degree of the granite rock and on the presence of unaltered blocks inside the profile
(Figure 20.17). All the previously cited authors apply these criteria to establish divisions
within the weathering profile. Gerrard (1988) made a wide compilation of classifications
of weathering profiles presented by different authors. Ollier (1984) proposed the most
widely used zonation (Gerrard, 1988):
1. Soil.
2. Structureless regolith.
3. Regolith retaining rock structure (Figure 20.18).
4. Regolith with rounded corestones.
5. Regolith with angular, locked corestones (Figure 20.19).
6. Fresh, unweathered rock.
Figure 20.13. Deep weathering distribution in metres, in a fine-grained, biotite granite, near Jos, northern
Nigeria. 1 Outcrops of unweathered granite showing contours at 15 m intervals. On the cross sections
(A-B, C-D) saprolite is ruled (after Thomas (1966), in Thomas (1994), Figure 3.16).
Obviously, the zonation of weathering profiles on sedimentary, metamorphic and

volcanic rocks is very variable, although the Ollier profile type is also applicable. In all
of them rock fissuration plays a fundamental role. Other profile divisions can be made
based upon chemical, mineralogical, hydrological, or engineering criteria, depending on
the objectives (Ollier, 1984).
Figure 20.15. Spheroidal weathering in acid volcanic rocks, Fort de France, Martinique Island, Lesser
Antilles.
Figure 20.16. Spheroidal weathering in basic lavas of the Gregory Rift Valley, Kabernet, Kenya.
Boden ~ / ' z / ' / ' / / ] ~ / J ' , / ~ z ~ ~ J / z / ' , / - / ( / z / Z , Tl'A Migratory

Colluvium U / / / / / / ~ / / / / / / / / ~ Layer
":':':':':':':':':'::':':':'::':':':':':':'::':':'i:i:i:i:i:i:i:!:i:i:
"oOoOoQeeo~176176176176176176176
Zooe, i!i!iiii!iiiiii!iii!iiiiiiiiiiiiiiii!iiiii!!i!i!i!!!i!i!iiiii!ii.....
9 ~,oO : .+g ~.. ~ o.,~ ~... : %.~ ~.o : ..~ %.,. I "o'g
o e B e Q o e e . o e e e o e s o ~ e e o e e o e *
ZonellA '1"i":':'2":'2"2"2"2"i":'2"2"2"2"""'''""'"" ""
.......... :~i.i.~-:.i-2-:-:-:'~'~+'i-"
l . + . . . . , . . , . . o , . . . . - - o + + . -
:::::::::::::::::::::::::::: ....... :g:

+ ~ o . . . + + . . . . . . . . . - - . - - o - -
+ . , e . . . , . + . + i . . + . . ~ 9
Zone liB .~'2"1"i .'O.'i'~'~'~':'~':'~ . ( ~ . ' ~ ' ~ ' ~ ":':'~:~"

:::::::::::::::::::::::::::::::::::
. ,,. . . . . . . . , ~ , 9 . . ... 9 . . .
Zone III Basal

Surface
Zone IV
Figure 20.17. Characteristic weathering profile over jointed granite (modified from Ruxton and Berry
(1961)).
Figure 20.19. Weathering profile where the transition between the rounded and the angular block zones
can be observed, Spanish Central System, Villacastin-Avila road.
3.7. Weatheringrates
Chemical weathering processes exist in other climatic zones, but in the humid tropics
alteration velocity is much higher (Muxart and Birot, 1977). In this sense, and in addition
to the study of these processes, the quantification of rock weathering is also necessary.
The methods applied for this purpose are varied. Quantitative data can be obtained from
the alteration of building and ornamental stones, whose age of installation is known.
Numerous authors have carried out these types of works under different climates (Ollier,
1984). Approximate data can also be obtained in the laboratory by accelerated aging.
Values recorded in the beginning of the weathering process are the highest. Therefore,
recorded weathering rates will be greater than those existing throughout the whole
alteration, and they have been estimated to be about 200 to 400 times higher than the ones
obtained in the field (Swoboda-Colberg and Drever, 1993). Nevertheless, one of the most
utilized methods is based on the quantification of the solid and solved discharge, measured
in experimental plots of some tens of metres long. The chemical composition of the
weathering residue has also been studied (Nahon, 1991). The experimentation period
is variable and normally fluctuates between 1 and 5 years. In humid tropical savanna zones
of Cameroon a surficial lowering rate has been obtained of 1 to 2 mm/ka (Birot, 1978), and
1.2 mm/ka in Ivory Coast (Roose and Lelong, 1976). In Burkina Faso, Senegal and Ivory
Coast values fluctuate between 1 and 8 mm/ka (compiled by Goudie, 1995). Another
method consists in the study of the bedload, suspended, and dissolved load discharges
from small basins. In the study of three experimental basins in New England in western
Australia, it was found that the dissolved load is ten times greater than the suspended load,
and the bedload is virtually absent (Zakaria, 1977; in Ollier (1984), p. 209). Obtained
denudation values are 14.4 mm/ka for basalts and 4.5 mm/ka for granites (Ollier, 1984).
Data from Rhodesia indicate a rate of surface lowering in granites of 15 mm/ka in rainy
zones and 6 mm/ka in drier zones (Owens and Watson, 1979). Erosion values obtained
on ultrabasic rocks of New Caledonia are about 28 mm/ka (Trescases, 1973, in Muxart
and Birot (1977)). A compilation made by Thomas (1994) on alteration velocities gave an
estimation of 2 to 50 mm/ka. Finally, weathering rates over long periods of time can be
calculated from absolute dating of geologic materials (pyroclasts, lavas, etc.).
Values obtained from experimental methods indicate that most regoliths formed during
the Quaternary because average weathering rate values are about 20 m per million years.
Birot (1978) gave values for the alteration of 1 m granite: 50,000 years in a rainforest, and
70,000 years in a savanna. All of them are similar to that proposed by Thomas (1994).
When considering these long time periods, it must be kept in mind that the climatic
changes can have occurred during the profile development. In dry periods the weathering
rate slows, whereas denudation increases.
3.8. Laterites
Several denotations exist in the literature for the term "laterite," such as "duricrust,"
"plintite," "ferric laterite," "ferricrete," "latosol," and others (Thomas, 1974), and it is
difficult to define. In the 19th century the physical properties, strength, and colour of
laterites were already known. Buchanan (1807) described a material soft enough to be cut
into blocks with an iron instrument, although it hardened when dried. This is the reason
why it is used for making bricks, from which the Latin term laterite is derived. In the
transition to the 20th century laterite begun to be defined as a function of its chemical
composition, by using the content of iron, aluminium and the relationship between
SiO2, A1203, and Fe203. Although numerous laterite definitions have been proposed
(for example, Alexandre and Cady, 1962; Sivarajasingham et al., 1962; Norton, 1973;
Schellmann, 1981), little difference exists among them. A laterite is considered to be
formed in environments with intense chemical weathering, lacking alkali and alkaline
earth elements, enriched in iron and aluminium, and with important quantities of quartz
and kaolinite. It hardens by wetting and drying.
A controversy exists between the pedological and the geomorphological concepts of
laterite. Pedologists understand a lateritic soil as a residual one, iron-enriched in its upper
part under conditions of wetting and drying, whereas for geomorphologists laterites
generate under special geomorphic conditions and the precipitates are produced in the
water table fluctuation zone. For this reason, pedologists consider this type of laterite as a
podsolic tropical soil. All this has led to a differentiation between pedogenetic laterites
and the ones generated by a fluctuating water table (McFarlane, 1976). This controversy
transforms into a difficult agreement when trying to explain soil thickness of 60 m or more
(Hays, 1967), in which case it is hard to argue for a pedogenetic origin. Following Goudie
(1973), this last type requires a thin horizon, although in most cases this does not occur.
3.8.1. Constitution and structure
Chemical analyses of A1203, Fe203, SIO2, and clay minerals from laterites can be
represented in triangular diagrams, which permit the making of compositional
differentiations. Some authors use Fe203, A1203, and kaolinite percentages (Bardossy
and Aleva, 1990), while Dury (1969) and Schellmann (1981) substituted SiO2 in the
diagram for kaolinite (Figure 20.20). These studies show a wide compositional variety,
02
SiO 2
Figure 20.20. Ternary diagram with different types of duricrusts as a function of their chemical
composition (after Dury (1969), in Thomas (1994), Figure 4.2).
even within a single level. Thomas (1974) proposed a terminological simplification by

using three terms: bauxite for a duricrust rich in aluminium, silcrete where the silica
content is high, and the term laterite was restricted to materials of intermediate
composition.
Because laterite is generated from a residue resulting from the chemical weathering of a
rock, we can expect to find a direct relationship to the parent rock composition. At a
regional scale, some variations in the iron content have been recorded, while at a minor
scale no clear relations exist between the laterite chemical composition and the parent
rock. In this sense, the generation of bauxite is not always favoured by rocks rich in A1203
(McFarlane, 1983).
The main neoforming minerals in laterites are iron and aluminium oxides and
hydroxides, such as hematite (Fe203), goethite (FeO.OH), gibbsite [AI(OH)3], boehmite
(A10.OH), and diaspore (A10-OH), and clay minerals such as kaolinite and halloysite
[AlaSi401o(OH)18], both bearing the same composition. All the neoformed minerals tend
towards a higher stability according to the environmental circumstances and in humid
tropical zones they modify mainly by leaching. The kaolinite can proceed from feldspar
alteration or from a neoformed montmorillonite. The goethite directly derives from biotite
and hornblende, and the hematite from goethite by dehydration. Feldspar weathering can
directly originate gibbsite, although normally this mineral proceeds from the desilicifica-
tion of kaolinite.
Laterites present different structures, frequently bearing nodules without inner
structure, and pisoliths (diameter > 2 mm) and ooliths (< 2 mm) with concentric layers.
These can appear as scattered or packed concretions, forming a cemented oolithic or
pisolithic laterite. The concentric structure forms by enrichment in the water table
fluctuating zone. The pedogenetic pisoliths are composed of loose material and present
irregular form and size, with a manganiferous external layer (McFarlane, 1976). Other
structural types in laterites are vermicular, vesicular, cellular, tubular, and other types,
whose differentiation is based on the morphology of hollows. This terminology is not
clearly defined and McFarlane (1976, 1983) only uses the term vermiform for charac-
terizing laterites with inner tubes and cavities (Figure 20.21). The diameter of the conduits
is about 2 to 3 cm and they are commonly filled with kaolin. The origin of such structures
is complex. It is ascribed to termite activity (termite armouring of Erhardt, 1951) and
to plant root penetration. McFarlane (1976) considered that they result from precipitation
and they only appear in laterites formed by water table fluctuation. This author also
indicates that pisolithic laterites are immature laterites, while the vermiform ones are
mature laterites and derive from the former. Therefore, pisolithic structures can be found
alternating with vermiform ones.
3.8.2. Factors affecting generation
Laterites and bauxites appear in a wide variety of rock types (Mabbutt, 1961b; Goudie,
1973, among others), although rock mineralogy substantially influences the composition
of lateritic rocks (Thomas, 1994). Nevertheless, there exist zones where a given lithology
favours laterite development, whereas in other sites the same lithology inhibits its
generation. This seems to indicate the existence of other more determinant factors in its
development (McFarlane, 1976).
Figure 20.21. Vermiform laterite, Eastern Bouak6, Ivory Coast.
Topography is one of the factors that substantially affect laterite development. In steep
slopes runoff is important and erodes weathering products, whereas in gentle slopes water
easily penetrates into the soil, dissolving the most soluble constituents and accumulating
the less mobile ones. As a consequence, in its generation laterite is associated with zones
of low relief. Some laterites, however, can be seen crowning mesas and platforms, which
are interpreted as produced by incision of low-relief laterites. Detrital laterites can also
develop at the foot of these high-relief areas (Figure 20.22). Some authors think that
laterites form with the progressive lowering of the land surface, whose final product is a
planated surface capped by a thick layer of residual materials. Therefore, laterite delays
the formation of this planation (McFarlane, 1976, 1983).
The climate prevailing in the humid tropics is the most adequate for laterite generation,
although they also seem to form under temperate climates (Taylor et al., 1992). Soil
temperatures, close to 27~ are the most suitable for the development of lateritization
processes (Thomas, 1994) because at these temperatures percolating water disintegrates
rocks more easily and silicon dissolution is faster. The most suitable conditions for
precipitation are those of humid and dry tropical climates, because only humid climates do
not favour the oxidation-reduction conditions needed for iron mobility and precipitation
(Nahon, 1986).
The biological factors include the vegetation influence and the activity of organisms
and organic compounds. Laterites commonly appear associated with grasslands (savanna),
although they have also been recognized in rainforest zones (McFarlane, 1976). Thomas
(1994) indicated that bauxites are related to humid conditions of rainforest vegetation,
Figure 20.22. Lateriticbreccia over weathered gneiss, Machakos, Kenya.
whereas the generation of laterites is favoured under drier conditions. The rainforest
maintains soil humidity and the vegetation mass produces an important raindrop
interception and reduces sheet wash erosion. Plants absorb silicon, which is generally
stored in the leaves. At the same time, different micro-organisms are involved in
aluminium, iron, and silicon mobility (McFarlane, 1987). Aluminium and iron can also be
leached from chelates, as organometallic compounds.
Most laterites are associated with wide planations and the time needed for the
generation of these forms covers millions of years. Ahnert (1970) estimated the time
required for their development to be about 10 million years. In addition, several studies
have been made about the parent rock thickness necessary for the generation of a given
thickness of laterite. By taking into account that granites have an average proportion of
2% of FeO + Fe203, Trendall (1962) calculated that for generating a laterite 9 m thick,
weathering would be required of 180 m of the Buganda granite (Uganda), which would
take several million years. Humid tropical zones, especially those of the savanna type
closer to desert margins, have been subjected to important climatic changes. For this
reason, it is common to find extensive lateritic mantles in semi-arid zones. Moreover,
many lateritic deposits of Late Cretaceous and Paleogene ages have been recognized in
extratropical areas. The basal deposits of the Tertiary sedimentation in the Duero Basin of
the Zamora-Salamanca region in western Spain are constituted by transported laterites,
kaolinitic sandstones, and silcretes (Martfn-Serrano, 1988). Because the upper concomi-
tant levels have a Paleocene age, it is deduced that prior to Tertiary times the Hesperian
Massif was exposed to a humid tropical climate. The ferruginous Bu~aco sandstone in
Portugal (Birot, 1949a,b) must constitute a correlative sediment to this tropical climate.
3.8.3. The laterite profile
Laterite forms part of a group of weathered materials that are organized into zones or
horizons, forming what is known as the laterite profile. An ideal profile was elaborated
by Millot (1964) (Figure 20.23), inspired by differentiations made by Walther (1915) and
reported by Thomas (1974) and Faniran and Jeje (1983) in their respective treatises.
Above the lateritic horizon there lies a brownish loose material, slightly humic and with
ferruginous concretions and fragments. The laterite, with a dark ochre colour, presents a
thickness ranging between some metres to 60 m with dominant vermiform and pisolithic
structures. Below the lateritic horizon there appears the mottled clay zone with red
ferruginous blots 1 to 9 m thick, although in some profiles it is lacking. The transition to
the laterite can be sharp or gradual (McFarlane, 1976). The lowest horizon is the pallid
zone, also named lithomarge, constituted by kaolinitic clays and quartz grains, and of
variable thickness (5 to 60 m). Its white colour is due to iron leaching and, much like the
motted clay zone, it is absent in some profiles. In general, if the pallid zone presents an
important thickness, the laterite thins and vice versa (Loughnan et al., 1962). Downwards
there is a transition zone of variable thickness, constituted by weathered material rarely
containing fresh rock blocks, which lies upon the parent rock.
This standard or ideal profile serves as a guide for recognizing the different zones that
can exist in lateritic profiles although, as mentioned above, some horizons may be
missing in a given profile. Laterites or bauxites of great thickness in some cases lie
almost directly on fresh rock, whereas thin laterites commonly rest on very thick
regoliths (McFarlane, 1983).
PROFILE TRUNCATEDAFTER EROSION LATERITIC PROFILEWITH CRUST UNDER FOREST
~ u m i c horizon_ _ _ ~ 0
:":'~.:..'.'.'.'~."i':':.:;"."~".:.:".::::.'Beige
.: gravellysoil I I
;;L.~:~,:.".,,':'.:~".';;:'~-'~.'.,;,':~.aedconcretionary i i" 1
so, i t
-5
Mottledclay ..~
Whiteclay I l- 7
,\ (lithomarge) I I"E
\\ \\ \ \ \ ,, . . -8
'\ \ 6 \ ( ~ \ \ \ . \ Gr'tw'thc~ I I~ ~
\ \ ~ . \ \ . \ \ \~,x, I i -
~--~~~~ Parentrock ! rll

BO. HO,
Figure 20.23. Lateriticprofile under shade forest, and its equivalent truncated by erosion in Sudan (after
Millot (1964)).
3.8.4. Laterite genesis
One of the first hypotheses indicates that laterite originates from a weathering profile
where the pallid zone looses iron, which is transferred to the laterite. This iron
enrichment is produced by capillarity and by seasonal fluctuations of the water table. The
capillary rising seems to be ineffective and overemphasized (Goudie, 1973). Moreover,
when the pallid zone presents a thickness of about 60 m the capillarity mechanism
results are hardly acceptable. Water table oscillations, especially in climates with humid
and dry seasons, favour the movement of dissolved iron. When the groundwater table
rises, the iron moves and precipitates in the higher parts of the profile (McFarlane,
1976, 1983).
Another explanation for the iron enrichment results as a consequence of its lateral
movement from higher parts in the slopes towards the valleys (Maignien, 1966)
(Figure 20.24). By this mechanism iron concentrates in the lower parts of slopes and in
the valley floors. If erosion processes indicated in the figure act over a considerable
time interval, a relief inversion can be produced by which laterites existing in low areas
occupy divide positions (McFarlane, 1976) (Figure 20.25).
In order to explain the strong thickness of the pallid zone, Trendall (1966) proposed
topographic surface lowering through time, by which the profile is displaced downwards.
These ideas were developed by McFarlane (1983) and by Tardy and Roquin (1992) when
interpreting thick laterites lying upon a regolith. Instead of considering the weathering
profile as a static column, they estimated that it lowers by losing solutes and fine particles,
and then horizons are continuously changing. Hence, the weathered rock that supplies
iron locates in progressively lower topographic positions, as happens with the laterite.
After studies made in the lowlands of Uganda, McFarlane (1976, 1983) proposed a set of
A Marshy area and

, formation of a cuirasse
I-e . ........
~.~rosion Lowering
-- ~ ' - . . . . . l ~ m l ~ of the base
C ..... I level of
erosion
Cuirasse
" ~ .
,,::: . . . : ~
Fe ............:......................:.........
Grotto
Figure 20.24. Lateral migration of iron oxihydroxides and laterite formation. (A) Migration of Fe in
groundwater by organic acids and accumulation in marshy valley floors. (B) and (C) Incision of drainage
with a falling water table, which leads to induration and erosion of the duricrust (Maignien (1966), in
Thomas (1974), Figure 13).
(a) INITATION OF LATERITIZATION

Decomposing
9".. ock
~176
(b) LATERITE EXPOSED BY EROSION, DESSICATED AND HARDENED
(c) VALLEY FORMS BEHIND LATERITE 'REEF'
Reversal J ~ ... ,
of~~ ",,~~proc~
d r a i ~
. . . . .
(d) FURTHER EROSION ISOLATES REEF AND UNDERCUTS CAPROCK
Figure 20.25. Reliefinversion by erosion of slope-valley laterites (Clare (1960), in McFarlane (1976),
Figure 20).
several developing stages for laterites generated by water table oscillations (Figure 20.26).
In this evolutionary sequence, a topographic surface lowering is required, with conti-
nuous segregation developed during the time intervals of water table oscillation that
give rise to a pisolithic lateritic layer. When relief smooths, these pisolithic lateritic layers
transform into massive vermiform laterite. Therefore, profile and land surface evolutions
are interdependent.
W.T Water table
V//A
OoO Pisoliths presently
l:o:l ""i
...
~~
9o~
9 9
forming
Pisoliths no longer
forming
~ Massive vermiform
laterite
~176176
9 9 oOo
(a) Oo0 ...

(b) ~176 9~ ;~
(c) - - -
(~) ~i:i 0~ ~ .-...-
(e) !i;i!:i:!iii:i
,........
(f) '"':'.'".
(g)
(h)
(i) (J) (k) (I) (m)
Figure 20.26. Evolution of laterites generated by water table fluctuations. (a) Iron segregation to form
pisoliths within the narrow range of oscillation of the groundwater table. (b) and (c) The lowering of the
groundwater table lowers the locus of pisolith formation, leaving earlier-formed pisoliths in the vadose
zone. (d), (e), (f), and (g) The pisoliths accumulate at the base of the soil. (h), (i), (j), and (k) The water
table stabilises and continued leaching reduces the vadose zone to bring the residual sheet of pisoliths into
the zone of intermittent saturation. (1) and (m) The pisolithic residuum is altered to goethite-rich massive
vermiform laterite (after McFarlane (1983)).
3.9. Bauxites
They are also denoted as aluminic laterites and alcretes, and are considered to be the
final product of deep weathering because SiO2 and Fe z+ have been intensely leached
(Thomas, 1994). They can be recognized in the geological record of numerous places on
the Earth. Conditions for their development are related to humid tropical climates, porous
parent rocks, good drainage, and enough time for generation (Ollier and Pain, 1996).
Bauxites can be generated by alteration of rocks with aluminic silicates, such as basic
igneous (western Africa) (Valet 9 1972; Boulang6 and Mill 9 1988), metamorphic
(Malaysia) (Allen, 1972), and volcanic rocks (Hawaii) (Sherman, 1952). These bauxites
can be produced by the direct weathering of alkaline rocks, without any volume loss
(Boulang~ and Mill 9 1988), or by indirect weathering, where kaolinite desilicification
gives rise to gibbsite (Schellmann, 1977). Detrital bauxites have also been recognized,
originating as breccias or nodular deposits (Allen, 1972). Where tropical weathering
affects carbonate rocks, then karstic bauxites can be produced (Bardossy, 1981). Bauxites
also develop upon sediments (Tertiary arkoses of Weipa, Queensland, Australia)
(Loughnan and Bayliss, 1961) and can be generated by sedimentation of aeolian dust
(Brimhall et al., 1988).
Many bauxites can be considered as derived from leaching of old laterites. Tardy and
Roquin (1992) indicated that some aluminic laterites present two layers with gibbsite
enrichment: one near the saprolite profile base, almost in contact with the parent rock, and
the other at the profile top, near the soil surface.
Bauxitization increases with fluvial incision. It has also been observed that bauxites
show a better development along fault lines, which facilitate leaching processes. In many
cases the bauxite outcropping on an escarpment laterally wedges towards the lateritic
mass. This indicates that the bauxite does not form a continuous layer and is genetically
related to the escarpment evolution (McFarlane, 1983).
About 10% of the bauxites belong to the karstic type (Ollier and Pain, 1996). In these
cases the bauxite fills fissures, depressions, and cavities of the karstified carbonate rock.
Initially bauxites were thought to constitute insoluble residues from the dissolution of
calcareous massifs. It is difficult, however, to explain the source of important aluminium
quantities. Following Bardossy (1981), karstic depressions produced by differential
dissolution receive aluminium-rich sediments supplied from surrounding zones. More-
over, bauxitization is favoured by the good drainage conditions of these zones, which
are normally protected against erosion. Bardossy also indicated that, while the karstic
depression progressively grows in size due to carbonate dissolution, aluminium content
increases by leaching of the renewed external supplies.
Bauxites appear in the geological record of a great many sites in the world (Valeton,
1972). In Spain, bauxites have a karstic origin and are mainly located in the Jurassic-
Lower Cretaceous limit of the South Pyrenees, in Late Triassic sediments of the central
Catalonian Coastal Ranges and in the Lower Cretaceous of the northeastern Maestrazgo
region (Combes, 1969).
Formation of karstic bauxites has progressively increased since Paleozoic times
(Bardossy, 1981), with a notable increment during the Cenozoic probably related to an
increase in carbonate rocks. Neogene karstic bauxites constitute the fourth part of the total
existing in the world.
Chapter 21
Tropical landforms
1. Introduction
Tropical zones affected by the Alpine Orogeny are constituted of high relief with great
volcanic cones like Nevado del Ruiz in Colombia, Cotopaxi in Ecuador, La Soufri~re
in Guadalupe (Figure 21.1), Pinatubo in the Philippines, Merapi in Java, and many others.
These tropical regions correspond to the mountains of Central America, the Antilles,
southeast Asia and Indonesia. In the three last areas dissolution processes affect great
extensions of carbonate rocks, giving rise to characteristic tropical karst landforms with
towers and conical hills. These tropical zones present a remarkable tectonic control
because they still undergo Quaternary crustal instability. Weathering products generated
in these areas are easily exported by fluvial action. Hence, these regions are characterized
by undergoing vertical tectonic movements that generate relief, important volcanic
activity of the explosive type, and an important erosive fluvial activity (Wirthmann, 2000).
Instead, regions on shields or cratons of Africa, Brazil-Guyana, India and Australia
are characterized by tectonic stability, at least during recent geologic times. These
circumstances have favoured the development of extensive planation surfaces mainly
upon Precambrian rocks, typical of these tropical shield regions. As a consequence, relief
is commonly smooth and weathering products undergo weak water erosion. The result
of these conditions is the development of thick weathering profiles that sometimes can
exceed 100 m. Base-level lowering produces deep incisions in the regolith, originating a
set of landforms related to differential weathering and erosion processes (Faniran and
Jeje, 1983). Finally, the rift valley system is located in eastern Africa and is associated with
important volcanism. The most outstanding examples of this igneous activity are Mount
Kenya (5194 m), and Kilimanjaro (5895 m) (Figure 21.2) in Tanzania.
2. Erosion and sedimentation in fluvial systems
Water courses in the humid tropics transport little coarse bed load. This is due to intense
alteration and that the regions were not glaciated in the recent past as were so many
of the temperate areas of the Northern Hemisphere. The regolith is constituted of sand,
silt, and clay particles and, therefore, rivers mainly evacuate materials of these sizes
and dissolved products. These sediments have little abrasive power and small erosive
capability. Little research exists on the bed load of tropical rivers and hence very few
data are available about this topic. More research work was produced for mountain
zones, where sediments are rapidly removed by major flows which are able to transport
all the solid discharge to lower parts of the basin by incising into the existing rocks.
Figure 21.2. The great Kilimanjaro strato-volcano (5895 m high) in Tanzania, rising over the plains of
the Amboseli National Park of Kenya.
Tropical landforms 535
This behaviour sharply contrasts with the morphodynamics of rivers flowing through
low relief zones, where sediment evacuation by rivers is mainly done by suspension and
dissolution (Thomas, 1994).
In the preceding chapter the main characteristics of precipitation were analysed. In this
chapter hydrological, erosional and depositional aspects of fluvial systems are discussed.
During rain the soil can reach its infiltration capacity and then overland flow occurs.
The run-off coefficient expresses the percentage of precipitation flowing on the surface,
and knowledge of it is very important for understanding the dynamics of geomorpho-
logical systems. Calculation of the run-off coefficient is usually made on plots, micro-
basins, and fluvial catchments. Results vary substantially through time due to the important
role that storm duration and intensity play on the water infiltration.
2.1. Soil-loss quantification
By using our data about soil loss and other data compiled from other authors, results
indicate they strongly differ depending on the studied zones and are controlled by several
factors (Fournier, 1960, Walling, 1984, Douglas and Spencer, 1985b). Nevertheless, in
small plots (Figure 21.3) on slopes with rainforest or savanna vegetation and surfaces
sloping from 5 to 15~ similar values of soil loss are obtained. Data obtained by Kesel
(1977a) in Guyana and by Nortcliff et al. (1990) in Ivory Coast fluctuate between 0.2 and
8.5 T/ha/yr. Where vegetation undergoes an important removal by human cutting or by
fire, these values strongly increase (Lal, 1986). This type of vegetation totally covers
Figure 21.3. Experimentalplots for studying water erosion. Katumani Experimental Station, Machakos,
Kenya.
the soil with a continuous and dense layer of fallen stems and leaves that can reach 25 cm
thick (Ruxton, 1967), preventing the effects of raindrop impact and splash. Therefore,
particles are removed by water flow under the layer of fallen leaves.
Research carried out on fluvial basins reveals that suspended load varies with rain
regime, and from one storm to another. Moreover, suspended load rapidly increases with
water discharge (Douglas, 1977). Data about suspended load obtained from experimental
plots are not easily comparable with values obtained from small basins, due to the
implication of other types of process. In Nigeria, for basins smaller than 20 klTl 2 in
rainforest zones, values obtained of suspended load fluctuate between 0.4 and
29.5 T/kmZ/yr (Ogunkoya and Jeje, 1987). In areas of high relief, such as eastern and
southern Africa, sediment loss considerably increases to values of 100 to 1000 T/kmZ/yr
(Walling, 1984). On the other hand, where basins exceed 10,000 km 2, obtained values on
transported sediments decrease substantially (Pilgrim et al., 1982). In this way, the Congo
River only records 11.3 T/kmZ/yr and the Niger River 33.1 T/km2/yr (Milliman and
Meade, 1983).
Data compiled by Douglas and Spencer (1985b) about suspended load, run-off, basin
area, and vegetation cover percentage on tropical rivers of different locations in the world
can be used for making some considerations. The Amazon River begins in the Peruvian
Andes and at the basin head it has numerous and important affluents excavating the
Andean Cordillera. These tributaries have a great erosive power and denudation velocities
are very high, whereas in the medium and low courses erosion is moderate. Recorded
values of suspended load (Figure 21.4) fluctuate between 0.3 and 115.9 m3/kmZ/yr
Figure 21.4. The confluence of Negro and Solimoes rivers, affluents of the Amazon River. The Solimoes
River transports an important suspended load, while the one of the Negro River is very small. This
confluence is called "encounter of the waters" and is located some kilometers East of Manaus, Brazil.
(expressed in volume) (Gibbs, 1967). The greatest erosion values correspond to islands
of SE Asia, Indonesia and Papua-New Guinea, where slopes and runoff are very high.
The Cilulung River of Indonesia annually transports 8000 m3/km 2, the highest value ever
recorded. In areas of low relief, represented by the cratonic regions of Brazil, Africa, India,
and Australia, recorded values are much lower, about 30 m3/km2/yr. Finally, in the lower
Amazon tributaries, where relief is very low, suspended load is remarkably lower and
ranges between 0.3 and 3.8 m3/km2/yr.
Studies on plots and small basins carried out in Malaysia indicate that dissolved load
represents 27.4 per cent of the total sediment discharge (Douglas and Spencer, 1985b). In
fluvial basins chemical denudation varies between 143 T/km2/yr for the Yamuna River,
affluent of the Ganges River, and 9 T/km2/yr for the Congo or Zaire River. The Amazon
River presents intermediate values, of 35 T/km2/yr (Sarin et al., 1989). Nevertheless,
average values of dissolved load for the affluents of the low Amazon basin are of 5 to
8 mg/1, very small if compared with the global average concentration in rivers, which is
about 120 mg/1 (Walling and Webb, 1986). The highest dissolved concentrations appear
in rivers crossing evaporitic formations, with values of up to 60,000 mg/1.
By considering the average values from all rivers of the world, the resulting relationship
between suspended and dissolved load is 3.5:1 (Meybeck, 1979). The highest values are
those of the Ganges-Brahmaputra rivers, which bears a relationship of 11:1 because of
the great supply given by the Himalayan range. On the other hand, the Congo River
basin has little altitudinal contrast and hence the relationship of this river is of 0.9:1 and
dissolved load exceeds that transported in suspension (Walling and Webb, 1986).
2.2. Slopes and stone lines
Despite the great extension occupied by slopes within these environments (Young, 1972),
few rigorous studies exist about them. This can be due to the high vegetation density,
which hampers their study, and that a great part of the observations made on them are
based on visual estimations rather than on accurate measurements of angles and forms in
the field (Chorley, 1964).
Relief in shield areas is sometimes interrupted by long and deep escarpments linked
to different factors and processes. Differential erosion is recognizable on escarpments
on sandstones, of which the most spectacular ones are those of Roraima in southeast
Venezuela, with plunging walls several hundred metres high. In Africa, some of the great
escarpments are associated with the stepping of extensive erosion surfaces of different
ages (King, 1962). Escarpments of tectonic origin, generated by endogenetic activity
during the Cenozoic period, reach their utmost expression along the rift valley system of
East Africa (Figure 21.5) with vertical flights greater than 1500 m. In general, they form
on basic volcanic rocks and have been mostly generated in recent times because they
appear relatively unaltered by erosion (Nyamweru, 1980). Within the prelittoral Tertiary
graben system of the region between Sao Paulo and Rio de Janeiro in Brazil important
escarpments can be recognized in the Da Mantiqueira and Do Mar ranges, with heights
close to 1000 m (Petri and Fflgaro, 1983).
In the cratonic areas, bared hills on crystalline rocks are plentiful, commonly forming
inselbergs. In some cases they constitute grouped hemispheric domes, named half-oranges
Figure 21.5. View of the Gregory Rift Valley bottom and its western border, seen from the Menengai
Caldera. Nakuru, Kenya.
(in Portuguese, meias laranjas), or the sugar loafs (bornhardts), that are very common in
several states of eastern Brazil. Their prevailing morphology consists of convex forms in
the upper parts of the slopes and concave profiles in the lower parts; rectilinear elements
or, in some cases, hanging faces, develop between them (Savigear, 1960; Thomas, 1965).
Where the hill is constituted by blocks the slope profile is complex and forms a stepped
microrelief (Pugh, 1966).
Slopes developed on flat zones on weathered materials mainly bear a convex profile,
although it can change to a convex-concave one by valley incision. In his study made on
the Mato Grosso of Brazil, after using the methodology proposed by Savigear (1952) and
later extended by Young (1964, 1970) differentiated among five profile types (Figure 21.6).
Profile A has a continuous convex element with low slopes, is associated with minor-
order valleys and represents 44% of all the measured profiles. Profiles B and C present
gentle convex crests interrupted by steps and incised by valleys. Profiles D and E bear
laterites on their crests and show more evident concavities. Following Swan (1972),
profiles developed on coarse-grained regolith present convex forms due to their high
porosity, while in fine-grained laterites slope morphology acquires a concave form.
The main processes that produce slope accumulations are surficial water flow and mass
movements. The vegetation and precipitation regimes considerably influence colluvium
removal, by which a high intensity rainfall can trigger failures. One of the most important
problems, also common in slopes of other morphoclimatic zones, is the absence of dating
of slope deposits. Some ages have been obtained from archaeological remains. Relative
dating has been established by studying soil sequences, in which different slope deposits
stack vertically. At the same time, these colluvial sequences can alternate with alluvial
sequences or interdigitate with them. Climatic changes tending to more humid or drier
+0.6 ~ S
2.5~ Max
S
B s
0.8 ~ X Max ~ V
, , ~ 2.7 ~
7.8 ~ Max
C s
~ ~,~3.0 o Max
4.3 ~ x
~ ~ MaXMin
2~53.5 ~ Max .
3 1 , , ~ : 0~ Mm Max
",..,.~k,~.5 ~
S
S 3.4 ~ S V
x ~3o ~ o
+2.7 ~
~ 3 . 2 ~ s
E "'"--'-4"'-- 5"2~ X
" ~ ~ ~ " ' ~ . 3 o
S Segment
Max o
Max MaximumSegment
Min MinimumSegment
X Convexelement
m 4.0 ~ S
V Concaveelement
0 . . . . . . . . . . v
0 lOOm
vert.exagg, x 3
Figure 21.6. Slope profiles in the Mato Grosso region of Brazil (after Young, 1970).
periods are recorded in slope deposits of rainforest zones (Thomas and Thorp, 1992) and
more commonly on those of savanna areas (Watson et al., 1983, among others).
Stone lines constitute a continuous or discontinuous layer of little weathered
clasts, normally made of quartz, quartzite, and laterite fragments, with a thickness
between 0.1 and 1 m. They are located within the slope deposit and occur 0.5 to 2 m
below the surface (Figure 21.7), although in SW Nigeria they can occur at 5 m
depth. Stone lines usually have a rectilinear outline, despite the fact that in some cases they
present many undulations. In general, there only exists a single stone line, although
F61ster (1969) recognized three layers in slopes of Nigeria. Stone lines are mainly
described for only humid tropical zones although they have been also cited in temperate
climates.
Several hypotheses have been proposed about their genesis. One of them considers
stone lines as a residual layer resultant of intense sheetwash erosion under a dry period
during which fine material is exported. Stone accumulation on the slope takes place at
the same time weathering deepens. Afterwards, during the transition to a humid period,
the stone line is covered by fine material (De Ploey, 1964; Vogt, 1966; De Ploey and
Poesen, 1989). It has been also proposed that the evacuation of fine and dissolved
material occurs by underground waters under the stone line (Young, 1976; McFarlane
and Pollard, 1989).
Figure 21.7. Stoneline located a meter below the surface. The weathered granitic substrate can be seen in
the foreground. Machakos, Kenya.
Other authors indicate that creep, which affects soil and quartz clasts coming from the
abundant small dikes, removes and redistributes the stones. These are concentrated at
the saprolite-soil contact, forming the stone line (Berry and Ruxton, 1959). Creep is more
active in some soil discontinuities, developed at 0.5 m depth, and can be accompanied
by small laminar slides and interstitial flow, which helps in evacuating fine material
(Moeyersons, 1989). All this produces clast removal towards the discontinuities, giving
rise to the stone line.
Another hypothesis, supported by geomorphologists and pedologists, is based on the
activity of termites, which remove the fine fractions of the soil (< 2 mm). These particles
are transported from the underground to the surface, leaving behind the coarse fragments,
which constitute the stone line (Williams, 1978; De Dapper, 1978, 1989). Several
arguments exist in favour of this interpretation: the size of the particles located over the
stone line is the same as the one of termitaria; termite galleries are usually observable
under the stone line, and the stone lines present synformal outlines under a termite tumulus
(Figure 21.8). Nevertheless, there exist regions such as Madagascar, where stone lines are
common whereas termites are rare or do no exist (Young, 1972); in other areas, rounded
stones, allocthonous fragments or prehistoric artifacts contradict this interpretation.
Thomas (1994) compiled data about the speed of generation of the upper soil parts by
termite bioturbation. It was estimated that a layer 3 to 5 m thick is formed in 6000 to
10,000 years, although this was estimated without knowing the account of surface erosion.
After studying the Northern Territory of Australia, Williams (1968) calculated a genera-
tion speed of 0.03 mm/yr, a much lower value after considering the overland flow erosion.
3m
~ active needle
~,...t.~i'...iiii.:iii:i.iT:i~~ ~ termite tumulus
2
zooe +
~ :..:.:~:i:i".::?i:;.:::i:.:::/.i..:.:i:!:.
:i.~:::i:;":::
:i:!:i:.i:i.i.i:::i:~:i!:~ii.!!i.,.~::i:~
:: ~:i:.~:":.i:9i.::::~
: sapro,~t~o~ a,e r~a,
..::....::..::..::.;:.:.:.:
.. ...;.".;,..~.~'.-'-..-..y.7.-.~:~.~=..
...............g.,..,,,.......... ........m...~,... .~......... ...: ..........,~
~,:.=.~".: ... ~. .....: ~.......... organic matter
9;-4,,~ -,-:" ,~ ' :" " .-.'",.:.-.:.-.-'.~ ~.~,, - " ; ~ . . . . . . ,.
.~s~ir|~'.;.&_
~...~.~:~
9 ' ' ". ,:,9 .!..~,:
! i : ~: . .~. : . , ;: : . ~ ~ , .. 9
stone layer
fragmented sandstone bed
Ill!',',',',',',',I',i i i i i I11i !i i',II11l--sha,e

IIIIIIIIIIIIIII
"te
I!!!!!!!!lmiiiill!miill!!!lmiiiiiii!!!llli!lli!l[iiiiiiiii
I',llllll!!!mliiill!miii',l!!!mlllml!!!lliil!lim!llimlml~ll~
l~,.lllllllllllllll!,~ll',il~.,~lllllll
0 1 I I ! I ! I i
o 1 2 3 4 5 6 7m
Figure 21.8. Termite activity under a stone layer near Lubok Perong, Perak Terap, Malaysia (after De
Dapper, 1989, in Thomas, 1994, Figure 8.33).
Stone lines interpreted as a residual layer lead to the past development of a climatic
change towards more arid conditions. Bigarella and Andrade (1965) made these
circumstances evident in Curitiba, Brazil, where they interpreted stone lines as
palaeopavements. Values about erosion and formation of the upper soil parts by
termites vary between thousands and hundreds of thousand years. Within these time
intervals important climatic changes occur, which affect the regional geomorphic
processes, fauna, and vegetation. It seems that in order to explain the genesis of stone
lines, all the above-cited processes do not act separately nor maintain the same intensity
during these long periods of time. Nevertheless, an order of actuation can be established,
which begins with an aridification that leads to the evacuation of fine material to the
valley bottoms. In the meantime, the coarser fragments form the stone line that will be
subsequently buried by fine material of mechanical or biological origin (De Dapper,
1989).
2.3. Fluvial erosive landforms
In some fluvial basin heads the vegetation cover may have diminished due to natural or
anthropic causes. These circumstances commonly trigger gullying processes on slopes,
which progress with great celerity (Morgan, 1986) (Figure 21.9). The gully head retreats
by sapping and if the material under erosion is relatively uniform the gully walls remain
vertical. Where resistant rocks constitute the wall top, the base is eroded more rapidly and
even the water, while falling in the channel head, exerts a cascade effect that emphasizes
wall undercutting. This produces falling of material from the top and the subsequent gully
head retreat. The piping process is relatively common in gully systems and actively
collaborates in their development (Jones, 1981). Collapse of subsurface tunnel roofs in
Figure 21.9. Meandering gully developed on non-consolidated sediments. Kathorin Basin, Baringo
District, Kenya.
the channel head produces gully retreat (Crouch, 1983) (Figure 21.10). Collapsing can also
occur in the gully walls and bottom, leading to its deepening and widening. Active piping
can generate gullies as well (L6ffler, 1974; Jones, 1990, among others). A sum of factors
combine to promote piping, such as soil cracking, hydraulic gradient, exchangeable
sodium percentage (ESP), presence of expansive clays, and other factors. Their respective
role has been analysed when considering piping in the previous section on arid regions.
Gully-head retreat has been studied by comparing aerial photographs of 1956 and 1972 in
Nigeria (Jones, 1990; in Thomas, 1994) and values of 20 m/yr were obtained for big
gullies. By using microtopographic profilemeter in New South Wales, Australia, Crouch
(1990) calculated a 1 m/yr retreat. Oostwoud and Bryan (1994) applied the sediment
budget concept to the study of gullies in Lameluk of the Baringo District, Kenya, and
obtained retreat values fluctuating between 0.5 and 10 m/yr.
All the gullies mentioned above correspond to cratonic areas, where relief is commonly
smooth or is associated with planation surfaces. Areas related to lithospheric plate
interaction during the Alpine Orogeny, however, present a very uneven relief. In these
areas fluvial activity shows a great erosive potential and an important transport of
materials towards lower zones. A net of deep gullies with abrupt walls (Figure 21.11)
represents this dynamics.
Rivers crossing the extensive plains of tropical cratonic areas flow upon an alteration
mantle of varying thickness. For this reason, fluvial incision makes the rock substratum
outcrop in some cases (Figure 21.12). Rivers mainly transport suspended fine particles,
which conditions channel morphology. Straight channels are commonly developed,
belonging to system 1 of Schumm (1981), with low gradient, relatively uniform width
and scarce bed load. Stream velocity and energy are low. Channels are usually narrow and
Figure 21.10. Collapsesdue to piping affecting a gully head, walls and bottom. Njemps Flats, Baringo
District, Kenya.
deep and the margins are quite stable due to their clay-silty nature. As a consequence,
these are stable channels with little lateral migration. Meandering rivers (system 3a of
Schumm (1981)) also develop, where bed load is very small in relation to the total load
and channel gradient is somewhat greater. These are sinuous channels (Figure 21.13)
with suspended load, similar width and stable margins due to their constitution, although
meanders can be cut off.
In some cases fiver longitudinal profiles show clear changes associated to a gradient
increase, which are represented by rapids on the rocky channel bed. Another type of
modification consists in a profile leap represented by vertical waterfalls, cascades, and
cataracts of diverse origin. These abrupt changes in the river longitudinal profile constitute
knickpoints, which tend to migrate upstream until their disappearance (Petts and Foster,
1985). Most rapids are structurally controlled by predominant fracture systems (Howard
and Dolan, 1981). Pits and potholes form on rapids when water flow concentrates at a
given point during a considerable time span. They are generated by water eddies and
Figure 21.12. The Mara River, a straight channel cut in the weathering mantle of Loita plains. The
amphibolite rock substratum can be seen in the foreground. Masai Mara National Reserve on the Kenya-
Tanzania frontier.
Figure 21.13. Flat plain where a meandering channel cuts in the regolith. Several tor morphologies can
also be observed. Campo de Az~ilvaro, Southern Duero Basin border, Avila, Spain.
develop considerable depths, abrupt walls, and spiral furrows. Pebbles can be found in
their bottom.
Cascades correspond to a sudden interruption of the river longitudinal profile, in
which water falls vertically (Figure 21.14), while cataracts can be considered as a group
of cascades (Figure 21.15). Undoubtedly, they are the most spectacular fluvial forms
and are the object of tourist attractions. Besides, they can supply hydroelectric energy.
Their origin is related to eustatic oscillations, tectonic uplift and different resistances to
erosion of the channel bedrock. The biggest cascade in the world is the Salto del Angel
(Angel Falls) in Venezuela, at 980 m high. The Iguaz6 cataracts, located in the intersecting
frontiers of Brazil, Paraguay, and Argentina, have an 84 m fall and cut Upper Jurassic-
Lower Cretaceous basalts of the Paranfi plateau. In the humid tropics the Victoria cataracts
of the Zambezi River, on the frontier between Zambia and Zimbabwe, are also important.
Schwarzbach (1967) elaborated a genetic classification between destructive cascades,
in which water flow erodes the valley head, and constructive cascades, in which water
precipitates CaCO3 and the cascade advances with the course of time. The cascades
and cataracts mentioned above belong to the destructive type and have originated on
horizontally arranged materials with different resistances to erosion.
Erosion on cascades of tropical rivers is much less than the ones of higher latitudes
(as, for example, the Niagara Falls), due to the small bed load of tropical rivers and, hence,
reduced abrasion power. Bed load percentage considerably increases in tropical zones
with a marked rough topography relief. Clasts are only transported small distances, how-
ever, due to the size reduction by chemical weathering and abrasion (Douglas, 1977).
Cascade and cataract evolution depends on knickpoint recession, which is a function of the
water discharge variations and of rock resistance. Sometimes cascade or cataract erosion
progresses more rapidly following weakness lines (Figure 21.15). Studies on the evolution
Figure 21.14. Thompson Cascade, developed on basalts of the Eastern Gregory Rift Valley margin,
Nyahururu, Kenya.
of the Niagara cataract show that its head progressively attained a curved morphology,
which is believed to be the most stable form (Philbrick, 1970). Escarpment retreating rates
measured on different cascades and cataracts of the world fluctuate between 0.25 and
1.64 m/yr (Douglas, 1977). Deep plunge pools develop at the cascade foot on the rock bed.
They have varying size as a function of waterfall height, solid load size and rock
resistance. The depth of these closed depressions is an inverse function of the retreat rate
of the cascade head (Philbrick, 1970). When hitting the bottom, water partly flows
retrogressively and produces an intense erosion at the wall base that can even cause
collapse.
2.4. Fluvial sedimentary landforms
In regions of low relief within the humid tropics many valley heads present flat-bottom
elongated depressions, without any evident channel, seasonally flooded and covered by
Figure 21.15. Iguazti Cataracts during low waters, on basalts of the Paran~i plateau. Note the curved
border form. Garganta del Diablo (Devil's Throat), located on the frontier limits of Brazil, Paraguay, and
Argentina.
herbaceous vegetation (M~ickel, 1974). These areas are named dambos in central Africa
and receive other names in different tropical regions of the world (Thomas and Goudie,
1985; Boast, 1990; Thomas, 1994). They are typical of savanna climates, with a
precipitation between 600 and 1500 mm, although they have also been recognized in
rainforest zones.
Dambos have concave margins of gentle slope, usually between 2 and 6 ~ (Acres et al.,
1985). The substratum weakness lines influence their plan morphology and they align
following these prevailing directions. Soils developed on dambos are hydromorphic or
gley soils (Young, 1976), with a higher organic matter content than the watershed soils,
which increases towards the dambo centre.
No virtual shrub vegetation or trees exist on dambos, due to the close proximity of
the water table to the surface. A wide variety of herbaceous vegetation is recognizable on
dambos and their percentage increases towards their centre as a consequence of the higher
persistence of the saturation zone. Vegetation changes through the year as a function of
the existing precipitation regime, which leads to water-table oscillations (Miickel, 1974).
A vegetation zoning is observed, from tree species in the surroundings, sharply limited by
herbaceous vegetation and the prevalence of grass and other hydrophile plants on sandy
soils outcropping in the central zone with a fluctuating water table (M~ickel, 1985).
Dambos mainly develop in central and southern Africa upon the African and post
African erosion surfaces (King, 1962), developed on granitic and metamorphic rocks.
Land flatness favours dambo formation and promotes the water table to flow out as springs
or remain close to the surface (Boast, 1990). Moreover, a plane relief favours water
penetration and the development of thick saprolites, which following Raunet (1985)
constitute the location of a great part of dambos. In the Taubat6 Basin of southeastern
Brazil, Coltrinari and Nogueira (1989) indicated that slopes limiting dambos are stepped
and present small rounded depressions, with or without hydrophilic vegetation. Possibly
their origin may be related to slides in the Tertiary sediments.
For some authors, the origin of dambos is related to the drainage network dynamics and
evolution, while for others it is associated with differential chemical and biochemical
weathering. The former argue for an incision phase in the valley head, followed by a filling
up with a variety of alluvium and colluvium (Figure 21.16). As fluvial activity diminishes,
slope washing supplies fine material to the valley bottom. Particle removal seems to
coincide with the rainy season, when vegetation cover is smaller (M~ickel, 1974, 1985).
Meadows (1985) ascribed valley incision and infilling phases in Malawi to climatic
changes.
Another hypothesis states that dambos formed independently of fluvial networks as
some of them are not integrated into the drainage system. It is believed that the main
mechanisms are chemical and biochemical weathering (McFarlane, 1989; in Boast, 1990).
The existence of topographical irregularities is thought to be originated by differential
weathering. These are controlled by a higher fracture density in the rock substrate, which
D' C' B' A'
., "- - ..... " __ _ 6m

,, ,, \ __~- ----_ _-ajm
U J Jc |B - 3m Clay Isopach (m) 040 100m
A' A B' B
0
2.5
:i!!~::... 83 ...: ""~"" 104~"

5.0
7.5
10m
C' C
D' D
- 14.0
[-'-"] Clay ['77"] Decomposed Shale (soft)
Sand/Silt Weathered Shale (hard fragments)
Laterite Unweathered Shale
Figure 21.16. Dambo plan morphology and cross-sections in Kankamo (Zambia). Note the buried
channel morphology and the sedimentary infill of sand, silt and clay (Clark, 1974, in Thomas, 1994,
redrawn by Clarke, 1974, in Fookes and Vaughan, 1986, Figure 11.17).
favours water penetration and, hence, rock alteration. Solutes are leached towards the low
parts in the profile and the underground flow runs towards the water courses. All this leads
to a slow land-surface lowering, which gives rise to zones with poorly contrasted highs
and lows. In this sense, a fluvial network rejuvenation affects dambos because incision
produces water-table lowering and increases leaching due to the rising hydraulic gradient
(McFarlane and Whitlow, 1990). McFarlane (1989) in Boast (1990), established an
evolutionary sequence starting from a small rounded depression that transformed into a
dambo due to surface lowering. Drainage network rejuvenation brings about a transfor-
mation of big dambos into smaller linear forms. Finally, if the water table keeps on
lowering, dambos increase their size individually.
Although alluvial fans reach their maximum expression in arid and semiarid zones,
they also appear in the humid tropics. Nevertheless, in these environments they usually
acquire a lesser development. Their study also becomes difficult due to the vegetation
cover. Fan sedimentation originates by changes in the flow hydraulic geometry. This is due
to a sudden decrease in the channel slope, which transforms the flow from a confined state
to an unconfined one. Fans present a triangular plan shape with convex transverse profiles
and concave radial ones. They can coalesce with other nearby fans.
Alluvial fans occur in some African areas, although in general they are of a small size
(Faniran and Jeje, 1983; Thomas, 1994). Instead, fans from the Pantanal region (Brazil)
occupy a considerable extension and one of them reaches up to 50,000 km 2 (Klammer,
1982a). Fans can develop at the foot of mountains, such as in the high regions of eastern
Papua-New Guinea (Blake and Paijmans, 1973), where the coarsest fan deposits become
altered very rapidly and with later erosion are transformed into fine particles. Some fans are
located at the base of plateaux, such as the ones described by Hill and Rackham (1978) in the
Jos Plateau of central Nigeria. These authors also cite fan development within the plateau,
between granitic residual forms. Fans have been also observed at the foot of labyrinthic
platforms in eastern Zambia and SW of Sierra Leone, where they form iron-cemented
breccias (Thomas, 1983). Small fans appear in the limit between residual hills and plains,
although they have a little sediment supply due to the scarce development of the regolith
upon the hills (Thomas, 1994).
One alluvial fan type cited by Tricart (1974a) is the one located at the foot of gully
zones as a consequence of anthropic tree felling, and occur in Madagascar, Brazil, and
Hong Kong. These deep gullies are called lavakas in Madagascar and vogorocas in Brazil.
In Madagascar alluvial fans are basically constituted by clay and micaceous sands coming
from the erosion of thick saprolites. In the Betsiboka basin, 60,000 km 2 in area, sediment
loss is as much as 15,106 m3/yr.
Areas with high relief, developed during the Alpine Orogeny and many of them located
on insular zones, normally present sharp topographic gradients. The large gullies, with a
radial pattern in many volcanic islands, rapidly remove great amounts of sediments that
are deposited in zones of changing slope, usually on the coastal fringe. This is the case
of the alluvial fan produced by the Rivi~re S~che on Martinique Island, Lesser Antilles
(Figure 21.17). A natural cut allows observation of a sediment thickness of about 12 m,
constituted by stratified heterometric deposits. At the base there are several car tires inside
a laminated deposit, possibly from the 1960s or 1970s. It can be deduced from all this
that erosive activity is very intense and rapid, due to the heavy precipitation and frequent
cyclones (700 mm/day in August 1970, D'Ercole, 1994) and to the presence of pyroclastic
Figure 21.17. Alluvialfan deposit including a group of tires near its base and incised afterwards. This
disposition indicates the great celerity of geomorphic processes. Rivibre Sbche, Martinica Island, Lesser
Antilles.
material with an important chemical weathering in the drainage basin. After the fan
sedimentation a deep incision took place, which permitted investigation of the deposit.
This alternation of intense aggradation and incision has occurred in only a few decades.
Forms resulting from sedimentation in flood plains of tropical rivers are similar to those
generated on other morphoclimatic zones. As they are covered by vegetation, these areas
are difficult to study and hence few data exist about them. Most solid load of tropical rivers
is made of sands, although when river slope is important the bed load is mainly formed
by gravels (Figure 21.18). Fine sediments deposited by vertical accretion can generate
very stable channels due to cohesion of silt-clay particles in the margins. Most of the time
the flood plain is not affected by fiver dynamics, but during big floods, frequent in these
areas, intense erosion and sedimentation take place.
Braided fluvial systems develop gravel and/or sand bars separated by low sinuosity
channels. As bars have low height, they are completely covered by large floods, which
modify their morphology and location. Meandering rivers (Leopold et al., 1964) develop,
Figure 21.18. Gravelbed load on a braided fiver flowing from the Iraz6 volcano slope (3432 m). Sucio
River, Braulio Carrillo National Park in Costa Rica. Photo: F. Guti6rrez.
by lateral accretion of sand and gravel, scroll bars separated by elongated muddy
depressions or swales, which fill up with well stratified fine material. Both elements
constitute point bar deposits. Bank overflow by a meandering river produces natural
levees, which constitute sediment wedges of fine sand gently dipping to the flood plain.
During an important flooding episode, water distributes following small channels that cut
previous sediments forming sandy tongues or crevasse splay deposits. In the lower parts of
the flood plain silt and clay sediments accumulate in ponds and shallow lakes, forming
backswamp deposits. Figure 21.19 shows a map with the distribution of some of these
subenvironments in the Rufiji River of Tanganyika. Regarding the extension of
backswamp zones, the most important region is the Tonle Sap or Great Lake, in Sri
Lanka, which is kept flooded by the Mekong River from May to October (Douglas, 1977).
Sedimentary processes are very complex and vary abruptly, partly due to the important
seasonal variations of these rivers. The Auranga River, in northwest India, develops
under a monsoon climate with four rainy months (June to September) and storms with an
intensity of up to 200 mm/h (Gupta and Dutt, 1989). During the dry season the river
becomes braided with low sandy bars. During floods braided bars are destroyed and the
river transforms into a meandering one with the generation of gravel and sand point bars.
2.5. Mass movements
These comprise all the material motions induced by gravity on slopes. They can be
produced under any climatic environment, though in the humid tropics and also in some
Flood basin
Levee and crevasse-splay 0 miles !0

9 . . . . . ,4, 4 , A J
Fine grained channel-fill

Bluffs
Channel sands
Figure 21.19. Map of part of the Rufiji (Tanganyika), showing the distribution of several fluvial
subenvironments (Anderson, 1961, in Allen, 1965).
periglacial zones, they constitute the dominant geomorphic process (Peltier, 1950; Wilson,
1968, 1969). In some regions mass movements are not very common, whereas in
others their great abundance makes them an important factor in landform development
(Figure 21.20). In the mountainous Taiwan Island, which experiences high intensity
tropical precipitation, 7810 slides occurred in a 104 km 2 area between 1965 and 1977
(Lee, 1981; in Thomas, 1994).
The term slide refers to a rapid mass movement separated from the underlying stable
part of the slope by a slip plane. In the high parts of the slope small and slow creep
movements can take place, affecting unconsolidated materials. Sliding surfaces serve as
indicators of mass movements. In the humid tropics, however, they are rapidly covered
by vegetation (Figure 21.21), which hinders their recognition. Many slides have a latent
or dormant character and can reactivate under adequate conditions. Human activities like
tree felling, incorrect constructions, and so forth, can trigger slides. In this sense, mass
movements constitute an important geomorphic hazard when affecting human populations
and their activities. Finally, it must be pointed out that, despite the great number of slides
produced in humid tropics, there is a lack of deep knowledge about them. The reasons for
this are the difficulty of their study and the high cost of monitoring devices.
Many factors influence slide generation: mechanical properties of rocks and regolith,
precipitation, vegetation, slope gradient, seismicity, anthropic action, and other factors
(Varnes, 1958, 1978; Zaruba and Mencl, 1969; Hutchinson, 1968, 1988; Crozier, 1986). It
is necessary to understand the role of each factor in order to know the sensitivity of a given
area to undergo slides. When several factors increase their intensity at the same time, slide
triggering is considerably favoured. This wide variety of factors responsible for the origin
of mass movements is an indicator of the high diversity of slide types.
The most important characteristics regarding the mechanical behaviour of regolith
are inherited discontinuities, texture, and clay mineral types. Obviously, the intensity of
weathering and saprolite structure also have an influence, because slopes covered by
Figure 21.20. Numerousslides and debris slide-flows on steep slopes of the Central Andean Cordillera,
affecting the railway track between San Lorenzo and Ibarra, Ecuador. Photo: F. Guti~rrez.
weathered boulders can evolve by the means of the rocks moving. Grus presents a high
porosity with low shear strength and where saturated by water it looses its cohesion. In
laterites on granitic rocks the sliding plane at depth is normally close to the contact
between fresh rock and regolith (Tricart, 1974b). Mass movement is commonly produced
through rotational slides or slumps and block glides (Durgin, 1977). These slide types
prevail on the "meias laranjas" dome landform of southern Minas Gerais, Brazil (Tricart,
1974b). The existence of relict discontinuities in the regolith, including the basal
weathering surface and old sliding planes, favours water penetration which reduces shear
strength and can trigger translational slides. Some examples have been cited in the Sierra
de la Costa, Venezuela (Garner, 1960; in Tricart, 1974b).
Kaolinites existing in the regolith have a low shrinkage limit, by which they are
not cracked during dry periods. The presence of clay minerals of the 2:1 type
(montmorillonite), however, gives rise to an important contraction with fissure generation,
which allows rapid infiltration and the change to a liquid state, favouring slide generation.
Figure 21.21. The crown of a slide covered by vegetation. Saint Denis, Martinique Island, Lesser
Antilles.
On the other hand, kaolinitic saprolites present high plastic and liquid limits, whereas in
montmorillonitic saprolites these are lower. As a consequence, when the liquid limit is
attained during heavy rainfall, numerous failures of different types are triggered, as it
happened in the Andean piedmont of Venezuela, where 899 mm was recorded at La Fria
during February and March of 1972 (Tricart, 1974b). Where slope gradient is significant,
dissection prevails and clay minerals are of the 2:1 type because a long time period is
needed for kaolinite generation. The presence of significant quantities of swelling clays
exacerbates heaving processes and reduces shear strength in the slope material. Besides,
when the dip angle of bedding in sedimentary rocks is similar to the slope angle, trans-
lational slide triggering is favoured (Figure 21.22).
In the humid tropics a major annual rainfall is produced, sometimes through rainstorms
of variable intensity. Water penetrating into the regolith moves towards a zone of lower
potential gradient, by the means of saturated or non-saturated flows. The former takes
place when all the regolith pores are filled by water, which is common in the lower parts of
the saprolite and also in the upper parts after an intense precipitation. The latter prevails
during most times and is characterized by the presence of water in the small pores with a
very slow flow (Selby, 1993). An increase of pressure in the pore water produces a change
in strength by diminishing cohesion and internal friction, which favours mass movements.
A rapid increase in the water pore pressure can produce a sudden liquefaction, especially
in fine and silty sands (Zaruba and Mencl, 1969).
High intensity precipitation is not common in the humid tropical climates (Lal et al.,
1981), but it is evident that mass movements are triggered by low frequency and intense
rainfall. Moreover, deep failures (Figure 21.23) are clearly influenced by preceeding rain
Figure 21.23. Deep slide that cut a road for a long time in Saint Denis, Martinica Island, Lesser
Antilles.
or by long-lasting precipitation (Thomas, 1994). In March 1974 it rained 742 mm in 16

days in Tubarau in southeast Brazil, of which 240 mm were recorded on the last day.
Catastrophic consequences occurred, with numerous landslides and flooding, with 25
people died in one single slide (Bigarella and Becker, 1975).
Brand et al. (1985), in their studies on the relationship between precipitation and slope
failure in Hong Kong, indicated that if rain intensity exceeds 70 mm/h important failures
are triggered, whereas if it does not reach 100 mm/day failures are very scarce. These
authors also emphasised the role played by preceeding rains, because they condition the
development of surficial or deep slope failures.
Research on the influence of vegetation on slope stability was begun in the decade of
the 1960s. Until then many authors had considered it as a factor with a low incidence, but
at present it has been demonstrated under several circumstances that vegetation has an
important influence (Greenway, 1987). Nevertheless, tree and shrub afforestation has been
used for more than a century for controlling erosion and trying to stabilise slopes. So
(1971) studied 700 failures triggered by an intense rainstorm in Hong Kong and suggested
that such mass movements are more commonly located on the more vegetated slopes.
When studying the role of vegetation on slope stability, Greenway (1987) established a
distinction between hydrological and mechanical mechanisms. Among the former, the
beneficial role of raindrop interception can be mentioned, which produces water loss by
absorption and evaporation and hence reduces infiltration. On the other hand, roots and
stems increase infiltration capacity and subsurface flow. Roots also take water from the
soil and diminish pore water pressure. An adverse factor is the loss of soil moisture, which
leads to soil cracking and to a higher infiltration capacity. Physical mechanisms include
reinforcement by roots, which increase soil shear strength. Factors that also must be taken
into account include root-type, penetration depth, lateral extension, and root strength.
Tsukamoto and Kusakabe (1984, in Greenway, 1987), in their study of the influence of
vegetation on slopes in Japan, developed a simple slope classification in four types
(Figure 21.24). Type A slopes have a thin soil thickness where roots do not penetrate into
the bedrock; the soil-rock interface can act as sliding plane. Type B slopes are similar to
type A except that roots penetrate into the bedrock. Type C slopes have a thicker regolith
and contain a transition layer in which soil density and shear strength increase with depth.
Roots penetrating this layer provide a stabilizing force to the slope. Finally, type D slopes
have a very thick regolith and roots are "floating" within the soil, with little mechanical
influence of stability. Below the root zone deep slip planes can be produced.
Roots also reduce susceptibility to erosion. Tree weight depends on the species,
diameter, height, and separation. Overload on a slope increases shear stress and pore water
pressure. In a broad sense, if the slope angle is small then overload increases stability.
Finally, forces of instability derived from wind action can be transmitted from the
vegetation to the soil. Strong winds can pull out trees, especially those with superficial
roots.
In slopes with low inclination very few failures occur. In the African humid tropics
several authors (Thomas, 1994) have indicated that slope instability appears with gradients
higher than 26 to 28 ~ although these values can substantially differ in other regions of the
world. According to different authors, most slides develop between 28 and 45 ~. When
slopes are very steep failures are very numerous, especially in relation to heavy rainfall
(Figure 21.20 and Figure 21.25). Slopes can modify their inclination by fluvial or marine
Tropical landfo rms 557
TYPE A '~'-~'~3 TYPE C ;,"~~

" Soil ~~ [J~'~
.~:~.~:.~:~ ,'~,~." - Bedrock ~.'..~ 9. ~
Transi
layert~ion~ ~
..~i- L ,
TY P E B ~t~ ~
...~., ,,"~
, ~.4 .It',
4~.,'"... ;- :'_~,e.'
~ ~ I~
,,~.~,~ ' ..."'
T Y P ;'~.~'~~.~. , x '.f~" x~,
Figure 21.24. Slope classification scheme based on root reinforcement and anchoring (Tsukamoto and
Kusakabe, 1984; in Greenway, 1987, Figure 6.14).
sapping and exceptionally by tectonic activity; the increase of topographic gradient

induces a change in the stress state, which can trigger failures.
Humid tropical zones located in alpine orogenic belts are exposed to sporadic
earthquakes, which affect slope stability. Earthquakes produce rock oscillations with
different frequencies and, hence, the sudden stress modification can disturb slope
equilibrium. Liquefaction can take place in silts and loose sands. In Papua-New Guinea,
Simonett (1967) studied slide frequency and pointed out that the number of slides
decreases logarithmically with distance to the epicentre. When earthquake magnitude is
important, the number and size of slides increase.
Investigations on plots and experimental basins indicate that zones bared of vegetation
are exposed to higher erosion than vegetated areas. Deforestation resulting from abusive
tree felling or man-made fires, the latter very frequent in the Amazon rainforest, influence
the rate of mass movement. This anthropic action basically affects saprolites, because in
dry periods desiccation cracks favour water infiltration during rainfall. These processes
have been interpreted as partly responsible for the triggering of numerous slides in the
widely deforested area of Serra do Mar, between Sac Paulo and Rio de Janeiro (Tricart,
1974b). Similarly, the absence of forest mass brings about the lack of root reinforcement in
the regolith, the deprivation of bedrock anchoring and the loss of water absorption. As a
consequence, slopes undergo a shear strength decrease. These circumstances favour slide
development during intense rainfall. On 23 February 1970 a 3-hour rainstorm of 100.7 mm
in the Uluguru mountains (Tanzania) led to the generation of 1000 new superficial failures
and the reactivation of other previous ones (Temple and Rapp, 1972). The work done by
So (1971), however, posed doubt on the role of vegetation as an essential factor in failure
Figure 21.25. Slide on a steep slope with rainforest (pluvisilva) vegetation. Pitones del Corbert,
Martinica Island, Lesser Antilles.
development. This author indicated that 35% of the 700 slides produced in June 1966 in
Hong Kong developed on forested areas, which occupied 8.4% of the region affected by
mass movements.
3. Landforms developed on laterites
Landforms related to laterites appear in the humid tropics and also in semiarid zones
adjacent to savanna regions. Landforms mostly result from fluvial incision of lateritic
deposits, which produce tabular forms capped by laterites. These can last for long periods
of time due to the high resistance of laterites to erosion.
Tabular forms are platforms or mesas of limited dimensions, standing out in the
landscape and with angles lower than 0.5 ~. Their borders are formed by steep scarps that
retreat basically by weathering processes, leading to a reduction of the platform extension.
Many examples exist in the literature about these tabular landscapes (Figure 21.26)
(Maignien, 1966; Thomas, 1974, 1994; McFarlane, 1976). Fluvial incision in a weathered
zone can produce a new valley. Fe 2+ flowing to the valley gives rise to an iron enrichment
that leads to the generation of a new laterite level. Successive fluvial incisions generate a
terrace or bench morphology (Figure 21.27). Destruction of the laterite scarp supplies
fragments to the lower parts, forming slopes with loose laterite rubble (Figure 21.28) and
newly cemented colluvial deposits that can connect with pediments (Figure 20.22).
In general, laterites harden as a consequence of fluvial incision. When dried they become
very resistant to external agents and after long erosional periods they occupy dominant
";"~'~; "~~
..'~:
e,L
r ,o,
~II dt dl
"it de
dt d ~
de d,,
dt
!!if:i:. ~ ";"
'd,' dt d l . . " ""
..
,I, "'"~.
. ,,."'
,d, do il ....''~l~-.. ..
~1, ,It ,Ioi
LAKE
VICTORIA
2
de
~1/ ~lt 4
0 1 2 miles
) L i
0 i 2 3 kms
ft Laterite
4500 I 9 Hill "duricrust"
..... Pediment slope o ~ o , / " ' ~ " ~ " " ~ 2 . . . . .
r ow~,,,p ~ ~9___~-~ ~ ~ ~ o~.~.~ ~'ealmem
4000 I ~ / r 7~ t:iwamp
3500 i
1000 2000 3000 4000 5000 6000 7000 ft
Figure 21.26. Landforms resulting from ferricrete duricrust dissection in Uganda. 1, flat summits capped
by laterites (possible remnants of the African Surface); 2, intermediate benches; 3, alluvial silts and
lacustrine beds; 4, papyrus swamp (after Pallister, 1956).
Figure 21.28. Slope with loose laterite rubble in a forested savanna. Eastern Bouak6, Ivory Coast.
Unconsolided Slope 5%
ground Slope 2% Erosion 70 cm
__ [ . . . . . . . . -. . . . , tL~,_,~.__ ..~_____ . . . . . . . . : ~=~
Over an0in0 cu,rasse

Grollag / --
,... t . ~,, Talus slope
~ulrasse unaer '~
dissection
i ~ eam
Subsidence
C Remains of former
Ledge Cuirasse Cuirasse
~ , " " . "......T., . . . . . . . . . ~'~7,~ . ~ "0
F " ..... Gravelly ground ---,-.~ z,,#j,~.
Figure 21.29. Some characteristics of the morphology of lateritic hillslopes (from Maignien, 1966, in
Thomas, 1974, Figure 15).
positions, whereas initially they were restricted to valleys. This variation in their relative
topographical situation is a case of relief inversion (Figure 20.25).
The process of laterite fragmentation (Figure 21.29) is mostly due to undermining by
mechanical and chemical removal produced by the underground water that flows below
the laterites. This leads to the development of a slight bending in the border of the platform
or mesa, called cambering (Moss, 1965). Subsurface water flow, favoured by laterite
fracturing and regolith porosity, can generate passages and caves of variable size (Ollier,
1965). This subsurface network can evolve producing ductile or brittle subsidence and
giving rise to closed depressions of metric to hectometric size. These pseudokarstic forms
are scattered upon the horizontal surfaces of the tabular hills and can develop diffuse or
sharp borders (De Chetelat, 1938; Goudie, 1973).
4. Tropical planation: etched plains
A great part of the tropical zone is covered by extended plains (Figure 20.2) upon which
detached hills (inselbergs) or mountain ridges stand out. There exist two main types of
plains: erosive and depositional; the former develop on very variable lithologies and
structures, whereas the latter constitute accumulation zones of alluvial and lacustrine
sediments.
Throughout the history of geomorphology different theories have been proposed,

giving rise to terms such as peneplain (Davis, 1899), for humid zones, and pediplains, by
extension of pediments, for arid zones (King, 1942; 1953). Such plains form part of the
present landscape. These terms and their associated genetic implications have been
analysed in previous chapters.
In general, peneplains do not form perfect flat surfaces, but rounded and undulating
hills of low height (Vidal Romanf and Twidale, 1998). They develop on unaltered rocks
and some think that they can be recognized in South Africa, Transvaal, Namibia, and
Brazil. As will be discussed later, they are not exclusively limited to temperate zones.
Twidale (1983) indicated that peneplains are being progressively degraded and should be
denoted as ultiplains. Changes are so slow that the survival of antique forms occurs within
the present landform (Twidale, 1976a,b). Some authors consider that "pediments, and by
extension pediplains, are not genetically nor temporaly distinct from peneplains. They
usually coexist ... (and) are simply parts of a great group of landforms called planation
surfaces" (Vidal Romanf and Twidale, 1998, p. 101).
In the humid tropics there also exist exhumed plains or surfaces, generated by the
stripping of geological units that fossilize erosion surfaces. A classical example occurs in
the Valley of the Thousand Hills, Natal, South Africa, where the erosion of the
horizontally-bedded Table Mountain sandstone uncovers the underlying surface and some
dome inselbergs (Thomas, 1978).
Where the alteration mantle is eroded by fluvial action, the weathering front
outcrops as an etchplain surface (Wayland, 1933), referred to by other researchers as a
chemical corrosion surface (Vidal Romanf and Twidale, 1998). Numerous denomina-
tions have been proposed for these forms, which were compiled by Thomas (1989a).
Wayland estimated that the regolith thickness fluctuated by approximately 3 m and,
hence, during the next erosional cycle triggered by a rapid uplift, the regolith was
completely stripped giving rise to a new surface more or less parallel to the former.
These alteration and subsequent erosion processes can take place many times.
Wayland's ideas mainly differ in the weathering mantle thickness, which at present is
estimated in tens or hundreds of metres.
Btidel's (1957) concept of double planation surfaces (run-off surface and basal
weathering surface) is related to the etching idea. He applied his theories to zones with a
tropical seasonal climate and tectonic stability (Figure 21.30). His model starts from a
plain whose weathering depth fluctuates between 30 and 60 m that is eroded giving rise to
lower plains with a scatter of small rock domes (ruwares = whalebacks = knobs) 1 to 3 m
high or shield inselbergs. Rounded or parallepipedal blocks from the lower parts of the
weathering profile can remain detached during the surface lowering. Valleys with gentle
slopes or dambos drain these plains. In summary, it is deduced that soil erosion operates on
the run-off surface, whereas the development and deepening of the saprolite takes place in
the basal weathering surface. Under seasonal climates, where Btidel (1957, 1965, 1977,
1982) focused his research, these plains have a nearly null gradient (0.2 ~ in the Tamil Nadu
plains, southern India). Rainwater hardly finds routes for hierarchical drainage
organisation and hence a high water amount infiltrates into the regolith. During dry
seasons clay soils get cracked; at the same time, roots and other bioturbation conduits
constitute excellent ways for water penetration with the arrival of the rainy season. This
"cycle" favours weathering development along the whole saprolite and its basal deepening;
Figure 21.30. Deep weathering and denudation in humid tropics (simplified from Btidel (1957)).
similarly, surface run-off associated with rainfall can export a sheet of sediments, which
lowers the run-off surface. Btidel' s ideas were soon assimilated by scientists and applied to
all the cratonic humid tropics of the world (see Thomas, 1974; 1989a,b; 1994), and even to
extra-tropical zones with thick alteration profiles in temperate and cold regions (Bakker
and Levelt, 1964; Bakker, 1967; Gellert, 1970; Dury, 1971; Biidel, 1977, 1982; S6derman,
1985; Hall, 1986, 1988). The latter indicate the survival of etchplains whose age for
Europe is considered as Upper Mesozoic-Paleocene (Thomas, 1994). Nevertheless, in
post-Miocene times a sandy alteration profile was generated under temperate climates
(Bakker and Levelt, 1964; Hall et al., 1989), which can induce significant confusions to the
interpretations.
Oilier (1965, 1984) has pointed out that saprolite has very variable depths and
non-parallel mantles as suggested by Wayland and Btidel. The regolith forms deep
depressions separated by bulges constituted by unaltered rocks. The erosion of the saprolite
may produce low zones surrounded by fresh rock highs. Saprolite plays an important role in
relief configuration due to the slow and continuous lowering of the etchplain, while
surrounding hills become relatively higher (Bremer, 1971, 1985), as a consequence of the
progressive removal of weathered rock. This process of depression genesis is related to
densely fractured rocks, which facilitate water penetration through discontinuity planes.
Btidel (1977, 1982) referred to this process as "divergent weathering."
Thomas (1965, 1974, 1994) proposed a classification of plains and etchplains as a

function of denudation rate of the lateritic profiles. In his last version he introduced some
modifications, although this differentiation has not received great acceptance. Thomas
(1989, 1994) and Pavich (1985) compared the basal surface lowering and the differential
erosion of the regolith with the dynamic equilibrium concept of Hack (1960, 1975), by
which there is presumed to exist a balance between regolith erosion and generation. On the
other hand, other authors (e.g., Ollier, 1991) estimated the landform genesis as a group of
events developed through time. Ollier (1981) considered the Uganda territory as formed
by a single etchplain (Figure 21.31). On the Buganda surface, which is equivalent to the
Gondwana surface, thick alteration profiles have been preserved. In the north, the regolith
Figure 21.31. Strippingof regolith in Uganda. The Buganda surface has erosion-surface remnants and
deep weathering. The African surface is partially stripped. The Acholi surface is stripped of old regolith
and modem soils are forming on fresh rock rather than saprolite (after Ollier, 1981).
has been completely stripped and the Archoli surface developed upon fresh rock, with
numerous inselbergs along its rim. On the African surface, the most extensive one, a deep
saprolite is eroded and several inselbergs outcrop, indicating a very irregular basal
weathering surface. Several Miocene volcanoes are located upon some points of this
surface. In the opinion of Oilier, this last surface was formed, at least, in the Lower
Tertiary period while the Buganda Surface must be of a Mesozoic age or older.
5. Inselbergs
This is a German word that means island mountain and was first used by Bornhardt (1900)
to describe abrupt rocky hills rearing from the tropical plains of Tanganyika (today
Tanzania). They basically develop in cratonic tropical and subtropical regions, especially
on granites and metamorphic rocks. Nevertheless, they can also occur on sedimentary
rocks like the arkoses of the famous Ayers Rock (Uluru) monolith or the Olgas
conglomerates, both in the central desert of Australia (Twidale, 1978, 1982a; Vidal
Romanf and Twidale, 1998). The study of these forms is complex because it comprises of
factors like climate, lithology and rock structure, which at the same time determine the
degree of rock weathering. Similarly, they raise basic genetic questions, such as those
inherent to the exhumation of the basal weathering surface or to models of slope retreat.
Research on inselbergs must also include the effects of climatic changes and the
persistence or duration of these forms. As a consequence, many authors have focused their
studies on diverse aspects of inselbergs.
Inselbergs have raised numerous problems related to different definitions and
terminology. Twidale (1968a,b) defined inselbergs as isolated residual hills that stand
out in the tropical plains. Later, Twidale (1982a) and Vidal Romanf and Twidale (1998)
considered that inselbergs can occupy any topographic position. When studying the
morros and sugar loafs of the Rio de Janeiro zone, Birot (1958) already indicated that
crystalline domes are located in variable positions on the landscape. Thomas (1974, 1978)
agreed with the idea of not necessarily relating inselberg and plain. If this opinion is agreed
with, the initial definition of inselberg becomes rather distorted. Twidale (1968) also
indicated that inselbergs are located on tropical plains, although the existence of inselbergs
on the European temperate climates was known long ago (Penck, 1924). Therefore, either
inselbergs form under different climates, or they are paleoclimatic indicators. In order to
accurately define the situation of inselbergs in the world, Kesel (1973) analysed existing
publications and calculated the percentages of inselberg distribution under different
climates: 40% in savanna climates; 32% in arid and semiarid zones; 12% in humid
continental and subarctic climates and 6% for each of the humid tropics, subtropics and
Mediterranean zones. This author concluded, as did King (1957a), that inselbergs can
occur under nearly all climates. He recognized, however, that the humid and dry tropics
are favourable regions for inselberg formation, as previously suggested by Cotton (1961,
1962). The application of numerical parameters to the identification of inselbergs, as
compiled by Thomas (1978) from several studies, does not provide determinant values
and, following this author, can be applied to numerous hills.
The definition proposed by Young (1972) has a more general character and the
inselberg denotation is applied to isolated hills with steep slopes that arise abruptly in
zones with moderate relief. This author distinguishes several types: mesas on horizontal
strata or laterites, hills with rectilinear slopes (typical of arid zones), and convex-concave
slopes covered by regolith, rock domes, and boulders (tors) or castellated forms (castle
koppies). Thomas (1976, 1978, 1994) made similar differentiations. The German authors
distinguished between inselbergs of position (fernlinge) and inselbergs of resistance
(hartlinge). The former have been preserved because of their situation on water divide
zones, where there is little erosion, whereas the latter originate because of the great
resistance to erosion of the rocks. Nevertheless, two main types of inselbergs can be
differentiated: boulder inselbergs or tors and domed inselbergs or bornhardts.
5.1. Boulder inselbergs (tors)
The morphology of boulder inselbergs is varied and controlled by fracture systems. They
develop on gneisses, schists, quartzites, diabases, sandstones, and arkoses, although they
are more abundant on granites (Gerrard, 1988). The boulders (spheroidally weathered
corestones) occur in diverse topographical positions and their size and form depend on the
fracture type and spacing. In Australia they vary in size from 0.25 to 33 m, although the
most frequent sizes are 1 to 2 m. Joint blocks commonly develop a spherical or ellipsoidal
corestone form from preferential decomposition of the comers and edges over the faces,
and in some places, present parallepipedic trends (Figure 21.32). Some blocks remain in
an unstable position and are named "perched" or "balanced blocks" (Vilaplana, 1987;
Pedraza et al., 1989; Vidal Romanf and Twidale, 1998), which frequently end up by falling
down the slope. Groups of angular blocks sometimes appear on summit zones, simulating
Figure 21.32. Tors developed on granites. Northern side of the Serra da Estr~la, Portugal.
an old fortress, and are named castellated forms (castle kopjes = koppies - rock towers).
Some castle koppies are dominated by spheroidally-weathered corestones. On the other
hand, rounded boulders can occur in any climatic environment.
The origin of the boulders supposes a deep differential weathering, which deepens with
higher intensity in zones of intense fracturing. Afterwards, an environmental change
brings about the saprolite erosion. As a result, the weathering profile base outcrops,
characterized by the presence of boulders that originate the above-mentioned forms
(Figure 21.33) (Linton, 1955). This author referred to this sequence of processes as a two-
stage hypothesis. Boulder generation can be observed on roadcuts and lineal works. The
balance between weathering and denudation must be kept in mind. If weathering is more
intense than erosion, then neither the weathering front nor the associated blocks outcrop.
On the other hand, when erosion is more intense than weathering, then the profile is
dismantled and basal blocks appear (Vidal Romanf and Twidale, 1998). Outcropping
boulders are exposed to subaerial weathering. The regolith can be removed by fluvial
action and give rise to block channels more than 10 km long, such as the Homen River of
north Portugal. Saprolite exportation can also take place by marine, glacial and eolian
erosion (Vidal Romanf and Twidale, 1998).
In general, the development of many boulder inselbergs is related to the decay of blocks
located within the inselberg. The main processes are block-weathering, grus erosion and
land surface lowering, which can lead to the disappearance of the boulder inselberg
(Figure 21.34) (Thomas, 1978).
As indicated above, boulders have a wide distribution all over the world and, because
deep weathering is required, in many zones they are assumed to be a very old alteration
(a) A 9 I'
__ A
(b) ^
(c) A ~176
........ 9 , . . . . . . . . . . . . . . . . ~ . . . . . . . . . . . ~ . . . . . . . . . . . . . . A
~ 1 7 6 1 7 . 6 . 1. 7. 6. . . .
9 ,,
Major Joints Growan Corestones
Figure 21.33. The evolution of the tors and granite landscapes based on the drawings of Linton (1955).
Figure 21.34. A sequence of development for boulder inselbergs, involving decay of interior joint blocks
within the zone of circulating vadose water, together with lateral weathering penetration leading to slope
retreat. The first stage (A) may be reached without recourse to deep preweathering, but this is not precluded.
The stripping involved in the transition A-B can be related to a climatic change (after Thomas, 1978).
of the bedrock. Linton (1955) suggested an intense chemical weathering during Tertiary
times for the Dartmoor granite of southern England. In this case regolith removal took
place by periglacial processes, in the same sense as proposed by Demek (1964) in the
Bohemian Massif (Czechoslovakia). Later on, Czudek and Demek (1971) interpreted
tors as originated by weathering and removal under periglacial envronments.

Mineralogical analyses of the Dartmoor grus make evident the absence of alteration
of feldspars and a low clay content (Doornkamp, 1974). Consequently, in European
latitudes a "sandy weathering product" can form (Bakker, 1967), which can involve
producing blocks.
King (1948) proposed that tors, the same as dome inselbergs, originate by slope retreat
in arid and semiarid climates, giving rise to a pediment at their feet. Similar ideas,
although developed for periglacial environments, were elaborated by Czudek (1964)
(Figure 8.11).
From the above discussion, there is no doubt that there exist several hypothesis for
explaining the origin of such boulders. Two main tendencies can be deduced: one related
to deep weathering and later erosion (the two-stage hypothesis of Linton) and the other
related to slope retreat in warm and cold deserts.
5.2. Domed inselbergs (bornhardts)
As previously noted, the term bornhardt is due to Willis (1936), for defining crystalline
domes in the honour of the German explorer Bornhardt (1900), who first described
inselbergs. They have a dome form and develop on resistant massive rocks, mainly
granites, although also on sandstones and conglomerates in central Australia. Twidale
(1982a) considered the Mallos de Riglos conglomeratic monoliths in the Spanish
Southern External Pyrenean ranges (Figure 21.35), to be bornhardts. This assumption
can be extrapolated to the conglomeratic mountains of Montserrat, Montsant,
Figure 21.35. Monoliths developed on massive horizontal conglomerates in the southern limit of the
external Pyrenean ranges, in contact with the Ebro Basin. Mallos de Riglos, Huesca, Spain.
Iregua River, and elsewhere, which form part of the marginal Tertiary units of the Ebro
Depression in Spain.
Bornhardt slopes are commonly steep and with a very variable size, from some metres
to several hundreds of metres (Figure 21.36). Their plan form presents circular or elliptical
morphologies, commonly affected by fracturing (King, 1948; Twidale and Bourne, 1978).
Where elongated and with low height, they are called "whalebacks" or "ruwares." Some
bornhardts appear isolated, while others cluster to form hemispheric groups, known in
Brazil as half-oranges (Birot, 1958). Contacts between the rock monolith and the
bornhardt foot can be sharp, with a clear knickpoint, or transition with no evident slope
modification. Several causes exist for explaining a sudden slope change: contact between
rocks of different mechanical properties, contrast between fresh rock and saprolite (Ollier,
1960; Thomas, 1965) and intense weathering at the slope base (Ruxton, 1958; Twidale,
1962; Mabbut, 1966; Twidale and Bourne, 1978).
Bornhardts can occur in different climatic environments (Kesel, 1973), from equatorial
rainforest zones, or intermediate latitudes (Figure 21.37 and Figure 21.38), to the north of
Norway (Btidel, 1978). For the Hercynian Massif of the Iberian Peninsula, numerous
works exist about granite bornhardts (Birot, 1949b; Birot and Sol& 1954a,b; Brum
Ferreira, 1978; Coud4-Gaussen, 1981; Sanz, 1988; Pedraza et al., 1989; Vidal Romanf and
Twidale, 1998, among others). They have also been recognized in granites of the Catalan
coastal ranges (Roqu4 and Pallf, 1998).
A set of factors exists that influences the location of domed inselbergs. Most studies
focused on checking bornhardt location in relation to rock type have concluded the
absence of lithological variations in their piedmont. Some studies, however, highlight a
Figure 21.36. Domedinselberg several hundreds of metres high. Abuja, central Nigeria. Photo: J. Ldpez
Martfnez.
Figure 21.38. Bornhardt with summit exfoliation joints. Yosemite National Park, California. Photo:
C. Sancho.
relation between inselberg situation and rock type (Shroder, 1973). Jeje (1973) indicated
that in zones of southeastern Nigeria, inselbergs are located on biotitic porphyritic granites
(Figure 21.39). In his research near Johannesburg, Brooks (1978) pointed out that tors and
castle koppies appear in a transitional zone between gneisses and migmatites and
homogeneous porphyritic granodiorites, whereas in other zones they prevail on biotitic
porphyritic adamellites. In summary, this author indicated that inselbergs are more
common on rocks that have suffered potassic metasomatism or on potassium-rich plutons,
or on those containing primary low-temperature quartz. Such relations, however, do not
occur between inselberg location and lithology.
Most inselbergs are mainly located in areas of crustal stability (Kesel, 1973), although
they can also appear on tectonically active ranges. One of the most classic examples is the
Pic Para, in southeastern Brazil, which forms part of a tectonic horst (Birot, 1958).
Twidale (1973, 1982b) and Vidal Romani and Twidale (1998) indicated that sheeting
produces lateral compressive stresses that exceed those generated by the loss of loading.
The acceptance of this hypothesis implies that domes controlled by sheeting must be
generated by this mechanism. Numerous data on stress measurement in tunnels and wells
confirm this idea (Twidale, 1982b).
Figure 21.39. Residualhills in southwestern Nigeria. 1: porphyritic granite with biotite; 2: gneisses and
schists; 3: inselberg, small domes and koppies; 4: fault; 5: fracture (after Jeje, 1973).
At the end of the 19th century inselbergs were interpreted as generated by glacial,
marine and aeolian processes (Twidale, 1982; Vidal Romanf and Twidale, 1998). Later on,
fluvial erosion was considered as the main cause. At present, two main theories explain
bornhardt generation. Several authors (Cotton, 1942; Howard, 1942; Pugh, 1956) have
proposed the scarp retreat hypothesis, although it was King (1942, 1948, 1949, 1953,
1962) who led this idea with the most emphasis. More recently, King' s points of view have
been supported by Kesel (1973, 1977b) after studying deserts of the southwestern USA,
and by Selby (1977) in the Namib Desert. Criticism against King argues that none of his
work was done in humid tropics, zones with a great profusion of bornhardts.
The hypothesis deals with inselbergs of position (fernlinge) that for extended periods of
time have experienced parallel scarp retreat. This process has led to a pediment
enlargement that can finally give rise to pediplains (Figure 21.40). Inselberg development
by pediplanation involves a progressive decrease in size, until its complete disappearance
occurs. Most researchers restrict these processes to arid and semiarid environments, where
their effects are most profound (King, 1957a). For this author, weathering at the scarp foot
leads to an increase in its steepness and to slope retreat. In warm deserts like the Namib,
however, two opposite interpretations can be postulated. Selby (1977) suggested that
inselbergs originated by pediplanation, while Ollier (1978) considered they have a great
antiquity and form part of an exhumed topography. Consequently, there is a need for a
deep knowledge of past geological times, especially about climates, in order to avoid such
controversial interpretations.
The other theory about the origin of domed inselbergs is based on the exhumation of a
deeply altered rock massif. When the basal weathering surface outcrops, domical forms
appear indicating an origin under the alteration mantle. The hypothesis is quite similar to
Figure 21.40. Inselberg standing out over extensive plains. Narok Plateau (1825 m), southwestern
Kenya.
the "two-stage" one proposed by Linton for boulder inselbergs and tors. This theory has
been checked in quartz diorite pit fronts in southern Cameroon, where a 50-m dome is
surrounded by weathered material (Boy6 and Frisch, 1973). Some roadcuts also permit the
observation of these incipient domical forms (Thomas, 1978). Domes are located on the
less fractured parts of the weathering front. Stripping is progressively carried out by fluvial
action and domical forms slowly appear on the land surface as small cupolas,
which announce an embryonic or incipient domed inselberg (Thomas, 1978; Twidale,
1982) (Figure 21.41).
In general, small bornhardts are related to a single episode of deep regolith excavation
that can reach 100 m. When trying to explain great dome inselbergs, however, difficulties
arise (Thomas, 1965; 1974). Nevertheless, if a continuous weathering at the slope foot is
admitted, then stripping produces a surface lowering of the surrounding plains and
bornhardts several hundred metres high can be satisfactory explained (Figure 21.42).
Consequently, it is a matter of repeated alternations between weathering and denudation
phases throughout a long period of geological time (Willis, 1936; Btidel, 1957; Twidale,
1964, 1982; Thomas, 1978, 1994). Thus, many great domed inselbergs are geologically
very old.
The scarp retreat theory does not explain some forms related to the contact rock-plain
and rock-saprolite. Instead, the exhumation hypothesis justifies the existence of embryonic
domes and microforms generated at the weathering front (polygonal weathering, tafoni,
gnammas, pseudokarren, caves, and so forth) (Twidale, 1982a; Thomas, 1994; Vidal
Romanf and Twidale, 1998). It also explains the genesis of stepped slopes (Twidale and
Bourne, 1975, 1978) and mushroom profiles (Twidale, 1962), which together with tafoni
lines at different heights in Ayers Rock (Uluru) (Twidale, 1978), indicate an older
Figure 21.41. Embryonicdome forms on a savanna region. Masai Mara, Kenya.

Figure 21.42. Domed inselberg development by continuous weathering of mantled rock slopes during
dissection (after Thomas, 1978).
subsurface weathering. These forms also reveal the amplitude of lateral corrosion on
slopes of massive bornhardts during different stages of base-level standstill.
Regardless of whether bornhardts may have originated by scarp retreat or by
exhumation, they require a long time to be formed. Their age, however, is a function of
the velocity of geomorphic processes involved in their genesis and of the size of the
form. Hence, we can find dome inselbergs recently exposed as well as bornhardts tens
of millions of years old. Another problem arises when trying to explain the survival of
great bornhardts for long periods of geological time. It has been interpreted that, since
the inselberg domical form is controlled by sheeting, precipitation water flows as if on
an umbrella, producing very weak alteration and low erosion (Thomas, 1974).
Mesozoic or Early Tertiary ages have been proposed for many dome inselbergs.
Savigear (1960) considered that inselbergs in Nigeria were exhumed from a Cretaceous
sedimentary cover. Barbier (1967) indicated that domes (sugarloafs) in the Tassili
Massif, South Algeria, have also been exhumed from below an Upper Paleozoic
sandstone. Chaput (1971) pointed out that the pre-Triassic surface in eastern Sierra
Morena in Spain is partially exhumed, with the outcrop of quartzitic inselbergs.
Twidale and Bourne (1975) and Twidale (1978) used stratigraphical and geomorpho-
logical arguments to suggest that the famous Ayers Rock (Uluru) monolith dates back
to the Mesozoic age.
As the age of inselbergs is sometimes very old, these forms have been exposed to
endogenic and exogenic activity during long periods of time. They may have been
affected by the motion of lithospheric plates, with the modification of their latitude and
even climate, as for example, in India. As far as the origin of inselbergs is concerned,
climatic changes affect the balance between weathering and erosion velocities. Periods
during which weathering prevails correspond to geomorphological and ecological
stability (biostatic stages) and when this equilibrium is interrupted then erosion sets in
(rhexistatic stages) (Erhart, 1967). In the European middle latitudes a wide and deep
ferralitic weathering has been recognized, especially during Eocene times (Millot,
1964). This alteration is confirmed in the sedimentary record of Tertiary basins and
gives convincing arguments for an old deep weathering of zones where inselbergs have
been formed.
6. Tropical karst
Landforms developed on limestones present clear analogies with forms developed on

granites in humid tropical climates (Twidale, 1990). The activity of the etching processes
can be also recognized in carbonate rocks and is extensive in karstic corrosion plains and
their tower forms. There is even a clear similarity between the half-orange morphologies
on granite landforms and the domical forms of the tropical cone karst. These
circumstances induce one to consider a matter of form convergence or equifinality
(Thomas, 1994).
Lehmann (1936) was probably the pioneer in studying the climatic differentiation of
karstic landforms. In the Karst Symposium held in Frankfurt, Germany, in 1953, Lehmann
indicated that every climate has its own karstic development, although factors like
composition, thickness and carbonate rock fracturing must be also taken into account.
During the symposium a karst climatic zonality was recognized, including the
differentiation of karst landforms of humid tropical zones, among others. These forms
mainly develop on tropical zones with carbonate outcrops associated with alpine ranges;
the Greater Antilles, some regions surrounding the Caribbean Sea, southeast Asia,
Indonesia, New Guinea and several African zones.
Some significant differences exist between tropical karst and other karst types
developed in other climates. Firstly, carbonate dissolution is more rapid due to the higher
content of biogenic CO2 (Birot, 1954, 1966; Ford and Williams, 1989). Secondly, strong
precipitation produces intense slope gullying, giving rise to irregular closed depressions.
Finally, high evaporation rates induce calcium carbonate precipitation forming very hard
caprocks on the steep slopes, which often is considered as the main cause of their great
inclination (Sweeting, 1972).
Tropical karst landforms are usually divided into two types: kegelkarst (cone karst) and
turmkarst (tower karst). Differentiation is made as a function of hill morphology. In the
future it is possible that the classification be based, more logically, on the morphology of
closed depressions (Sweeting, 1972). Balazs (1973) proposed a tropical karst differen-
tiation mainly based on morphometric considerations and distinguished three types: doline
karst, cone karst and tower karst. Cone karst hills are 30 to 120 m high and have a density
of 15 to 30 hills/km 2, whereas in tower karst heights fluctuate between 100 and 300 m and
their density is about 5 to 10 hills/km 2.
Many factors influence the development of these forms. Limestone types seem to play
an important role. Jamaican cone karst hills are located on very fissured limestones
(Sweeting, 1958; 1972), whereas karst towers in Puerto Rico (Monroe, 1964) and those of
Sarawak (Malaysia) (Wilford and Wall, 1965) develop on crystalline limestones with soft
interbedded levels. Bedding strike and dip affect form development in such a way that hills
and depressions opened at their toes (dolines, uvalas, poljes, and valleys) elongate
following those directions. In the same sense, splitting planes affect landform
development.
No clear relationship exists between total precipitation and type of tropical karst. In his
studies on New Guinea, Verstappen (1964) indicated that a precipitation of at least
1500 mm is needed for the cone karst to develop. In China, the cone karst region of
Schuicheng has an average annual precipitation of 1230 mm and the tower karst zone of
Guilin, the most spectacular one in China, has an annual rainfall of 1870 mm (Sweeting,
1995). In order to make comparisons, seasonality, intensity of precipitation, and
temperature also need to be known, because altitudinal variations clearly influence cone
karst development.
Water infiltration is favored by the open-joint network, which controls surface
drainage. Despite allogenic rivers, which usually have large discharges, the drainage
pattern is constituted by dry valleys, blind valleys ending in a ponor, and steephead valleys
that begin in a spring (cul-de-sac). Superficial associated forms are bare and covered
karren, dolines of different types, uvalas, and poljes. The rapid and intense infiltration of
water into the karst massif gives rise to an important cave system, mainly horizontal,
embracing a long and complex history of tropical karst development (Barbary et al., 1991;
Sweeting, 1995). Caves generate more easily in the zone of fluctuating phreatic level,
characterised by a greater dissolution (Jennings, 1985).
6.1. Cone karst
This is constituted by numerous conical or hemispheric hills (Figure 21.43), separated by

closed depressions called cockpits in Jamaica (Sweeting, 1958). In Java, a hill density is of
about 30 cones/km 2 (Lehmann, 1936). Cone height fluctuates depending on the sites and
the highest ones at 100 to 130 m occur in Jamaica. Slopes present inclinations between 40
and 60 ~. Cones can be symmetrical or asymmetrical, due to control by dipping strata
dipping or to fracturing. Commonly slopes are covered by a quite resistant duricrust with
thickness of up to 5 to 10 m in Puerto Rico (Monroe, 1966). This deposit originates after
brief storms during which water enriched in CO2 dissolves calcite, which then evaporates
and CaCO3 precipitates, especially on the windward slopes. Cone summits can be affected
by an elongated depression of up to 100 m, named zanj6n in Puerto Rico. It results from
TROPICAL COCKPITS
Topographic
/ /, // ~ ..-- -- ~.~,...,.~ divide,
0 1/2
I, . . . . .
Kilometer
Figure 21.43. Conicalkarst and cockpit (from Williams, 1969).
the dissolution and enlargement of a splitting plane, for which it can be considered as a
structural karren.
Cockpits are closed depressions with a star plan-form that develops at the cone foot
and are separated from each other by hill passes. This morphology is due to the intense
tropical heavy rains that cut gullies on slopes. Morphometric studies on the cone karst of
New Guinea seem to confirm this interpretation (Williams, 1972). In some cases,
cockpits are oriented following fracture lines. Where several cockpits coalesce, a glade is
generated (Sweeting, 1958; 1972), a Jamaican name equivalent to the uvala of the
temperate climate karst. Glades can present an irregular bottom, with small depressions
inside.
The origin of cone karst is problematic and the ideas of Lehmann (1936), developed
by the Java karst, are partially admitted (Figure 21.44). This process is thought to start
with a plane surface that suffers upwarping by endogenous activity. This relief variation
produces drainage rejuvenation and strong gully incision on limestones. A later stage
consists of drainage disintegration, with the development of closed depressions and cone
generation. This hypothesis is mainly applied the last stages of evolution. Following
Sweeting (1972), it is possible that a better knowledge of these depressions might supply
new data for solving the genesis of cone karst.
--_
Figure 21.44. Evolution of conical karst in Goenoeng Sewoe, Java (Lehmann, 1936).
6.2. Tower karst
This is also called "pinnacle karst" and in the Spanish speaking Greater Antilles, mogote
or pepino karst (Figure 21.45). This karst type develops in the Caribbean zone, Indonesia,
New Guinea, and especially in southeast Asia, where it reaches an extent of 600,000 km 2
in southern China and northern Vietnam (Sweeting, 1972). Tower karst is constituted by
steep hills, isolated or grouped, that can exceed a height of 300 m. Towers are usually
surrounded by a fiver, lake or alluvial plain.
Two main tower-karst types are recognized (Sweeting, 1972). The first one is formed
by alluvial plains with underlying limestone over which carbonate towers stand out
(Figure 21.46). In the second one, towers arise on a plain over non-carbonate rocks, like
granites in central Jamaica (Sweeting, 1958; 1972) and Malaysia (Jennings, 1985). The
Guilin tower karst in China is the most spectacular one and develops on a 2600-m thick
limestone unit of middle and upper Devonian and upper Carboniferous age (Sweeting,
1995). Two varieties have been distinguished (Yuan, 1987): the fengling type is the most
extensive one and is represented by isolated towers with abrupt walls arising from plain
limestone surfaces covered by Quaternary alluvium. The fengcong type is formed by a
scatter of tower groups, which stand out upon a rocky basement. Slope form and inclination
depend on limestone type. Shear-isolated peaks prevail on crystalline limestone, whereas
domical forms are associated with dolomites. The statistical study of towers in the Guilin
area results in an average height of 74 m and a density of 1.23 to 1.59 towers/km 2. In the
fengcong type, towers are located on high zones and slope inclination fluctuates between
30 and 70 ~ depending on the rock type on which they form. Depressions of varying size
Figure 21.45. Mogotesin the karst of Sierra de los Organos, the westernmost ranges in Cuba. The plain
forms part of the Vifiales polje. Photo: J.L. Pefia.
active high-level abandoned river caves

river alluvium with gravels and guano cliff foot
cave I r \ .,~" "
I I ,~ / ". __.,'-" cave
river
river
Figure 21.46. Schematic section through tower karst of Kinta valley, Malaysia (Jennings, 1985).
appear between peaks, and some of them are several kilometres long. They present steep
slopes and are interpreted as old fluvial valleys (Sweeting, 1995).
In a broad sense, the different tower karst types are controlled by the fracturing
network, which affects these monoliths through the development of very deep corridors
and bogaz (Figure 21.47). Dolines of abrupt walls are also recognized, possibly related to
cave collapse. Precipitation of CaCO3 takes place in the tower surface and generates hard
calcretes, resistant to erosion. This deposit protects towers against karstic corrosion and
hence slope evolution is very slow, favouring the endurance of form.
Lakes, swamp zones and fluvial plains develop at the foot offengling types of towers.
The contact with the tower base is characterized by a great geomorphological activity.
Horizontal grooves can occur close to the scarp foot, hundreds of metres long, which
represent old lacustrine or fluvial levels. Overhanging slopes sometimes appear in relation
C C
Figure 21.47. Characteristics of tropical karst hills (towers) of Sierra de los Organos, Cuba. a: active
hillfoot cave; b: abandoned caves with speleothems; c: deep karstified fissures; d: vadose cave systems,
related to previous levels of plains; e: alluviated valley floor; f: suballuvial phreatic cave systems;
g: limestone knobs; h: detached limestone blocks; i: flat surface of intervening plain, frequently flooded;
k: limestone tower (Lehmann, 1955).
to grooves, which usually collapse, but the most common feature is the presence of caves
in the contact tower-plain, with increases of phreatic or vadose water that accelerates
limestone dissolution in this zone and as a consequence slopes become steeper. This zone
of basal undermining can be emphasized by fluvial erosion. If this activity decreases then
the angle of the basal slope segment declines (Gerstenhauer, 1960) and the tower degrades.
In the Guilin karst of southern China, caves develop with greater profusion in
crystalline limestones. They normally form nearly horizontal cavities with large collapse
chambers that distribute with varying numbers at diverse heights. The first three cave
levels, located at _+0; 7 to 15 and 25 to 35 m over the water level, occur all along the
Guilin region, although other cavity systems develop at 50, 90, and even 300 m. Lower
caves are recent, while the ones located at higher altitudes are older. This indicates that
towers have been growing at the same time the plain lowered, mainly by karstic corrosion.
It is believed that their evolution started in the upper Pliocene period. They present middle
Pleistocene hominoid remains in the 90 m cave (Gigantopithecus). Many data exist on
absolute dating, basically made by U/Th and C-14 techniques. Palaeomagnetic studies
indicate that sediments from one cave are at least 900,000 years of age (Sweeting, 1995).
Plains originate by limestone dissolution planation under alluvial sediments, ponds, or
lacustrine zones. Planation surfaces are controlled by phreatic level oscillations (Ford and
Williams, 1989). They are equivalent to the marginal or border plains of polje bottoms in
temperate regions. Following Thomas (1994), their evolution is quite similar to the one of
etchplains, where surface plain lowering is accompanied by tower growth.
Chapter 22
Applied geomorphology in tropical regions
1. Introduction
Tropical zones embrace a great variety of landforms and ecosystems that include
important resources, although they also bear numerous hazards (Reading et al., 1995).
Tropical climates control the life and economic activity of people at a higher level than in
middle latitude climates. In the last 50 years a great demographic explosion has taken
place in tropical regions. At present some 2400 million inhabitants live in these areas,
representing some 45% of the world's population. About 60% live in southern and eastern
Asia. Many of these nations are developing countries with a mainly agricultural
subsistence economy (McGregor and Nieuwolt, 1998). The need for food has led to
agricultural extension and intensification, together with an increase in deforestation
practices. All this has resulted in an increase of soil erosion (Figure 22.1), flooding and
mass movement hazards. These circumstances make it necessary to apply soil
conservation practices (Douglas and Spencer, 1982).
The thick regoliths developed in the humid tropics have been studied in detail due to
their implications for engineering activities (Lovegrove and Fookes, 1972). They have a
varying depth, structure, and mineralogical composition, and abrupt changes are common
within the regolith (Douglas, 1986). All these variations considerably complicate
engineering works.
Humid tropical areas include numerous mineral, hydrocarbon, forest, and agricultural
resources whose exploitations affect different geomorphic processes, usually by triggering
or accelerating their activity. Furthermore, many areas are located on tectonically active
zones and hence can be affected by earthquakes and tsunamis. Volcanic hazards are also
very important, mainly the explosive type, with the emission of burning clouds, ash falls,
and generation of lahars. Finally, hazards with possibly the highest occurrence are those
associated with tropical cyclones, with strong destructive winds, floods, and slides related
to the cyclonic rains, as well as onshore storm surges caused by wind and atmospheric
pressure changes. An example of many of these hazards is represented by the great human
and material losses produced by the Mitch hurricane in Nicaragua and Honduras at the end
of October and beginning of November 1998.
2. Regolith characteristics. Some implications for engineering geology
A wide variety of rocks outcrop in these areas, although granites and other chemically
resistant quartz-rich rocks are among the more abundant in outcrop. For this reason, most
researchers have focused their studies on rocks of granitic compositions. After rock
Figure 22.1. Gully head developed on a partially deforested area, weakly affecting the settlement and
crops, Muranga, Kenya.
weathering, the resultant regolith includes two main particle sizes: one is constituted by
non-altered quartz grains and the other is represented by the clay fraction resulting from
feldspar and mica weathering. The regolith presents a porosity of up to 60%, high
permeability and angles of internal friction between 20 and 30 ~ (Dearman and Shibakova,
1989). During heavy rains saturation can be attained, giving rise to catastrophic rock falls.
Towards their base, alteration profiles include blocks of non-altered rock of varying size.
These blocks pose problems in road cuts and also in slope stabilisation (Douglas, 1986).
Weathering depth is quite variable. Prevalent clay minerals in the saprolite are kaolinite
and halloysite under high precipitation conditions. Montmorillonite appears in the
transition zones to semiarid regions or in poorly drained areas. High clay contents
contribute to steep slope instability and landslide generation during heavy rain periods
(Thomas, 1986).
Chapter 20 included the analysis of the weathering profile and its differentiation into
zones (Oilier, 1984), based on grus structure and composition, and on fresh rock
morphology. The same criteria have been adapted to design a classification of alteration
profiles with an engineering purpose (Dearman et al., 1978). The Tropical Soil Committee
of the International Society of Soil Mechanics and Foundation Engineering (ISSMFE,
1985) developed a classification based on these principles, suitable enough for engineering
works and especially useful when estimating loading capacity for foundation purposes
(Dearman and Shibakova, 1989). This classification appears in Table 22.1. The six
different grades greatly resemble the zones proposed by Oilier.
Three-dimensional knowledge of regolith is of prime importance for making
engineering works. Drilling and geophysical exploration are carried out with this
Table 22.1. Classification of residual tropical soils by degrees of weathering (ISSMFE, 1985).
~~
Grade Degree of decomposition Field recognition Engineering properties
O~
VI Soil No recognisable rock texture; Unsuitable for important foundations. ~
surface layer contains humus Unstable on slopes when
and plant roots cover is destroyed
V Completely weathered Rock completely decomposed by Can be excavated by
weathering in place but hand or ripping without ~
texture still recognisable. In use of explosives. Unsuitable
types of granitic origin, for foundations of concrete
original feldspars completely dams or large structures.
decomposed May be suitable for ~"
to clay minerals. Cannot foundations of Earth dams
be recovered as cores and for fill. Unstable
by ordinary rotary drilling in high cuttings at E"
methods steep angles. Requires erosion
protection
IV Highly weathered Rock so weakened by Similar to grade V. Unlikely to
weathering that fairly large be suitable for foundations of
pieces can be broken concrete dams. Erratic presence
and crumble in the of boulders makes it an
hands. Sometimes recovered unreliable foundation stratum for
as core by careful large structures
rotary drilling. Stained by
limonite. Less than 50%
rock
(Continued)
Table 22.1. (Continued)
Grade Degree of decomposition Field recognition Engineering properties
III Moderately weathered Considerably weathered throughout. Excavated with difficulty without
Possessing some strength use of explosives. Mostly
- large pieces (e.g. NX crushed under bulldozer tracks.
drill core) cannot be Suitable for foundations of
broken by hand. Often small concrete structures and
limonite-stained. 50 to 90% rockfill dams. May be
rock suitable for semipervious fill.
Stability in cuttings depends
on structural features, especially
joint attitudes
Slightly weathered Distinctly weathered through Requires explosives for excavation.
much of the rock fabric Suitable for concrete dam
with slight limonite staining. foundations. Highly permeable
Some decomposed feldspar in through open joints. Often more
granites. Strength approaching permeable than the zones
that of fresh rock. More above or below. Questionable
than 90% rock as concrete aggregate
Fresh rock Fresh rock may have Staining indicates water percolation
some limonite-stained joints along joints; individual pieces
r,,~~
immediately beneath may be loosened by
weathered rock blasting or stress relief ~,~~
and support may be

required in tunnels and
shafts
c~
Applied geomorphology in tropical regions 587
objective (Ollier and Pain, 1996). Seismic refraction profiling gives good results for
obtaining regolith depth but only barely defines some of these zones (Kesel, 1976). The
hammer seismic technique is the most adequate for calculating regolith thicknesses less
than 30 m and has been used to distinguish regolith from colluvium in slopes, in order to
know their evolution (Mills, 1990). The electric method allows the differentiation of layers
with varying electrical resistivity and their thickness within the regolith. Both land and
airborne electromagnetic devices permit the obtaining of three-dimensional maps of the
regolith by the means of measuring the vertical conductivity distribution (Street and
Anderson, 1993). Ground-penetrating radar techniques (GPR) also allow three-dimen-
sional mapping up to depths of about 40 m (Mellet, 1990).
3. Laterites for construction
Laterite is a rock of very diverse uses and is resistant to erosion. It has been used in the
construction of temples in Cambodia, in the building of the San Marcos Spanish Castle in
Saint Augustine, Florida, and in the construction of many routes in tropical zones, such as
the resistant roads of Gabon (Persons, 1970).
As far as engineering is concerned, it is worth differentiating between laterite, which is
an irreversibly hardened rock, and lateritic soil (latosol), which softens when moistened
and hardens when dried (Persons, 1984). Lateritic soils are easily extracted with axes and
pickaxes in quarries. Bricks obtained are dried in the open air and need at least 2 years for
hardening (McFarlane, 1976). Hardened laterite bricks have been employed for wall
construction, revetments, small dams, sewers, channels, and pavements. Its nice texture
and resistance make this rock an attractive material in architecture and engineering
(Persons, 1970, 1984). Its use as a construction material has declined in most tropical
countries, however, and only persists in those where laterite is the only existing resistant
rock. Numerous remains of ancient civilisations have been preserved in which springs,
breakwaters, moats, sewers, and so forth, which were constructed of such materials, have
functioned for hundreds of years (Persons, 1970).
Laterite is also used in the construction of road and airport subgrades. In the streets and
radial roads of Lagos, Nigeria laterite is commonly used due to its resistance and proximity
to quarries. In Brazil and different countries of the Guinea Gulf the construction technique
is as follows (Figure 22.2): oblong blocks of laterite are placed on an excavated base and
the voids are filled with crushed laterite, whose sizes vary between 0.1 and 10 mm
(Persons, 1970, 1984). Afterwards, it is compacted and a 5-cm layer of crushed laterite is
added. In the humid tropics it is necessary for the road subgrade to have good drainage in
order to prevent deformation in the bearing surface. Some lateritic soils are sensitive to
moisture and lose strength with an increase in humidity. For this reason, it is necessary to
ventilate the lateritic soil for enough time before its completion.
Dams made with rock fills and constructed with ground and compacted laterites present
great resistance. This is in part because laterite particles are able to cement together and so
produce an impermeable surface. On the other hand, where no laterite is available, dams
are filled with diverse rocks that are still permeable after compaction and need a later
impermeabilisation (Persons, 1970, 1984).
2.INCH COVER OF CRUSHED

COMPACTED LATERITE
WEARING SURFACE
INTERSTICES FILLED WITH
EXCAVATED
OF 1 FOOT
DEPT~
CRUSHED LATERITE
IRREGULAR OBLONG LATERITE

COMPACTED SUBGRADE / STONES 10 INCHES HIGH BY ABOUT
6 INCHES IN WIDTH
Figure 22.2. Schemeof laterite use as road subgrade (after Persons, 1970).
4. Supergene mineral ores and placers
Many mineral deposits are a consequence of weathering processes. Metallic elements are
easily removed during alteration and can be transported to the sea or precipitate within the
weathering profile. If they concentrate they can generate a mineral ore with economic
interest. This process is known as supergene enrichmentand is of prime importance with
sulphide ores (Ollier, 1977, 1984; Ollier and Pain, 1996).
In a supergene ore several zones can be differentiated (Figure 22.3). The upper part of
the weathering profile constitutes the oxidized zone. It can be subdivided into a lower zone
with oxides that enriches towards the base, and a leached zone. In the upper part of this
leached zone residual iron minerals can concentrate to form a caprock (gossan), which
constitutes an excellent prospecting sign. Beneath the oxidized zone there appears the
sulphide enrichment zone that forms by alteration of the upper material, which precipitates
below the water table under reducing conditions. The limit between oxidation and sulphide
zones corresponds to the water level location at the time of their formation. The slow
lowering of the water table favours the development of an important thickness in the
supergene sulphide zone. Water-table lowering can be due to climatic and tectonic
changes or to drainage modifications by volcanic eruptions. Due to all these lowering
processes, the supergene sulphide zone, of variable thickness, can be located at some
metres to hundreds of metres depth, and can even reach more than 600 m depth in several
places (Ollier, 1984; Ollier and Pain, 1996).
Laterites can form exploitable ores of iron, aluminium, nickel, manganese, gold, etc.
(Figure 22.4). Within the alteration profile the metallic concentrations increase with depth
and in the upper parts aluminium oxihydroxides prevail. These metals with high ion
charge are transported within the regolith as small amorphous particles that afterwards
ground surface
" \ \ \ '
,gossan / I #' / / /
w, . ,, -, ,, \ \ \ \ \ '
,, o,.,
i'" "'.i"" , \ / \ /\ \/ \ / \ / \ / \ /' /

", , .,,*. , 9
o~ " N / ,, I ,, / , / ,, / \ / \ / \ , /
~ 1 796 o ~ 197 6 " . 9
9 / / /
leached zone \ \ \ ,
9"..'...." / I / I / / I I
9 .'. . / / / / / / / /
.'.........-
... . . . . .t, \\ \\ \\ \\ \\ \\ \\ ,"
t i i ,- ,, / / /
oxidized ore \ \ \ "
9 " " ater-table
supergene sulphide ore

- \~.\~.\ ~ .~ .~ .~
i/ \\ t.\groundwater base 4
.f / t. / ,,e
\ \ \ \ \ \ \ ~
I / / I' (, / / / / / /
NNN~,
primary ore / /
/N/N/N/,, , ,, ,, ,, \ \ ',
I / / I
I
\\\',
I I /
\_
I
IN
~'/\/\/\/\/\/"
/ / / / / /
Figure 22.3. Different zones in an ore body formed by supergene enrichment (Ollier and Pain, 1996).
dehydrate and crystallise (Ktihnel, 1987). Particle removal towards the lower profile parts
and their ulterior crystallisation were confirmed by Aleva (1991) for nickeliferous laterites
and by Nahon et al. (1985) for manganesiferous laterites.
Gold from auriferous laterites is very pure (>95% in Au) and precipitates in quartz
grain fissures and also around iron oxide nodules (Thomas, 1994). It is believed that gold is
released from the sulphide crystalline structure under very low pH conditions. It is
afterwards solubilised under the sulphate form and when it deepens into the profile pH and
Eh conditions change, giving rise to its precipitation near the phreatic level (Wilson,
1984). The activity of continental brines has also been considered, because when entering
into the profile, chlorine waters reach a high acidity. Gold solubilises in the form of
chlorides and, under reducing conditions, the increase of Fe z+ content makes gold
precipitate in the form of small crystals of great purity (Mann, 1984). Similarly, the role of
humid acids has been also considered in the removal of colloidal gold.
Many iron ores exist in relation to karstification. They are mainly constituted by
limonite, hematite, goethite, and siderite, which are an alteration residue of limestones and
dolomites. They occur forming discrete accumulations, filling fractures or karstic
depression of varying size. Contacts between carbonate rock and iron are very irregular.
The carbonate massif usually develops underground caves linked to phreatic flow. In
Spain several iron ores exist related to carbonate karst processes; Alquife in Grenada,
Sierra Menera in Teruel, Cabfirceno in Santander, and Cerro del Hierro (Iron Hill) in
Seville, all of them developed on carbonate rock of different ages. The iron ore of Sierra
Menera is of the interstratal type and karstification of Ashgillian mixed iron/magnesium
. . . . . . . . . . . . . . . . . . . . "~ "l'~f .......... ~~ . . . . . . .
J r . o
km "
Equatorial Scale o ~, 3000
OD
,. ,
Placer deposits Weathering deposits

o G o l d , silver, lead, zinc, copper 9 Lateritic bauxite
Tin C ' Karstic bauxite
v Titanium 9 Lateritic nickel
O Diamonds 9 Residual manganese
9 Kaolin
v L a t e r i t i c Iron
Figure 22.4. Major mineral weathering deposits and placers formed under humid tropical conditions
(Reading et al., 1995).
carbonates gave rise to minerals of secondary iron (Figure 22.5) (Fernfindez-Nieto et al.,
1981).
Nearly 10% of the world's aluminium production proceeds from karstic bauxites
(Caribbean region and southeast Asia). They mainly appear filling karst depressions that
have been slowly lowering at the same time that the bauxite deposit increased. They are
recognised from the Cambrian through the Quaternary, always related to warm climates
(Bardossy, 1981).
In the Malayan tower karst, staniferous alluvium penetrates the endokarst and fill voids
and fissures. It constitutes an important tin ore where processes of previous karstification
and later placer infillings are involved (Sweeting, 1972).
Placers originate by mechanical concentration due to the gravity separation of heavy
minerals (denser than quartz), by water, or by wind (Bateman, 1950). For this process to
occur, it is required that minerals be released from the altered rock and become
concentrated. Most placers are generated in fluvial and littoral environments. The most
important exploitable minerals are gold, diamond, ilmenite, rutile, cassiterite, zircon,
monazite, gemstones, and others. Although placers have a wide distribution all over the
world (Figure 22.4), humid and subhumid tropical environments present optimum
conditions for their formation (Thomas, 1994). Most placers result from the activity of
present (Figure 22.6) or recent geomorphological processes, such as diamonds in the
Central African Republic or cassiterite in Indonesia. Other deposits were generated in
former geological times, such as the Tertiary auriferous placers of Las Mrdulas in Le6n
Figure 22.6. Prospecting a gold placer by the use of trays and a suction dredge. Nzi River affluent,
Bouak~, Ivory Coast.
the same time, they can be transported to lower terrace levels. Fluvial channel beds are
affected by a continuous reworking and channel migration in such a way that coarse gravel
sedimentation favours trapping of small diamonds, such as those in Sierra Leone and
Ghana (Thomas and Thorp, 1993). In Indonesia and Malaysia cassiterite placers derive
from the stripping of weathering profiles on granites and andesites, and appear in
colluvium, alluvial fans and fillings of endokarstic cavities (Aleva, 1985; Gupta et al.,
1987). A wide variety of economic minerals can concentrate on beach placers, such as
ilmenite, zircon, and monazite. Their concentration is made by waves and littoral drift
currents. On the other hand, removal of beach sands by wind can give rise to aeolian
placers on littoral dunes.
The commercial extraction of minerals can produce important environmental
modifications, usually tolerated by the governments of developing countries. A great
part of the cassiterite exploitations of alluvial placers in SE Asia are near the coast.
Extraction is carried out by dredging and pumping fluvial channel beds. This operation
produces intense water turbulence and an important part of the suspended load is spilled
into the sea, affecting the life of coral reefs. The Madeira Basin of the SW Amazon River
represents another example of impact derived from some metallurgical treatment
practices. Dredged gold is amalgamated with mercury and some 5 to 30% mercury is
lost in the fiver. Water pollution has serious effects on the fauna existing in this fluvial
basin (Reading et al., 1995).
5. Slope failures and hazards
Mass movements constitute a hazard when threatening people's life and properties. During
the last century this hazard has increased due to population growth and to the expansion of
areas needed for obtaining essential resources, which on occasion makes people exploit
hazardous zones (Crozier, 1986). In addition, other factors have converged, such as
population migration to the great cities and the construction of numerous engineering
works that modify the landscape and introduce slope instability (Figure 22.7 and
Figure 22.8). Slope failures can produce deaths and injury, material losses and
environmental damage. In humid tropical zones failures present a very important danger
due to the existence of more or less thick saprolites and to the high precipitation. These
circumstances favour the removal of surface or deep altered layers, which takes place
especially in relation to rainstorms. For this reason, it is necessary to draw detailed
geomorphological maps and make boreholes for terrain recognition (Tricart, 1974a).
Heavy rainfall commonly gives rise to slope failures and floods. From the first to the
fourth of October 1968, several states of northwest India underwent major monsoon rains
that produced slides and flooding, burying or partially submerging the cities of Darjeeling
and Jalpaiguri. About 1000 people died and many material losses were recorded,
especially oil pipes, roads, railways, and houses (Davis, 1992). This combined action of
flooding and slope failure also took place in Honduras and Nicaragua, as a consequence of
the rain associated with the Mitch hurricane, between the end of October and beginning of
November 1998. Some 85% of the land was flooded, more than 10,000 people died, 15,000
disappeared, and enormous damage was made to the infrastructure (Figure 22.9), houses
and crops. The slide produced on the Casitas volcano was channelled by the hill slopes and
Figure 22.8. Bellefontaine slide affecting pyroclastic materials, which have fallen from the roadcut
slope, Western Martinique Island, Lesser Antilles.
Figure 22.9. Bridge fallen into the Cocos River as a consequence of floods associated with the Mitch
hurricane, during the end of October and beginning of November of 1998, North of Nicaragua, near the
Honduras frontier. Photo: A. G6mez Sal.
water addition transformed the pyroclastic deposits into a debris flow, which spread out
through the croplands existing at the volcano foot. Numerous settlements were buried,
causing around 1500 deaths. The debris flow thickness was estimated to be about 2 m
(personal communication from Prof. A. Cendrero) (Figure 22.10).
Slope failures constitute periodic and catastrophic erosive processes in some zones. The
region between Sao Paulo and Rio de Janeiro in Brazil is crossed by the abrupt ridges of
Serra do Mar and Mantiqueira, separated by the Val Paraiba valley. Catastrophic failures
have been recorded in 1942, 1956, 1966, 1967, and 1988. As a consequence of the floods
and slides of 1942, governmental measures were established in relation to construction
practices, forest conservation, and soil use. In March 1956, the city of Santos was affected
by heavy rains, with a maximum intensity of 250 mm in 10 h, that triggered numerous
rotational and translational slides (Pichler, 1958). In January 1966, Rio de Janeiro
experienced rainfall of 472 mm in 72 h and produced mass movements such as rock falls,
mud flows and rotational slides (Tricart, 1974a). The disaster produced 1000 deaths. The
most spectacular slope failure development in this region took place in the steep Serra das
Araras, which forms part of the Serra do Mar and is located 50 km West of Rio de Janeiro.
On the 22nd and 23rd of January 1967, 275 mm of precipitation occurred with an intensity
of 114 mm/h, accompanied by strong thundering, which triggered more than 10,000
failures in a 180 km 2 area. There were debris slides, debris avalanches, and mud flows that
affected water and power supplies and road traffic, cutting the Sao Paulo-Rio de Janeiro
highway. About 1700 people died and material losses were estimated at 1600 million
dollars (Jones, 1973). Finally, in February 1988, debris avalanches in Rio de Janeiro killed
200 people, and 20,000 lost their homes (Smith and de Sfinchez, 1992).
Figure 22.10. Slope failure on the Casitas volcano and subsequent debris flow, triggered by rains
associated with Hurricane Mitch, during the end of October and beginning of November of 1998. Debris
flow materials buried about 1500 people. Northeastern Le6n city, Maribios volcanic ridge, Nicaragua.
Photo: A. G6mez Sal.
In the geomorphological literature many examples exist of catastrophic events of

different magnitude produced by slides. One of the most important took place on the night
of 29th October, 1959, in the Minatitl~in city of southeast Mexico, when massive mudflows
buried numerous houses and caused the death of 800 people while sleeping. On the 13th
and 14th of November 1963, significant mud and rock movements originated in the Grand
Rivibre du Nord in Haiti in relation to intense tropical rainstorms. These slope failures
affected many villages and killed some 500 villagers and tourists (Davis, 1992).
6. Catastrophic flooding
Floods constitute the most common geomorphic hazard in humid tropics, due to the
intense and prolonged rainfall that happens in these zones. The use of floodplains for
obtaining agricultural resources leads to a considerable hazard increase. As a consequence
of flooding, in Asia about 4 million hectares of land annually receive significant damage,
affecting 17 million people (Reading et al., 1995).
Thunderstorms are the major rain producers in low-latitude tropical regions.
Figure 22.11 represents the global distribution of thunderstorms with different
frequencies: more than 100, between 50 and 100, between 20 and 50, and less than 20
storms per year (Lamb, 1972). A band of higher thunderstorm frequency occurs in areas
near to the equator.
Precipitation in the tropics is associated with the convergence of air masses between
both hemispheres (Intertropical Convergence Zone), with tropical cyclones, and with
~~ "~,.---o
..... ~
k
,, "~
J
Figure 22.11. Annual thunderstorm frequency. Regions with 100 or more thunderstorms per year are
black. The dotted line encompasses regions with more than 50 thunderstorms per year and the dashed-
dotted line encompasses regions with at least 20 thunderstorms per year (after Lamb, 1972).
orographic effects (Hayden, 1988). In the first case, air convergence and elevation produce
a cloudy strip around the equator. The seasonal variation of this cloud belt gives rise to
monsoon rains, which in Asia develop between June and September, with an average of
seven monsoon depressions per year (Gupta, 1988). Sometimes these rains can last up to 9
days. Cyclones move from east to west and maximum precipitation takes place on islands
or on continental margins, where the land surface is warmed up and topography enhances
air uplift and increases convectivity (Hayden, 1988). In Cilaos, Reunion Island, 1870 mm
rain fell during a single day (Gupta, 1988). Tropical cyclones with winds greater than
116 km/h are called "hurricanes" and in eastern Asia are known as "typhoons." The
orographic effect consists in the rise and condensation of air masses, which generate the
maximum diurnal precipitation during the period of higher sun warming.
Gauging stations in humid tropics are scarce and in general they are located on big
rivers. Consequently, records of great floods are very limited. Recurrence periods of
intense precipitation are much better known, however, and these data can be related to
great floods, by which some approximate values can be available (Gupta, 1988).
Drainage basin hydrology experiences deep modifications in urban zones, mainly due
to the impermeable surfaces of streets and roofs, which increase flood frequency (Leopold,
1968). This urban-related hydrological change affects urban areas when channel and sewer
networks are insufficient for water evacuation. Then flooding of cities occurs, as it
sometimes happens in Kuala Lumpur (Malaysia) and Singapore (Gupta, 1984; 1993).
Bangladesh, former East Pakistan, is the country that has suffered the worst
catastrophic floods in the humid tropics. It is a small and poor country, with an area of
q/~
9~w,~ 89~ 92~
. "~ r.,..
F,w%
BANGLADESH
N26 ~
0 50 100
,) I_, I
KM
I
BARIND TRACT ".'. Bahadurabad

SYLHET
: "
:::.:.:..
.:::.:
i"~ Rajshahi "":
MADHUPUR
'; ;: TR AC:T .
Hardinge
Bridge
Bhairab
Bazar
L"%, \
i ~
,v),
FARIDPUR I
23 ~ TROUGH
( Khulna
o
CALCUTTA (
o
Chittagong
"1
IIIIIIIIIIi
I EASTERN
"o :::*....
::':-::: Pleistocene Terraces
Tertiary Hills N
21
Structural Depressions
II Gauging Stations
92OE i o
I . . . . . . |
Figure 22.12. Major physiographic features and rivers of Bangladesh (Rasid and Pramanik, 1993).
144,836 k n ] 2 and more than 110 million inhabitants, of which a greater part live upon
floodplains and have a per capita income of about 140 dollars (Brammer, 1990; Khalil,
1990). The Ganges-Brahmaputra delta is located in the southern region, and it is also fed
by the Meghna River (Figure 22.12). The delta contains numerous islands where
population density is up to 1000 people/km 2, and agricultural land exploitation is very
intense (Carter, 1988). About 60% of the country presents a flooding hazard. Two great
closed and shallow structural depressions develop on those wide deltaic plains and
floodplains. Elevated zones correspond to Pleistocene fluvial terraces and Tertiary hills
(Rasid and Pramanik, 1993). Maximum annual precipitation is 11,615 mm, recorded at
Cherrapunji, to the NE of the country. Rainfall decreases from this region towards the
west, where it reaches an average value of 1250 mm (Brammer, 1990). In their low
courses, the Ganges, Brahmaputra and Megna rivers have valley widths of about 10 km
and during floods maximum discharge is 160,000 m3/s, three times the average value of
the Mississippi River. These rivers transport a sediment load of 2.5 x 109 t/year (Coleman,
1969) and have a marked seasonality. They record high levels from the beginning of spring
to the end of summer, due to the Himalayan snow melting and to the June-July monsoon
rains, which can give rise to important flooding.
The Bay of Bengal has a tropical cyclone frequency of 20% (Barry and Chorley, 1987).
When cyclones approach the coast they produce a sudden sea level rise due to strong winds
and low atmospheric pressure (storm surge), and high storm waves. These waves can reach
6 to 7 m in height and penetrate deeply into the coastal fringe, which is constituted by
deltaic zones of very low height, and hence houses and crops get flooded. Sea penetration
effects are emphasised when the cyclone passing coincides with high tide conditions, as
happened in November, 1970 (Smith and Ward, 1998). Strong destructive winds that can
exceed 200 km/h and strong cyclonic precipitation adds to the marine flooding (Reading
et al., 1995).
The Bangladesh territory has suffered nearly 20 important floods since the middle-20th
century (Brammer, 1990). Some have only produced material damage, but in most of them
the number of deaths has been very high. The biggest catastrophe happened on the 12th of
November, 1970, and it is considered as the worst natural disaster of the 20th century. It
was generated by a cyclone that arrived at the coast at 11 p.m. and produced 300,000
deaths (Burton et al., 1978; Carter, 1988). In successive years new catastrophes also took
place. In 1987 and 1988 fiver floods were caused by strong monsoonal precipitation and,
according to official data, they produced 1657 and 2379 deaths, respectively
(Brammer, 1990). On 30 April 1991, as a consequence of a new catastrophic cyclone,
70,000 people lost their lives in SE Bangladesh and millions lost their homes (Davis, 1992;
Reading et al., 1995).
Protection measures against these natural catastrophes have been quite scarce due to the
great poverty of the country. Earth walls and tree fences have been constructed, although
the former are designed for fighting against salinization. During the 1970 cyclones, some
5% of people were saved in official refuges and about 38% by climbing into the trees. This
cyclone was identified 3 days before and monitored by satellite, although no intensity was
calculated (Burton et al., 1978). Despite alarms, very few people were evacuated and most
of them remained on their lands, perhaps due to religious reasons (Carter, 1988). In the
1991 cyclone, however, meteorological predictions and alarms led to a major evacuation.
Chapter 23
Environmental change
1. Introduction
In the course of Earth's history numerous changes in the climate, sea level, vegetation
cover, animal population, soils, and landforms have taken place (Goudie, 1992), but it is
Quaternary time during which those constant environmental changes, which occasionally
take place in short spans, can be detected with more accuracy. In recent times, from the
geomorphological perspective these environmental changes are due to the three basic
causes that are geological, climatic and anthropic. The difficulty in knowing in detail the
factors that trigger these changes derives from the interactions among causes, as often they
act jointly, and it is therefore, difficult to interpret the contribution of each separate one
to the environmental change.
Geological environmental change is mainly due to neotectonic activity (Figure 23.1),
which is significant in areas located on active plate margins, where isostatic compensation
and relative displacement between plates of the lithosphere produce significant relief
modifications. For example, the deserts of mountains and depressions of the western USA
and central Asia (Mabbutt, 1977) that show large crustal instability are affected by such
activity. On the other hand, shield and platform deserts, such as the Sahara, India, and
Australia, are not significantly influenced by endogenic causes. On the latter continent
climate change can be more easily studied in the absence of tectonics.
Even though humans have dwelt on Earth for about 3 million years, their interaction
with the environment began to be significant only about 10,000 years ago with the
agricultural revolution, and especially with the more recent industrial and medical
revolutions (Goudie, 1981b). Humans can significantly affect the vegetation, fauna, soil,
water, and climate, which is why is they are considered a geomorphological agent
(Dov Nir, 1983), as well as the fact that human activity modifies the Earth's surface
(Figure 23.2) and interferes with non-human geomorphological processes.
Climate change seems familiar, because we are used to temporal fluctuations in
precipitation and temperature in our lifetime. In addition to these short-term variations,
longer and deeper fluctuations occur. These features are illustrated in the different climate
change scales shown by Goudie (1992) and Boulton (1993). In those scales climate
fluctuations are reflected from intervals of tens of years to the scale of geological time
(Figure 23.3).
Long-term climate changes (Goodess et al., 1992) become apparent in temporal scales
above 20,000 years, as with Quaternary glaciations. These changes are produced by
modifications outside of the climate system, such as changes in the geometry of the orbit of
the Earth around the sun (Milankovitch cycles). Short-term climate changes range from
100 to 20,000 years. They are related in part to external causes, but some mechanisms can
Figure 23.1. Nigtielas Fault, located in the northwestern border of the Lecrfn valley in the Betic Range,
Spain.
Figure 23.2. Open-pit mine works exploitation of Cretaceous lignite in the Val de Arifio, northern Teruel
Province, Iberian Range, Spain. Photo: B. Ler~inoz.
Environmental change 603
be relevant, namely changes in the oceanic circulation, volcanic activity, and in the gas
content of the atmosphere. On the other hand, the term "climate variability" is used for
those changes of less than 100 years, which are due basically to internal mechanisms
(Goodess et al., 1992).
In most cases, in the landscape there is a collection of landforms resulting from the past
activity of the geomorphic processes, which acted under a wide spectrum of environmental
conditions that were different from the present ones (Figure 23.4). As indicated above,
these processes, coming from tectonic, anthropic, and climatic activity can interact.
Therefore, for the study of the climate change, it is necessary to discern the contribution
of each one and try to eliminate the "noise" attached to the activity of the first two.
Undoubtedly, human action during the last millennium has been very significant in certain
areas, as in the Mediterranean countries (Goudie, 1981b, 1990a; Mannion, 1997), and
presents very complex problems for determining the real influence of climate change on
the genesis of the current landscape. Therefore, the best conditions for researching climate
change are those with a history of long crustal stability and sparse habitation. In this way,
those areas are a reflection of landform generation due exclusively to geomorphological
processes triggered by climate changes.
It is necessary to know the age of the forms related to climate change. Early researchers
used relative dating in relation to stratigraphic series, morph genetic sequences, degree of
weathering, and soil development (Birkeland, 1984). Recently, above all during the last 20
years, many dating methods have been developed, providing direct or indirect absolute
ages for buried or exposed particles. Examples of the first group include radiocarbon,
thermoluminescence, uranium series, cosmogenic isotopes, amino acid racemization, IR
luminescence, and electron spin resonance. The indirect techniques are mainly based on
paleomagnetism, and the isotopes of O and C. They are used depending on the material
available, and their precision varies according to the technique utilized. Some of these
methods are under experimental development, but in the future they promise to revolu-
tionize our knowledge and comprehension of environmental changes (Stokes, 1997).
2. Some considerations on climate change
As indicated by Chorley et al. (1984), climate change is one of the subjects about which
many geomorphologists feel rather ignorant. According to these authors, this is due to the
following; in the course of Earth's history continuous climate changes have occurred,
change mechanisms act in complex ways, climate changes are apparent under different
patterns, and, finally, the lack of knowledge of the most significant aspects of the climate
producing erosive and sedimentary processes hinders understanding.
Another important fact comes from the knowledge of the required time for the
adaptation of landscape forms to new climate conditions. Some authors point out that
some landforms can remain during long periods of the history of the Earth. Twidale (1978)
indicated that the well-known Ayers Rock (Uluru) inselberg in the Australian desert
has remained from upper Mesozoic-lower Cenozoic to present. On the other hand, areas
with rills and active dune fields experience constant and rapid evolution. This persistence
or modification of landforms is a function, among other factors, of their spatial location
within the desert morphoclimatic region. Thus, landforms located in hyperarid
. . . . . . . . . . . .
-, g
I . . . . .
=,
8~
9 m,/._
~(D (
) i-
0
,,--<,%
L
o
0
0
.,'t~
t
0
V--
,,---
"r" t
! E
i
tI >
9
(I
\ 1
1 1 1 , i 1 i~ ,
,,--
i
I..0 0
~ I
LO
I
',--
/
I
04
1
O cxl
i
",a"
i
d ~ cS o ~ ~ o
(0o) eJn:leJedLue~,u! eOueqo ~ (0o) eJn:leJedwe:lu! eOueqo ~ (0o) eJnl.eJedwe:i,u! eOueqo
0
Y
_~i eueoOlOH--
....~'~~~.C~ --ocMO
.,..,
E
m i:1.
,....J ~------ ~-- CO
~--...,...._~, E m
~- 8o
E "- I--
U.I
o
,e--
7 o
) 0
i I ! l i 1 ~l E:)
wJeM PloO o c~ ~ ~ CXl O
I ~"I cOI (3OI O
o4
eJnleJedwe~,leqOl6 ueelAI ~ (0o) aJnJ,'eJedwa~,u! e6ue40 ~ (0o) eJnleJedwel u! e6ue40 i
Figure 23.4. Foreground:terraces of the Jachal river, which present a braided morphology. Background:
structural landscapes corresponding to the Precordillera of Andes, from where a pediment system
develops. The area is located between Rodeos and San Jos~ de Jachal, San Juan Province, Argentina.
environments have undergone little or no change through Quaternary climate fluctuations

(erosive paralysis of Oberlander, 1997b). In comparison to this, landscapes existing in
transition regions will be clearly affected by temporal variations in the global atmospheric
circulation, which cause the migration of the boundaries of the climatic zones. The latter
are the most suitable for the reconstruction of the climate changes that have occurred in
recent periods of the history of the Earth.
Many landforms are in a dynamic equilibrium (Hack, 1960), in which processes act
but the active elements of the system are balanced. If any modification takes place, for
example, a climate change, the geomorphic threshold can be exceeded (Schumm, 1973,
1979), causing instability in the environment. These thresholds can be sharp or gradual
(Begin and Schumm, 1984) and can act in any time interval, from minutes to millions of
years. Geomorphic thresholds can be triggered by extrinsic or intrinsic causes. An example
of the first, or extrinsic type is an abrupt change that eliminates vegetation that leads to
Figure 23.3. Climate change during the last 180 ma. (a) Climate change on a 10v to 108 year timescale,
reflecting long-term climate evolution as a result of plate tectonic changes. (b) Change with 105 to 106
year timescale of glacial/interglacial periods, showing the 100,000 year cycle. (c) Change on a 104 to 105
year timescale showing the structure of change within a glacial/interglacial cycle, dominated by 20,000
and 40,000 year Earth orbital changes. (d) Change over the last 104 years and the transition from a
glacial to interglacial period. (e) Change over the last 103 years showing the Medieval Warm Period and
the Little Ice Age. (f) Change during the last century, with a possible anthropic influence on climate
(Boulton, 1993).
intense erosion. Intrinsic thresholds operate inside the system, as when there is a continuous
supply of particles on a slope, which then reaches a critical threshold of instability. In
some cases, these geomorphologic thresholds can be quantitatively defined and used as
the basis for the recognition of potentially unstable landscapes (Schumm, 1979). On
the other hand, because connections inside a geomorphic system are so complex, when
an external change takes place, it spreads over the system in a complex way (complex
response of Schumm, 1973). Bull (1991) analysed a collection of important concepts in
relation to the response of the geomorphic systems to the climate change (Figure 23.5).
This author showed a reaction time of the system when faced with stimulus, followed by a
relaxation time, which is necessary to reach the new conditions of equilibrium. The sum
of reaction and relaxation times constitute the response time. The time during which
the geomorphologic system remains in equilibrium is the persistence time. Therefore,
systems can have different thresholds, short to long times of persistence, and very variable
times of response. The latter indicate the system sensitivity to change (Brunsden, 1990).
The resistance of the natural systems to climate changes is not well known, so
landscape sensitivity to climate changes is a field of research of special interest. "The
sensitivity of a landscape to change is expressed as the likelihood that a given change in the
controls of a system will produce a sensible, recognizable and persistent response"
(Brunsden and Thornes, 1979). Sensitivity may vary in space and time and also depends
upon the scale (Thornes and Brunsden, 1977). Geomorphologic systems of small areas
,,
R e l a x a t i o n time :
Reaction
time ~ Rx =-:
aa .-.....~.,
i Rt ~:
R $ s p o n s e time . ~ Persistence
,,.._,
Rt
L
iaal ! time
I Ps -
, i
!Ra
, ~
,
,
E T
Figure 23.5. Changesin the height of a fluvial channel produced by aggradation and degradation are used
to show the components of response time. Rt is the response time, which is the sum of reaction time (Ra)
and relaxation time (Rx). Ps is the persistence time in the new equilibrium conditions. T and E represent
threshold and equilibrium, respectively (Bull, 1991).
reach a balance with the new conditions faster and permit carrying out the analysis of short
time spans. Temporal changes can be slow and gradual or sudden and catastrophic.
In general landforms present a slow response to climate change, so they commonly can be
analysed only for long time periods. To understand the more or less rapid changes in the
development of landscapes, it is necessary to know as well as possible the magnitude of
the processes and their temporal record. All this can lead to the understanding of the
sensitivity of geomophic systems (Allison and Thomas, 1993). It is possible to distinguish
those landscapes of high sensitivity to external impulse which respond rapidly to climate
change, such as fluvial channels. Insensitive systems are those whose response is slow and
present a strong resistance to change such as interfluve areas or the plains of Africa and
Australia. Insensitive areas commonly correspond to landforms originated on resistant
lithologies, whereas in sensitive zones small changes take place more rapidly, as is shown
in rilling areas (Brunsden and Thornes, 1979).
In desert areas, activation of the systems, which is necessary to produce a response on
the landscape, is generally much smaller than in other climatic zones. This indicates that
climatic sensitivity for arid zones landform generation is high because the relaxation time
necessary to reflect changes in the forms is low (Shaw and Thomas, 1993). These models
of short relaxation time are termed labile landforms by Trudgill (1976) (Figure 23.6).
to "~'~,.~..~-_.,,,, tf
(a)
__ ..,~._~n~,., ~ ,
TIME
-----.-- Climaticchange ........ Labile~ -t LANDFORM

RESPONSE
(b) Sluggish
Figure 23.6. Labileto sluggish landform responses over time. (a) Corresponds to a cyclic climate change
and (b) a pseudo step-functional change of climate (modified after Trudgill, 1976).
Labile landforms are rapidly modified under extreme events or they have limited
resistance to change. Their response is almost parallel to climate change. On the contrary,
sluggish landforms are subordinated to weak processes or present a large resistance to
change.
Gradual climate change is assumed and the hydrologic response is of the same type.
Nevertheless, Knox (1972), in a work on southeastern Wisconsin, USA, indicated that
A B C D E
Relative Precipitation Regime
I I
]
Humid
Arid .. 1 i
B
Relative Vegetation Cover
Maximum
Minimum
~ , , ,,,, ,
LIF-- L_ I
Relative Hillslope Potential for Erosion
Maximum I III I I II
Minimum
D
Relative Geomorphic Work
(e.g., Sediment Yield)
Maximum
Minimum
Figure 23.7. Vegetational and geomorphic responses to abrupt changes in climatic regimes. The curves
would seem to be most applicable to climatic regions that have annual precipitation between 250 and
1520 mm (Knox, 1972; in Schumm (1977), Figure 3.3).
climate change can be abrupt and, as a result, a response is triggered. This response,
although short, can be opposite to the curve of soil loss of Langbein and Schumm (1958).
This author (Knox) pointed out that if there is an intense and maintained increase in
precipitation, this causes erosion of the channel and a rise in slope gradient (Figure 23.7C).
Therefore, in this short period of time an important increase in the loss of soil is also
produced, followed by a fall (Figure 23.7D). Nevertheless, these fluctuations are difficult to
distinguish from the effect of large flooding with long return period (Chorley et al., 1984).
3. Future climate and the greenhouse effect
The study of the climate trend is nowadays the purpose of many researches, as a result of
the increased warming predicted by scientists due to the effect of greenhouse gases. These
gases are water vapour, carbon dioxide, nitrous oxide, methane, and chlorofluorocarbo-
nates. They have increased during the last decades, affecting the global climate. The most
significant effect is due to the rise of carbon dioxide produced by the combustion of
an ever-increasing quantity of coal, petroleum, and other fossil combustibles, as well as
woodcutting and forest fires. The CO2 is almost transparent to short-wave solar radiation
and absorbs a major part of the long wave. Therefore, it is one of the factors that produce a
greenhouse effect, raising the temperature of the Earth's surface and lower atmosphere.
Thus an increase in the quantity of CO2 leads to global warming (Budyko, 1982).
In the meeting that took place in 1990, sponsored by the United Nations, the Inter-
governmental Panel on Climate Change (IPCC) produced a report (Houghton et al., 1990).
It was pointed out that according to different models (Figure 23.8) temperatures can raise
0.3~ per decade in the higher assumption, as an effect of the greenhouse gases increase.
BUSINESS-
uJ
n" 4 - AS-USUAL
v
rru3
uj~3 SCENARIO B
13_ v..-
SCENARIO C
oI---- ~o 2
SCENARIO D
LM .
J ~
<rr ]
LM
rr .
0 --
1850 1900 1950 2000 2050 2100
YEAR
Figure 23.8. Simulations of the increase in global mean temperature from 1850 to 1990 due to observed
increases in greenhouse gasses, and predictions of the rise between 1990 and 2100 resulting from IPCC
Scenario B, C, and D emissions, with the business-as-usual case for comparison. The rise in temperature
is 0.3~ per decade for scenario A; 0.2~ (scenario B); slightly more than 0.1 ~ (scenario C) and 0.1~
(scenario D) (after Houghton et al., 1990).
As a result of it, a partial ice cap melting could happen and see level could raise 5 to 6 cm
on average in one decade.
These variations will generate intense modifications of natural systems producing
important socioeconomic implications that would seriously affect human life. For this
reason, during the summit in Rio de Janeiro in 1992 a Covenant on Climate Change was
signed (CCC). In the agreement a group of programs was proposed to mitigate future
climate change, as well as to promote new technologies to reduce greenhouse gases.
From IPCC in 1990 to the one in 1995 (Houghton et al., 1996), relevant progress has
been made. This lies in the application of more powerful simulation models, as well as the
consideration of the effects of sulphate aerosols and the reduction of the stratospheric and
anthropic ozone, which highly affect climate change (Santer et al., 1996). All that research
led IPCC in 1995 to the development of new scenarios for the period 1990 to 2100, on
the basis of different hypothesis. For the average scenario (IS92a) an increase of 2~ in
the temperature of the air in 2100 is estimated, and 1 to 2.5~ in the case of extreme
suppositions (IS92c and IS92e). The sea level rise due to ice masses melting and the
thermal expansion of the oceans are calculated for the year 2100 as 50 cm (scenario IS92a)
and 15 to 95 cm in the case of other extreme scenarios (Houghton et al., 1990). These new
values reduce quite a lot the most catastrophic forecasts of the IPCC in 1990. This
important decrease is basically due to the consideration of sulphate aerosols in predictions
and to the application of new atmosphere-ocean models.
Nevertheless, Houghton et al. (1996, p. 5) indicated that in the predictions of climate for
the future:
important uncertainties remain which come from vague estimates of future gas emissions,
from the representation in climatic processes models (as those associated to clouds, ocean,
marine ice, and vegetation) and from the systematic capture of climatic variables. Further-
more, short- and long-term climate changes are difficult to predict, because the nonlinearity of
the climatic system, so that, numerous "surprises" can occur.
The most important finding of the IPCC third assessment report, was the statement:
"There is now new and stronger evidence that most of the warming observed over the last
50 years is attributable to human activities." This represents a significant strengthening
of the analogous statement issued by the IPCC (1996): "The balance of evidences suggest
a discernible human influence on global climate." The globally-averaged surface air
temperature has increased by about 0.6 + / - 0 . 2 ~ over the 20th century. Most of this
warming has occurred during two periods: 1910 to 1945 and 1976 to 2000, with gradual
cooling during the intervening period. The range, reported in the IPCC Third Assessment
Report (IPCC, 2001), is higher than the 1.0 to 3.5~ range reported in the Second
Assessment Report (IPCC, 1996) probably because a greater range of scenarios are now
being used. In the 1966 report only six scenarios were used, whereas now 40 scenarios
are used.
Although there is a general consensus on the fact that global warming will occur in the
near future, some opinions disagree or express doubts about discrepancies. Dickinson
(1986) showed that general models of circulation (GCMs) have many limitations and
should be carefully used. Sundquist (1993), in a study about the emissions from ice during
the last interglacial period and the absorption balance in the oceans, indicated that there are
significant uncertainties in relation to anthropogenic CO2, considering its later effects.
Pat6-Cornell (1996) pointed out that the models used for the study of the effects of global
climate change are based on a limited knowledge of the main phenomena (for instance, the
role of clouds on oceans). Demangeot (1996) considered that proof of the actual quantity
of warming of the planet is not convincing and asked specialists for more precise data on
icecap melting, the role of deforestation, and so forth. "Since 1940, the Greenland coastal
stations data have undergone predominantly a cooling trend. At the summit of the
Greenland ice sheet the summer average temperature has decreased at the rate of 2.2~ per
decade. This suggests that the Greenland Ice Sheet and coastal regions are not following
the current global warming trend" (Chylek, 2004).
Embleton (1989) indicated that climate changes forecast for the next decades or
centuries is extremely difficult. Obviously, CO2 has significantly increased from 1750,
according to data provided by the analysis of ice cores in Camp Siple in Antarctica and
from air samples taken since 1958 at the observatory of Mauna Loa in Hawaii. There is no
doubt that CO2 produces an increase in temperature. As indicated above, predictions show
an increase of up to 3~ per century (Houghton et al., 1990). If these predictions came
true a "super-interglacial" would occur (Figure 23.9), which would then be the event of
greatest magnitude in the Quaternary. According to Embleton (1989) and Goudie (1990b),
_
SUPER INTERGLACIAL I \
INDUCED BY I
O
o
"~ 2 m GREENHOUSE GASES I
z\
i.iJ
rr
LAST I
cr INTERGLACIAL PRESENT I
ILl I INTERGLACIAL I
8..
.. 0 -
._1
o-1
LAST
_z-2 / GLACIAL /NATURAL
D
O
-4
-5 I ! I I I 1 I 1
-140 -120 -100 -80 -60 -40 -20 0 20
THOUSANDS OF YEARS
; A S T l FUTURE
=
Figure 23.9. Prognosis of a "super-interglacial" induced by greenhouse warming compared with the
natural temperature trend of the last 140,000 years (Lamont Newsletter 6, 1984; in Embleton (1989)).
oceanic water, which constitutes an enormous reservoir of CO2, and marine biota play an
important role on the regulation dynamics of atmospheric CO2. Carbon dioxide flux in
the oceans depends on CO2 pressure, atmospheric concentration of CO2, and on surface
winds. Carbon dioxide pressure is controlled by oceanic transport processes, geochemical
processes, and biological processes in the top part of the oceans. Furthermore, solar flux
and latent calorific flow have an influence, as do the above-mentioned factors, in climate
modelling. Therefore, the system is very complex and the models of the flows in the
ocean-atmosphere inter-phase can be subject to large errors (Dickinson et al., 1996).
Moreover, there is limited knowledge on the temperature and salinity of deep waters, close
boundaries between currents, ocean eddies, and on thermohaline circulation (McBean
et al., 1996).
The statistical analyses of the daily difference of temperature between maxima and
minima from 2000 stations all over the world for a period of 40 years show that nights are
warmer than days, 0.84~ and 0.28~ respectively. This means that the principal warming
is due to the increase of night temperatures. Statistics reveal that cloud cover seems to be
the cause of this circumstance (Henderson-Sellers, 1992), because clouds regulate night
temperatures by controlling long-wave emissions. During the 20th century, clouds have
increased 3.4 to 9.4%, possibly because of the increment of condensation nucleus due to air
pollution. This could explain the smaller percentage of clouds above oceans with respect
to continents. Without doubt, it is necessary to know the variations of cloud cover in order
to carry out further analysis of climate change in the future.
Volcanic activity, mainly in the case of acid composition eruptions, which are more
explosive, can occur sporadically and has a marked influence on climate. If the eruption
is large, volcanic ash is driven into the stratosphere, affecting general atmospheric cir-
culation. In June 1991, the Pinatubo volcano, located in the Philippine islands, erupted.
Volcanic aerosols extended all over the planet and the models developed indicated that the
average global temperature decreased on the surface up to 0.5~ for 2 years and 0.6~ in
the troposphere for several months. The effects of the eruption were less evident in 1994.
The consequences of the volcanic activity, which produced opacity in the stratosphere, had
a significant influence on climate, but only sporadically.
Man, by means of industrial activity and agriculture, increases atmospheric dust
significantly. Also, large quantities of aeolian dust can be deposited on to continents and
into oceans from long distances. Dust reaches the upper part of the troposphere and restricts
solar radiation. As a result, there is a global cooling (Bryant, 1997).
The importance of anthropic sulphate aerosols on climate change is becoming
accepted. These aerosols come mainly from coal and petroleum combustion. Records
indicate that during the last century the exponential increase in sulphate aerosols is parallel
to the rise of CO2. The high reflectivity of the sulphate aerosols produces a cooling in the
troposphere with both clear and cloudy skies (Pearce, 1994; Santer et al., 1996).
Houghton et al. (1996) point out the difficulty for the prediction of long- and short-term
climate changes. This is what some authors call "climate surprises" (Jouzel and Lorius,
1999). Rapid climate variations have been found for the last glacial age. In cores obtained
from the Greenland icecap, several warming phases of about 10~ have been found. Those
phases lasted several tens of years and gradually turned into colder conditions. Such
sequences, the duration of which range from 500 to 2000 years, recurred about 20 times in
the last glacial epoch (Bender et al., 1994), and not only in Greenland. On the other hand,
during the last 70,000 years, several glacial sedimentary layers coming from icebergs
have been identified from cores in the northern Atlantic Ocean. These six synchronic
stages are called Heinrich events (Heinrich, 1988; Bond et al., 1993, 1997). The increment
of freshwater and the temperature decrease in the upper part of the oceans produced
significant changes in some of the climatic parameters. Therefore, the occurrences of
rapid climate variations in the past make feasible the triggering of these types of climate
modifications in the future.
Without doubt, climate change modelling is uncertain and complex, due to the quantity
of parameters needed and the difficulty in comprehension of their relationships. This
debate is not about if climate change is due to human activities, but rather how quickly, to
what magnitude, and with what regional implications (Weaver, 2003). An alternative
methodology for the study of climate changes is the study of the geological records (Petit-
Maire, 1999), wherein valuable data about climate change can be obtained.
In order to study the consequences of warming on landscape, researchers can go into
the past, mainly to the Climate Optimum or Altithermal, which occurred about 7000 BP,
and compare it with the last maximum interglacial at about 120,000 BP, so that the
response of the landscape in warmer phases can be analysed. Thus, landscape changes can
show the geomorphologic setting for the future and the attached consequences. This
prognostic approach is an important contribution from geomorphology.
If the change predicted by IPCC becomes evident, it would affect areas very differently
depending on the climatic environment. For example, in arid regions an increase of the
existing problems is expected (Boer and De Groot, 1990), such as soil degradation,
hydrologic and aeolian erosion, decrease of soil fertility, salinization, sedimentation, and
flooding. Moreover, an expansion of semiarid areas, an increase in fire risk, and a decrease
in the quality and quantity of subterranean water would be expected.
Chapter 24
Climate change in glacial and periglacial regions
1. Introduction
Glacial and periglacial regions constitute morphoclimatic zones related to cold. As Tricart
(1967) showed; "cool countries are those in which the geomorphological action of the
water is controlled by its presence as solid state, permanently or periodically." Glaciers are
developed in areas in which the snow does not melt completely from year to year, so that
snow accumulates and in the end it is converted into ice. Limits of the glaciers are clear and
correspond to ice accumulation. Periglacial areas are characterized by frost and thaw, and
snow does not remain from year to year. Frozen soils can develop under the Earth' s surface.
During the last glacial epoch some 30% of the Earth's surface was overlain by ice,
covering vast areas of North America and Eurasia. Also the Antarctic expanded. In
mountainous areas, small icecaps were formed and valley glaciers advanced. Nowadays, ice
occupies 10% of the Earth' s surface. The big advances and recessions of the ice masses have
sculpted wide areas, eroding rock massifs and depositing thick accumulations of glacial
materials. Movements of the glacier fronts respond to climate changes, so they constitute a
source of information for the study of those changes. On the other hand, ice sheets and
glaciers have an important influence on the general system of atmospheric circulation.
As a result of the important advances and recessions of the masses of ice during Quaternary,
the periglacial domain underwent major fluctuations. During glaciations the extent and
location of the periglacial zones were very different from the present. Thus, vast areas that
were covered by ice today were periglacial then. Also, the areas in the surroundings of the ice
that were periglacial, nowadays are temperate zones. Therefore, changes from glacial to
periglacial conditions and from the latter to temperate can be recognized.
On the other hand, there are regions in which the periglacial domain has remained during
the Quaternary. They correspond to areas in which the precipitation is so small that they
cannot generate an icecap, as is the case of wide areas in Siberia (King, 1976a). These
important climatic fluctuations affecting cold areas have generated relict forms, which remain
under new climatic characteristics, and other landforms that correspond to current conditions.
Within both situations there are many inherited and partially eroded landforms that are
difficult to recognize. This aspect adds to doubts on the differentiation of relict and current
landforms.
2. The paleoclimate information provided by drill cores from ice sheets

and deep sea sediments
The studies of ice and sediment samples from boreholes makes possible the knowledge
of past climate changes that have affected the whole planet. Many of these changes have
been produced very rapidly. These important knowledge advances on climate change have
been developed mainly over the last three decades.
From drill cores made in ice sheets, it is possible to obtain paleoclimate records that
contain annual snow precipitation data. Records from the upper part of the core are
alternate layers coming from annual precipitation, whereas in lower layers the
identification of the alternation is not so simple and the records have to be dated by
indirect methods.
Valley glaciers and mountain icecaps have also been cored in Tibet, the Peruvian
Andes, and elsewhere, but the data only provide information on recent climate changes,
which is different from the results obtained from deep drilling in polar ice sheets. The first
deep core was drilled in 1963 to 1966, in Camp Century, Greenland (Dansgaard et al.,
1969). In Antarctica, the Vostok drill in 1983 (Jouzel et al., 1987) provided a climatic
record of the last 160,000 years, exceeding the ice of the last interglacial. The final section
of this core has reached back to 400,000 yr BP Recently the ice has been drilled at
Dome C, located 500 km away from Russian station Vostok, in the East Antarctic ice
sheet, by a consortium called the European Project for Ice Coring in Antarctica (EPICA).
A 740,000 years record is being researched in this project (EPICA, 2004). It is thought
that the final section could reach back to 900,000 years or even further (Walker, 2004).
Between 740,000 and 430,000 years ago, the climate was characterized by less pronounced
warmth in interglacial periods in Antarctica, but a higher proportion of each cycle was
spent in the warm mode (EPICA, 2004, McManus, 2004). The result of this research may
imply that without human intervention a climate similar to the present one would extend
well into the future (EPICA, 2004).
The drill cores, GRIP and GISP2, were carried out in the centre of Greenland, 28 km
apart. In the first one no horizontal movement in the ice was recognized, while in GISP2,
from 2750 m depth on, disagreements in the interpretation exist, due probably to ice flow.
This increases toward the boundaries of the ice sheet.
Oxygen atoms, which are constituents of the glacial ice, have three types of isotopes,
160, 170, and 180. The relative abundance in nature is 0.2, 0.04, and 99.76%, respectively.
During water evaporation on the surface of the sea, the heaviest isotope enriches mainly
in the liquid phase and decreases in the vapor phase. This process is called fractionating.
The resulting vapor has become impoverished in 180, compared to the former water.
Condensation depends on the temperature and the isotopic composition of the precipita-
tion reflects the temperature of the condensation. Therefore, the isotopic composition of
the glacial ice can be used to reconstruct the temperature of the original precipitation. The
isotopes existing in the ice are determined by melting the ice in the laboratory and then
analysing it utilizing a mass spectrometer. The difference between 180/160 and a sample
of standard water (Standard Mean Ocean Water, SMOW) is expressed as ~ 180. If values
are small, the paleotemperatures were low, and vice versa. Because of this, detecting
global climate fluctuations is possible (Dawson, 1992; Ehlers, 1996; Benn and Evans,
1998). Figure 24.1 shows data obtained for GRIP, almost 3000 m depth. The table
represents the interglacial conditions at 130,000yr BP (Eemian or last interglacial),
several interstades (short warming phases) and the rapid termination of the glacial
conditions around 11,000 yr BP. Holocene climate changes have a poor representation,
according to Dansgaard et al. (1993), whereas in the later work of Bond et al. (1997)
a detailed subdivision has been carried out.
Climate change in glacial and periglacial regions 617
-36-34 (%,)
_
10 1500
AGE (103 yr)
~-11
- -12
1_-14 1 Belling
-16
IS No.
"- -18
-20
2
500 2000
25
30
E Denekamp E
35
v v
8
i 9 i
F- F-
CL
W
40 I? CL
W
0 2 Hengelo 0
13
50 .14 Glinde
Oerel
1000 60 18 2500
70 19
20
80 -21 Odderade
22
-100 23 Brerup
24
i~ii i s a E : ~
1500 b).~3000
-45 -40 -35 -30
~180 (%0)
Figure 24.1. Continuous 6180 record, corresponding to the GRIP Summit, plotted in two sections on a
linear depth scale. (a) From the surface to 1500 m deep; (b) from 1500 to 3000 m deep. Measures are taken
at each 2.2 m. Glacial interstades are indicated to the fight of curve b (Dansgaard et al. 1993).
The interstadials are rapid and short-term events that occurred several times between
110,000 and 10,000 years ago. They are called interstadials to be distinguished from
cool phases or stadials. In Figure 24.1, 24 interstadials can be recognized. They are called
" D a n s g a a r d - O e s c h g e r events" (Bond et al., 1993). Each warm interstadial is attached to
a cool one, and the pair, which last about 1500 years, is called a "Dansgaard-Oeschger
cycle."
The analysis of the Vostok drill core in eastern Antarctica has revealed important
information regarding CO2 variations in upper Pleistocene and Holocene (Figure 24.2)
(Lorius et al., 1988). The temporal variability in CO2 existing in air bubbles within the ice
A . . . .
BEE D
II
,- 9 i ii
C1 A1 B1 E1 D1
CO 2ppmv 280-
260-
\
240-
AT ~
220 - - 2
200 -
180- "t -2
0
-4
-6
-8
_
- -10
'" I '' I' . . . . . . I
0 50000 100000 150000
years B.P.
Figure 24.2. C O 2 variation curve for the last 160,000 years (upper part) obtained from the Vostok ice
core in Antarctica, and a smoothed curve of the temperature record from oxygen isotopes for the same core
(bottom curve) (after Lorius et al. 1988; in: Dawson (1992), Figure 2.5).
presents a good correlation with the fluctuation of the oxygen isotope, which matches the
changes in temperature.
The study of the deep sea sediments has provided very relevant data in relation to
climate fluctuations in the upper Tertiary and Quaternary eras. In these sediments, there
are planktonic foraminifera, which float on the sea when alive, and benthonic foraminifera
that live at the bottom of the sea. Their shells are made up of CaCO3 and the oxygen is
made up of 180 and 160. These calcareous shells incorporate oxygen isotopes depending
on their concentration in the seawater. When water evaporates, the vapor enriches in
160 and then precipitates on icecaps, the 160 content is higher than in the oceans, which
are enriched in 180. Therefore, analysing the isotopes of oxygen, it is possible to
reconstruct the fluctuations in the volume of the ice sheets and thus, the glacioeustatic
changes of the sea level. Emiliani (1955) was the first to use this technique. He interpreted
that the observed fluctuations mainly reflect the changes in temperature corresponding
to glacial and interglacial periods. The analyses of the deep drill cores in the Pacific
Composite 8180 Record (0- 2.5 Ma)

3.0
3.5
~" 4.0
4.5
5.0 8
I !. . . . . . O
5,5 ............. 1 ,
0.0 0.5 1.I0 1.5 2.0 2.5
Age (Ma)
Figure 24.3. Variations in the oxygen isotope compositions in planktonic foraminifera from the
equatorial Indian Ocean for the last 2.5 million years (Ma). It can be observed that the global ice volume
has increased mainly during the last 750,000 years and that the mean duration of the cool periods has also
increased (Crowley and North, 1991; based on data in Raymo et al., 1990).
(Schackelton and Opdyke, 1973, 1976) clearly denote that the record of isotopes of oxygen
reflects the whole climate history of the Quaternary (Figure 24.3). There exist several
difficulties in obtaining suitable records of oxygen isotopes in those oceanic areas in which
the velocity of sedimentation has been relatively small. Also the effect of bioturbation,
submarine landslides and turbidity currents hide and make the interpretations complicated.
Other significant changes that are recorded in deep cores from the northern Atlantic to
Lisbon latitude during the last 70,000 years are the so-called Heinrich events (Heinrich,
1988; Bond et al., 1993, 1997). In contrast to the interstades these seem to be cool events,
but also for short periods of time. These events are recorded in the cores by clasts that were
transported by the ice, which implies a great presence of icebergs during those events.
When the ice melts, icebergs deposit solids on the marine bottom (Table 24.1).
Values for H O - H 3 have been obtained from Bond et al. (1997), and those corresponding
to H 4 - H 6 from Bond et al. (1993). It is not well known whether or not these processes
have been originated by the intern dynamic of the icecaps (MacAyeal, 1993) or by periodic
external climate changes (Broecker, 1994). The Heinrich events have been correlated
with records of pollen from Europe and North America (Grimm et al., 1993), with loess
Table 24.1. Distribution of the principal Heinrich events (H) and

Younger Dryas (YD) for the last 70,000 years (ages in thousand years
from present) (Adams et al., 1999).
Event Age Event Age
YD or HO 12.3 H1 16.8
H2 24.1 H3 30.1
H4 35.9 H5 50.0
H6 66.0
columns from the centre of China (Porter and Zhisheng, 1995), glacial deposits in western
USA (Clark and Bartlein, 1995), and with sequences of talus flatirons in northeastern Spain
(Guti6rrez et al., 1998b). All this seems to indicate that there is a global climatic influence
of the effects of the Heinrich events, at least in the Northern Hemisphere.
3. Fluctuations of quaternary ice sheets and resulting landforms
The advances and recessions of the glaciers have a bearing on morphogenesis, so it is

fundamental to know accurately the oscillations that occurred in order to understand
the resulting glacial landscape. These glacial landforms are produced by erosion and sedi-
mentation processes and crop out when ice masses retreat. On the basis of existing
landforms, areas and flow directions of the ice in ice sheets have been calculated.
The maximum expansions of the last ice sheets have a similar age. Dawson (1992),
compiled global data that indicate an age of 15,000 to 20,000 years for the Euroasiatic
ice sheet (late Valdai), 18,000 to 20,000 years for the Fennoscandian (late Weichselian),
for the Laurentide ice sheet 20,000 to 25,000 years (late Wisconsin), and for the Antarctica
ice sheet 17,000 to 21,000 years. The distribution of glaciers and icecaps for the last glacial
maximum, with variations of the equilibrium line depending on the latitude and altitude,
are shown in Figure 24.4 (Broecker and Denton, 1990).
The warming subsequent to the glacial maximum caused the disappearance of the
Eurasian, Fennoscandian, and Laurentide ice sheets, and many valley glaciers and small
glacier caps. The response can be observed in the data that show the modification of the
general circulation of the atmosphere, in the rise of the sea due to the incorporation of
fresh water, and thus, in the modification of the oceanic circulation. Melting events are
not synchronous; the Cordilleran ice sheet disappeared 10,000 years ago, the Laurentide
ended 6500 years ago, the Fennoscandian melted about 8000 yr BP, and the Eurasian had a
shorter life, 9000 yr BP (according to some authors, in Dawson, 1992).
It is difficult to establish a chronology for the Fennoscandian ice sheet due to the fact
that moraines are very broken and cannot be properly correlated in fiord areas. After
13,000 yr BP, a rapid slimming and retreat occurred, which corresponds to a warm inter-
stadial that lasted until 11,000 yr BP. In southern Scandinavia a big lake was formed
between northern Europe and the southern part of the ice. This occurred between 12,800
and 10,300 yr BP, and it is called the Baltic frozen lake (Figure 24.5) (Benn and Evans,
1998). About 10,000 yr BP, the retreat of the ice made possible for seawater to come into
the Baltic Sea (Yoldia Stage). After that, there was a rapid recession of the ice, between
10,000 and 9000 yr BP, so that the ice sheet was reduced to a small extension in northern
Sweden and a big fresh water lake was originated (Ancylus Lake). Finally, and probably
due to the glacioisostatic rebound and to the Flandrian transgression, marine waters went
into the Baltic Sea and Botnia Gulf (Litorina sea).
The Laurentide ice sheet extended from the Canadian archipelago in the Artic Ocean to
the northern states of USA, in the south. It bordered on the Atlantic Ocean to the east and
the Rocky Mountains to the west. From this boundary to the Pacific Ocean, the Cordilleran
ice sheet developed. As a result of the glacioisostatic depression, large proglacial
lakes were originated in the southern part. Ice recession was characterized by sudden
retreats that produced catastrophic floods and spectacular modification of the landscape.
Figure 24.4. On the left: the distribution of the principal ice sheets and mountain glacier complexes
during the last glacial maximum. On the right: the worldwide lowering of equilibrium line altitudes during
the last glacial maximum is depicted along a north/south transect (open boxed line) (after Broecker and
Denton, 1990).
Figure 24.5. Stages in the recession of the Fennoscandian ice sheet and associated marine and lake
events. (a) Baltic Ice Lake; (b) Yoldia Stage; (c) Ancylus Lake; (d) Litorina Sea. (According to Eronen,
1983; in: Benn and Evans, 1998).
The history of the melting these two ice sheets is not well known. The southern limit
of the Laurentide ice sheet presented several ice lobes (Figure 24.6a) whose advances
and recessions seem to be nonsynchronous. The first recession started in the eastern part
17,000 years ago, whereas the east remained active until 14,000 yr BP. The Cordilleran
ice sheet started to move back after the many big j6kulhlaup that broke the ice barrier
of Lake Missoula 13,000 years ago (Figure 5.9 and Figure 24.6a). The thinning of the ice
continued until its disappearance, 10,000 years ago. In the Laurentide ice sheet, the period
13,000 to 11,000 yr BP marked the disappearance of the ice in the Great Lakes. The most
important periglacial lake was Lake Agassiz (Figure 24.6b), which overflowed even into
the Gulf of Mexico. Later, the ice sheet recession caused a change in drainage into the
western Atlantic Ocean, between 11,000 and 10,000 yr BP. The maximum extension took
place between 9900 and 9500 yr BP, covering an area of 350,000 km 2. Later than this
period and until approximately 8000 yr BP, many j6kulhlaup occurred, some of which
were larger volumes than the multiple ones from Lake Missoula. The Laurentide ice sheet
Figure 24.6. Reconstruction of the Laurentide ice sheet. (a) Situation for 14,000 BP, showing positions
of major ice divides. C.R. = Columbia River. (b) position for 8400 years ago (after Dyke and Prest, 1987;
in: Dawson (1992), Figure 5.6A and Figure 5.9A).
was receding gradually to the northeast and about 7000 years ago only small remnants
remained (Dawson, 1992).
The study of the principal groups of glacial erosive and depositional landforms is very
relevant and complex, and can indicate the fluctuations and processes produced by the
activity of the large icecaps during the Quaternary. These interpretations must be based on
precise analogies with the actual environment. Despite the difficulties for interpretation,
nowadays there is a detailed knowledge of the principal groups of landforms, the processes
involved and the geomorphologic history of the Laurentide and Fennoscandian ice sheets.
In the last three decades, important progress on the knowledge of climate change has
been made on the basis of core records, deep sea sediments, and ice sheets. This implies a
substantial scientific advance concerning the interpretation of the landforms produced by
ice sheets (Hughes, 1992).
Erosive morphologies resulting from the activity of ice sheets are a function of a
group of basic variables. The factor of greatest effect is the temperature at the base of
the ice, because when it reaches its melting point it slides and erodes. Geology plays a
variable role; on the one hand, the erosion develops more easily in fractures, whereas on
the other hand, the presence of permeable rocks facilitates the loss of water and the
decrease of erosion. Topography controls the ice thickness, and also favours the
convergence and divergence of the ice flow (Chorley et al., 1984). Nevertheless, it is
necessary to take into account that some zones, as in Buchan, Scotland, there was not any
glacial erosion, because thick weathered profiles remain intact after having been covered
by ice (Hall, 1986). This is one of the arguments used by the followers of the hypothesis
of glacial protectionism.
Some authors have calculated the volume eroded by ice in ice sheets of intermediate
latitudes. White (1972) considered that its erosive capacity has been very high, because it
has eroded and mobilized a rock thickness of more than 1000 m in the Hudson Bay region.
The erosive action of the ice sheets is helped by the existence of thick regoliths, as in
northern Scotland where there is a regolith thicker than 50 m (Fitzpatrick, 1963). When the
front of the glacier advances, it runs into loose material, which is easily evacuated, pushed
as does a bulldozer (Bakker, 1965; Thomas, 1994). These regoliths, resulting from
preglacial chemical weathering, constitute a good reason for explaining the fine-grained
tills, in comparison to those made up of fresh rocks (Feininger, 1971).
Regional erosional landforms are marks of subglacial erosion affecting bedrock. They
consist of elongated hills, roches moutonn6e and bedrock basins. The erosion is more
intense along joints, faults and dikes. This landscape is known as a topography of hills and
lakes, so termed in Scotland by Linton (1963). Also large areas exist with this morphology
in the Canadian Shield and western Greenland. This topography is characterized by a
landscape in a variety of forms, controlled by weak substrate structures, and with widths of
no more than 100 m.
The deposits generated by the activity of glacier masses occupy 10% of the Earth's
surface. Depositional landforms, which are related to icecaps, often are difficult to inter-
pret, because they are the result of the superposition of materials corresponding to several
glaciations, in which dates are commonly difficult to obtain.
Moraines formed by icecaps present a simpler pattern than the ones generated in valley
glaciers. At first glance, the landforms look straight but their main form is arched, as from
the Laurentide ice sheet, where this form is the result of numerous lobes on the frontal side
of the icecap (Figure 24.7). In many places, these large moraines are thrust moraines, with
hundreds kilometres of length and thicknesses of up to 400 m. Locally, they can exceed
200 m height. The interpretation is complex when deposits of different ages but similar
lithology are superimposed. The C-14 dates indicate the different ice-sheet lobes have
(a)
Driftlessarea
LakeMichigan
9 ', ' " 1 - . 9 ,,
'i.il.!ii i !i !i;i iIi!:i!i i :i
//
~ Prominentmoraines I ~
~ Drift ~" ,': :..:'..;..'~:::.:: .'.~'.->~'.'::.::. :':.: :.'.. >:

:.' .'.'. () "'..'.':"':'." ".""'" 50" ;:.'"': :'.'. :',' :100miies
~ Illinoianand older drifts
.";'.i . . . . . . .:".'..'
.'.':'. 0 ". . . . . . .'..." 100km :.'.''.'.': ..
- 300
(b) w E
0"
-250 ~"
0-
O.
200 []
0-
O- 150
Figure 24.7. Glacial deposits at the southern end of Lake Michigan. (a) At least four lobes in the front of
the Laurentide ice sheet have formed this configuration (Frye and Willman, 1973) 9(b) Profile across some
till sheets in northern Illinois (Frye and Willman, 1973; Wickham and Johnson, 1981, in Chorley et al.
(1984), Figure 19 9
experienced numerous cyclic advances and recessions (Clark, 1994). The retreat of the ice
sheets produces abundant landforms, such as kames and kettles, fluvioglacial landforms,
(eskers and pitted sandar) and glaciolacustrine, drumlins. All these landforms occupy their
largest extension within the successive boundaries of the icecap. Excellent examples exist
in central Canada (Figure 24.8), where the dividing line of the ice was located in the
eastern part and was covered by ribbed moraines, drumlins and eskers. The western
periphery is characterized by outcrops of the substrate and a thin layer of glacial materials
(Aylsworth and Shilts, 1989). These spectacular examples can be recognized in Labrador
in eastern Canada, Scandinavia, Finland and in the Kola Peninsula of Russia (Figure 5.13).
Benn and Evans (1998) tried to make a generalization, at least for the Laurentide ice
sheet, of the evolution of these depositional landforms for the last glacial retreat. The first
times were characterized by successive advances and retreats of the ice, producing the
deposition of a great quantity of sediments and their restructuring, as well as lakes
dammed by ice. After that, ice masses suddenly receded and moraines and discontinuous
eskers were generated on bedrock. Meanwhile, new dams produced marginal lakes.
During the last retreat stages, moraines were more extensive and an important network of
eskers, which indicate the flow direction, were produced.
4. Valley and cirque glaciers retreat
Alpine glacier valleys are characterized by very deep vertical slopes, where cryoclastsy
acts intensely, providing clasts to the margins and bottom of the valley. Generally, these
troughs were developed from periglacial valleys.
In historic times, valley inhabitants worried about the fluctuations of the glaciers,
because they produced many disasters affecting human life and properties. The cata-
strophic events were caused by advances of a glacier tongue and mainly by bursts of lakes
dammed by ice or landslides (Tufnell, 1984). This information about the historical events
in alpine valleys provides relevant data on the evolution of the valleys (Grove, 1979,
1988). Comparison of paintings and engravings of the 18th and 19th centuries with recent
and current photographs supplies relevant data on the fluctuation of the glaciers (Lamb,
1977; Stroeven et al., 1989). On the other hand, during the past few decades precise works
on mass balances have been carried out. In those works the variations of glaciers have been
related to climate change (Oerlemans, 1989a,b).
Alpine regions are characterized by a landscape with pyramid peaks (horns), edges
between glaciers, cirques, and longitudinal profiles with frequent over-excavations. In
areas that have undergone large retreats, or that also have limited supply, the predominant
landscape is the one with glacial cirques (Figure 24.9). These landforms were developed
in all aspects, although shady places are the most prone, because they are sheltered from
solar radiation and, therefore, from the thaw of the ice. Often after ice thaw, rock glaciers
arise in the cirques.
Valley glaciers are the most spectacular landforms, which reflect the great capacity for
erosion of these masses of ice. One of the most significant features is the lateral moraines,
which join with terminal ones and generate moraine ridges. These ridges reach a greater
thickness if the glacier is stable, although generally, melt water erodes the ridges little by
little. In some glaciers, as in the Nigardsbreen glacial valley of Norway, the terminal parts
Figure 24.8. Distribution of the glacial landforms in Keewatin, central Canada. (a) Location of the
drumlins and ribbed moraines (Rogen moraines). (b) Distribution of eskers and rock substrates (after
Aylsworth and Shilts, 1989).
Figure 24.9. Cirque glaciers, Vilcanota Range, southern Peru. Photo: J.L. Pefia.
that correspond to moraine ridges present a good state of preservation and there are
numerous dates by means of lichenometry (Figure 24.10).
Some of the most interesting effects of mountainous environments are the lakes
dammed by the ice, which have a very ephemeral life. In such cases, the mass of ice of the
principal glacier blocks the outlet of a secondary valley. Glacier snout recession triggers
the water evacuation from the lake and a new advance can produce another lake. Thus,
it is possible to study the fluctuations of the ice, which are related to short-term climate
changes. Many lakes dammed by ice are known in high-altitude and high-latitude regions.
In Twidale (1976a) many examples are presented.
A particular case of glacier sedimentation occurs when a glacier snout reaches a lake
(Figure 2.16 and Figure 2.26) or the sea, a circumstance commonly associated with the
effects of climate changes (Figure 24.11). Figure 24.12 shows two of the facies associa-
tions proposed by Powell (1981) for the glacier-fiords of Alaska. The depositional model
represented in the top part of the figure is characteristic of the margins of a stable glacier
slowly receding. In such conditions large fans or moraines are deposited. The bottom part
represents the margins of a fiord that receives sediments from a terrestrial glacier. A delta
is generated where melt water goes into a fiord, producing a system of top, frontal and
bottom layers (Figure 24.13).
Figure 24.10. The moraine ridge sequence documenting the retreat of Nigardsbreen, Norway, from its
Little Ice Age maximum position. Moraines are labelled A-X and numbered 1-19 on the north-east and
1-18 on the south-west sides of the foreland. The historical dates are based upon lichenometry, which
was undertaken at sites marked (+). (From Bickerton and Matthews (1992)).
Figure 24.11. Glaciertongue in the Antarctica ice sheet, flowing into the sea. Whisky Glacier, James
Ross Island, Antarctic. Photo: A. Martfn-Serrano.
The Glaciar Coronas are the most extensive ice mass on the southern slopes of the
Maladeta Range, located in the Central Pyrenees of Spain. Using early and modern
photographs and aerial photographs a decrease in length of 325 m, from 1928 to 1998 has
been calculated (Chueca et al., 2003). Similar researches focused on glacier retreat have
been carried out in the Peruvian and Bolivian Andes as well as in the Italian Central
Apennines (Georges, 2004; Ramirez et al., 2001; D'Orefice et al., 2000).
5. Glacioisostasy and glacioeustacy
Glacioisostasy is a consequence of the growth and disappearance of icecaps. When an

icecap increases, the lithosphere becomes deformed because of the weight of the ice mass.
The magnitude of the depression depends on the thickness of the icecap and the ratio
between ice and rock densities, which is about 0.3, so the depression generated is about one-
third of the icecap thickness. This depression reaches its highest values under the maximum
thickness of the ice and decreases toward the margins. The ice sheet is surrounded by a
peripheral depression, whose height results from the ratio H/11.5, where H is the height of
the icecap (for 1800 m, the depression is 155 m). Farther away from the depression a slight
convexity is generated, whose height is H/100; that is, 18 m (Figure 24.14) (Walcott, 1970).
Its origin is due to a lateral transference of subcrustal material under the icecap. The
glacioisostatic depression disappears with thaw, by an elastic rebound, weak at the begin-
ning (restricted rebound), followed by an accelerated uplifting period (postglacial rebound)
and, finally by a residual rebound of minimal magnitude (Andrews, 1970).
The information on glacioisostatic rebound is obtained from studies on beaches and
uplifted deltas, that contain seashells and silt and clay deposits with fossils that can be
Figure 24.12. Depositional models for retreat based on Alaskan examples. Above: slowly retreating
glacier in shallow water; below: terrestrial glacier margin supplying sediment to a delta. (Drawings by
R. W. Tope, from Powell, 1981).
Figure 24.13. Bertrand Glacier which supplies sediments to the Argentine Lake, Patagonia, Argentina.
Photo: C. Sancho.
dated by m e a n s of C- 14. It is possible to d e v e l o p isobases, w h i c h are isolines that r e p r e s e n t

the g e o m e t r y of the surfaces g e n e r a t e d at sea level, resulting f r o m the effect of the ice load
in a particular period. T h e s e isobases m a k e it possible to verify that the L a u r e n t i d e ice
sheet, during the glacial m a x i m u m , had several d o m e s associated with a central d e p r e s s i o n
(Andrews, 1970, 1975), w h e r e a s with the F e n n o s c a n d i a n ice sheet there was only a central
d o m e (M6rner, 1980) (Figure 24.5 and F i g u r e 24.15), with the highest uplifting in the
north of the G u l f of Bohnia. T h e isobases shape an elliptic m o r p h o l o g y r e v e a l i n g that it
d e v e l o p e d separately f r o m the d o m e on the British Isles.
Altitude
i I I
(m) I I I
600 I Peripheral depression I Forebulge I
Glacial surface ! I I
200
~ I
I
I
I
I
I
I
I
I
I
I
I
0-- ...... ...................:.-.:.:.:.:.:.:.:.:.:.:.:.:.:..:.:-:-:.:--:-:-:-:--:-:-:
Land surface ......:.::•:i:i:i:!:i:!:i:i:i:i:i•i:i:i:i•i•i•i•i•i:i:!:i•i•i:!:i:i•!•i•!:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:i:

-200 -
F.--:-.-.-.:::.-;-.-.-.-.-.:.-.::-.-.-.-:-.-'-:-:-:-:-:-:-:-:'
: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : b a : e : : n"e" : " :•:•:•::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::•:•:::•:•:::::•::::
:•:•:•:i•:•••:•:•••:•••••••••••••••••.••••••••••••••••••••••••••••••:•:•:•:•:.:•:•:•:•:.:
"'" F ~" " i ..... i .............~........ ~..... ~..... "i ~'" """~...."'" ~. . . . i . . . . .'" ~'~/" . . . . , . . . . . i ~'" ~i ~~''"'~,~~. . . . . . .
0 40 80 200 400 600
Distance from ice sheet centre (km)
Figure 24.14. The principle of glacioisostasy, showing the depression of the lithosphere below an ice
sheet. (Modified from Benn and Evans (1998); reproduced by permission).
Figure 24.15. Regional isobase maps. (a) Isobase map showing shoreline emergence in eastern Canada
since c. 6000 yr BP. Note the evidence for multiple domes. (b) Patterns of absolute uplift in Scandinavia
during the Holocene. (c) Generalized isobases for the Main Postglacial Shoreline (7000 to 6000 yr BP) in
Scotland. (Andrews, 1970; M6rner, 1980, in: Benn and Evans (1998), Figure 1.40).
These data obtained from the studies of glacioisostasy have been used to obtain
information on the viscous nature of the mantle. Two models have been proposed; the first
indicates that the viscosity of the mantle is relatively constant with depth, and the second
suggests an increment of the viscosity on the top of the upper mantle. Values of the second
model fit well with the loss of load of the Laurentide ice sheet (Peltier and Andrews, 1983)
and for Fennoscandia, the first model fits better (M6rner, 1980). This implies that the
processes of the icecap construction and thaw are associated with plastic deformation
within the asthenosphere. As a consequence of the increments and decreases of the ice
masses existing on the continents, important variations of sea level are produced, called
glacioeustatic changes. These fluctuations are recorded by means of the study of isotopes
of oxygen in foraminifera from deep sea cores (Figure 24.3).
Other mechanisms coexist with glacioeustasy: the tectonic factors related to the
capacity of the oceans; isostasy; hydrostasy, which represents the response of the litho-
sphere to the load of water in the oceans, and the geoidal effect. The most important are
the glacial and tectonic factors, although during upper Quaternary, the glacioeustatic
variations were more significant. Without doubt the reconstruction of the ancient sea levels
for some areas is quite complex, because several factors are involved.
Fairbanks (1989) carried out a study on the corals of the Barbados Islands. Corals die
when they are exposed above sea level. If the sea level rises, new corals grow on older
ones. Thus, by means of radiometric dating, it is possible to know the date of the corals,
and so, a record of the variations of the sea levels in the postglacial period. This record,
together with that carried out for the Indian Ocean (Figure 24.3), can represent the
fluctuations of the sea level for a great part of the Earth' s surface, because both of them are
similar, but cannot be applied in areas close to ice sheets (Boulton, 1993).
The importance of eustasy and isostasy depends on the location in relation to the
icecaps. At the boundaries of the caps of ice, isostasy is the principal effect. According
to the characteristics of the postglacial contour lines, the surface of the Earth has been
divided into six sea level zones (Clark et al., 1978, 1980) (Figure 24.16). Zone 1 is
located on the ice sheet margins, experiencing constant uplift due to isostatic recovery.
Zone 1 to 2 constitutes a transition area, with an initial rise and a later subsidence due
to the marginal rebound of the ice sheet. Zone 2 is far from ice sheets and sea level
continuously rises, as in Zone 3, but an emerging beach 1000 years ago formed the
latter. Zone 4 corresponds to tropical areas where sea has been rising since the
glaciation, while in Zone 5 a rise occurred when the large ice sheets melted and then
surfaced due to a hydrostatic effect. Zone 6 includes all the continental platforms,
except those of Zone 2, and is characterized by a slight rise after the thaw, as a result
of the isostatic uplift of the coastal margins. These sea level lines can be affected by
tectonic movements.
6. Reconstruction of periglacial environments
Relict periglacial landforms play an important role in the reconstruction of the climates
of the upper Quaternary. The degree of modification of the landscape depends on the
location and persistence of the periglacial areas, which are more pronounced in the
proximity of the ice sheets. Lithology has a considerable influence on the preservation
100
m
O, , , , J I , , , -80 ............... -,
18 1 0 0 0 y r B R 0 18 0 18 0
0 .... 4 0 5
0
-30
-80 .... ,--~ ......--, - 2 0 I -, ..... 1 -5 ,
18 0 10 0 10 0 10 0
Figure 24.16. Distribution of sea-level zones and typical relative sea-level curves. (In Clark et al. 1978).
of periglacial landforms. There are rocks very susceptible to gelifraction, whereas the
periglacial processes easily produce loose materials. These processes vary according to
the climate that can be continental, as the one in European Russia, or marine, such as
the Iceland climate. The paleoclimatic reconstructions should depend on the actual
formational characteristics, but this is not always valid, because there are many
differences between the periglacial environments in the Pleistocene of the intermediate
latitudes and the ones existing nowadays at high latitudes (French, 1996). The principal
reason is due to the insolation. Both latitudes present summers and winters, but a great
contrast and speed exists in the change between both seasons. The day-night rhythm is
the main difference (Arctic day or night). On the other hand, in the Pleistocene the
daytime insolation was higher in intermediate latitudes than nowadays in high latitudes,
so gelifraction, frost creep, and ice wedge action probably were several times more
intense than in the actual latitudes. Also, in the intermediate latitudes, the effect of the
orientation with regard to the sun is very important, so on slopes facing the sun many
more ice-thaw cycles occur than on the shaded side. On the other hand, during the
Pleistocene the wind action was more significant at intermediate latitudes than at high
ones, because large gradients existed at the boundaries of the ice sheets and vast plains
of till and sandar, which provide fine sediments, were developed. Finally, the fluvial
activity was more intense in intermediate latitudes due to a higher number of melting days.
In short, periglacial environments of intermediate latitudes during the Pleistocene period
were fundamentally different from the continental high latitudes, as pointed out above.
Therefore, the principal problem is to find precise indicators of past periglacial
conditions. Perhaps the most realistic and best trace is the presence of wedges of ice and
sand, which indicate permafrost. Landforms inherited from the thermokarst of the pingos
can be also considered as indicators of permafrost, as well as fossil rock glaciers. The
rest of the periglacial paleoforms are uncertain indicators of periglacial conditions
(French, 1996).
7. Relict periglacial landforms
French (1996) differentiated between the features related to the development of ancient
frozen soils in intermediate latitudes and those related to the existence of landforms
originated by the action of the frost. The identification of the latter depends on the
interpretation, whereas the features denoting ancient permafrost cannot be doubted.
Fossil wedges are relatively common and constitute the most important features for
recognizing ancient permafrost. They can constitute ice wedge casts filled by sediments
(Figure 7.4), which commonly form polygonal systems and sand wedges (Figure 7.12).
It is estimated that winter temperatures on the surface of active ice wedges vary between
- 15 and - 2 0 ~ (P6w6, 1983). For this author, the mean annual temperature of the air
to produce ice wedge ranges from - 6 to -8~ In central and western Europe they are
abundant in a wide band and the width and depth increase toward the south of the ice sheet
margin. This is understood as a intensification of the periglacial conditions, which seem to
be more severe (Poser, 1948). In North America wedge casts are located in the proximity
of the Laurentide ice sheet in a relatively narrow band, ranging in their age from 15,000 to
20,000 yr BP, according to several authors.
Ancient pingos can be recognized by a ramp, generally ring-shaped, that includes an
interior depression. This ramp is the difference from other depressions generated by the
thermokarst, as with the palsas. The height of the ramps depends on the size of the initial
pingo and varies between 0.5 and 5 m. The diameter of the depression ranges from
200 to 300 m. Remains of pingos occur mainly in western Europe and North America.
A classic example is the one studied by Pissart (1965) in the Ardennes of Belgium
(Figure 24.17). The irregular form and the interference among pingos show a complex
distribution, which is interpreted as pingos originating within an open system. The
date of most pingos in Europe and North America is very late glacial, approximately
10,000 to 12,000 yr B P.
Fossil rock glaciers, which can be located below the limit of present-day arboreous
vegetation, indicate a periglacial climate with permafrost in their origin. This reveals
the importance they have as paleoclimate indicators (Barsch, 1988).
Relict morphologies generated by frost are numerous, but only a few provide
unequivocal evidences of periglacial conditions. At intermediate latitudes commonly
( ~ Pingo ramparts ,'.'.'.', Probablepingo ramparts ~ Contours(m)
Figure 24.17. Map of the remaining pingos in Hautes Fagnes Moor, Ardennes, close to Malmedy, eastern
Belgium (Pissart, 1965).
there are nonharmonically deformed soft sediments, termed cryoturbations or involutions.

They have been considered indicators of permafrost by some authors, but these defor-
mations are azonal and can be generated by different processes (see the section of
patterned ground in Chapter 8). Periglacial involutions are formed by thrust of the frost or
by thermokarst.
The fields, slopes, and streams of rocks (blockstreams), as well as the screes or talus,
constitute landforms generated in high areas by the action of the gelifraction. The larger
the number of freeze-thaw cycles, the more important the development of these landforms
is. In rock accumulations, the activity or inactivity of these landforms must be considered.
Lichen colonisation, as in the slopes and river of blocks of the Tremedal massif (Guti6rrez
and Pefia, 1977), indicate that a deposit is inactive.
Cryoplanation terraces, generated by snowy processes, are common on high European
mountains and, above all, in the central region of Alaska (P6w6, 1983). Periglacial tors and
the related cryoplanation terraces do not clearly indicate permafrost conditions (French,
1996), although some authors think the contrary (Dawson, 1992).
The grkzes litdes and the protalus ramparts are related to seasonal accumulations of
snow, according to most of the authors. Finally, solifluction deposits are poorly classified,
and are very difficult to differentiate from those originating in warmer climates.
8. Climate fluctuations in periglacial regions during the late quaternary
Most of the periglacial regions are over permafrost. It disappears when summer warming
exceeds cooling in winter, and vice versa. At present, the periglacial environments occupy
25% of the Earth's surface and during the last glaciation could reach 20 to 25 per cent or
more (French, 1996).
Possibly, the most relevant criteria to estimate the periglacial area during the late
Quaternary are the distribution of plants and the extension of ecozones. All this implies
that there were tundra, steppe and forest fluctuations at this time. Therefore, palynological
research is crucial to these effects and has made possible to establish pollen regions and
climate periods (French, 1996). Also, faunal remains, such as molluscs and insects, help
with the interpretation of the environmental conditions. Nevertheless, the studies leading
to this are scarce.
Permafrost evolves very slowly with respect to warming and cooling stages. In the late
Pleistocene and Holocene the thaw of the ice sheets and the subsequent sea level rise are
processes much more rapid than the loss of thickness of the permafrost (Dawson, 1992).
For example, the permafrost in northeastern Siberia has always been present since the late
Pleistocene (Washburn, 1979).
The most detailed knowledge that exists on the climate and vegetation fluctuations that
have occurred during the late Quaternary exists in Europe. Btidel (1951) differentiated
three vegetation zones in Europe (tundra, steppe, and forest) and indicated that the
majority of the European surface to the south of the ice sheets was affected by periglacial
conditions, except for the southern Iberian Peninsula and the Mediterranean coasts.
Nevertheless, nowadays it is calculated that the extension of the periglacial band in central
and western Europe was about 50 km during the maximum of the last glaciation
(Weichselian). This width has been obtained mainly by the analysis of ice and sand
wedges and annular ramps of pingos. The most important climate record corresponds to
the pollen diagram of Grand Pile (Mook and Woillard, 1982), which shows the first stage
of the development of the permafrost was about 90,000 years ago. Later, other permafrost
periods, separated by warm interstadials, follow each other.
In European and Asiatic Russia an important increase of permafrost occurred during
the late Quaternary, reaching up to 700 m thick. The continuous permafrost was located
up to 50~ having in total a dimension of 400 to 600 km (Figure 24.18) (Baulin and
Danilova, 1984). The evolution of the permafrost is similar to that in Europe, with growing
periods that coincide with cool stages, alternating with warmer periods.
Several data sets reveal that in North America there was a relict permafrost
corresponding to the last glacial maximum (upper Wisconsinan), because for previous
periods the information is very scarce. The permafrost band to the south of the Laurentide
ice sheet is narrower than around the Euroasiatic ice sheet. This can be explained by a
faster advance of the Laurentide ice sheet toward more southern latitudes, and therefore,
areas with intense periglacial conditions are more limited than in Europe. In Figure 24.19
the areas of permafrost are indicated, together with the landforms related to fossil
permafrost and other inactive periglacial morphologies that do not require permafrost
conditions. On the other hand, at present the permafrost occurs under the 82% of the
surface of Alaska and northwestern Canada (P6w6, 1983). On the Alaskan coast, many
pingos corresponding to a closed system occur, whereas in central Alaska pingos of
9
. ,..~
~ ,.~
9
>
9
. ,.~
. ,.~
~D
+.a
9
. ,..~
+.a
~ ,.~
0
9
,4
r
640
Climatic Geomorphology
9
9
dD
r~
9
(D
~ ,..~
d::
0
0
<
o ,..~
0
. ,..~
~ ,..~
0
o ,..~
=
0
3:
~D
9
+.a
~D
~D
. ,.~
r
the open system are abundant; all of them were generated in the Holocene. Also the vast
cryoplanation terraces in the centre of the territory are significant.
It seems that in the Southern Hemisphere the permafrost did not develop so much,
because the described morphologies do not need permafrost conditions. Only ice-wedge
casts in southern Patagonia, Argentina, have been mentioned (Grosso and Corte, 1989).
9. Some considerations on global climate change in periglacial areas
The predicted temperature increase due to the greenhouse effect will seriously affect
permafrost areas, which at present occupies approximately the 25% of the Earth' s surface,
as mentioned above. It is reasonable to state that permafrost will decrease in depth and
extension. An important increase of the active layer will imply substantial modifications
to the hydrologic, geomorphologic and biologic processes of the permafrost regions.
Moreover subsidence problems, affecting the infrastructure will happen. By utilizing
general circulation models (GCMs) and a scientific paleoreconstruction, assuming an
increment in temperature of 2~ it can be concluded that the continuous permafrost
will reduce its extension 29 to 67% during the 21st century in the Northern Hemisphere
(Anisimov and Nelson, 1996). Permafrost areas are the most susceptible to global
warming (Boer and de Groot, 1990; Smith, 1993). Nevertheless, the results obtained from
GCMs do not provide precise forecasts, but indicate the trend of future changes in
permafrost regions (Pissart, 1990).
Several studies indicate that permafrost temperatures have been rising in the last 20
to 30 years. Temperature records taken between 1970 and 1990 in arctic Europe and
subarctic Russia, show an increment of 0.6 to 0.7~ at 3 m depth (Pavlov, 1994). Similar
values have been obtained in Tibet and in North America. The warming is interpreted, as
in other areas of the planet, as the result of the global climate change or due to a natural
trend of climate warming.
Chapter 25
Climate change in arid and tropical humid regions
Rainforest, savannah, and semiarid areas, although presenting different morphogenetic

processes and being studied separately, have been affected by important climate changes
during the Cenozoic period, as is shown by the anomalous paleoforms existing in the
different environments. Thus, for example, aeolian deposits in the Congo rainforest or
laterites in the arid zones in southern Niger Republic, bear witness to these important
climate fluctuations.
The purpose of this chapter is to show the evolution of these regions during late Tertiary
and Quaternary, from the analysis of erosive and depositional landforms. Dating results
(isotopic, palaeomagnetic, thermoluminescence, and others) make it possible to elaborate
a sequence of events, adequately time-ranked, which provide the means for correlations
between different areas. From these, an approximation can be obtained of the climate
changes that occurred in the recent history of the Earth. Furthermore, it is possible to
provide data and valuable interpretations for the prediction of future climate (Petit-Maire,
1999).
1. Arid regions
The study of the landforms in arid regions that are tectonically stable and without
significant anthropic activity, reflect that many of the existing landscapes have been
generated under geomorphologic conditions that differ from the current ones. Therefore,
those landforms should be considered as paleoforms or inherited landforms, originating
under morphoclimatic conditions different from the present ones. For instance, the
extensive dune fields of the Sahelian zone, covered by grassland vegetation indicate
a higher aridity during the time of formation. Similarly, no water currently flows in the
complex network of ouads (wadis) in the central Sahara, indicating periods of higher
precipitation in the past. (Tricart, 1969, 1979). These landscapes show significant
modifications to the distribution of the arid regions over the surface of the planet.
From this evidence came the concept of the pluvial period or lacustrine phase, that can
be defined as that in which hydrological and biogeographic conditions were more humid
than present conditions. In comparison to the one before, an interpluvial or arid period
occurred under dryer conditions (Rognon, 1980). This definition has taken into account
water availability, which not only is conditioned by the precipitation, but also by
evapotranspiration. Thus, for a given precipitation, a decrease in temperature increases the
water availability of soils. Therefore, this double factor of precipitation and temperature
makes more difficult a definition of pluvial period. The alternation of pluvial and
interpluvial periods can be recognized for deserts by a collection of indirect data of diverse
nature: geomorphologic, archaeologic, pedologic, palinologic, hydrologic, and hydro-
geologic (Tricart, 1969, 1979; Thomas, 1977c; Demangeot, 1981; Goudie, 1992).
1.1. Fluvial systems
Although precipitation is scarce in arid environments, stream flow paradoxically plays an

important role in the construction of the landforms of these areas. This is due to the lack of
vegetation cover and to the shortage of soils, so that the geological substrate is unprotected
and exposed to rain drops. Precipitation, although scarce, is frequently very intense and
with an irregular temporal distribution. Thus, in Saharan semiarid areas there is a flooding
per year in average terms, but in more arid regions up to 10 years can happen without any
water circulation in the ouads.
Many authors have dealt with the influence of climate change on the modification of the
geomorphic processes in fluvial systems. In Schumm (1977) and Knox (1984) the direct
and indirect effects of climate on these systems are presented in detail. In arid regions
where vegetal cover is scarce, the variations of the vegetation, as a result of climate
change, can produce significant modifications to the hydrology. Knox (1984) established a
critical threshold in the vegetation cover that can be of 70 %, corresponding to an annual
average precipitation of 400 to 500 mm. For values lower than 70 %, the quantity of soil
subject to erosion increases very rapidly. These circumstances can be clearly seen in the
curve (Figure 25.1) of Langbein and Schumm (1958), wherein annual average
precipitation variations are presented relative to annual sediment yield. In deserts, the
higher the precipitation, the greater the runoff and the erosion increases to reach a
maximum of about 300 mm in precipitation. Where this precipitation increases the
vegetation cover becomes more extensive and the sediment yield decreases rapidly.
350
E0
m
.m
~ 300
,_ 250
t~
"
t- 200
9 15o
~,
t-"
tl)
E
~5 100
m
= 50
c
r t" D e s e r t t" . . . . Grassland .... k . . . . . . . . . . Forest . . . . . . . . . . .
0 i A J , 9 , = ,
0 200 400 600 800 1000 1200 400

14
Mean annual (effectiveprecepitation)in millimeters
Figure 25.1. Variationof sediment yield with climate in fluvial systems (Langbein and Schumm, 1958).
Climate change in arid and tropical humid regions 645
1000 . . . .
~- I I I I I I
9 ~ tO r I~
I I I I
g~ 800
o 600 I
"o
"~,
400
~ 2oo
o l l, l l l l
0 10 20 30 40 50 60 70
Mean annual precipitation (inches)
Figure 25.2. The effect of temperature on the relation between mean annual sediment yield and mean
annual precipitation (Schumm, 1965).
Meanwhile in very extreme deserts there is little water erosion, but in semiarid regions the
stream flow is very frequent due to stormy precipitation, and the sediment yields reach
higher values. This graph shows also the importance of vegetation cover as the most
effective factor in reducing stream flow and soil erosion. But these maxima of sediment
yield vary depending upon the temperature (Schumm, 1965) (Figure 25.2), in such a form
that the higher the temperature, the erosion maxima coincide with higher precipitation.
That is, for higher temperatures, with a higher evapotranspiration, more precipitation is
needed to maintain the percentage of the vegetal cover, there is a lesser stream flow, and
the sediment yield maximum moves to the right.
In semiarid regions, which correspond to areas where great erosion is produced,
flooding is also much more frequent, because the water from precipitation flows rapidly. In
basins with different drainage areas (Figure 25.3), the magnitude of the flooding increases
significantly as average annual rainfall decreases from 500 mm to approximately 375 mm
(Knox, 1972). This indicates that a climate change from relatively wet to dry will produce
a widening of the channel to adapt to bigger floods.
Similarly, the density of the drainage network tends to increase in semiarid regions and
normally decrease with the increment in precipitation (Gregory, 1976) (Figure 25.4). The
graph indicates a sharp decrease in density from 800 mm precipitation. Obviously, there
are variations as a result of the different types of rocks, relief and vegetation.
All these examples illustrate the sensitivity of the landscape of these areas to climate
change. In arid regions, the vegetation is more sensitive to climate and a slight
modification in the vegetation cover can produce significant hydrologic changes.
Therefore, in these regions the intensity of erosion can vary very much as a result of
slight climate changes.
20 ,
I ...... I ........ I 1 ""
O")
tO
100 sq.mi.
o 15
"O
c"
00
O
0 sq.mi. ~
"0
0
0
0
,~, 5
U_
0 . . . . . . ! . . . . . . l l T T . . . .
10 15 20 25 30 35
Mean annual precipitation (inches)
Figure 25.3. Relation between flood magnitude and mean annual precipitation for three sizes of drainage
basins (after Knox (1972)).
25
- 35
20 - - 30
E
. .~?-
E 15
"-.4
>,
CO
O
a
;
O
10
e-
Humid ~
Tropical / " Tropical//it" 10
l-lumlcl,/
r----------r~---'---T T 0
0 500 1000 1500 2000 2500 2700
Mean Annual Precipitation (mm)
Figure 25.4. Relation between drainage density and mean annual precipitation (modified from Gregory
(1976), in Chorley et al. (1984), Figure 18.5).
The oscillations in temperatures and precipitation that occurred during the Holocene
(the last 10,000 years) have been of low magnitude (+ 2~ and 10-20% in precipitation)
(Lamb, 1977). Even so, these variations have affected the intensity of many
geomorphologic processes, which control the morphologic stability of the landscapes,
especially in fluvial systems (Knox, 1984).
1.1.1. Hillslopes and their evolution
The best typical profile of hillslopes in arid regions can be recognized in sequences of
stratified rocks. This profile consists of a free face at the top, a slope covered totally or
partially by detritus in the middle, and a basal concavity from where a pediment develops
(King, 1962).
In a study of the hillslopes of a great part of the territory of the USA, the influence of the
climate on the evolution and morphology of the profiles is stated (Toy, 1977). In order to
minimize the influence of the geology and the structure, the author carried out the analyses
on marine claystones with an angle of dip of less than 5~ According to this work, slopes in
arid zones tend to be shorter, steeper and with a smaller radius of curvature than in humid
regions (Figure 25.5).
In some places, as in the Judean desert of Israel, hillslopes developed on limestone and
marls of the upper Cretaceous, present smooth profiles, with concave-convex shapes, and
are covered by a thin layer of detritus (Figure 25.6). This morphology indicates an
environment more humid than the existing one (260 mm), when the morphology was
generated. The higher precipitation is corroborated by the development of hohlkarren,
kluftkarren (Figure 11.11) and kamenitzas on Cretaceous limestones.
1.1.2. Triangular slope facets
The knowledge of the processes affecting slopes in arid and semiarid environments has
progressed considerably in the last few decades. Nevertheless, works related to their
evolution are scarce, due to the difficulty in dating the different relict hillslopes. In
semiarid environments where there is a higher percentage of vegetal cover and there has
Arid . . . . . . . . . . . . . . . . . . . . . . ~ .......... ; Humid
Figure 25.5. Conceptualizedhillslope profiles (Toy, 1977).

Figure 25.6. Regularizedslopes on limestones and marls of the upper Cretaceous, outcropping as a result
of the incision by gullies. The convex-concave profile seems to indicate a previous stage more humid than
present (260 mm). Mishor Adumim, Judean Desert, Israel.
been an old human presence, however, it is possible to find carbonaceous remains and
archaeological materials within the deposits, making dating possible.
Slope accumulations can be later affected by a period of incision that can destroy all the
accumulation, or can generate relict slopes termed talus flatirons (Koons, 1955). They are
also known as tripartite slopes (Gossmann, 1976) and triangular slope facets (Btidel,
1982), although this latter term can be confused with the terminology of eroded fault
scarps. These morphologies present a triangular or trapezoidal ground plan and are located
around and separated from the scarp (Figure 14.15 and Figure 25.7). Therefore, the
generation of the flatirons requires the alternation of accumulative and erosive periods
during the slope evolution. The development of several of these stages can produce a talus
flatirons sequence, with the most recent flatirons being those which are close to the scarp.
These relict landforms are not very common due to the fact that they are located on
stratified formations with horizontal disposition, in which the cornice overlies easily
eroded material (Koons, 1955; Everard, 1963; Btidel, 1970; Blume and Barth, 1972;
Figure 25.7. Schematiccross section of the San Pablo Mesa showing a flatiron sequence with five stages
of slope evolution.
Gerson, 1982; Gerson and Grossman, 1987; Schmidt, 1989a,b, 1994; Arauzo et al.,
1996a). Furthermore, the free face has to be thin to avoid an excessive accumulation of
detritus (Schmidt, 1987), allowing the rapid destruction of the cornice and retreat of the
slope. On the other hand, the alternation of hard and weak layers below the free face makes
difficult the formation of talus flatirons.
Once the factors favouring talus flatirons are indicated, it is necessary to know their
causes. Koons (1955) interpreted the features from the discontinuous balance between
sediment yield and removal in the slopes (Figure 14.16). This noncyclic origin is against
the interpretation based on the modification of the dominant morphogenetic processes due
to climate changes (Everard, 1963; Gerson, 1982; Gerson and Grossman, 1987; Sancho
et al., 1988, Schmidt, 1989, 1994; Arauzo et al., 1996a,b; Guti6rrez et al., 1998a,b) (Figure
25.8). In recent periods, the possible influence of anthropic action on the modification of
the geomorphic processes has also been pointed out (Everard, 1963; Sancho et al., 1988;
Guti~rrez and Pefia, 1989, 1992; Arauzo et al., 1996a,b; Guti~rrez et al., 1998a,b).
Figure 25.8. A generalized model of the effects of major environmental changes on escarpment retreat.
D -- dry, arid to extremely arid, climate; W = relatively wet, moderately arid to semiarid, climate (after
Gerson, 1982).
The importance of climate changes has been discussed at great length for several types
of landscapes. In desert regions with scarce land cover, the variations of vegetation can
produce important modifications of the hydrologic conditions. Climate characteristics in
the centre of the Ebro Depression in Spain agree with maximum sediment yields of the
curve proposed by Schumm (1965). According to this graph, small variations of
temperature and annual precipitation can modify the percentage of land cover and, so, the
sediment yield.
On the other hand, in periods with more humidity (pluvial) the processes that generate
sediments are more active on the free face, mainly undercutting and slumping. This is due
to the increase of infiltration and groundwater flow (Ahnert, 1960). When a decrease in the
temperature occurs, the number of frost days per year raises and, therefore, the sediment
yield in the scarp will be higher due to frost action.
The climatic interpretation of alluviation and trenching stages leads to the establish-
ment of a correlation with known climate sequences. The two most recent accumulation
stages are well represented in all the slopes in northern Spain (Guti6rrez and Pefia, 1989,
1992, 1998) and have been dated in many places by means of archaeological and
radiometric techniques. The most recent, post-medieval accumulation can be correlated
with the Little Ice Age (1450 to 1850) (Grove, 1988). The accumulation corresponding to
the stage previous to the latter was formed during the Cold Period of the Iron Age (2900 to
2300 BP) (Lamb, 1977; Gribbin and Lamb, 1978; Grove, 1979; Harding, 1982). Guti6rrez
and Pefia (1989, 1992, 1998) correlated these two accumulation stages with those cold
periods for the northeastern Iberian Peninsula (Figure 25.9).
At present there are only two dates for accumulations older than the above mentioned,
both obtained by means of C-14 at Mezalocha and Chalamera in Zaragoza province and at
Monteagudo, Sofia province in Spain (Guti6rrez et al., 1998a; Guti6rrez and Mart/nez,
2001). In principle, there is a good correlation with the Heinrich events established from
the analysis of deep-sea cores in the North Atlantic (Guti~rrez et al., 2005). Ages obtained
for the accumulations could correspond to H3 (27 ka BP) and H4 (35 ka BP) (Heinrich,
1988; Bond et al., 1993, 1997). A Heinrich event is represented by glacial sediments
coming from the melting of icebergs with scarce foraminifera. The stages of formation of
these events are related to movements of a great abundance of icebergs, resulting from the
detachment from northern ice sheets. These circumstances produce a widespread cooling
of the superficial waters of the ocean. As a result, cool winds from the Atlantic came into
the Ebro Depression. This fall in temperature produced a decrease in evapotranspiration
and therefore, an increment in the availability of water in soils, causing an increase in land
cover. These conditions seem to have been suitable for the accumulation of sediments on
slopes. These Heinrich events occur in periods of extreme atmospheric cooling and are
followed by net and short interstadials (Dansgaard et al., 1993).
From the above, it may be deduced that accumulation stages coincide with cool-climate
epochs, whereas the stages of trenching correspond with warm ones in which talus
flatirons are formed. Therefore, the sequences of talus flatirons record climate changes.
In semiarid environments, those changes in slope systems that generate talus flatirons,
are produced in relatively small time spans. In dryer areas (Figure 25.10), however,
climatic changes on the order of 100 kyr are needed to generate these relict slopes (Gerson,
1982; Gerson and Grossman, 1987; Bull, 1991; Schmidt, 1994, 1996).
Iron Age Little Ice Age

Colder
and wetter
Secondary Optimui
Warmer
and drier
3.000 2.000 1.000 0 BP
Post-Bronze Post-Medieval
accumulation
incision
B.C. 1.500 1.000 500 0 500 1 . 0 0 0 1.500 2.000 A.D.
Figure 25.9. Relation between accumulation and erosion stages in slope deposits in NE Spain and
climatic epochs of the Upper Holocene (Guti~rrez and Pefia, 1989).
Current available chronological data are very scarce and it is necessary to have many
more dates available to validate this climate origin. If the research that is being carried out
corroborates this hypothesis, it will be possible to obtain a climate curve for this region for
the late Pleistocene and Holocene periods. On the other hand, given the position of the
Iberian Peninsula between Europe and Africa, these works can be very useful to correlate
it with the pluvial periods in northern Africa and climate fluctuations identified in central
and northern Europe.
1.1.3. The ramblas or arroyos
Rambla is a term used mainly to name ephemeral fluvial courses in the semiarid areas of
the Mediterranean zone of Spain. These courses present a flat-floored channel and
vertically cut banks, and are only active when heavy rains occur, producing floods that
transport a large amount of sediment load (Segura, 1990). Arroyo is the word introduced
by the Spanish colonizers of the deserts of the western USA. The term is applied to courses
Figure 25.10. Hillslope triangular facets sequence. Makht6sh Ramon, Negev Desert, Israel.
from 5 to 200 km with deep transverse cross sections and flat floors. They are intermittent
courses that can be affected by sporadic storms, incising into unconsolidated material
(American Geological Institute, 1972; Bull, 1977). In these alluvial systems it is possible
to differentiate several stages of alluviation and trenching (Figure 25.11 and Figure 25.12).
Furthermore, the reader will no doubt recognize the strong similarity or even identity of
these features with the ouads, wadis, or nullahs of much of North Africa, the Middle East,
and southwest Asia, respectively.
North American deserts constitute the region in which these geomorphologic arroyo
systems have been more deeply studied. The most advanced knowledge in the fields of
hydrology, paleohydrology, geomorphology, and climate changes come from there.
Sedimentation of valley fills by ephemeral intermittent courses exists with precipitation
ranges from 100 to 500 mm (Bull, 1977). Temperature regulates discharge, because it
controls the evapotranspiration and the type and density of vegetation. These effects are
more pronounced in arid regions than in subhumid and humid ones (Langbein, 1949).
Sediments in these ephemeral courses come mainly from sand, silt, and clay supply from
slope erosion. Later, these particles are easily mobilized, except clay layers, in which
infiltration decreases and cohesion rises.
On the other hand, anthropic action is important, because humans have modified land
cover in many ways; overgrazing, excessive construction, deforestation, fires, and so forth.
The increment in biomass affects the resistance to erosion of the channel (Graf, 1979). In
this sense, Cooke and Reeves (1976) showed that the introduction of herbivorous animals
into New Mexico and Arizona triggered an important increment in the erosion of soils.
In ramblas or arroyos several stepped levels can be recognized (Figure 13.36, within the
sediments filling these landforms, which result from different stages of aggradation and
C l i m a t e c h a n g e in a r i d a n d t r o p i c a l h u m i d r e g i o n s 653
7.~.5,,~, . . . ~ ~ , - , ' ~ - . ' : 7 . . . - . ~ . ' , ' ~ ~ " - ~ . - - r _ - ..... ~ - - - . . - , - - - . - ~ , . . . . . . . . . -...--.-.,.-'~...--."~.

:',;"s." .'>"'Z':@?~r,' ' . :'.,e<.;;..,;;-.-'-.,':-~'.~>'-...::-?
.-'...~ . . . . . . . "Z...- .... ~,-. - . s "- ; ~ ' - -.".,,:~,':-~,'~" --
~., ~ : ~ : ". " 99 - "..- -";.6'~"
v - . . ,.-a" .
'= " .... "~-'.'..~.~,
. . a 9 .~, . . . .
~.~"~'7... ,,'"
~.: ~,,- r
5";"':' ":.~:~':'.'~;"'
' . . . . : . . . . '~ . L , . . . ' . - : .
...... ,~,. ..... .~ -~.,,;.-. - ~\..,,~:..,;%.. ~ ~ ,, ~ . : . . . . . 7~....,,~ ...,....-~ ..... ~,:...:.... ....
.9-.". . . .- . .
--" -
i-~_~.'_,~.~ . . . . . . ~:~[~_.r ...... - . , ~ , ; . : ~ ; , ~ : ~ ~
-~7~.a-~',.%"~.~.*,:;. 9.. ?" , , . ' l ? ~ - , c p ' . , - - . " . ;~'2b"~ff~i'~',~,,r~:~:,, ~,~.~ , . . ~ : .... ~ . _ . r ".- .~':,~.",~,'_
..... . . ~ , : , ~
":.,::~- -' .;
'. , ..... : . ~ . . , , , - . ~ . . -
. -,. _ , ~ . ' ~ . ; " ... ..-..m; ; ~~),'~r " - " ";:: ; I ~ , : , ~".~, "" " . . . . ~. _ ~ I ~ "9 9~ .,.,,~
9 . : - ..- . -., -. ~. -. - .: -. ~. ' . _ , -,,~-...~..'r~_':~-.
..... . . . . ".. -,:a. ~ . " ~ , y ~ , ~,R~,.',",,
; ~.[ . , ,-,
.'. .- .-....-"
. 9 .....,.:'.",. . ' . r . ~ .~,:'~1~1~ ~ '
,..
"v~r-.."
'..~:'--'-..~:,.
, -.,,.-.
,"
. . . ., :
--:" 7 : 7 '--...%r'.'.,:. :. "-~:2.~' ~.~:;~V/,g~., ; '-:'~.~ '" .- ...... " ? : ~ . - . - . ~ ' . ' , ' . ' ~ ; : . ' ~ . ' . & . ~ " "'. :,.:,~.:: ..... :i
~.
-. " "
" " - " ""."..:/":',v.
.',.:.'.'?',
, : ~ ',"o~"..'-~;- ".a. ,' ",'"-~ ;,,"
. . . . . ~..,.6.':.'.':.".:.~.:,.,,"',., ,- :.~'~;~~.
':"" "' ~ ~..,,.,.
". . .~. -. .' ~. ." : :"' ' , ' " * " " .... ... ...s'.~,~"~'7
. ~ a . . ~ , ' r ~ - - . . , ""
~ , ' .,_~,,,..~,-~:,,- . .~. . ,..
." . . . . " ' e ' ' , : ; . , " , ' ~ ' , . " . / / ~ _ % % .. 9 ." ' - _ ~ ' ~ , ' ~ 9 ~ . ' g / ' : ~ ' K ~ ' . ~ i ~ r - 7 _ ' ~ " \'~""" .,."~.'" 9 ' ldk .'4. "e; ,'.~.; ~ ' : ."
. ." .- '.". . : ' ; '9".," ' ..... ,~., :~z /...," .x "<. . ."~. . .". " " ~ e ' ~ "~ & ~ , ~ " .....
"'"'"~"/:':';'';" ~''
~ '" .-.-., ,.' : , Z 2 ..".,. ,:,':.'.
.... ,.,' "-. #,.:~-: ~ "
... . ~.~,,;',
" "**',,,,',,:,,:,, , " , . :.
~, ,~..,.....,.,.
-.~-,,.;.,..,.,,.,.,..z~.~,. ,.,~..,.. . . . . . t- 9 ~.~.,.,;: . ~,-.a~.,...'.,.-:v: .... % ~ .. 9 ' ,,-"..>l.~'."x ,, , .~
. ' K . . , , " . , '~~ ;"',~' ," ~ ~', ,.,'..
9 . " ~ - ~ '-':"..~.~'~ ' ~ . ; - .'l',' .' '..'~...-.,~.~
' ' ~ ~'_~,4
'~ -." " . ~a"~'~" , . 4 ." ' .'." 'I. ~,'.,
, ' ~ ,~,..
,: ,'~. .. '. .. t. *. ..., " x : -
."..* ,..,
.... .',,.
- ~,' " ,'~ " " . ' " ~ , ", 'q'/.;~'9~, ,,.' 'V ~ " " ~ : ~ " ~;>~,':/. V q.7. : t: ...... ' ' ~"5: 4', ,~ ' : . ' ~' " . ~ '
.9' . , .".~.."-. -, ..,: ,' , . ..'~'-q'_

~ .,,,,
:.1~. ~ , ~ ,.. ~.~
." . ,r, . .j.:~;_
. ,,.',r , . : "~.~ ,.~t. . . . ~:., ., ~ . .. . . . , ~ ' ,. , ., .,.,.t
? , ~ . . . :"5';' "~~. " ." . . . . . ' . . " '." "" ' "" . ..
.:,..'(.~~',;~;.:.. ' ....... . .." .,,:~W" ,;~,,.=-,.,,,~'~'~~~;"h;~,:~"~--"~'~ . '
4, 6. , .,, ~ . . . . . . ... . _.,,..,- ,..~. :':a.. ::,"..~."~

' . 9 . ....... :.::'::: ', .'....'.,,. ',.' :"' :.:1
~;~,~.,RL~,~'. ' " ' - . - 4 ~ . . . ~ ~ . / ~ - .." ' ."" ~..':..":::','..'., '.::: ::.'".:':b:..:',.:2'.',".:::'.,"l
.,~..a~,~ - - ~ :2L,,,.....,..,,e, ;. ...' . . . . . . ." ,,,.,,..'.. , ' . , . . . . , : ..... . . ,.'.'.." ~ ....a.. : " ' v . : " ; , % .' . ' . ' " . . . ' " I
~~.....,,,..<,:7': -........ ~ , , ~ _ ,.~,'.;,,~ ' ~ ....... . .,,:.: .:...":.,:::. .. ,.: . ,- ,.'~
~m'r'V",Y,15;..,,,,~~- ~'.-~, . . . . . . . ~...-,-,',,,,,,',,,,..'r'..,,,r,a .-,..,,,:',,,~ .." ,"~,~ .,.,, ,,,;,, .'. .... '..," ....... . '~,.I
-~-"-"~'~'-"~'-'"',,"~"{. ' " ' ',"'': ""~, " r ~ " . " * " ' " " . , : , , - " ' 2 ..... ":"" " " " ' : " ".,,". ~ ",.'..."'""."'; ..... I
Figure 25.11. C o m p o s i t e s k e t c h , b a s e d o n field s k e t c h e s a n d p h o t o g r a p h s , s h o w i n g the d i f f e r e n t v a r i e t i e s

o f v a l l e y - f l o o r g u l l i e s . A c t i v e l y e r o d i n g s c a r p s are i n d i c a t e d b y d a r k e r s h a d i n g . D i s c o n t i n u o u s g u l l i e s (1)
f o r m in t r i b u t a r y v a l l e y s . T h e y w i l l c o a l e s c e w h e n the h e a d c u t at 6 r e a c h e s t h e m . In this v a l l e y t h e r e are
t w o k n i c k p o i n t s (2, 3). T h e i n c i s i o n o f the v a l l e y floor f o r m s t e r r a c e s (4, 5). T r i b u t a r y g u l l i e s e r o d e
a l l u v i u m t h a t fills t r i b u t a r y v a l l e y s (7) a n d t h e y w i l l e v e n t u a l l y i n t e g r a t e the d r a i n a g e n e t w o r k ( B r i c e ,
1966, in S c h u m m et al. ( 1 9 8 4 ) , F i g u r e 4.1).
Figure 25.12. T e r r a c e t r e n c h e d in a v a l l e y fill b y i n c i s i o n o f the V i l l a f r a n c a g u l l y or V i l l a f r a n c a de E b r o ,

Zaragoza province, Spain. Ebro Depression.
Figure 25.13. Aerialview of a dendritic drainage network consisting of flat floored valleys excavated in
Miocene gypsum and marls. Central part of the Ebro Depression of Spain.
incision. In the centre of the Ebro Depression in Spain, an intricate dendritic network of
flat valleys that appears incised, showing several levels with sediments from the Holocene,
has developed over Miocene marly gypsum material (Figure 25.13), (Pefia et al., 1993;
Arauzo and Guti~rrez, 1994). In the ephemeral courses, it is also possible to recognize
stages with a shortage of sedimentation that are recorded in incipient soil profiles.
Therefore, these ephemeral courses are characterized by lengthy periods of sedimentation,
alternating with a brief hiatus with shallow incised channels and the development of
entisols (Leopold, 1994). The aggradation processes described above are followed by a
stage of degradation that corresponds to the incision of the rambla that starts with a rapid
incision followed by an exponential decrease, due to the approximation of a new base-
level of erosion (Bull, 1997). For southern Arizona, Cooke and Reeves (1976) proposed a
model on the different causes of the beginning of the arroyo incision in which many
variables interact. Land-use changes, random variations in the frequency and magnitude of
the processes and the secular climate changes are the principals. The first and the last seem
to be the principal causes of the initiation of the arroyos. Nevertheless, the origin is a
classic example of equifinality or convergence of landforms, in which it is very difficult to
identify a unique mechanism (Cooke and Reeves, 1976).
In the analysis of the origin of trenching in arroyos it is very complicated to distinguish
between anthropic actions and secular or short-term climate changes. Another way to deal
with the problem is based on the fact that there is an increase of the erosion produced by
waters and a decrease in the resistance to erosion in the sediments of the arroyo. In any
case, vegetation seems to be the key to the initiation and development of the arroyos.
Figure 25.14. Rambla or arroyo with the channel covered mainly by shrub-like vegetation. Vertical
scarps are composed by erodible materials. Barranco Grande, Las B~irdenas Reales, Navarra, Spain.
Vegetal cover, on slopes and riversides (Figure 25.14), responds rapidly to the changes
in temperature, precipitation and anthropic actions. Among the modifications to land use
are the over-pasturage in arid zones with silty soils that produces a decrease in the vegetal
cover, intense trampling, and increments of rainsplash, with collectively a decrease in
infiltration capacity. All of these factors, together with construction works (roads,
railroads, bridges), deforestation, and fires, are the principal triggering factors of the
human actions. Two examples illustrate these anthropic triggering factors.
Thornes (1976), in a work on the Almaz~in basin of Soria, Spain, indicated that the
excessive pasturage due to the in-place livestock and to the en-route transhumance to
summer pastures in Urbion was an important factor for intense erosion in these areas, due
to the decrease of a great part of the grassland vegetation. As a result, gullies were
triggered, which incised into the Tertiary terrains that consisted of easily eroded materials.
A similar problem occurred during the colonization of the southwestern USA. Settlers
came in great numbers, building numerous farms and roads and put many thousand head of
cattle into this area. All these circumstances produced the disappearance of a great part of
the existing vegetation, triggering an intense arroyo action that generated gully trenching
of more than 50 m in depth. This erosive process occurred between 1850 and 1920 and
especially from 1870 to 1890. Settlers had to migrate to more favourable areas. This
anthropic environmental change was known from a long time ago and Cooke and Reeves
(1976) studied this in depth in an excellent monograph.
Other authors (Bryan, 1941; Antevs, 1952) held that trenching had a double cause and
pointed out that although pasturage also contributed, vegetation was weakened at the end
(a) (b) (c) (d)
,: I_ 13_,'!
", /
Figure 25.15. The cycle of trenching and alluviation in a semiarid valley. Dotted line indicates
alluviation within a channel. Solid line indicates trenching or a well-defined channel (Schumm and
Hadley, 1957).
of 19th century due to a climate change. This hypothesis is supported by dendrochro-

nologic studies, which show a period of lower precipitation (Stockton and Fritts, 1971).
Another model proposed by Cooke and Reeves (1976) is the one related to the random
variations in the frequency and magnitude of the processes. One factor that can contribute
to incision is the tectonic fall of the base level (Bull and Pearthree, 1988). Another is
related to erosion capacity, which rises exponentially with the increment in discharge
(Leopold et al., 1964). This can produce flooding when storms occur. These events
produce trenching, while in periods of less discharge alluviation is predominant. Schumm
and Hadley (1957) suggested that incision is due to internal adjustments within the fluvial
system. For small basins less than 50 km 2, the sedimentation can be produced by flow
dissipation, due to infiltration, evaporation, and so forth. These authors generalized those
changes and proposed a cycle for semiarid erosion in which the evolution of trenching and
alluviation is considered for small basins (Figure 25.15).
Minor and secular climate changes constituted the last model for trenching. For short-
term periods there is enough evidence of the fact that erosion and sedimentation result
from climate changes, as well as from human activity. Leopold (1976) put instruments for
monitoring ephemeral water courses in New Mexico for a 15-year period and deduced that
the arroyos that had been eroded during the century changed their trend toward
aggradation, from 1961 on. The author ascribed the change to the cooling after 1940.
Hereford (1984) described a similar situation in Arizona. On the other hand, the
hydrologic and geomorphologic responses to E1 Nifio oscillations can produce droughts
(Andrade and Sellers, 1988) or large floods in southern USA. It is also necessary to take
into account that this climate perturbation is persistent.
Without any doubt, during the colonization epochs in the semiarid zones that affected
the southwestern USA, South Africa, India, and Australia, the trenching of the fluvial
courses increased markedly (Cooke et al., 1993). Nevertheless, in prehistoric periods the
alternation of aggradation and incision seems to have been due to climate changes
(Antevs, 1952). Other researchers, on the basis of dendrochronologic studies with records
of thousands of years, indicated that there was an alternation of dry and wet epochs that
were related to trenching and sedimentation. Similar results show the studies carried out
by means of C-14 and archaeological remains.
Leopold (1994) carried out a synthesis of the climate history following the last
Wisconsinan ice retreat. The synthesis was based on much research carried out on the
alluvial deposits of southwestern USA. A period of aggradation can be recognized from
the last glaciation to the Altithermal Phase or Climate Optimum, some 4000 to 6000 years
ago. During dry conditions in this time period intense erosion was triggered, affecting
valley fills, which produced terraces up to 10 m above the channel. Later, another
aggradation period developed, which finished about 1200 to 1400 yr AD (Medieval Warm
Period or Secondary Climate Optimum). In this dry period an incision of the fluvial
deposits previously deposited occurred again. During the wet and cold stage
corresponding to the Little Ice Age (1450 to 1850 AD) aggradation occurred yet again.
From 1880 to 1920 there was a change toward aridity that produced erosion in the valleys.
This general proposal is based on existing data, but is not universal. This sequence of
aggradation and erosion stages can be different for tributary headwaters than for principal
channels (Balling and Wells, 1990). Furthermore, it is possible that a synchronism exists
among degradation stages within a vast region, because each fluvial system has a different
sensitivity to climate perturbations (Bull, 1997). This author calculated that a 100 years of
time were needed for the trenching, but at least 500 years were necessary to complete the
aggradation in incised channels.
In conclusion, it is too simplistic to ascribe channel trenching to only a single factor, as,
for instance, an increase in temperature. The changes in the density of vegetation of slopes
or riversides play an important role in the incision initiation. Vegetation can be modified
by changes in precipitation, temperature, overgrazing, fires, and other human activities.
Incision begins when these variables create an imbalance within the system, which can
trigger and maintain the process of trenching. Some explanations such as climate change,
however, have a regional significance, whereas others can only be locally referenced.
There are defenders and detractors of each hypothesis regarding the complicated problems
about rambla and arroyos, but it is necessary to take into account that different changes in
an area can produce similar geomorphic results.
1.1.4. Alluvial fans
These are semiconical forms that are produced when water flows and confined channel
sediments reach the mountain foot, where there is a reduction in the stream capacity,
absence of confinement, and sedimentation of the charge, mainly by the larger ephemeral
rivers. Although they are located in many climate environments (Rachocki, 1981), here
only those actually located in arid regions are considered in order to deduce the climate
changes that have occurred.
The study of alluvial fans has progressed a lot through the application of
sedimentological techniques (Mayer et al., 1984; Grossman and Gerson, 1987),
the analysis of palaeosoils (Bull, 1991), charcoal dating and the rock varnish method
(Dorn, 1994a). From that, it is possible to identify the climatic features from deposits
and palaeosols, as well as from correlations within the fan and in the region.
Thus, the objective is to establish the influence of the climate on the genesis of alluvial
fans.
For some authors, the fundamental stages of aggradation in alluvial fans occur in wet
periods (Glennie, 1970; Dorn et al., 1987; Maizels, 1987), whereas others think that
sedimentation increases in drier stages (Wells et al., 1987; Bull, 1991). An intermediate
solution is adopted by other researchers who have indicated that aggradation can
be produced in both climate conditions (Williams, 1973; Ponti, 1985). All this shows
the complexity of the interpretation and the extreme difficulty in carrying out correlations.
Four conceptual models have been proposed, however, to relate climate changes with
alluvial fans (Dorn, 1994b). When a transition to a drier climate takes place, there is
produced a greater denudation of the accumulated deposits in slopes and floodplains in
the upstream basin, previously retained by a greater land cover, creating an aggradation
stage (Knox, 1983; Wells et al., 1987; Harvey, 1990; Bull, 1991; Harvey, 1997). The
paraglacial model is significant due to the great quantity of erodible materials and the
large volumes of water released during melting epochs which affect alluvial fans. In this
way, glaciations produce great quantities of sediments that supply alluvial fans, lasting
for thousands of years after the glacial stage. In the deserts of the western USA and
central Asia, many examples can be recognized in which upstream basins of alluvial fans
have been affected by glaciers. Another alternative, previously described, is the direct
relation between a wet stage and aggradation. Some authors relate the latter with glacial
conditions in which water availability is higher (Bull, 1991). Finally, the periglacial
model implies an important production of gelifracts, which when they are eroded,
generate a stage of aggradation in the alluvial fan (Williams, 1973; Wasson, 1977).
These periglacial processes are very marked in some mountains, in which upstream
basins of alluvial fans develop. In the southeastern part of Spain, as indicated, the main
aggradation periods coincide with the cold phases of the Quaternary and the incision
periods with times of scarce sediments, which correspond to warmer phases (Harvey,
1978, 1990, 1996, 1997).
The construction of an alluvial fan requires a long period of time and, therefore, during
its formation there will be changes in the climatic variables. Despite the diversity of the
models proposed, it is really difficult to differentiate the climate influence from other
"noise," such as the tectonic or the intrinsic factors of the system.
Nevertheless, analysing the paleo-vegetation record of the Mohave Desert, climate
correlations between different areas with alluvial fans have been carried out (Spauldin,
1990). In this type of climate correlation it is necessary to take into account that the areas
analysed usually have highly variable tectonic histories of fairly recent age.
At present, researchers agree that the construction of the alluvial fans in arid regions is
due to storms of high magnitude and low frequency (Beaty, 1974; Baker, 1977; Blair and
McPherson, 1994a,b; Dorn, 1994b; Harvey, 1997), which can have occurred in any
climatic period of its history. Therefore, it is difficult to forecast the effects of the climate
change (Harvey, 1996). As pointed out by this author, threshold conditions and sensitivity
to environmental change (Brunsden and Thormes, 1979) vary from one alluvial fan to
another one. In the semiarid zone of the southeastern part of Spain, during the period
1985 to 1990, a significant increment in the number of storms occurred. The erosion was
limited to the alluvial fan apex, where marked trenches were produced, while eroded
sediment was deposited downstream, without any evidence of new incisions in the middle
and distal zones of the alluvial fans (Harvey, 1996). Therefore, it is very complicated to
reconstruct the evolutionary history or, even ascertain data on the climate changes that can
have affected the alluvial fans (Blair and McPherson, 1994a,b).
1.1.5. Desert rivers
In order to deduce environmental changes, the modifications in the form of the channel and
in the type of the deposit are analysed. During historic periods many examples of human
intervention are known, which trigger important variations in morphology, as well as in
fluvial sediments (see Chapter 19). Furthermore, neotectonic activity is another factor that
can influence the modification and intensity of the fluvial processes in some areas. Due to
this, geomorphological works are carried out in cratonic areas where there is no tectonic
activity, and so the "noise" produced by such activity can be avoided. Rivers are relatively
insensitive to climate changes if these are not very marked (Reid, 1994). Rivers in arid and
wet regions, however, present important differences, however, not only in form but also in
the transported solid charge (Wolman and Gerson, 1978). In arid regions few hydrologic
records are available, or do not exist, so the recent variations and their relations with
climate are not well known. Some researchers, such as Schumm and Lichty (1963), have
studied temporal variations of the channel parameters to deduce climate changes. One
difficulty in drainage basins is the response of the variables of the fluvial system to climate
changes (Schumm and Lichty, 1965). These variables are interrelated and climate affects
mainly the weathering, pedogenesis, vegetation and hydrology (Schumm, 1977).
Furthermore, the variables susceptible to change can be more active in one basin than
in another.
The modifications that the fluvial system can undergo may be the result of changes in
the average characteristics of the climate and also of the modifications in the common
values of magnitude-frequency of the different events (Knighton, 1998). Similarly, small
climate changes have produced modifications in fluvial systems (Knox, 1983). The studies
of the floods of the Mississippi River for the last 7000 years reveal that there was a warm
and dry stage between 5000 and 3000 yr BP, with minor flooding, and a abrupt climate
change of a colder and wetter nature that produced important floods (Knox, 1993).
Moreover, this author indicated that the most important floods in the last 9000 years are
due to periodic changes in the atmospheric circulation systems (Knox, 1995).
When a channel increases due to a climate change, the width, depth and the inclination
increases and vice versa. If the sediment load increases, the erosion will rise in the
drainage basin. This increment in sediments can be produced by a climate change toward
more aridity or by deforestation. A fall in the solid load can be due to a climate change
toward wetter conditions or to soil conservation practices. In general, any modification in
the channel is attached to a change in the type of sediment charge. When a change into a
more humid climate takes place, an increment of increase is produced in the percentage of
the vegetal cover and, therefore, there is a decrease in the loss of soils. This increase in
precipitation turns ephemeral courses into perennial ones, the density of drainage grows
and erosion in the tributaries rises, which produces an increase in the solid load, mainly
during storms. If climate returns to semiarid conditions, soil loss and erosion rise,
meanwhile discharge falls. All these variations in discharge and solid load, which in this
case can be a climate change from subhumid to semiarid and vice versa, produce important
hydrologic modifications and changes in the morphology of the channels, termed channel
metamorphosis (Schumm, 1977; Chorley et al., 1984).
Many large drainage basins go through different climatic zones. The Nile fiver rises in
central Africa, where it presents an important discharge and transports a small load
Figure 25.16. Diagrammade from an aerial photograph of a portion of the Riverine Plain near Darlington
Point, New South Wales. The sinuous Murrumbidgee River, which is about 200 ft wide, flows to the left
across the top of the figure. It is confined to an irregular floodplain on which a large oxbow lake (youngest
paleochannel) is preserved. The oldest paleochannel crosses the lower part of the figure (after
Schumm (1969)).
of sediments. In middle and lower parts, the discharge falls but sediment load increases,
because it goes through the arid countries of Sudan and Egypt. The history of the climate
variations that occurred during the Quaternary is complex and difficult to study (Williams
and Faure, 1980).
One of the best known examples of fluvial metamorphosis is the meandering river
Murrumbidgee, which flows through the almost flat riverine plain in New South Wales,
Australia (Schumm, 1968, 1969). The channel, 60 m wide, is surrounded by a floodplain in
which several abandoned meanders occur (Figure 25.16) indicating that the river had a
higher discharge. In the lower part of the figure an old channel of low sinuosity is shown.
Different parameters in existing channels were measured, in recent and in old ones (width,
depth, sinuosity, slope gradient, meander amplitude), as well as the discharge and the solid
charge, to compare the hydrologic variations and the changes in the sediment charge when
active. These geomorphological observations, together with the pedologic ones, indicate
that the climate was drier when the paleochannels were active, while younger
paleochannels occurred under more humid conditions.
In a dry stage the river decreases its discharge, land cover gets reduced and solid
transportation rises. The change toward wetter conditions produces an increment in the
discharge, width, depth, and in the amplitude of the meanders. The result represents a
complete transformation of the fluvial system. An example of fluvial system abandonment
and replacement by another system occurs in Oman, in western Arenas Wahiba (Maizels,
1987). There it is possible to recognize a five paleochannels system consisting of cemented
gravels, which is part of an alluvial fan, (Figure 25.17). The erosion of the fan produces
lowering, emphasized by aeolian deflation, leaving an exhumed landscape in the shape of a
cord, which sticks out about 30 m above the plain. Older channels can correspond to
humid stages in the early Pliocene-Pleistocene, whereas the exhumation is related to
semiarid stages during the Pleistocene, corresponding to lower sea levels.
In arid environments overlaps and interferences between aeolian and fluvial domains are
common. Thus, southwestern Egypt close to the frontier with Sudan and Algeria, is
constituted by the large sand layer of Selima. That is a hyperarid zone, in which the study of
radar images has shown that there is a system of big fluvial channels overlain by the sand
layer, having an age that probably corresponds to the early Miocene (McCauley et al., 1982).
The most classic and complex example of interference between alternating aeolian and
fluvial stages occurred in the middle course of the fiver Niger in Mali (Figure 25.18).
Fluvial accumulations took place in the humid period of the early Quaternary and totally
stopped during the subsequent stage of highest aridity, producing longitudinal dunes, or
trade winds ones (ENE-SSW), in which watercourses disappeared. Later, in a period
wetter than the present one, they crossed dune fields and filled the existing enclosed
depressions, producing the expansion of lakes, as demonstrated by the high quantity of
ceramic remains from the Neolithic. Nowadays, the discharge is enough to make the Niger
have an exorheic character, although present precipitation is lower than Neolithic and
much higher than the one existing during the stage of dune formation (Tricart, 1969;
1979). In the section about dune systems this aspect will be discussed in more detail
(Figure 25.18).
1.2. Paleolakes
The study of the shorelines of ancient lakes, together with the analysis of the sediments
deposited, provide very valuable data on the hydrologic modifications that have occurred.
Many desert lakes present clear geomorphological and stratigraphic evidence of having
had, during the late Quaternary in times with a higher effective humidity (Mabbut, 1977),
dimensions much larger than the present ones.
Shoreline mapping of lakes requires a detailed photogeomorphologic interpretation and
meticulous field work (Sack, 1994). Measures taken at different points around the basin of
the lake for a given shoreline normally present different heights. This can be due to coastal
geomorphologic processes and in deep lakes to hydroisostatic rebounds (Gilbert, 1890).
With these data, it is possible to determine the area and depth of a paleolake for different
periods. The existence of subbasins, separated by thresholds, as in Lake Bonneville
(Gilbert, 1890), controls the levels of the paleolake (Benson, 1978; Benson and Paillet,
1989) and makes their study difficult, because of the complex dynamics of each subbasin.
Another difficulty is related to the paleohydrologic variations that occurred within the
Figure 25.17. Channel pattern and lithology changes associated with successive generations of
paleochannels on Fan 1, in the western Wahiba Sands of Oman. Channel sinuosity and mean width
decrease progressively with time (after Maizels (1987)).
9
~g
E
0
9,..~
9
9
~D
~ ,.~
0
~, ~
9
0
,s=
9
+.a ,.
~ ,.~
~ ,..~
~ ,..~
Figure 25.19. Smalltrench excavated in the boundary of the sebkha E1Melah. The base corresponds to
gray rhythmic sediments and the top consists of a fine sand level, Zarzis, southern Tunisia.
fluvial basins that supplied the paleolake. Any important variation that modifies the water
discharge can be incorrectly interpreted as a climate change.
The study of the stratigraphic record of a lacustrine depression provides relevant data to
know its history (Figure 15.19). Geomorphologic, sedimentologic, mineralogic,
geochemical, paleontologic and palynologic research provide keys for obtaining suitable
paleo-environmental information on the different events through which the lacustrine
system has undergone (Figure 25.20).
Lake expansions are due to a rise in precipitation or to a fall in temperature that
produces a decrease in evapotranspiration. These circumstances correspond to the old
concept of the pluvial period or lacustrine phase (Rognon, 1980). Nevertheless,
conclusions obtained do not have the same verisimilitude that conclusions do based on
the analysis of dune systems, because the relationship between precipitation and lake
levels get complicated because of the temperature and other nonclimatic factors (Goudie,
1992). Some lakes, as with those related to the Rift Valley in eastern Africa (Figure 25.21),
have a marked tectonism during the Quaternary, which can modify the geometry and
capacity of the lake. Similarly, volcanic activity, mainly effusive, produces very
pronounced modifications. Despite these circumstances, works carried out by numerous
authors on lakes in eastern Africa and the southern Sahara indicate that they are excellent
paleoclimatic indicators (Figure 25.22). In the deserts on shields and platforms (Mabbutt,
1977) large lakes develop in shallow structural depressions during relatively wet periods.
Lake Chad is the largest known endorheic basin (Figure 25.23), whose boundaries are on
the Guinea ridge, Congo watershed and in the massifs of central Sahara. This lake has been
studied by many researchers (Barbeu, 1961; Grove and Warren, 1968; Pias, 1970; Servant
Figure 25.21. Lake Nakuru, located in the Rift Valley Gregory, and therefore, of tectonic origin, even at
present. This lake, of alkaline water, is famous because it contains the highest number of pink flamingos in
the world. Nakuru, Kenya.
Figure 25.22. Fluctuationsin lake levels in east Africa and the southern side of the Sahara. Data: different
authors (compiled by Selby (1985)).
and Servan-Vildary, 1980; Nicholson, 1981) who have spatially and temporally analysed
dry and wet phases that have occurred in the basin. During several wet periods the
extension of the lake was larger than the present one. The flatness of the lacustrine basin is
expressed by the depth, 3 to 7 m deep, for an actual area of 20,000 km 2. In one of the
expansive stages (7000 to 5400 yr BP) (Pias, 1970), the lake, whose level was 320 m,
occupied an area of 350,000 k m 2 ("Megachad") and overflowed into the Atlantic Ocean
and the river Nile in dry periods the lake was partially covered by dunes of ENE direction
(trade wind dune). This type of lake is characterized by overlapping between lacustrine
and aeolian landforms, as indicated in Figure 25.23.
Figure 25.23. The Chad basin, showing dune systems and the shoreline of "Megachad" at about 320 m
(Grove and Warren, 1968).
In mountains and desert depressions there are numerous lakes, frequently deep, that
originated during stages of intense fracturing. In those lakes, shorelines of different height,
beach ridges, spits, small deltas, saline mud plains, and so forth can be recognized (Gracia,
1995). In the Great Basin, located in the northern part of the Basin and Range province in
USA, exists a large number of "pluvial" lakes (Morrison, 1965); about 110 to 120 lakes,
formed by extensional tectonic block movements during the upper Pliocene and lower
Pleistocene (Figure 25.24). In the stage of highest expansion Lake Bonneville was
generated, which was studied in detail by Gilbert (1890), one of the founders of the
discipline of geomorphology. This lake reached an area of 51,640 km 2 and a depth of
340 m. At present, the Great Salt Lake of Utah, which is the largest lake resulting from its
desiccation, has an area ranging from 2600 to 6500 km 2. Toward the west, Lake Lahontan
at 22,900 km 2, also developed. Some basins were connected by overflow connections and
others overflowed directly in the ocean through the rivers Snake and Colorado.
Figure 25.24. Maximumextent of Great Basin lakes during post-Sangamonian time. Arrows indicate
overflow connections. The heavy line shows the present drainage divide of Great Basin (after Morrison,
1965).
The aridity of these North American deserts is due to an orographic effect. The high
ranges (Coast Ranges, Sierra Nevada and Cascades) are close to the Pacific coast and
obstruct the cyclonic circulation coming from the west. As a result, deserts in the
rainshadow are generated. In this area, temperatures have been very low, mainly during
glacial periods, because the Laurentide ice sheet was close and as in high mountain areas,
around the coast, many smaller glaciers also developed. As temperatures fell,
evapotranspiration decreased and lowlands in the Great Basin received large amounts
of water from glacier melting. Therefore, large lakes were formed within endorheic zones,
which nowadays show spectacular lacustrine terraces. This area is an ideal region to
recognize the relationships between the mountain glaciations and lake phases. The work of
Smith and Street-Perrott (1983) analysed many of these basins and discussed the
chronology of the lake phases. They established a period of expansion between 24,000 and
14,000 yr BP; another between 14,000 and 10,000 yr BP with wide fluctuations that can be
synchronous or not. Finally, between 10,000 and 5000 yr BP, an arid or dry phase can be
recognized, with a fall of the lake levels. Recently, Oviatt (1997) pointed out that the
different stages of lake-level fall have ages of 21, 18.5 to 19, 17.5, 16 to 15.5, 14 to 13, and
10 kyr, all of them obtained by means of C-14. They are synchronous with the end of the
Heinrich H1 and H2 events and with other smaller stages of iceberg drift (a, b, c, and
Younger Dryas) in the northern Atlantic, suggesting a global change.
The study of shorelines and their interpretation in deep lakes is complicated due to the
fact that they have been deformed because of the hydrostatic unloading. The study of lake
levels is crucial, however, to determine the magnitude of the subsidence in particular areas
of the Earth's surface, as a result of the load due to the water volume and the subsequent
rise resulting from the removal of lacustrine water. Figure 25.25 on the left shows isobaths
Figure 25.25. Pleistocene Lake Bonneville, Utah. Left: Maximum water depth (in feet) averaged over
circles of 25-mile radius. The white area shows the extent of the lake at the Bonneville shoreline and the
stippled areas indicate the present lakes. Drainage of the Pleistocene lake occurred as the result of an outlet
being cut at Red Rock Pass. Right: Isostatic rebound (in feet) resulting from the removal of lake water, as
measured by the deformation of the Bonneville shoreline. Recent displacement along the Wasatch Fault is
shown. (Crittenden, 1963).
of the lake Bonneville during the period 25,000 to 11,000 yr BP and the one on the right
indicates the isostatic rebound produced by lake desiccation, which is 20 % of the column
of water (Crittenden, 1963).
The Aral-Caspian Sea system is one of the most spectacular cases of lake expansion
during the Pleistocene era, which is formed by wide shallow deformations. Huge amounts
of glacial melt water went into this system from Syr Darya and Amu Darya rivers, to the
Aral Sea, and from Volga and Ural to the Caspian Sea. When the level of the lake reached
76 m above the Caspian Sea, it met the Aral Sea, generating the largest "pluvial" lake in
the world, flooding an area of 1,100,000 klTl 2 and penetrating 1300 km into the river Volga
(Goudie, 1992). The Caspian Sea also reached the Black Sea through the Mantych
depression.
Despite the difficulties of interpreting climate on the basis of geomorphologic studies of
paleolakes, these are crucial for paleoclimatic reconstructions, especially if they are based
on multidisciplinary paleoenvironmental studies.
1.3. Dune systems
1.3.1. Introduction
Large aeolian accumulations or ergs, together with the erosive landforms existing in
intertropical deserts, occupy 20 to 25% of the Earth's surface (Livingstone and Warren,
1996). The extension of these dune fields is well known in most of the deserts in the world
(Figure 18.2) (Thomas, 1997b), from aerialphotos and satellite images. Dunes have been
generated in periods of 1000 to 100,000 years (Lancaster, 1995). Expansions and
contractions of the principal ice sheets, which considerably affected lower latitudes,
changing climate and sea level, occurred during dune-field formation. The most significant
variations become evident through changes of the boundaries of great sand deserts, in
sand-particle supply and in the mobility of the dunes. These sand deserts are not
exclusively Pleistocene, but many of them also became larger during the global climate
cooling (Figure 25.26) (Shackleton and Kennet, 1975; Goudie, 1992); this is what
Williams (1994) call "upper Cenozoic desiccation."
Many periods of dryness can be identified in the course of the Quaternary. The study of
the ice cores dated in the upper part of the Greenland ice sheet (GISP2: Greenland Ice
Sheet Project 2), show fluctuations in the electrical conductivity of the ice, at scales of < 5
to 20 years. These reveal rapid oscillations in the atmospheric dust content (Figure 25.27)
in periods in which the atmosphere was very dusty, separated by others without dust. The
analyses of continental sediments within deep oceanic cores can also be used to interpret
the extension, intensity and duration of the arid zones during the Quaternary. Continents in
intertropical regions provide fluvial and aeolian sediments to the marine bottom.
Furthermore, there is a relatively constant supply of organic carbonates, and thus, those
cores with a high content of carbonates indicate a scarce supply from continents. Aeolian
supplies are represented by high percentage of silt-sized quartz grains, which can be
differentiated from fluvial ones because of their tendency for reddish staining (Diester-
Haas, 1996). Increments of aeolian dust took place during Quaternary cold periods, in
which the quartz content at the bottom of the Atlantic Ocean moved toward the south,
Age (Ma)
0 0 0 0 O
0
1 1 1 1 .... I 1
O_ Palaeocene 1 Eocene I Oligocene Miocene

o4
oo
v
l_
.1--,
-7
G)
_.L
E
(1)
.i-,
0g LO-
+
Figure 25.26. Cenozoic sea surface temperatures in the Southern Ocean deduced from changes in oxygen
isotopic composition of planktonic foraminifera at DSDP sites 277,279, and 281 (Shackleton and Kennett,
1975, in, Abrahams and Parsons, (1994), Figure 26.4).
at about 8~ indicating an expansion of the aridity and of the Saharan sand desert
(Figure 25.28) (Kolla et al., 1979). The existence of fresh water diatoms in cores
corresponding to the maximum of the last glaciation can be interpreted as particles
deflated from desiccated desert lakes, which also supports the idea of this expansion
(Parmenter and Folger, 1974; Pokras and Mix, 1985). Finally, important data can be
obtained from the analysis of deep-sea drilling cores on the basis of the study of feldspar
weathering, pollen and phytoliths, and salinity and temperature of the marine water from
oxygen isotopic composition in foraminifera.
This research provides indirect data on the important dry periods that can be distinguished
in the recent geologic record. The detailed study of the different active or inactive dune fields,
HOLOCENE PLEISTOCENE
..................................................
~'100 _ Aller~d B~lling
._
._>
~ 60
~ , ~ ~ ~ l Younger Dryas .....
~9 40
n~ 0 __J_
E
10.5 11.0 11.5 12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5
Date (1000 yr BP)
Figure 25.27. Yearly averages of electrical conductivity (related to dust content) of a core at the summit
of the Greenland icecap (Taylor et al., 1993). The core shows remarkable (and sudden) alternations
between the dustiness of cold periods and the relatively clear skies of interglacial periods.
Figure 25.28. Difference in quartz percentage in late-glacial and Holocene sections of Atlantic sediment
cores. Enhanced late-glacial concentrations in the Northern Hemisphere tropical zone are interpreted as the
product of dust inputs deflated from an expanded Sahara and/or due to stronger circulation systems (after
Kolla et al. (1979)).
within the different deserts in the world, however, provides valuable information on the
extension, duration and distribution of the periods of marked aridity, as well as on the changes
in wind directions and circulation systems through time (Lancaster, 1995).
The dating of sandy accumulations has been one of the most important problems. Until
a short time ago, dating was difficult and dates were relative. Thus, for example, C-14
O_ f i k
g
5-
h o =
,... <
10-
b e ~ ~
..::, ~
.,.,..'-, :::::.: I-
; ".:.:v C :."74,1 ,.,.., c"
:,'.v ~.",,'i,: ," .:, "_~_ ,-"
v
o
~t.','.: ::',...". " :-
,_o _.. ...:,..:; .... C Y':"..!.
15- ..:::-.
....
._
I..-
.. :......
..--v...
,.-.
..-. , -
~ .;;5'.
:-v'/..
__ ....,.
,~!~! ~ ......... ~..:: ~.
:.:',~,. ~ .:v;-.
..,..,
~-:;::::: ~ ~ ....... ~::-:.,::: .E: ::;:::;::::
,,,,., z = .... :::"::,;::,:.:
20- ~ ..= ...-, ,.,,, :.:~:.::
..-, ._o ~ "5 ..-,
L'~ .--
x:: & m 1.1.1
25- E ~ .-9- ~ .-g_
Iii
30.
a: Rognon (1987) g: Butzer e t a L (1978)
b: Talbot (1980), Pachur e t al. (1987) h: Magaritz and Enzel (1990)
c: Michel (1973), Grove (1985) i: Wasson e t a L (1983)
d: Deacon e t a L (1984), Lancaster (1989) j: Sung-Chiao (1984)
e: Lancaster (1981 ) k: Bowler and Wasson (1984)
f: Shaw and Cooke (1986)
Figure 25.29. A summary of late Quaternary dry periods for the tropical and subtropical continental
deserts (Africa, Asia, and Australia) for the last 30,000 years. Dune building episodes most likely took
place during the above periods (compiled from several authors by Tchakerian (1994)).
dating of the organic matter existing in paleosols developed on dunes reveals an age pre- or
post-. Another difficulty is to have available reliable dates, because many dune fields are
formed during several epochs. There was extraordinary progress at the beginning of 1980s,
with the use of many new techniques such as thermoluminescence (TL), and optically
simulated luminescence (OSL), which permitted the dating of sand in dunes and loess
(Wintle, 1993). Ages up to 500,000 years have been obtained. These dates made possible
correlations with paleoclimatic data by using other techniques. With all these data, it was
possible to make paleoclimatic correlations and interpretations. As a result of this progress,
actually it is now possible to know the different dry periods that occurred in all deserts in
the world (Figure 25.29), for which different chronologic histories can be obtained.
Dunes are sensitive to variations in atmospheric parameters, such as wind direction and
intensity, changes in precipitation that influence evapotranspiration, soil humidity,
vegetation percentage, and particle mobility. The areas of dunes tend to emphasize
Figure 25.30. Possible responses of dunes to changes in climate (after Rognon (1982)).
the effects of dry and wet phases (Rognon, 1980) and rapidly respond to climate changes.
If precipitation increases a little, vegetation grows up mainly in the lower parts of
the dunes, whereas in the upper parts sand mobilizes producing leeward avalanches
(Figure 25.30). When precipitation increases, vegetation colonizes the entire dune, except
the crest, and calcification starts in sandy accumulations. Finally, when periods of
important and durable rainfall occur, the groundwater table appears on the surface
generating marsh deposits in the interdune areas; furthermore, vegetation of large size
settles, stabilising the dunes and generating soils. In the cases in which precipitation is
very intense, dunes can be eroded (Talbot and Williams, 1978). On the other hand, in dry
periods the aquifers are low and, therefore, water availability is virtually zero, hampering
the growth of the vegetation and favouring the aeolian activity during these periods.
1.3.2. Active and inactive dunes
The knowledge of the degree of activity of dunes is problematic and imprecise. From
numerous data, it is estimated that when average precipitation exceeds 100 to 300 mm
(Goudie, 1992), vegetation is effective enough to restrict the movement of the dunes. The
degree of activity can be obtained from climatic indexes collected by Cooke et al. (1993),
and elaborated by Chepil et al. (1962), Wasson (1984), and Lancaster (1988). Climate
indexes are imprecise due to limited meteorological stations and to great year-on-year
variability in semiarid climates (Livingstone and Warren, 1996), in such a way that in
some years the dunes are active and in others with less precipitation, they lose most of the
mobility or even become stable.
The transport and sedimentation of active dunes (Figure 25.31) become apparent in
surfaces with ripples and leeward avalanches. Nowadays, overgrazing and other human
activities in deserts can trigger the reactivation of the dunes, as occurs in the very
populated area of Rajasthan, India, in the desert of Thar (Goudie, 1992). Dormant dunes
(Lancaster, 1995) or episodically active dunes (Livingstone and Warren, 1996) are those
in which the movement of the sand is slow or does not exist for long time spans
Figure 25.31. Largebarchanoid dunes. Moving sand can be observed in the foreground. A1Liwa, Rub al
Khali Desert, United Arab Emirates.
Figure 25.32. Dormant longitudinal dunes with well-developed shrub-like vegetation. The mobility of
the sand is limited to the crests of dunes, while flanks are stabilized by biocrusts and shrubs. Experimental
Estation of Nizzana, western Negev Desert, Israel.
(Figure 25.32). In general, they are located in areas with 100 to 300 m m of precipitation,
great variability in year-on-year precipitation, in such a form that in dry periods land cover
decreases and sand mobilizes. The effect is the opposite in wet years. In these dunes,
primary sedimentary structures are affected by bioturbation. Vegetation shows a good
state of development with high percentages of perennial plants. Inactive, fixed or relict
Figure 25.33. Inactive dunes with shrubs and trees cover. Beer Sheva region, Negev Desert, Israel.
dunes are remains from past drier climates and, therefore, are indicators of paleodeserts
(Figure 25.33). These dune systems are located in areas that nowadays have rainfalls of
250 mm, and even 2000 mm, as in Zaire (Thomas and Goudie, 1984), the Orinoco Basin
(Tricart, 1985), Pantanal of Brazil (Klammer, 1982b), and elsewhere. They consist of
dunes and sand sheets that are stabilized by structured soils. Within these soils, caliches
and ferralites develop due to the precipitation increase (Tricart, 1979; V61kel and Grunert,
1990). Alternatively, dunes can be cemented, generally variably, which produces
aeolianites that commonly have weathering alveoles and karstic micromorphologies
(Figure 25.34). On the other hand, dunes become modified by nonaeolian processes and
elluviation. Crests become rounded and reduced, flanks eroded by water erosion and the
resulting sediments fill up interdune depressions (Mabbutt, 1977). Nowadays, flanks of
longitudinal dunes in the southern Sahara do not reach 5 ~. During the initial degradation
stages of a dune, rilling, gullying, and piping processes can be triggered in wet seasons,
which can destroy former aeolian structures. Flooding in wetter stages is another cause of
inactivity, as has occurred in the desert of Thar (Singh, 1971) and in the Chad Basin
(Grove, 1958), where dunes are situated below lacustrine deposits. In all these relict dunes
the return to an active state is only possible through important environmental changes.
An aspect-causing controversy and, on the other hand, also common in other branches
of geomorphology, is the estimation of the time needed for the development of a landform
such as a dune. Some researchers have pointed out that actual dunes are probably relict
(Glennie, 1970; Besler, 1982), because only very strong winds, as the ones that occurred in
upper Pleistocene, are capable of generating these landforms. It is also said, alternatively,
that dunes require long periods of time to be formed. Thus, on the basis of empirical
observations, Wilson (1972) estimated that the 100-m high dunes in eastern Great Erg
Figure 25.34. Aeolianite with cross-stratification affected by alveolinization controlled laminae. Sila,
United Arab Emirates.
(a) . / " ,50o..,~2E.~9~0o.,,~3,~o. d~ , ~ 0 o .40 o' ~0o i , . ~ o o ,~0o~. 1 i 0 o - ~ _
C ~176 ",4.',-! /"'~ ~,~&,z~> ~,~,,~,.-._m_~-,,,/~ ,oo~
i., .... ~_~:."__~ ~

,~~~:~ ~~_'_........ 20 ~
[~176 ,~, '~ i",, '( "~.2'~-'-..~.,., oo
--'--2--0
i
40~ . . . . . . . . . . ~ I i " , ~ 4 /
60 ~ 30 ~ 0~ 30 ~ 60 ~ 90 ~ 120 ~ 150 ~ 180 ~ /
~0> ~-I,0o.,,0o.,o.,~0.,~o./. io , ,oo. 40o , ~ o o" i~o o" 1"soo" s ~
. . . . . . . . . . . ~,~_ .- - _ _ 60o",~
~"" . FEW DATA. ""'i" "
; ~
-~
:, '"" l I ~'~" ,
;';:,~
,,~,~
"::;:
': ,
,,
91% ,. . . . . . .
__ 20 ~
t o
0 o-. , , -- # ,', " l 9 6 o

0 1000 2000 3000 (km)
. ,,, ~I~6 ~ \
_)0 o _ - . ),,,,
. . . . . . . . . . 'J v. : -.~ .-..'- ' - - -
_ 150 ~ 120 ~ 90 ~ ,. .... ~ 390 0o 3.0~ . 6,0~ . 99 ~ , 1~ o , , o., ~ o /
(C) ~ i , 5 0 ~ ~ . ~ e ' ( ~ l ~ O ~ ' {jo j,,,..,,3bo " .6"0 o " 9"~0 o " 1.50o" ii~0 o
~,~.:----,~...~ . ~ . ~ _ i . f , ~ r _~~,~..o~
/400 #-':. " -~ %';,~-.,,^~ o j ~ 400",,
I ~',.:.", /" ~-&~.:.,-~'.-...~~ 9
Lo ....
. .
-'~,.-/~-.~_ .......
. .
I:!',,,.."-~~,",, ~, .. ..,_% ., , - . . ~ , _;,,..-
...
.... ....
,0o.
l .,,.,. ,, ' '':" '" ' ,%
[,,o =,,=o=,,,sc,,~ )/% "-% r " ~,,,/,;s .....
t % '( /LL '#~ -~'" ;

....
.... ,,!,/. i ........ " , oo,
k ~ 150 ~ 120 ~ 90 ~ 60 ~ 30 ~ 9 (:to 9 3,0~ . 60 ~ . 90 ~ , 1;~0~ 150 ~ 180 ~ /

have an age of 10,000 years, and Lancaster (1989a) estimated in the age of the large dunes
of the Namib Desert to be some 42,000 years. Dunes of 70 m height exist on the coast of
the Huelva province, in E1 Asperillo of southern Spain, showing an age, by means of
archaeological and historical data, from the 18th century (Borja et al., 1999; Zazo et al.,
1999).
Today it is well known that there are extensive areas of stabilized dunes on all the
continents, in tropical zones as well as at high latitudes, in relation to periglacial climates
(Sarnthein, 1978) (Figure 25.35). Dunes formed, during the last glacial maximum,
2.5 million klTl 2 in Africa, which means the largest desert in the world (Thomas and
Shaw, 1991). Comparing the extension of these dune fields with the current area with
active dunes, it can be deduced that great changes in have occurred precipitation and in
the percentage of land cover. At present, active dune deserts, located between 30~
and 30~ occupy only 10%, while 18,000 years ago their extension was 50% (Goudie,
1992).
In the boundaries of most zonal deserts there are stabilized longitudinal dunes. The
most extensive ones are located in the Sahara, over 5~ (Figure 25.18) and stretch from the
Atlantic coast of Senegal to Sudan. They extend toward the south over 500 kin, in areas
that nowadays have a precipitation of 500 mm. Three generations of dunes can be
recognized (Grove and Warren, 1968; Talbot, 1980). The oldest group is older than 20,000
years, the second between 20,000 and 13,000 years and the most recent has an age of about
5000 years. Most of them were stabilized between 11,000 and 7000 years ago. The last
droughts and deforestation have reactivated some dunes and have generated new ones
(Nickling and Gillies, 1993). Other examples can be found in the Kalahari, northern
Arabia, Thar Desert, Australia, Venezuela, and elsewhere. All these dunes are completely
colonized by vegetation and have soils developed on them.
1.3.3. Paleoclimatic information provided by dune systems
Active dunes that developed on continents are located around and below the isohyet of
150 mm (Mainguet et al., 1980), with the southern Sahara as an excellent example of that
spatial distribution (Figure 25.36). The existence of these relict dunes represents an
irrefutable evidence of previous dryer stages. The situation of these dunes in areas of high
precipitation, as for example the selva of the Congo, reveals the drastic change that
occurred. Moreover, it is necessary to take into account that the reactivation or formation
of dunes can be the result of changes in the velocity of the wind and their interactions with
humidity (Thomas, 1992).
Comparing wind directions that mobilize sand with dune alignments, valuable
paleoclimate information can be obtained. There are wind systems that at present blow in
the same direction as during the late Pleistocene, as in the southern Sahara (trade winds).
Figure 25.35. Active dune fields and long-term resultant effective wind regimes for the present,
18,000 yr BP (glacial maximum), and 6000 yr BP Climatic Optimum. Continental outlines at 18,000 BP
represent sea-level lowering of 120 m. Hatched areas indicate dune fields and major dune contours.
Arrows are long-term resultant surface-wind directions. H signs stand for fossil wet conditions excluding
dune formation (after Sarnthein (1978)).
e,I ,-
I '..i': ' .'.:~:" I I --
- _ .... .....:
: :;- ::!i;,: .::-.?- , ~ .:--
"~5 ,
G)
(D
:g -o
._>
.c
,.1- ,.1--
o o ,,-k
E E
/t.i:.!ii;~ I @ __1 _1
~. I ::!:i:i:i:!::
Ik:;~i!!ii!i~i /
~: ~ -~
i ~
- .....E. ~1 ~ '~,~_~ -
"":':!ii::i::ii!!:. II ~r~.'E
"::iii!iiii!
9 11 ~-~./ "-
/
!
E
/
E E ~.o
~176176176 o
k, _ .:!:i:i::"" .~ /'
.:: ::::!ii:." 9:. ......
.'.'.:-:::::~ I
.v-- 1l) (1) r
~!!!ii!iiiii!
,Iit1111111111111
"~ 9
e.
o o
IIIIIiti!t1111111111t ~ ~ I ,-...
~iiil!ilili!iiiii!~:
~...... ~~ ~ ,-~ , I ".
9
..+.- : 9 I " ..~
,.. I i -
,,~176 ~,'."
k, - ~ ":::':::" s
!::..
'
:.,..
.,..~
. ,...~
t
t I'" . '
o
~ 9
g
9
-I/~ .,..~
_~ 9 "~ ::,:': . . . . <
9 ." :..:,... , ,..
~1 m ..:':.:.".'.." '
~5
.'.."".','" .
1
a,
Northern
Dune~,.,
\, ~. ~ 600 mm
Namib \ \ \ ~~\~.-~ Southwestern

Dunes Kalahari dunes
\
-30oS
400
Atlantic
Ocean
Figure 25.37. Isohyets and relict and dormant dune systems in the Kalahari region. Northern Dunes:
<20,000-30,000 years ago; Southwestern Kalahari Dunes:18,000 years and southern dunes started in
the early Holocene (after Lancaster, 1981).
On the other hand, there are regions in which present and past directions are very different.
In the desert of the Kalahari the dunes are principally inactive and in the main are covered
by acacia forest, shrubs and grass. Three dune systems can be recognized on it, generated
by winds with different paleodirections (Figure 25.37) (Lancaster, 1981).
Figure 25.38. Star-shapeddunes overlaying laminated marsh deposits. Maharaga, Rub al Khali Desert,
United Arab Emirates.
The majority of present sand deserts have been formed slowly during the climate
changes of the Quaternary and subsequent variations of the sea level. As a result, different
surfaces of regional extension can be identified (Kocurek, 1988), which separate different
stages of accumulation. Over them, pedogenic processes, water erosion, marsh and
lacustrine sedimentation, and so forth can be recognized to have occurred.
A great part of the sand deserts are fed by deflation of marine, lacustrine (Figure 25.38)
and fluvial sands. The more important and constant through time is the source area, the
major transport, the accumulation, and the migration of dunes. Due to the sea level
lowering during glacial epochs, the Persian Gulf was mostly emergent (Kassler, 1973) and
acted as a source area, supplying the construction of the Rub al Khali desert, as indicated
by the directions of the dune systems (Glennie et al., 1994). These dunes were formed
between 20,000 and 9000 years ago (McClure, 1978). In the Mojave Desert of California,
the sand sources came from the bottom of desiccated lakes and from fluvial systems. At the
Pleistocene-Holocene boundary there was an important increase in the aridity and the
sand supply fell sharply. The aeolian action was limited to the remobilization of dunes
(Lancaster, 1995).
In many sand deserts, interdune deposits are common, as in the A1 Liwa region in the
Rub al Khali Desert of Arabia (Figure 25.39). Shallow lakes develop, which are supplied
by springs, rises of the phreatic level, or by water flowage. The ephemeral character,
sometimes seasonal, with an alternation of expansion and desiccation stages, is the main
characteristic. The importance lies in the fossil content, which makes possible dating and
paleoenvironmental interpretations. Thus, interdune lacustrine deposits of the Rub al
Khali desert show wet conditions between 2400 and 3200 yr BP and also between 5000
and 8500 yr BP (Lancaster, 1995).
Figure 25.39. Interduneplaya with shrubs in the back. These playas can become floodedby precipitation
or water emanation. A1 Liwa, Rub al Khali Desert, United Arab Emirates.
The areas in which there are successions consisting of alternations of aeolian

accumulations corresponding to arid stages, and marsh and lacustrine sediments produced
in wetter stages, are locations ideal for the study of climate changes. One of those places is
located in the Erg Akchar of western Mauritania, which constitutes an excellent example
of periodicity produced by climate changes. The stratigraphic work and dating, by means
of C-14 and prehistoric remains, were carried out by Kocurek et al. (1991) (Figure 25.40).
The substrate is formed by longitudinal dunes with an age ranging from 13,000 to
20,000 yr BP. It is also possible to recognize two other stages of aeolian reactivation,
separated by surfaces that can be identified on the basis of specific stratigraphic
characteristics. Finally, the other periods correspond to phases of aeolian inactivity
because of the growth of the land cover. Marsh and lacustrine systems also develop
between dunes, with sedimentation of sand, carbonates and gypsum. In interdune areas
close to the ocean, marine transgressions turn these depressions into coastal sebkhas.
The southeastern part of the River Niger was supplied during the late Pleistocene, by
tributaries coming from the Hoggar massif (Figure 25.18), which are currently inactive.
Downstream the discharge increases with other tributaries, flowing into its delta. The upper
part of the Niger, which flows toward the northeast and is fed by Guinean mountains, flowed
into the Senegal Gulf during the late Pliocene and early Pleistocene (Goudie, 1992). Later,
during a dry phase, vast dune fields were formed, which acted as a barrier blocking the
former watercourse (Tricart, 1979). As a result, many ephemeral lakes were generated.
After that, by capture or overflowing, this part of the river flowed into the Niger, forming a
wide bend in the region of Timbuktu, Mali. This capture occurred 5000 to 6000 years ago.
The River Niger, during dry phases, was incapable of going through the extensive dunes and
disappeared inside and into the zone of lakes located in southern Timbuktu. In "pluvial"
(a) Ogolian (13,000-20,000 yr BP)
(b) Tchadian-Nouakchottian (4,000-11,000 yr BP)

- Surface I . ,It I ~ ! U '
.......... _.. - -,- ~-~- . ~ . ~ . ~ - . - , - - - . . ~ ~ ' . ' . . . " t ~ - - " : : .

......... ........ ." .-'..-.-.'~.:::,:',..-~ ! - :..~<.. .... "
(c) Tafolian-Modern ? (?-4,000 yr BP)
"'. : ": . : .... r.

Surface I...~.
J~'~-~-'~--~
~
(d) Modern (?) Surface 2
.~.... o i Oe ilia i
(e) Modern (1990) Surface 2

Surface I ' ~
o ~
,w/. 9 " ~ ~
9 ~ ~ ~ o ~ o o. oo
stages, the higher supply of the watercourse made it possible to cross the aeolian
accumulations, transforming the fluvial basin from endorheic into exorheic. All lakes in this
area expand during "pluvial" stages, while in dry periods increase the salinity by
evaporation, becoming even desiccated, as occurred in 1957 to Lake Fagnibine (Tricart,
1979). The present climate allows the Niger to flow into the Gulf of Guinea, which indicates
that actual precipitation is higher than that of the past, when the river was endorheic.
1.4. Causes of the paleoclimate fluctuations during the late Quaternary period
This is one of the subjects that has most fascinated experts on arid zones and has been
restricted mainly to the African continent, probably because it is the most well known by
many geomorphologists in different countries. According to Tricart (1969), however, the
reconstructions of the mechanisms that produce paleoclimate oscillations are different.
Areas located within tropical high pressure zones generate zonal deserts, whereas
extrazonal deserts are formed under specific geographic configurations, such as
continentality, orographic effect, and cold oceanic currents. The latter are normally
more sensitive to variations in the heat balance of the Earth and their evolution is
determined by local factors.
Theories on general circulation during the Pleistocene are presented in the excellent work
of Nicholson and Flohn (1980), in which the numerous authors theorize about the trajectory of
cyclones in middle latitudes, on the heat contrast imposed by the presence of large ice masses
with respect to the glacial-pluvial synchronism, and so forth. Nicholson and Flohn (1980)
have carried out the most relevant and synthetic work with regard to climate and
environmental changes in Africa and the general atmospheric circulation in the Pleistocene
and Holocene periods. During the biggest extension of the ice (18,000 yr BP) in the Northern
Hemisphere, the seasonal variations were minimal. Therefore, there was a displacement of the
cyclonic zones toward the south and an increase in the temperature gradient. On the other
hand, the Southern Hemisphere showed characteristics close to present ones.
For these authors, in this epoch and during the greater part of the interval from 20,000 to
12,000 yr BP, the aridity increased in the regions located in the southern Sahara and eastern
Africa. This contradicts the data of Sarnthein (1978), who held that the period of aridity was
about 2000 to 3000 years, around 18,000 years ago. But this supposition does not explain
that, from that date to 12,000 years, wet conditions occurred in northern Africa, in the
massifs of the interior Sahara and in southern Egypt. Nicholson and Flohn (1980) pointed
out that the biggest environmental changes in tropical Africa coincided with the accelerated
growth of the ice sheets (previous to 22,000 yr BP). Circulation systems meant, for most of
the period, aridity in tropical Africa and relatively wet conditions in northern Africa
(Figure 25.41), as a result of the displacement toward the south of the atmospheric
circulation systems.
Figure 25.40. Evolution sequence of the aeolian activity and stabilization periods in the erg Akchar,
Mauritania. (a) Longitudinal dunes formed in the Ogolian phase. (b) Dune stabilized by vegetation and
interdune lake formation during Tchandian and Novakchottian. (c) Dune reactivation and lake desiccation
in the Tafolian. (d) Formation of a new revegeted surface during the most recent period. (e) Modem
reactivation of dunes, which remobilize former ones (after Kocurek et al. (1991)).
Figure 25.41. Inferredpattern of atmospheric circulation over Africa during the last glacial maximum.
Dark shading indicates those areas more humid than today; lighter shading indicates areas considered drier
than today. The inset shows the present position of the ITCZ in summer and winter months (slightly
modified after Nicholson and Flohn (1980)).
Approximately 10,000 to 8000 years ago, there was a change in the circulation system
that produced the first subtropical lacustrine event, as a result of an increase in precipitation
that caused the expansion of the lakes in the southern Sahara, from eastern Africa to Senegal
and Mauritania. Meanwhile, aridity took place in northeastern Africa. These circumstances
are related to a gradual warming in the north of the Atlantic Ocean, a fall in the gradient of
temperature and a maximum drift toward north that displaced the aridity toward the
northeast, as indicated by the advances of dunes in Morocco and western Algeria.
Latter, about 7000 years ago, the Fennoscandian ice sheet disappeared and northern
Europe reached the Climatic Optimum over 6000 yr BP, while the Southern Hemisphere
remained cold. During the period 6500 to 4500 yr BP, both northern and southern Sahara
were more humid than at present, but with an intensity smaller than in the period 10,000 to
8000 yr BP, perhaps due to the fact that the Northern Hemisphere was warmer during this
third period.
The model of Nicholson and Flohn (1980) is the most complete among existing ones,
but presents several contradictions, mainly regarding the interpretation of the Southern
Hemisphere in Africa (Thomas, 1997c). The problem reaches high degrees of complexity
when interpretations are made globally and do not only refer to the African continent.
Without a doubt, a long and hard job must be done, because it is necessary to obtain new
data, reinterpret the past, and to have at one's disposal the highest number of accurate
datas. It is also essential to know the Quaternary climate evolution in the Mediterranean
areas, because this represent the transition zone between Europe and hot Africa. Therefore,
a detailed exploration in arid zones is needed.
2. Tropical regions
2.1. Introduction
Tropical climates include a wide variety of precipitation regimes and vegetation

associations, from selva to savanna and the transitions to semiarid zones. On the other
hand, the activity and intensity of the geomorphological processes change depending on the
zone. The variations in time of these morphoclimatic limits produce the modification of
geomorphologic processes and, therefore, the existing landscapes. Nevertheless, according
to Douglas (1969), humid tropical cores are stable ecosystems in which the energy is used by
vegetation and do not present much water erosion in the rainforest plains.
The most significant studies on climate change in the humid tropics have been carried
out since 1950. Brtickner (1955) in Ghana, and De Ploey (1965) in Zaire, on the basis of
the analysis of superficial deposits, found clear evidence of stages of aridity in the
Pleistocene period. In South America there are ancient dunes in the Llanos (plains) of
Venezuela, in Bahia state of Brazil and in central Amazon (Tricart, 1974a, 1975, 1985),
which show arid conditions in these humid tropical areas. Bigarella and Andrade (1965)
and Bigarella et al. (1969) developed the hypothesis of peneplanation in southeastern
Brazil. The works of Aubreville (1962) and Butzer (1978) analysed the spatial variations
of the limits of the savanna and rainforest because of climate changes. During the last 20
years there has been an important increment in the number of publications on this subject,
but there are not too many long-term chronological records (Thomas, 1994), which force
us to use temporal records of the lakes in the arid zones of Africa and Central America
(Stret-Perrott et al., 1985), because other long-term records are very rare.
For the study of climate change, there are available, apart from the record of semiarid
lakes, data from ice cores in ice sheets and from deep-sea drilling cores. There is an
important limitation, however, due to the C-14 and AMS C-14 techniques because they
can only be applied back to ---60,000 years ago.
Works carried out from cores in the equatorial Atlantic have analysed the content of
freshwater diatoms, opal phytoliths, pollen, and sapropel muds for the last 150,000 years.
From these analyses, up to seven different arid phases can be identified, depending on the
methodology used (Jausen and Van Iperen, 1991).
2.2. Biological modifications
Variations in the configuration of seas and lands, atmospheric circulation, and

physiography can produce important changes in the humidity, in such a form that the
rainforest can be replaced by the savanna and vice versa, and thus, the morphogenesic
changes (Tricart, 1974b). In the lower Congo and in the southern Ivory Coast, the forest
trends toward savanna (Tricart, 1974b). The opposite effect can be observed in Senegal.
On the basis of a review of the forest and savanna zones, Aubreville (1962) found
anomalous ecologic evidence in African humid tropics, as well as in the Guyana. These
anomalies are interpreted by this author as a lack of recuperation of the forest, after the
drier conditions occurred in the last glacial period. He pointed out that forest regions
drifted toward the south (the principal core in Angola), with some forest traces remaining
in the coastal areas of Guinea Gulf and Congo. This interpretation contradicts the idea of a
stable forest. A similar study was carried out by Butzer (1978), in which, during the glacial
optimum (18,000 BP) a large reduction of the extension of tropical forest occurred,
resulting in three bastions: Guinea-Sierra Leona-Liberia, Nigeria-Cameroon, and Zaire.
The work on the distribution of flora and fauna in the forest areas by Hamilton (1976),
provided important data about relevant variations of the central Africa forest in the last
20,000 year (Figure 25.42).
In the South American forest there are curious systems of speciation probably due to
changes in the extension of the forest in the upper Pleistocene era. The forest has been
fragmented into isolated sections in which speciation occurred (Figure 25.43) (Goudie,
1983a; Bradley, 1985; Crowley and North, 1991). In Brazil there is paleogeographic
evidence that supports the existence of xeromorphic plants in cerrados and tabuleiros in
central Brazil and in the western Amazon. The Barreiras Formation, consisting of
quartziferous sand and kaolin, which crops out in the area, was intensively incised during
the last glacial period as a result of the lowering of the base level resulting from the marine
regression. This erosion was favoured by the existence of a dry climate that produced the
disappearance of the forest in the lowlands along the Amazon River. This implies a
reduction of the area of forest during dry stages; there are forest refuges only in high areas
of the Guyana and Brazilian shield, where precipitation is higher. In the Holocene, climate
turned more humid and the Flandrian transgression raised the base level (Dias de Avila-
Pires, 1974; Tricart, 1985).
2.3. Geomorphologic evidences
2.3.1. Regolith and relict weathering profiles
The greatest weathering depths occur in the humid tropics and, normally, these profiles
were developed in the past, although actual climate conditions in the tropics can increase
weathering effects (Ollier and Pain, 1996). Some authors point out that the regolith
thickness in high areas is much smaller than in areas of gentle relief, while others, to the
contrary, indicate that in some cases the greatest depths are located in high areas, and in
the bottom of the valleys the bedrock crops out, because of the fluvial incision (Pain and
Ollier, 1981). From this, it can be deduced that the weathering depth depends on the
weathering intensity and erosion. On the other hand, very thick profiles need deep
groundwater to maintain an active weathering front (Ollier, 1988a,b).
In many known cases, weathering has been developed during the Tertiary and
Cretaceous periods, reaching thicknesses higher than 100 m. Most of these profiles are
j, b ,~
G u i n a ~ f '~~ - . ( ~ J ' /
Liberia - Camerou~ J ~ , J
r/" Limits ~ m~ sand ~ I ~ ~ ~'d'~~
I s~ '
i Main montaneforest refuges ~ ~ #v~.( ~
I ~ Improverishedmontaneforest (locally) ~000" "(a~ - - - - - E P
Breaker
Cho
~ Lowlandforest ~ ~ ~ ~ 1
I I Montaneforest / ~ 8000 B P
t) *
Cap R I
t "t Naturallimitof forest

f Southernlimitof desert
Lowlandforest
I I Montaneforest (C) Present
Figure 25.42. The probable distribution of forest in central Africa at 20,000 and 8000 BP compared with
the present (after Hamilton, 1976). Note that the existence of Greater Lake Chad at 8000 BP is now
questioned.
balanced with the present climate conditions, but have registered many climate changes in
the past, which makes the interpretation difficult. In Chapter 20, examples of different deep
weathering profiles, corresponding to humid tropics and extratropical regions, are shown.
The latter are located in Europe, Australia, India and USA; weathering can even be
Jurassic. From all these works, it can be stated that the principal period of deep weathering
took place in Mesozoic-lower Tertiary, when thick regoliths developed and later eroded as
(A) 80W 70W 60W 50W 40W

10N ,~J- ~---~_. , l
i' j~'/ ~ ~ " ~ r~~~ . ~ . w~176Angi~
lOS
20s
(B) 80W 70W --------l-----'r----

60W 50W 40W
10N
Butterflies
10S
20S
Figure 25.43. Proposed refuge areas for certain species of (A) woody angiosperms and (B) butterflies in
the Amazon Basin during dry climatic phases of the Pleistocene (after Bradley (1985), in Crowley and
North (1991), Figure 3.8).
,.;~~.s~ ' (a)

..:. t." ~.-.
Zq~'c-~'i,:,3~.',.'4~;~N
~ " / 4 , ' ~ . , ~ ' ~ i ~ Pliocene basalt
t~----~"~ ~ -~ ~.-.- ".... th- .~ ~ m . y . )
~Unalteredgranodiorite , ~-. "i.k..~.:..:"...::,..,.f " - "
e III (b)
~"'" ~ ~ " ~ . . ~ ~ , ~ ZoneII

' , ' ~ ....T-- ~"-~ "" \....~- ~ ~..,~.~ Zone I .__m.~.__
" C 1 , -" " ' -- "'""-
Figure 25.44. Diagrammaticrepresentation of conditions at one site in the Mojave Desert of California:
(a) Immediately after emplacement of the basalt. (b) Present conditions, showing the relict weathering
preserved beneath the basalt (Zone I); a partly-stripped surface (Zone II), and exposed relict core stones
forming the modern boulder surface (Zone III) (after Oberlander (1972)).
weathering decreased in intensity (Valeton, 1994). During the Eocene and several stages
in the Cretaceous, Earth was warmer and more humid than at present. Poles were warmer
and temperatures in the tropics were colder (Walker and Sloan, 1992).
The dating of weathering profiles is generally difficult, although there are some
illustrative examples. In the Mojave Desert of California, basalts of the upper Miocene
(8 to 9 million years) overlap weathering profiles developed in granites. At present there
exists a later stage, 9 more arid, eroded part of the basalt and weathered rock (Oberlander,
1972) (Figure 25.44). The analysis of oxygen isotopes in deep regoliths of Australia also
makes it possible to differentiate between regoliths generated in upper Mesozoic and lower
Tertiary and those formed later than middle Tertiary. These differences in the profiles can
be interpreted as the result of the drift of the Australia continent from high to low latitudes,
along the Mesozoic and Tertiary (Bird and Chivas, 1988).
2.3.2. Inselbergs
These landforms are magnificent and difficult to interpret. The main controversy is caused
by the genesis of these landforms. On the one hand, some authors such as King (1948,
1975) and Kesel (1973) consider that inselbergs as subaerial remains are generated from
scarp retreat. On the other, several researchers who stand out from the others; Btidel
(1957), Ollier (1960), Thomas (1965, 1994), Twidale (1964, 1982a,b), and Vidal and
Twidale (1998) maintained that inselbergs are landforms exhumed from thick chemically-
weathered mantles.
Pediplanation, which commonly occurs together with these landforms, in the first
hypothesis is due to the intrinsic development of the landform and, therefore, weathering
is superimposed on the inselberg-pediment system once this generated. In the
exhumation hypothesis, the development of the regolith is the dominant process in the
genesis of the inselbergs, playing in the accompanying pediment process a secondary and
belated role.
Obviously, the development of inselberg landscapes, according to the theory on slope
evolution, occurs in semiarid morphoclimatic zones. The genesis of these landscapes by
exhumation occurs in morphoclimatic systems of humid tropics. The latter explains the
small domal inselbergs (bornhardts), up to 50 m high, simply by denudation of a thick
saprolite. When this hypothesis is applied to the origin of large bornhardts many problems
rise. These difficulties can be solved if a continuous weathering is supposed at the foot of
the slope, as the weathering allows the surrounding plains to be lowered.
In savanna areas, the history is not monoclimatic (Birot, 1978), at least not in
Quaternary times, and, precisely, in those areas is where inselbergs proliferated (Kesel,
1973). The alternation of wet and dry climate stages is what facilitates the development of
these landforms. Furthermore, it is necessary to take into account that once the bornhardt-
plain was established, the contrast in the velocity of the weathering between both stages
assures the persistence of the system (Thomas, 1978).
Inselbergs, in some cases, can have been formed in quite old periods of the geologic
past and have remained until present. This interpretation has been proposed by Twidale
(1978) for the famous Ayers Rock (Uluru), in the desert plains of central Australia. The
author, on the basis of correlative sediment analyses and geomorphologic interpretations,
concluded that the development of that monolith started at the boundary of the Mesozoic-
Cenozoic period.
2.3.3. Fluvial systems
The morphology of the channels and fluvial sediments in the humid tropics records the
effects of environmental fluctuations, data on fluvial regimes as well as the evolution of the
system, because many of these sediments can be dated (Thomas, 1994; Thomas and Thorp,
1995).
The Amazon River, which represents 15% of the total discharge of all the rivers in the
world, has a slope gradient of 0.1 m/km near the Peruvian Andes, decreasing drastically to
0.03 m/km from Manaus, in the centre of the basin, to the mouth, 1400 km distant. These
contrasts in altimetry and gradient clearly affect the morphology of the channels. Rivers that
plough through the Andes, as with the Solimoes, the principal course of the Amazon River,
transport coarse-grained material but mainly sediments held in suspension (Figure 21.4)
and in the alluvial plains channels present low sinuosity. On the other hand, in rivers in the
centre of the fluvial basin, as with the river Juru~i, the sediment load transportation is almost
zero and the channels have a high sinuosity. All that can be interpreted is a function of the
Figure 25.45. Meander of the auriferous river Nzi crossing the woody savanna. The river transports a
great quantity of sediments held in suspension. Eastern Bouak6, Ivory Coast.
capacity of rivers to mobilize sediments in dry periods. These periods coincide with the
glacial periods, in which channels present low sinuosity (Baker, 1978).
In the past, it was thought that tropical rivers did not have terraces, but the exploration
for alluvial placers ended this opinion (Tricart, 1974b). In the majority of the rivers in
Ivory Coast, terraces occur that consist of rounded quartz alluvium, covered by about 2 m
of sticky red clays. These deposits are situated 3 to 4 m over the actual river, and there gold
and a higher proportion of diamonds are exploited (Figure 25.45) (Vogt, 1959).
In the diamond exploitations of the alluvial placers in the centre of Sierra Leone,
alluvial plains and channel fills have been studied by means of stratigraphic profiles and
drill cores. Materials have been dated by C-14, revealing a long history over the last
30,000 years. Moreover, older ferruginated terraces without dating occur in large rivers. In
the subsequent period, black clays of swamp environments can be identified in the margins
of the flood plains (36,000 and 20,500 BP) and gravel channel fills in latter periods. The
last two groups seem to have been formed in arid phases, while the excavation and filling
are related to wet periods in the Holocene (Thomas and Thorp, 1985). In southern Ghana,
a fluvial substrate consisting of weathered slates with filled polygonal cracks of 2 m depth
can be identified. These cracks indicate an intense contraction due to desiccation in dry
environments with almost no vegetation. Fluvial deposits with diamonds were deposited in
the channels, which have been dated at many levels, making possible the identification of
different dry and wet stages (Hall et al., 1985).
Two terraces at 5 to 6 m and 12 m above water level can be identified in the river Juru~i,
in the southwestern part of the Amazon basin. This deposit consists of sands and clays on
which a brownish-red soil has developed. The bottom of the valley is also formed by sandy
Figure 25.46. Basic scheme of slope evolution (Bigarella and Andrade, 1965). Stages depicted are:
(a) extensive intermontane surface formed by pediplanation under arid climate; (b) and (c) slight erosion of
the planated surface caused by a slight lowering of the local base level of erosion due to small climatic
fluctuations toward humid conditions within the arid epoch. (d) Generalized dissection of the planated
surface due to onset of humidity. (e) Valley-widening, alluviation, and colluviation accelerated by short
arid episodes within the humid epoch. (f) Escarpment retreat and formation of a pediment surface under
arid conditions. (g) Slight erosion pediment slopes during slight humid fluctuations within the arid epoch.
(h) Generalized incision under a new humid epoch. (i) Widening and alluviation of valleys caused by
episodic climatic fluctuations trending toward increasing aridity.
clays. As a whole, the data indicate that nowadays there is a phase of incision and,
previously, two stages of alluviation occurred. This alternation seems to be related to
climate changes. In the river Solimoes toward the northwest, the same sequence
of accumulation and incision can be observed and there are some abandoned meanders on
the terrace. Here, the important regression that occurred during the glacial period triggered
the dry climate in which an intense incision developed. Therefore, the origin of these
terraces seems to be eustatic and not climatic (Tricart, 1985).
The classic works of Bigarella and Andrade (1965) and Bigarella et al. (1969) deduced
dry stages in the Pliocene on the basis of analyses on pediments. The study area is located
in the Serra do Mar and its continuation toward southern coastal areas of Parana and Santa
i--.-I
i
Gulf of Guinea .... '~..-,. j i!
O ~ GABON
) '
D,,~O&,,~. X iJ" J ~"~~ZA!RE
10~
(')
--.,,.
20~
30~
[ " "11 Semiarid zone

Town
~ Area underlain by sanda
of the Kalahari System
Figure 25.47. The extent of desert, and deposits of desert origin, in southern Africa. Note that the
northward extent of the Namib Desert is controlled by the extent of the Benguela Current, and that the
Kalahari Sands reach across the Zaire River although they are now covered by forest and savanna
woodlands in the northern area (after Selby et al. (1979)).
Catarina states in Brazil. The authors pointed out that the margins of the valleys are
terraced due to an alternation of dry stages, in which pediments are formed, and humid
stages in which weathering and incision predominate (Figure 25.46). Dry phases
corresponded to glacial periods, whereas wet ones can be correlated with interglacial
stages.
2.3.4. Aeolian action
Aeolian accumulations and landforms generated by deflation in the humid tropics are
excellent indicators of the occurrence of ancient arid stages. Most works on this field have
been carried out in central-southern Africa and in northern and central South America.
In southwestern Africa the aridity is controlled by the cold current of Benguela, which
reached its highest energy during the glacial periods, producing more aridity. It is stated
also that this current produced the drift toward north in the Angola basin, as the cores from
deep Atlantic Ocean confirm. Therefore, the ideas of De Ploey (1965) are reaffirmed,
regarding the aeolian sands of the Kalahari in the western Congo (Figure 25.47). This
extension can be probably explained by the expansion of the Benguela current (Selby et al.,
1979).
The work by Tricart (1985) is a good state-of-the-art review of the information about
dry climates in the centre and north of South America. In the margin of the Llanos del
Orinoco (Orinoco Plains) extensive dune fields have been described (Tricart, 1974a). In
Landsat images it is observed that in those areas are formed by longitudinal dunes with a
N E - S W direction, parabolic dunes also occur. These aeolian accumulations come from
the alluvial cover deposited by piedmont Andean rivers. Their age can be related with the
last glacial period, in which a glacioeustatic lowering occurred. During the Holocene the
climate was warmer and more humid. Dunes have been fossilized by alluvium and become
flooded in rainy epochs. Dunes and landforms produced by deflation, which are partially
fossilized by sediments corresponding to the Flandrian transgression, can be also
recognized near the Orinoco Delta (Tricart and Alfonsi, 1981).
The depression of the Mato Grosso marsh of southwestern Brazil, where current
precipitation is about 1200 mm, have many big alluvial fans coming from Paleozoic
mountains. These landforms seem to indicate a period of dry climate, with scarce
vegetation and dune formation. In the fan Taquari, the greatest, it is possible to identify
deflation basins of up to 3 km diameter. At present, E1 Pantanal, as indicated by its name,
is flooded or colonized by marsh vegetation (Tricart, 1982).
References
Aber, J.S., Croot, D.G., Fenton, M.M., 1989. Glaciotectonic Landforms and Structures. Kluwer,
Dordrecht, p. 200.
Abrahams, A.D., Parsons, A.J., 1991a. Relation between infiltration and stone cover on a semiarid
hillslope, southern Arizona. J. Hydrol. 122, 49-59.
Abrahams, A.D., Parsons, A.J., 199 lb. Relation between sediment yield and gradient on debris-covered
hillslopes, Walnut Gulch, Arizona. Bull. Geol. Soc. Am. 103, 1109-1113.
Abrahams, A.D., Parsons, A.J., (Eds), 1994. Geomorphology of Desert Environments. Chapman & Hall,
London, p. 674.
Abrahams, A.D., Howard, A.D., Parsons, A.J., 1994. Rock mantled slopes. In: Abrahams, A.D.,
Parsons, A.J. (Eds), Geomorphology of Desert Environments. Chapman & Hall, London, pp. 173-212.
Acres, B.D., Blaire Rain, A., King, R.B., Lawton, R.M., Mitchell, A.J.B., Rackham, L.J., 1985. African
dambos: their distribution, characteristics and use. Z. Geomorphol. Suppl.bd. 52, 63-86.
Adams, F.D., 1938. The Birth and Development of the Geological Sciences. Dover, New York, p. 506.
Adams, J., Maslin, M., Thomas, E., 1999. Sudden climate transitions during the quaternary. Prog. Phys.
Geog. 23, 1-36.
Aghassy, J., 1973. Man-induced badlands topography. In: Coates, D.R. (Ed.), Environmental
Geomorphology and Landscape Conservation. Volume III. Non-Urban. Dowden, Hutchinson and
Ross, Stroudsburg, pp. 124-136.
Ahlmann, H.W., 1935. Contribution to the physics of glaciers. Geogr. J. 86, 97-113.
Ahnert, F., 1960. The influence of Pleistocene climates upon the morphology of cuesta scarps on the
Colorado Plateau. Ann. Assoc. Am. Geogr. 50, 139-156.
Ahnert, F., 1970. Functional relationships between denudation relief and uplift in large, mid-latitude
drainage basins. Am. J. Sci. 268, 243-263.
Ahnert, F., 1987a. Approaches to dynamic equilibrium in theoretical simulations of slope development.
Earth Surf. Processes Landforms 12, 3-15.
Ahnert, F., 1987b. An approach to identification of morphoclimates. In: Gardiner, V. (Ed.), International
Geomorphology 1986. Wiley, Chichester, pp. 159-188, Part II.
Ahnert, F., 1996. Introduction to Geomorphology. Arnold, London, p. 352.
Aires-Barros, L., 1975. Dry and wet laboratory tests and thermal fatigue of rocks. Eng. Geol. 9,
249-265.
Aitchison, G.D., Wood, C.C., 1965. Some interactions of compaction, permeability, and post-construction
deflocculation affecting the probability of piping failure in small earth dams, Proceedings 6th Inter-
national Conference on Soil Mechanics and Foundation Engineering, Montreal, Vol. 2, pp. 442-446.
Alberto, F., 1989. La desertizaci6n por salinizacirn en el Valle del Ebro. Azara 1, 45-53.
Albritton, C.C., Brooks, J.E., Issawi, B., Swedan, A., 1990. Origin of the Qattara depression, Egypt. Bull.
Geol. Soc. Am. 102, 952-960.
A1-Durrah, M., Bradford, J.M., 1982. New methods of studying soil detachment due to raindrop impact.
J. Soil Sci. Soc. Am. 45, 949-953.
Aleva, G.J.J., 1985. Indonesian fluvial casiterite placers and their genetic environments. J. Geol. Soc.
London 142, 815-836.
Aleva, G.J.J., 1991. Tropical weathering, denudation and mineral accumulation. Geol. Mijnbouw 70,
35-38.
Alexander, R.W., Calvo, A., 1990. The influence of lichens on slope processes in some Spanish badlands.
In: Thornes, J.B. (Ed.), Vegetation and Erosion. Wiley, Chichester, pp. 385-398.
Alexandre, L.T., Cady, J.G., 1962. Genesis and hardening of laterite in soils. Tech. Bull. 1282, 90, United
States Department of Agriculture, Soil Conservation Service.
Allen, J.G., 1972. Aluminium ore deposits. In: Fairbridge, R.W. (Ed.), The Encyclopedia of Geochemistry
and Environmental Sciences. Van Nostrand Reinhold, New York, pp. 23-27.
Allen, J.R.L., 1965. A review of the origin and characteristics of recent alluvial sediments. Sedimentology
5, 89-191.
Allen, J.R.L., 1971. Transverse erosional marks of mud and rock: their physical basis and geological
significance. Sediment. Geol. 5, 1-38.
Allison, R.J., 1997. Middle East and Arabia. In: Thomas, D.S.G. (Ed.), Arid Zone Geomorphology:
Process, Form and Change in Drylands. Wiley, Chichester, pp. 507-521.
Allison, R.J., Thomas, D.G., 1993. The sensivity of landscapes. In: Thomas, D.S.G., Allison, R.J. (Eds),
Landscape Sensivity. Wiley, Chichester, pp. 1-5.
Ambrose, J.W., 1964. Exhumed paleoplains of the Precambrian shield of North America. Am. J. Sci. 262,
815-857.
American Geological Institute, 1972. Glossary of Geology. American Geological Institute, Washington,
p. 805.
Anad6n, P., 1989. Lagos. In: Arche, A. (Ed.), Sedimentologfa. Consejo Superior de Investigaciones
Cientfficas, Madrid, pp. 219-276.
Anderson, B., 1961. The Rufiji Basin, Tanganyka. 7. Soils of main irrigatable areas. Rept. Govt.
Tanganyka, Prelim. Reconnaissance Surv. Rufiji Basin, 125 pp.
Anderson, R.S., 1986. Erosion profiles due to particles entrained by wind: application of an eolian
sediment transport model. Bull. Geol. Soc. Am. 97, 1270-1278.
Anderson, R.S., 1987. A theoretical model for aeolian impact ripples. Sedimentology 34, 943-956.
Anderson, R.S., 1989. Saltation of sand: a qualitative review with biological analogy. In: Gimminghang, C.H.,
Richtie, W., Willets, B.B., Willis, A.J. (Eds), Symposium: Coastal Sand Dunes, Royal Society of
Edinburgh, Proceeding, Vol. B96, pp. 149-165.
Anderson, R.S., Bunas, K.L., 1993. Grain size segregation and stratigraphy in aeolian ripples modelled
with a cellular automaton. Nature 365, 740-743.
Andersson, J.G., 1906. Solifuction; a component of subaerial denudation. J. Geol. 14, 91-112.
Andersson, R.S., Haft, P.K., 1988. Simulation of aeolian saltation. Science 241,820-823.
Andrade, E.R., Sellers, W.D., 1988. E1 Nifio and its effect on precipitation in Arizona and western
New Mexico. J. Climatol. 8, 403-410.
Andr6, M.F., 1993. Les Versants du Spitsberg. Presses Universitaires de Nancy, Nancy, p. 361.
Andrews, J.T., 1970. A geomorphological study of post-glacial uplift with particular reference to Arctic
Canada. Inst. Br. Geogr. Spec. Publ. 2, 156.
Andrews, J.T., 1975. Glacial Systems. An Approach to Glaciers and their Environments. Duxbury Press,
North Scituate, MA, p. 191.
Andrews, J.T., Dugdale, R.E., 1971. Quaternary history of Northern Cumberland Peninsula, Baffin Island,
N.W.T.: Part V: factors affecting corrie glacierization in Okoa Bay. Quaternary Res. 1,532-551.
Andrews, J.T., Smithson, B.B., 1966. Till fabric of the cross-valley moraines of north-central Baffin Island.
Bull. Geol. Soc. Am. 77, 271-290.
Anisimov, D.A., Nelson, F.E., 1996. Permafrost distribution in the Northern Hemisphere under scenarios
of climatic change. Global Planet. Change 14, 59-72.
Antevs, E., 1952. Arroyo cutting and filling. J. Geol. 60, 375-385.
Arauzo, T., Guti6rrez, M., 1994. Evoluci6n de los valles de fondo piano del centro de la Depresi6n del
Ebro. In: Arn~iez, J., Garcfa-Ruiz, J.M., Gdmez Villar, A. (Eds), Geomorfologfa en Espafia, Logrofio,
Vol. 1, pp. 277-290.
Arauzo, T., Guti6rrez, M., Sancho, C., 1996a. Facetas triangulares de ladera como indicadores
paleoclim~iticos en ambientes semifiridos (Depresi6n del Ebro). Geogaceta 20, 1093-1095.
Arauzo, T., Guti6rrez-Elorza, M., Sancho, C., 1996b. Retroceso de escarpes en ambientes semifiridos a
partir de facetas triangulares de ladera (Depresi6n del Ebro). Cadernos do Laboratorio Xeol6xico de
Laxe 21,405-416, Sada (A Corufia).
Aristarain, L.F., 1970. Chemical analysis of caliche profiles from high plains, New Mexico. J. Geol. 78,
201-212.
Aristarain, L.F., 1971. On the definition of caliche deposits. Z. Geomorphol. 15, 274-289.
Arnborg, L.B., Walker, H.J., Peippo, J., 1967. Suspended load in the Colville River, Alaska, 1962. Geogr.
Ann. 49, 131-144.
Artieda, O., 1993. Factores geol6gicos que inciden en el desarrollo de los suelos en un medio semifirido. E1
Caso de Quinto (Zaragoza). Tesis de Licenciatura, p. 305, Facultad de Ciencias. Zaragoza (In6dita).
References 699
Ash, J.E., Wasson, R.J., 1983. Vegetation and sand mobility in the Australian desert dunefield.
Z. Geomorphol. Suppl.bd. 45, 7-25.
Ashida, K., 1985. Debris Flow. Disaster on Alluvial Fans. Kokon-shoin, Tokyo, p. 224.
Aubreville, A., 1962. Savanisation tropicale et glaciation quaternaires. Adansonia 2, 16-84.
Aufrrre, M.L., 1928. L'orientation des dunes et la direction des vents. C. R. Acad. Sci. Paris 187,
833-835.
Aufr~re, M.L., 1935. Essai sur les dunes du Sahara algrrien. Geogr. Ann. 18, 481-550.
Augustinus, P.C., 1992. Outlet glaciar trough size-drainage area relationships. Fiordland, New Zealand.
Geomorphology 4, 347-361.
Aune, Q.A., 1983. Quick clays and California's clays: no quick solution. In: Tank, R.W. (Ed.),
Environmental Geology. Oxford University Press, New York, pp. 145-150.
Aylsworth, J.M., Shilts, W.W., 1989. Bedforms of the keewatin ice sheet, Canada. Sediment. Geol. 62,
407-428.
Bagnold, R.A., 1941. The Physics of Blown Sand and Desert Dunes. Methuen, London, p. 265.
Bagnold, R.A., 1953. The surface movement of blown sand in relation to meteorology. Res. Counc. Israel,
Spec. Publ. 2, 89-96.
Baker, A.A., 1936. Geology of the Monument Valley-Navajo Mountain region, San Juan County, Utah.
US Geol. Surv. 865, 106.
Baker, V.R., 1977. Stream channel response to floods with examples from Central Texas. Bull. Geol.
Soc. Am. 88, 1057-1071.
Baker, V.R., 1978. Adjustment of fluvial system to climate and source terrain in tropical and subtropical
environments. In: Miall, A.D. (Ed.), Fluvial Sedimentology, Canadian Society of Petroleum Geologists,
Memoir 5, pp. 211-230.
Baker, V.R., (Ed.), 1981b. Catastrophic Flooding. The Origin of the Chanelled Scabland. Dowden,
Hutchinson and Ross, Stroudsburg, p. 360.
Baker, V.R., 1993. Extraterrestrial geomorphology: science and philosophy of Earthlike planetary
landscapes. Geomorphology 7, 9-35.
Baker, V.R., Bunker, R.C., 1985. Cataclysmic late Pleistocene flooding from glacial Lake Missoula: a
review. Quaternary Sci. Rev. 4, 1-41.
Baker, V.R., Twidale, C.R., 1991. The reenchantment of geomorphology. Geomorphology 4, 73-100.
Baker, V.R., Kochel, R.C., Laity, J.E., Howard, A.D., 1990. Spring sapping and valley network
development. In: Higgins, C.G., Coates, D.R. (Eds), Groundwater Geomorphology, Geological Society
of America Special Paper, 252, pp. 235-265.
Baker, V.R., Benito, G., Rudey, A.N., 1993. Paleohydrology of late Pleistocene superflooding, Altay
Mountains, Siberia. Science 259, 348-350.
Bakker, J.P., 1965. A forgotten factor in the development of glacial stairways. Z. Geomorphol. 9,
18-34.
Bakker, J.P., 1967. Weathering of granites in different climates. In: Macar, P. (Ed.), L'Evolution des
Versant, Congrbs et Colloque Universit6 Lirge, Vol. 40, pp. 51-68.
Bakker, J.P., Levelt, Th.W.M., 1964. An inquiry into the probability of a polyclimatic development of
peneplains and pediments (etchplains) in Europe during the Senonian and Tertiary period. Publication
Service Carte Grologique, Luxembourg 14, 27-75.
Balazs, D., 1973. Relief types of tropical karst areas. In: Jacucs, L. (Ed.), Symposium on Karst-
Morphogenesis. Attila Joseph University, Szeged, pp. 16-32.
Ballantyne, C.K., Kirkbride, M.P., 1987. Rockfall activity in upland Britain during the Loch Lomond
stadial. Geogr. J. 153, 86-92.
Balling, R.C., Wells, S.G., 1990. Historical rainfall patterns and arroyo activity within the Zuni River
drainage basin, New Mexico. Ann. Assoc. Am. Geogr. 80, 603-617.
Barbary, J.P., Maire, R., Shouyuz, Z., 1991. Grottes et Karsts Tropicaux de Chine Mrridionale, Vol. 4.
Karstologia, Mrmoire n~ Venissieux, p. 232.
Barbeau, J., 1961. Morphologie du Quaternaire des abords orientaux du Lac Tchad. Bulletin Institute
Equatoriale Recherches Geologiques et Mini~res 14, 73-82.
Barbier, R., 1967. Nouvelles rrflexions sur le probl~me des "pains de sucre" h propos d'observations dans
le Tassili N'Ajjer (Algrrie). Travaux Laboratoire de Grologie. Facult6 des Sciences de Grenoble 43,
15-21.
Bardossy, G., 1981. Karst Bauxites. Akadrmiai Kiad6, Budapest, p. 480.
Bardossy, J., Aleva, G.J.J., 1990. Lateritic Bauxites. Elsevier, Amsterdam, p. 624.
Barry, R.G., 1997. Paleoclimatology, climate system processes and the geomorphic record. In: Stoddart, D.R.
(Ed.), Process and Form in Geomorphology. Routledge, London, pp. 187-214.
Barry, R.G., Chorley, R.J., 1987. Atmosphere, Weather and Climate. Methuen, London.
Barsch, D., 1977a. Eine absch/itaung von schottproduktion und schutttransport in bereich activer
blockgletscher der schiveizer Alpen. Z. Geomorphol. Suppl.bd. 28, 148-160.
Barsch, D., 1977b. Nature and importance of mass-wasting by rock glaciers in the alpine permafrost
environments. Earth Surf. Processes 2, 231-245.
Barsch, D., 1988. Rockglaciers. In: Clark, M.J. (Ed.), Advances in Periglacial Geomorphology. Wiley,
Chichester, pp. 69-90.
Barsch, D., 1990. Geomorphology and geoecology. Z. Geomorphol. Suppl.bd. 79, 39-49.
Barsch, D., 1996. Rock Glaciers. Springer, Berlin, p. 331.
Barsch, D., Caine, N., 1984. The nature of mountain geomorphology. Mt. Res. Dev. 4, 287-298.
Barsch, D., Hell, G., 1975. Photogrammetrische bewegungsmessungen am blockgletscher Murt61
Oberengadin, Schweizer Alpen. Zeitschrift fiir Gletscherkunde und Glazialgeologie 9, 111-142.
Bateman, A.M., 1950. Economic Mineral Deposits. Wiley, New York.
Baulig, H., 1957. Peneplains and pediplains. Bull. Geol. Soc. Am. 68, 913-920.
Baulin, V.V., Danilova, N.S., 1984. Dynamics of late quaternary permafrost in Siberia. In: Velichko, A.A.
(Ed.), Late Quaternary Environments of the Soviet Union. Longman, London, pp. 69-78.
Baumer, M., 1990. Agroforestry and Desertification. Technical Centre for Agricultural and Rural
Cooperation, p. 249, Wageningen.
Beaty, C.B., 1974. Debris flow, alluvial fans and a revitalized catastrophism. Z. Geomorphol. Suppl.bd. 21,
39-51.
Beaumont, P., 1993. Drylands. Environmental Management and Development. Routledge, London, p. 536.
Beckinsale, R.F., Chorley, R.J., 1968. Geomorphology, history of. In: Fairbridge, R.W. (Ed.), The
Encyclopedia of Geomorphology. Dowden, Hutchinson and Ross, Stroudsburg, pp. 410-415.
Beckinsale, R.F., Chorley, R.J., 1991. The History of the Study of Landforms on the Development of
Geomorphology. Vol. 3: Historical and Regional Geomorphology 1890-1950. Routledge, London,
p. 496.
Begin, A.B., Schumm, S.A., 1984. Gradational thresholds and landforms singularity: significance for
quaternary studies. Quaternary Res. 21,267-274.
Bender, M., Sowers, T., Dickson, M.L., Orchado, J., Grooters, P., Mayewski, P.A., Meese, D.A., 1994.
Climate connection between Greenland and Antarctica during the last 100,000 yr. Nature 372,
663-666.
Benedict, J.B., 1970. Frost cracking in the Colorado Front Range. Geogr. Ann. 52, 87-93.
Benito, G., 1987. Karstificaci6n y colapsos kfirsticos en los yesos del sector de la Depresi6n del Ebro.
Cuaternario y Geomorfologfa 1, 61-76.
Benito, G., 1997. Energy expenditure and geomorphic work of the cataclysmic Missoula flooding in the
Columbia River George, USA. Earth Surf. Processes Landforms 22, 457-472.
Benito, G., Guti6rrez, M., 1988. Karst in gypsum and its environmental impact on the Middle Ebro Basin,
Spain. Environ. Geol. Water Sci. 12, 107-111.
Benito, G., P6rez del Campo, P., 1991. Sinkhole evolution in alluvial deposits within the Central Ebro
Basin, Northeast Spain. In: Johnson, A.I. (Ed.), Land Subsidence, International Association of
Hydrological Sciences, 200, pp. 323-331.
Benito, G., Guti6rrez, M., Sancho, C., 1993. The influence of physico-chemical properties on erosion
processes in badland areas, Ebro basin, NE-Spain. Z. Geomorphol. 37, 199-214.
Benito, G., Machado, M.J., Sancho, C., 1993. Sandstone weathering processes damaging prehistoric rock
paintings at the Albarracfn Cultural Park, NE Spain. Environ. Geol. 22, 71-79.
Benito, G., P6rez del Campo, P., Guti6rrez, M., Sancho, C., 1995. Natural and human-induced sinkholes in
gypsum terrain and associated environmental problems in NE Spain. Environ. Geol. 25, 156-164.
Benn, D.I., 1995. Fabric signature of subglacial till deformation Breidamerkurj6kull Iceland.
Sedimentology 42, 735-747.
Bennet, M.R., Glasser, N.F., 1996. Glacial Geology: Ice Sheets and Landforms. Wiley, Chichester,
p. 364.
Bennet, M.R., Hambrey, M.J., Huddart, D., Glasser, N.F., 2000. Resedimentation of debris on ice-cored
lateral moraine in the high-Artic (Konsvegen, Svalbard). Geomorphology 35, 21-40.
Benson, L.V., 1978. Fluctuation in the level of pluvial Lake Lahotan during the last 40,000 years.
Quaternary Res. 9, 300-318.
References 701
Benson, L.V., Paillet, F.L., 1989. The use of total lake-surface area as an indicator of climatic change:
examples from the Lahotan Basin. Quaternary Res. 32, 262-275.
Berry, L., Ruxton, B.P., 1959. Notes on weathering zones and soils on granite rocks in two tropical regions.
J. Soil Sci. 10, 54-73.
Beskow, G., 1930. Erdfliessen und Structurb6den der Hochgebirge im Licht der Frostheburg. Geol. F6ren.
Stockholm, 52, 622-638.
Besler, H., 1982. The north-eastern Rub' al Kh~ili within the borders of the United Arab Emirates.
Z. Geomorphol. NF 26, 495-505.
Bickerton, R.W., Matthews, J.A., 1992. On the accuracy of lichenometric dates: an assessment
based on the "little ice age" moraine sequence of Nigardsbreen, southern Norway. Holocene 2,
227-237.
Bigarella, J.J., Andrade, G.O. de, 1965. Contribution to the study of the Brazilian quaternary. Geol. Soc.
Am. Spec. Paper 84, 435-451.
Bigarella, J.J., Becker, R.D., (Eds), 1975. International symposium on the quaternary, Boletim Paranaense
de Geoci~ncias, p. 33.
Bigarella, J.J., Mousinho, M.M., Da Silva, J.X., 1969. Processes environments of the Brazilian quaternary.
In: P4w6, T.L. (Ed.), The Periglacial Environment. McGill-Queen's University Press, Montreal,
pp. 417-487.
Bird, M.I., Chivas, A.R., 1988. Oxygen isotope dating of the Australian regolith. Nature 331,513-516.
Birkeland, P.W., 1974. Pedology, Weathering and Geomorphology. Oxford University Press, New York,
p. 285.
Birkeland, P.W., 1984. Soils and Geomorphology. Oxford University Press, New York, p. 372.
Birot, P., 1949a. Essai sur Quelques Probl6mes de Morphologie G6nerale, Instituto para a Alta Cultura.
Centro de Estudos Geogrfificos, Lisboa, p. 176.
Birot, P., 1949b. Les Surfaces d'erosion du Portugal Central et Septentrional, Rapport de la Commission
pour la Cartographie des surfaces d' aplanissement, Union Geographique International. Congr6s Inter-
national de G6ographie, Lisboa, pp. 9-116.
Birot, P., 1954. Probl6mes de morphologie karstique. Ann. G~ogr. 63, 161-192.
Birot, P., 1955. Les M6thodes de la Morphologie. Presses Universitaires de France, Paris, p. 177.
Birot, P., 1958. Les d6mes crystallines. Mem. Doc., Cent. Natl Rech. Sci. 6, 8-34.
Birot, P., 1960. Le Cycle d'Erosion sous les Diff6rents Climats, Centro de Pesquisas de Geograffa do
Brasil. Faculdade Nacional de Filosofia, Rio de Janeiro, p. 137.
Birot, P., 1966. Le Relief Calcaire. Centre de Documentation Universitaire, Paris, p. 238.
Birot, P., 1968. Contribution a l'lEtude de la D4sagr6gation des Roches. Centre de Documentation
Universitaire, Paris, p. 232.
Birot, P., 1973. G6ographie Physique G6n6rale de la Zone Intertropicale. Centre de Documentation
Universitaire, Paris, p. 280.
Birot, P., 1978. Evolution des conceptions sur la gen6se des inselbergs. Z. Geomorphol., Suppl.bd. 31,
42-63.
B irot, P., Dresch, J., 1966. P6diments et glacis l'Ouest des l~tats-Unis. Ann. G4ogr. 411, 513-552.
Birot, P., So14, L., 1954a. Recherches morphologiques dans le Nord-Ouest de la Peninsule Iberique.
Memoires et Documents du Centre National de la Recherche Scientifique 4, 9-61, Paris.
Birot, P., Sol~, L., 1954b. Investigaciones sobre la Morfologfa de la Cordillera Central Espafiola, Instituto
Juan Sebastifin Elcano. Consejo Superior de Investigaciones Cientfficas, Madrid, p. 87.
Bjornsson, H., 1979. Glaciers in Iceland. J6kull 29, 74-80.
Bj6rnsson, H., 1992. J6kulhlaups in Iceland: prediction, characteristics and simulation. Ann. Glaciol. 18,
95-106.
Black, R.F., Barksdale, W.L., 1949. Oriented lakes of Northern Alaska. J. Geol. 57, 105-118.
Blackwelder, E., 1925. Exfoliation as a phase of rock weathering. J. Geol. 33,793-806.
Blackwelder, E., 1928. Mudflow as a geologic agent in semiarid mountains. Bull. Geol. Soc. Am. 39,
465 -484.
Blackwelder, E., 1931. Desert plains. J. Geol. 39, 133-140.
Blair, R.W., 1986. Karst landforms and lakes. In: Short, N.M., Blair, R.W. (Eds), Geomorphology from
Space. A Global Overview of Regional Landforms. NASA, Washington, pp. 407-446.
Blair, T.C., 1987. Sedimentary processes, vertical stratification sequences, and geomorphology of
the Roaring River alluvial fan, Rocky Mountain National Park, Colorado. J. Sediment. Petrol. 57,
1-18.
Blair, T.C., McPherson, J.G., 1994a. Alluvial fans and their natural distinction from rivers based on
morphology, hydraulic processes, sedimentary processes, and facies assemblages. J. Sediment. Res.
A64, 450-489.
Blair, T.C., McPherson, J.G., 1994b. Alluvial fan processes and forms. In: Abrahams, A.D., Parsons, A.J.
(Eds), Geomorphology of Desert Environments. Chapman & Hall, London, pp. 354-402.
Blake, D.H., Paijmans, J.K., 1973. Landforms types of eastern Papua and their associated characteristics.
Land Res. Ser. CSIRO Aust. 32, 11-21.
Bland, W., Rolls, D., 1998. Weathering: An Introduction to the Scientific Principles. Arnold, London,
p. 271.
Blissenbach, E., 1952. Relation of surface angle distribution to particle size distribution on alluvial fans.
J. Sediment. Petrol. 22, 25-28.
Blissenbach, E., 1954. Geology of alluvial fans in semiarid regions. Bull. Geol. Soc. Am. 65, 175-190.
Blume, H., Barth, H.K., 1972. Rampenstufen und Schuttrampen als Abtragungsformen in ariden
Schichtstufenlandschaften. Erkunde 26, 108-116.
Bltimel, W.D., 1981. Pedologische und geomorphologische aspekte der kalkkrustenbildung in
stidwesafrika und Stidtospanien, Karlsruher Geographische Hefte, Vol. 10, p. 228.
Bltimel, W.D., 1982. Calcretes in Namibia and SE Spain - relations to substratum, soil formation and
geomorphic actors. Catena Suppl. 1, 67-95.
Boast, R., 1990. Dambos: a review. Prog. Phys. Geog. 14, 153-177.
Boer, M.M., de Groot, R.S., 1990. Landscape-Ecological Impact of Climatic Change. IOS Press,
Amsterdam, p. 429.
Brgli, A., 1980. Karst Hydrology and Physical Speleology. Springer, Berlin, p. 270.
Bolt, B.A., Horn, W.L., Macdonald, G.A., Scott, R.F., 1975. Geological Hazards. Springer, Berlin, p. 328.
Bond, G., Broecker, W., Johnsen, S., McManus, J., Labeyrie, L/, Jozuel, J., Bonani, G., 1993. Correlations
between climate records from North Atlantic sediments and Greenland ice. Nature 365, 143-147.
Bond, G., Showers, W., Cheseby, M., Lotti, R., Almasi, P., deMenocal, P., Priore, P., Cullen, H., Hajdas,
I., Bonani, G., 1997. A pervasive millenial-scale cycle in North Atlantic Holocene and glacial climates.
Science 278, 1257-1265.
Bondurant, D.C., 1951. Sedimentation studies at Conchas Reservoir in New Mexico. Trans. Am. Soc. Civ.
Eng. 116, 1292-1295.
Bonython, C.W., Mason, R., 1953. The filling and drying of Lake Eyre. Geogr. J. 119, 321-330.
Boothroyd, J.C., Nummedal, D., 1978. Proglacial braided outwash: a model for humid alluvial-fan
deposits. In: Miall, A.D. (Ed.), Fluvial Sedimentology. Canada Society Petroleum Geology, Calgary,
pp. 641-668, Memoire 5.
Bordonau, J., 1993. The upper Pleistocene ice-lateral till complex of Cerler (Esera Valley, central
Southern Pyrenees: Spain). Quaternary Int. 18, 5-14.
Borja, F., Zazo, C., Dabrio, C.J., Diaz del Olmo, F., Goy, J.L., Lario, J., 1999. Holocene eolian facies and
human settlements along the Atlantic coast of southern Spain. Holocene 9, 333-339.
Bornhardt, W., 1900. Zur Oberflachengestaltung und Geologie Deutsch-Ostafrikas. Reimer, Berlin.
Boulangr, B., Millot, G., 1988. La distribution des bauxites sur le craton Ouest-Africain. Bull. Sci. Grolog.
41,113-123, Strasbourg.
Boulton, G.S., 1970a. On the origin and transport of englacial debris in Svalbard glaciers. J. Glaciol. 9,
213-229.
Boulton, G.S., 1970b. On the deposition of subglacial and melt-out tills at the margins of certain Svalbard
glaciers. J. Glaciol. 9, 231-245.
Boulton, G.S., 1971. Till genesis and fabric in Svalbarg, Spitsbergen. In: Goldthwait, R.P. (Ed.), Till: A
Symposium. Ohio State University Press, Columbus, pp. 41-72.
Boulton, G.S., 1972. Modern arctic glaciers as depositional models for former ice sheets. J. Geol. Soc.
London 128, 361-393.
Boulton, G.S., 1974. Processes and patterns of glacial erosion. In: Coates, D.R. (Ed.), Glacial
geomorphology. State University of New York, Binghamton, pp. 41-87.
Boulton, G.S., 1975. Processes and patterns of subglacial sedimentation: a theoretical approach. In: Wright,
A.E., Moseley, E. (Eds), Ice Ages: Ancient and Modern. Seel House Press, Liverpool, pp. 7-42.
Boulton, G.S., 1976. The origin of glacially fluted surfaces-observation and theory. J. Glaciol. 17,
287-309.
Boulton, G.S., 1978. Boulder shapes and grain-size distributions of debris as indicators of transport paths
through a glacier and till genesis. Sedimentology 23, 15-37.
References 703
Boulton, G.S., 1982. Subglacial processes and the development of glacial bedforms. In: Davidson-Arnott,
R., Nickling, W., Fahey, B.D. (Eds), Research in Glacial, Glacio-Fluvial and Glacio-Lacustrine
Systems. Geobooks, Norwich, pp. 1- 31.
Boulton, G.S., 1987. A theory of drumlin formation by subglacial sediment deformation. In: Menzies, J.,
Rose, J. (Eds), Drumlin Symposium. Balkema, Rotterdam, pp. 25-80.
Boulton, G.S., 1993. Ice ages and climatic change. In: Duff, P.McL.D. (Ed.), Holmes' Principles of
Physical Geology. Chapman & Hall, London, pp. 439-469.
Boulton, G.S., Vivian, R., 1973. Underneath the glaciers. Geogr. Mag. 45, 311-319.
Bowen, N.L., 1928. The Evolution of Igneous Rocks. Dover Publications, New York, p. 332.
Bowler, J.M., 1973. Clay dunes: their occurrence, formation and environmental significance. Earth Sci.
Rev. 9, 315-338.
Bowler, J.M., 1986. Spatial variability and hydrologic evolution of Australian lake basins: analogue for
Pleistocene hydrologic change and evaporate formation. Palaeogeogr, Palaeocl. Palaeoecol. 54, 21-41.
Bowler, J.M., McGee, J.W., 1978. Geomorphology of the Malle region in semiarid northern Victoria and
Western New South Wales. Proc. R. Soc. Victoria 90, 5-20.
Bowley, W.W., Burghardt, M.D., 1971. Thermodynamics and stones, EOS. Am. Geophys. Union Trans.
52, 4-7.
Bowler, J. M., Wasson, R. J., 1984. Glacial ages environments of inland Australia. In: Vogel, J. C. (Ed.).
Late Cainozoic Palaeoclimates of the Southern Hemisphere, Balkema, Rotterdam, pp. 183-208.
Bowyer-Bower, T.A.S., Bryan, R.B., 1986. Rill initiation: concepts and evaluation on badland slopes.
Z. Geomorphol., Suppl.bd. 59, 161 - 175.
Boy6, M., 1949. Importance du d6fongage p~riglaciaire dans l'elaboration des formes glaciaires. Compte
Rendus de l'Acad6mie des Sciences de Paris 229, 723-724.
Boy6, M., Frisch, P., 1973. D6gagement artificiel d'un d6me crystallin au Sud-Cameroun. Travaux et
Documents du G6ographie Tropicale, Bordeaux 8, 31-62.
Boyle, R.W., 1979. The geochemistry of gold and its deposits. Bull. Geol. Surv. Can. 280, 584.
Bradley, R.S., 1985. Quaternary Paleoclimatology: Methods of Paleoclimatic Reconstruction. Allen &
Unwin, Boston, MA.
Bradley, W.C., 1963. Large-scale exfoliation in massive sandstones of the Colorado Plateau. Bull. Geol.
Soc. Am. 74, 519-528.
Bradley, W.C., Hutton, J.T., Twidale, C.R., 1978. Role of salts in development of granitic tafoni, South
Australia. J. Geol. 86, 647-654.
Brakenridge, G.R., Shuster, J., 1986. Late quaternary geology and geomorphology in relation to
archaeological site locations, southern Arizona. J. Arid Environ. 10, 225-239.
Brammer, H., 1990. Floods in Bangladesh: I. Geographical background to the 1987 and 1988 floods.
Geogr. J. 156, 12-22.
Brand, E.W., Premchitt, J., Phillipson, H.B., 1985. Relationship between rainfall and landslides in Hong
Kong, IV International Symposium on Landslides, Toronto, 1984, Vol. 1, pp. 377-384.
Breed, C.S., Grow, T., 1979. In: McKee, E.D. (Ed.), Morphology and distribution of dunes in sand seas
observed by remote sensing. United States Geological Survey, Washington, pp. 253-303, Professional
Paper, 1052.
Breed, C.S., McCauley, J.F., Davis, P.A., 1987. Sand sheets of the eastern Sahara and ripple blankets on
Mars. In: Frostick, L.E., Reid, I. (Eds), Desert Sediments, Ancient and Modem. Blackwell, Oxford,
pp. 337-360.
Breed, C.S., McCauley, J.F., Whitney, M.I., 1989. Wind erosion forms. In: Thomas, D.S.G. (Ed.), Arid
Zone Geomorphology. Belhaven Press, London, pp. 284-307.
Breed, C.S., McCauley, J.F., Whitney, M.I., Tchakarian, V.P., Laity, J.E., 1997. Wind erosion in drylands.
In: Thomas, D.S.G. (Ed.), Arid Zone Geomorphology: Process, Form and Change in Drylands. Wiley,
Bremer, H., 1971. Fltisse, Fl~ichen- und Stufenbildung in den feuchten Tropen. Wuzburger Geographische
Arbeiten 35, 194.
Bremer, H., 1983. Albrecht Penck (1858-1945) and Walther Penck (1888-1923), two German
geomorphologists. Z. Geomorphol. 27, 129-138.
Bremer, H., 1985. Soil and slope development in the wet zone of Sri Lanka. In: Douglas, I., Spencer, T.
(Eds), Environmental Change and Tropical Geomorphology. Allen & Unwin, London, pp. 295-302.
Bremer, H., 1996. Climatic, climatogenetic and tectonic geomorphology. In: Mc Cann, S.B., Ford, D.C.
(Eds), Geomorphology sans Frontibres. Wiley, Chichester, pp. 34-39.
Bretz, J.H., 1923. The channeled scablands of the Columbia Plateau. J. Geol. 31, 617-649.
Bretz, J.H., 1969. The Lake Missoula floods and the channeled scablands. J. Geol. 72, 505-543.
Brice, J.L., 1966. Erosion and deposition in the loess mantled great plains, medicine creek drainage basin,
Nebraska. US Geol. Surv., Prof. Paper 350 H, 255-339.
Briere, P.R., 2000. Playa, playa lake, sebkha. Proposed definitions for old terms. J. Arid Environ. 45,
1-7.
Brimhall, G.H., Lewis, C.J., Ague, J.J., Dietrich, W.E., Hampel, J., Teague, T., Rix, C., 1988. Metal
enrichment in bauxites by deposition of chemically nature aeolian dust. Nature 333, 819-824.
Brodzikowski, K., Van Loon, A.J., 1991. Glacigenic Sediments. Elsevier, Amsterdam, p. 674.
Broecker, W.S., 1994. Massive iceberg discharges as triggers for global climate change. Nature 372,
421-424.
Broecker, W.S., Denton, G.H., 1990. What drives glacial cycles? Sci. Am. January, 43-50.
Brooks, G.A., 1978. A new approach to the study of inselberg landscapes. Z. Geomorphol. Suppl.bd. 31,
138-160.
Brosche, K.U., 1971. Neue Beobacktungen zu vorzeitslichen periglaziaterscheinungen in Ebrobecken.
Z. Geomorphol. 15, 107-114.
Brosche, K.U., 1972. Vorzeitliche Periglazialerscheinunger im Ebrobecken in der Ungebung voz Zaragoza
sowie ein Beitrag zur Ausdehbung von Schuttund Bloakdecken in Zentral-und-W-Teil der Iberische
Halbinsel. G6ttingen Geographische Abh. 60, 293-316.
Brown, R.J.E., 1970. Permafrost in Canada; its Influence on Northern Development. University of Toronto
Press, Toronto, p. 234.
Brtickner, W.D., 1955. The mantle rock (laterite) of the Gold Coast and its origin. Geol. Rundsch. 43,
307-327.
Brum Ferreira, A. de, 1978. Planaltos e Montanhas do Norte da Beira. Estudo de Geomorfologia, Vol. 4.
Mem6rias do Centro de Estudos Geogr~ificos, Lisboa, p. 374.
Brunner, F.K., Scheidegger, A.E., 1974. Kinematics of a scree slope. Rev. Ital. di Geofis. 23, 89-94.
Brunsden, D., 1990. Tablets of stone: towards the ten commandments of geomorphology. Z. Geomorphol.
Suppl.bd. 79, 1-37.
Brunsden, D., Thornes, J.B., 1979. Landscape sensivity and change. Trans. Inst. Br. Geogr. 4, 463-484.
Bryan, K., 1922. Erosion and sedimentation in the Papago country, Arizona. US Geol. Surv. Bull. 730,
19-90.
Bryan, K., 1927. Pedestal rocks formed by differential erosion. US Geol. Surv. Bull. 790A, 1-15.
Bryan, K., 1941. Pre-Columbian agriculture in the southwest as conditioned by periods of alluviation.
Assoc. Am. Geogr. Ann. 31,219-242.
Bryan, K., 1946. Cryopedology-the study of frozen ground and intensive frost-action with suggestions on
nomenclature. Am. J. Sci. 244, 622-642.
Bryan, R.B., 1987. Processes and significance of rill development. Catena Suppl. 8, 1-15.
Bryan, R.B., Poesen, J., 1989. Laboratory experiments on the influence of slope length on runoff,
percolation and rill development. Earth Surf. Processes Landforms 14, 211-231.
Bryan, R.B., Yair, A., 1982. Perspectives on studies of badland geomorphology. In: Bryan, R.B., Yair, A.
(Eds), Badland geomorphology and piping. Geobooks, Norwich, pp. 1-12.
Bryan, R.B., Yair, A., Hodges, W.K., 1978. Factors controlling the initiation of runoff and piping in
Dinosaur Provincial Park badlands, Alberta, Canada. Z. Geomorphol. Suppl.bd. 29, 151-168.
Bryant, E., 1991. Natural Hazards. Cambridge University Press, Cambridge, p. 294.
Bryant, E., 1997. Climate, Process and Change. Cambridge University Press, Cambridge, p. 209.
Bryant, R.G., Sellwood, B.W., Millington, A.C., Drake, N.A., 1994. Marine-like potash evaporate
formation on a continental playa: case study from Chott el Djerid, southern Tunisia. Sediment. Geol.
90, 269- 291.
Buchanan, F., 1807. A Journey from Madras through the Countries of Mysore, Kahara and Malabar,
Vol. 2. East Indian Company, London, see also Vol. 3, pp. 436-461; pages 66, 89, 251,258, 378,559.
Budd, W.F., Jenssen, D., Radok, U., 1970. The extent of basal melting in Antarctica. Polarforschung 6,
293-306.
Btidel, J., 1944. Die morphologischen wirkungen des eiszeitcklimas im glitocherfreien gebiet. Geol.
Rundsch. 34, 482-519.
Btidel, J., 1948. Das system der klimatischen Geomorphologie. Verhandlungen der Deutschen
Geographentag 27, 69-100, trans. Derbyshire, E. (Ed.) 1973. Climatic Geomorphology. Macmillan,
London, pp. 104-130.
References 705
Btidel, J., 1951. Die klimazonen des Eiszeitalters. Eiszeitalter und Gevenwart 1, 16-26, Trad. Inglesa,
1959. Int. Geol. Rev., 1(9), 72-79.
Btidel, J., 1957. Die "doppelten Einebnungsflachen" in den feuchten tropen. Z. Geomorphol. 1,201-288.
Btidel, J., 1960. Die Frostschott zone Stidorst Spitsbergen, Colloquium Geographica, Vol. 6, p. 105, Bonn.
Btidel, J., 1963. Klima-genetische geomorphologie. Geogr. Rundsch. 15, 269-285, trans. Derbyshire, E.
(Ed.) 1973. Climatic Geomorphology. Macmillan, London, pp. 202-227.
Btidel, J., 1965. Die Relieftypen de Flachensptil-Zone Sud-Indiens am Ostabfall Dekans gegen Madras,
Colloquium Geographicum, Vol. 8, p. 100, Bonn.
Biidel, J., 1968. Geomorphology-principles. In: Fairbridge, R.W. (Ed.), The Encyclopedia of
Geomorphology. Dowden, Hutchinson and Ross, Stroudsburg, pp. 416-422.
Biidel, J., 1970. Pedimente, Rumpffl~ichen und Rtickland-Steilh~inge. Z. Geomorphol. 14, 1-57.
Btidel, J., 1977. Klima-Geomorphologie. Borntraeger, Berlin, p. 304.
Biidel, J., 1978. Das Inselberg-Rumpffl~ichenrelief des heutigen Tropen und das Schicksal seiner fossilen
Altformen in anderen Klimazonen. Z. Geomorphol. Suppl.bd. 31, 79-110.
Btidel, J., 1980. Climatic and climatomorphic geomorphology. Z. Geomorphol. Suppl.bd. 36, 1-8.
Bfidel, J, 1982. Climatic Geomorphology. Princeton University Press, Princeton, p. 443, transl. Fischer, L.
and Busche, D.
Budyko, M.I., 1982. The Earth's Climate: Past and Future. Academic Press, Orlando, p. 307.
Bull, W.B., 1963. Alluvial fan deposits in Western Fresno County, California. J. Geol. 71,243-251.
Bull, W.B., 1964a. Geomorphology of segmented alluvial fans in Western Fresno County, California.
US Geol. Surv. Prof. Paper 352E, 89-125.
Bull, W.B., 1964b. History and causes of channel trenching in Western Fresno County, California. Am.
J. Sci. 262, 249-258.
Bull, W.B., 1968. Alluvial fan cone. In: Fairbridge, R.W. (Ed.), Enclycopedia of Geomorphology.
Dowden, Hutchinson and Ross, Stroudsburg, pp. 7-10.
Bull, W.B., 1972. Recognition of alluvial fan deposits in the stratigraphic record. In: Rigby, J.K., Hamblin,
W.K. (Eds), Recognition of Ancient Sedimentary Environments, Society of Economic Palaeontologist
and Mineralogists. Special Publication, 16, pp. 63-83.
Bull, W.B., 1977. The alluvial fan environments. Prog. Phys. Geog. 1,222-270.
Bull, W.B., 1979. The threshold of critical power in stream. Bull. Geol. Soc. Am. 90, 453-464.
Bull, W.B., 1991. Geomorphic Responses to Climate Change. Oxford University Press, New York,
p. 326.
Bull, W.B., 1997. Discontinous ephemeral streams. Geomorphology 19, 227-276.
Bull, W.B., McFadden, L.D., 1977. Tectonic geomorphology north and south of the Garlock fault,
California. In: Doehring, D.C. (Ed.), Geomorphology in Arid Regions. Publications in Geomorphol-
ogy, Binghamton, NY, pp. 115-137.
Bull, W.B., Pearthree, P.A., 1988. Frequency and size of quaternary surface ruptures of the Pilaycachi
fault, northeastern Sonora, Mexico. Bull. Seismol. Soc. Am. 78, 956-978.
Burton, I., Kates, R.W., White, G.F., 1978. The Environment as Hazard. Oxford University Press,
New York, p. 240.
Butzer, K.W., 1976. Geomorphology from the Earth. Harper & Row, New York, p. 463.
Butzer, K.W., 1978. Climate patterns in an unglaciated continents. Geogr. Mag. 51,201-208.
Butzer, K. W., Stuckenrath, A. J., Bruzewicz, Helgrem, D. M., 1978. Late Cenozoic Palaeoclimates of the
Face Escarpment, Kalahari margin, South Africa. Quaternary Research, 10, 310-339.
Cailleux, A., 1952. Polissage et surcreusement glaciaires dans l'hypoth~se de Boy& Rev. G6omorphol.
Dyn. 5, 247-257.
Cailleux, A., 1961. Histoire de la G6ologie. Presses Universitaires de France, Paris, p. 128.
Caine, N., 1972. Air photo analysis of blockfield fabrics in Talus Valley, Tasmania. J. Sediment. Petrol. 42,
33-48.
Calaforra, J.M., 1996. Contribuci6n al Conocimiento de la Karstologia de Yesos. Tesis Doctoral, p. 350,
Facultad de Ciencias. Granada (In~dita).
Calvin, P., Cailleux, A., 1962. A quantitative study of cavernous weathering (taffonis) and its application
to glacial chronology in Victoria Valley, Antarctica. Z. Geomorphol. 6, 317-324.
Campbell, I.A., 1974. Measurements of erosion on badlands surfaces. Z. Geomorphol. Suppl.bd. 21,
122-137.
Campbell, R.H., 1975. Soil slips, debris flows and rainstorms in the Santa Monica Mountains and vicinity,
southern California. US Geol. Surv. Prof. Paper 851, 51.
Campbell, I.A., 1982. Surface morphology and rates of change during a ten-year period in the Alberta
badlands. In: Bryan, R., Yair, A. (Eds), Badland Geomorphology and Piping. GeoBooks, Norwich,
pp. 221-236.
Campbell, I.A., 1989. Badlands and badland gullies. In: Thomas, D.S.G. (Ed.), Arid Zone
Geomorphology. Halstead Press, New York, pp. 159-193.
Campbell, I.A., Honsaker, J.L., 1982. Variability in badlands erosion; problems of scale and threshold
identification. In: Thorn, C.E. (Ed.), Space and Time in Geomorphology. George Allen & Unwin,
London, pp. 59-79.
Campbell, S.E., Seeler, J.S., Golubic, S., 1989. Desert crust formation and soil stabilization. Arid Soil Res.
Rehab. 3, 217- 228.
Capot-Rey, R., 1957. Le vent et le model6 6olien au Borkou. Travaux de l'Institut de Recherches
Sahariennes 15, 155-157.
Capot-Rey, R., 1963. Contribution h l'~tude et la representation des barkhanes. Travaux de l'Institut de
Recherches Sahariennes 22, 37-60.
Carol, H., 1947. The formation of roches moutonn6es. J. Glaciol. 1, 57-59.
Carr, M.H., 1981. The Surface of Mars. Yale University Press, New Haven.
Carroll, D., 1970. Rock Weathering. Plenum Press, New York, p. 203.
Carson, C.E., Hussey, K.M., 1962. The oriented lakes of Arctic Alaska. J. Geol. 70, 417-439.
Carson, M.A., 1971. An application of the concept of threshold slopes to the Laramie Mountains,
Wyoming. In: Brunsden, D. (Ed.), Slope Form and Process, Institute of British Geographers. Special
Publication, 3, pp. 31-47.
Carson, M.A., Kirkby, M.J., 1972. Hillslope Form and Process. Cambridge University Press, Cambridge,
p. 475.
Carter, R.W.G., 1988. Coastal Environments. An Introduction to the Physical, Ecological and Cultural
Systems of Coastlines. Academic Press, London, p. 617.
Chaput, J.L., 1971. Aspects morphologiques du sud-est de la Sierra Morena (Espagne). Rev. G6ogr. Phys.
G6ol. Dyn. 13, 55-66.
Charlesworth, J.K., 1957. The Quaternary Era, with Special Reference to its Glaciation. Arnold, London,
p. 1700.
Chepil, W.S., Siddoway, F.H., Armbrust, D.V., 1962. Climatic factor form estimating wind erodibility of
farm fields. J. Soil Water Conserv. 17, 162-165.
Chester, D., 1993. Volcanoes and Society. Arnold, London, p. 351.
Chico, R.J., 1963. Playa mud cracks: regular and king size. Geol. Soc. Am., Spec. Paper, 76.
Chorley, R.J., 1964. The nodal position and anomalous character of slope studies in geomorphological
research. Geogr. J. 130, 503-506.
Chorley, R.J., 1978. Bases for theory in geomophology. In: Embleton, C., Brunsden, D., Jones, D.K.C. (Eds),
Geomorphology: Present Problems and Future Prospects. Oxford University Press, Oxford, p. 113.
Chorley, R.J., Beckinsale, R.P., 1980. G.K. Gilbert's geomorphology. Geol. Soc. Am. Spec. Paper 183,
129-142.
Chorley, R.J., Beckinsale, R.P., Dunn, A.J, 1964. The History of the Study of Landforms or the
Development of Geomorphology. Vol. 1: Geomorphology Before Davis. Methuen, London, p. 678.
Chorley, R.J., Beckinsale, R.P., Dunn, A.J., 1973. The History of the Study of Landforms or the
Development of Geomorphology. Vol. 2: The Life and Work of William Morris Davis. Methuen,
London, p. 874.
Chorley, R.J., Schumm, S.A., Sugden, D.E., 1984. Geomorphology. Methuen, London, p. 605.
Chueca, J., Julifin, A., L6pez, I., 2003. Variations of glacier coronas, Pyrenees, Spain, during the 20th
century. J. Glaciol. 49 (166), 449-455.
Church, M., 1972. Baffin Island Sandurs; a Study of Arctic Fluvial Processes, Geological Survey of
Canada, Bulletin, 216, p. 208.
Church, M., 1988. Floods in cold climates. In: Baker, V.R., Craig Kochel, R., Patton, P.C. (Eds), Flood
Geomorphology. Wiley, New York, pp. 205-229.
Chylek, P., Box, J.E., Lesins, G., 2004. Global warming and the Greenland ice sheet. Climatic Change 63,
201-221.
Clare, K.E., 1960. Roadmaking gravel and soil in Central Africa. Rd. Res. Overseas Bull., 12.
Clark, D.A., 1974. The Kankoma elzy deposit. Economic Report No. 49. Geological Survey, Zambia.
Clark, J.A., 1980. A numerical model of worldwide sea level changes on a viscoelastic earth. In: M6rner,
N.A. (Ed.), Earth Rheology, Isostasy and Eustasy. Wiley, Chichester, pp. 525-534.
References 707
Clark, J.A., Farrell, W.E., Peltier, W.R., 1978. Global changes in postglacial sea level, a numerical
calculation. Quaternary Res. 9, 265-287.
Clark, M.J., 1988. Periglacial hydrology. In: Clark, M.J. (Ed.), Advances in Periglacial Geomorphology.
Wiley, Chichester, pp. 415-462.
Clark, P.U., 1994. Unstable behavior of the Laurentide ice sheet over deforming sediment and its
implications for climate change. Quaternary Res. 41, 19-25.
Clark, P.U., Bartlein, C.J., 1995. Correlation of the late Pleistocene glaciation in the western United States
with North Atlantic Heinrich events. Geology 23,483-486.
Clarke, G.K.C., 1991. Length, width and slope influences on glacier surging. J. Glaciol. 37, 236-246.
Clayton, L., 1964. Karst topography on stangnant glaciers. J. Glaciol. 5, 107-112.
Cloos, H., 1939. Hebung, spaltung, vulkanismus. Geol. Rundsch. 30, 3A.
Clos-Arceduc, A., 1969. Essai d'Explication des Formes Dunaires Sahariennes, l~tudes de Photo-
Interprrtation. Institut Grographique National, Paris, p. 66.
Coates, D.R., (Ed.), 1974. Glacial Geomorphology. State University of New York, Binghamton, p. 398.
Coates, D.R., 1983. Large-scale land subsidence. In: Gardner, R., Scoging, H. (Eds), Mega-
Geomorphology. Clarendon Press, Oxford, pp. 212-240.
Coates, D.R., 1987. Subsurface impacts. In: Gregory, K.J., Walling, D.E. (Eds), Human Activity and
Environmental Processes. Wiley, Chichester, pp. 271-304.
Cole, K.L., Mayer, L., 1982. Use of packrat middens to determine rates of cliff retreat in the Eastern Grand
Canyon, Arizona. Geology 10, 597-599.
Coleman, J.M., 1969. Brahmaputra river: channel processes and sedimentation. Sediment. Geol. 3,
129-239.
Colombo, F., 1979. Introducci6 a l'Estudi Sedimentologic dels Cons de Deyecci6 (Alluvial Fans) i dels
Sediments Associats, Vol. 5. Seminaris d'Estudis Universitaris, Barcelona, p. 163.
Colombo, F., 1989. Abanicos aluviales. In: Arche, A. (Ed.), Sedimentologfa, Vol. 1. Consejo Superior de
Investigaciones Cientfficas, Madrid, pp. 143-218.
Coltrinari, L., Nogueira, F., 1989. Dambo-like landforms in southeastern Brazil, Abstracts of
Papers, Second International Conference on Geomorphology, Frankfurt/Main, Vol. 1, p. 58,
Ge6ko-plus.
Combes, P.J., 1969. Recherches sur la Genrse des Bauxites dans le Nord-Est de l'Espagne, le Languedoc
et l'Arirge (France), M4moires du Centre d'l~tudes et de Recherches Grologiques et Hydrogrolo-
giques, p. 342, Montpellier.
Common, R., 1966. Slope failure and morphogenetic regions. In: Dury, G.H. (Ed.), Essays in
Geomorphology. Heineman, London, pp. 53-81.
Cooke, R.U., 1970a. Stone pavements in deserts. Ann. AAG 60, 560-577.
Cooke, R.U., 1970b. Morphometric analysis of pediments and associated landforms in the western Mojave
Desert, California. Am. J. Sci. 269, 26-38.
Cooke, R.U., 1977. Applied geomorphological studies in deserts: a review of examples. In: Hails, J.R.
(Ed.), Applied Geomorphology. Elsevier, Amsterdam, pp. 183-225.
Cooke, R.U., 1979. Laboratory simulation of salt weathering processes in arid environments. Earth Surf.
Processes 4, 347-359.
Cooke, R.U., 1981. Salt weathering in deserts. Proc. Geol. Assoc. 92, 1-16.
Cooke, R.U., 1984. Geomorphological Hazards in Los Angeles. Allen & Unwin, London, p. 206.
Cooke, R.U., 1986. Hot drylands. In: Fookes, P.G., Vaughan, P.R. (Eds), A Handbook of Engineering
Geomorphology. Surrey University Press, New York, pp. 109-124.
Cooke, R.U., Doornkamp, J.C., 1974. Geomorphology in Environmental Management. An Introduction.
Clarendon Press, Oxford, p. 413.
Cooke, R.V., Doornkamp, J.C., 1990. Geomorphology in Environmental Management. An Introduccion.
Cooke, R.U., Mason, P., 1973. Desert knolls pediment and associated landforms in the western Mojave
Desert, California. Rev. Geomorphol. Dyn. 20, 71-78.
Cooke, R.U., Reeves, R.W., 1976. Arroyos and Environmental Change in the American Southwest.
Cooke, R.U., Smalley, I.J., 1968. Salt weathering in deserts. Nature 220, 1226-1227.
Cooke, R.U., Warren, A., 1973. Geomorphology in Deserts. Batsford, London, p. 394.
Cooke, R.U., Brunsden, D., Doornkamp, J.C., 1982. Urban Geomorphology in Drylands. Oxford
University Press, Oxford, p. 324.
Cooke, R.U., Warren, A., Goudie, A., 1993. Deserts Geomorphology. UCL Press, London, p. 526.
Coque, R., 1955. Morphologie et crofit dans le sud-Tunesien. Ann. G6ogr. 64, 359-370.
Coque, R., 1962. La Tunisie pr6saharienne, Etude G6omorphologique, Imprimerie Oberthur, p. 476.
Corbel, J., 1963. P6diments d'Arizona. Centre de Documentation Cartographique et G6ographique.
Memoires et Documents 9, 33-95.
Corbel, J., 1964. L'6rosion terrestre, &ude quantitative (M6thodes-techniques-r6sultats). Ann. G6ogr. 73,
385-412.
Corte, A.E., 1969. Formaci6n en el laboratorio de estructuras como pliegues por congelamiento y
descogelamiento mtiltiple. Cuartas Jornadas Geol6gicas Argentinas 1,215-227.
Corte, A.E., 1971. Particle sorting by repeated freezing and thawing. Biuletyn Peryglacjalny 15,
175-240.
Corte, A.E., 1976. Rock glaciers. Biuletyn Peryglacjalny 26, 157-197.
Corte, A.E., Higashi, A., 1964. Experimental research on desiccation cracks in soil. U.S. Army Material
Command Cold Region Research and Engineering Laboratory, Research Report, 66, p. 72.
Costa, J.E., 1984. Physical geomorphology of debris flows. In: Costa, J.E., Fleisher, P.J. (Eds),
Developments and Applications of Geomorphology. Springer, Berlin, pp. 268-317.
Costa, J.E., 1988a. Floods from dam failures. In: Baker, V.R., Craig Kochel, R., Patton, P.C. (Eds),
Flood Geomorphology. Wiley, New York, pp. 439-463.
Costa, J.E., 1988b. Rheologic, geomorphologic and sedimentologic differentiation of water floods,
hyperconcentrated flows and debris flows. In: Baker, V.R., Cochel, R.C., Patton, P.C. (Eds), Flood
Costa, J.E., Baker, V.R., 1981. Surficial Geology: Building with the Earth. Wiley, New York, p. 498.
Costard, F., Dupeyrat, L., Gautier, E., Carey-Gailhardis, E., 2003. Fluvial thermal erosion investigations
along a rapidly eroding river bank: application to the Lena River (Central Siberia). Earth Surf.
Processes Landforms 28, 1349-1359.
Cotton, C.A., 1942. Climatic Accidents in Landscape Making. Whitcomb & Tombs, Wellington.
Cotton, C.A., 1961. Theory of savanna planation. Geography 46, 89-96.
Cotton, C.A., 1962. Plains and inselbergs of the humid tropics. Roy. Soc. New Zeal., Trans. (Geol.) 1,
269-277.
Coud6-Gaussen, G., 1981. Les Serras da Peneda et do Germs. Etude G6omorphologique. Mem6rias do
Centro de Estudos Geogr~ificos 5, 254.
Coud~-Gaussen, G., 1991. Les Poussi~res Sahariennes: Cycle S6dimentaire et Place dans les
Environments et Pal6oenvironments D6sertiques. John Libby Eurotext, Montrouge, p. 485.
Coud6-Gaussen, G., Rognon, P., 1988. Charact6risation s6dimentologique et conditions pal6oclimatiques
de la mise en place de loess and Nord du Sahara ~ipartir de l'example du sud-Tunesien. Bulletin de la
Societ6 G6ologique de France 4, 1081 - 1090.
Coussot, P., Meunier, M., 1996. Recognition, classification and mechanical description of debris flow.
Earth-Sci. Rev. 40, 209-227.
Cowardin, L. M., Carter, V., Golet, F. C., La Roe, E. T., 1979. Classification of wetlands and deepwater
habitats of the United States: U.S. Fish and Wildlife Service FWS/OBS - 79/31.
Crandell, D.R., Mullineaux, D.R., Rubin, M., 1975. Mount St. Helens Volcano: recent and future
behaviour. Science 187, 438-441.
Crittenden, M.D., 1963. New data on the isostatic deformation of Lake Bonneville. US Geol. Surv. Prof.
Paper, 454-E.
Croot, D.G., 1987. Glacio-tectonic structures: a mesoscale model of thin-skinned thrust sheets? J. Struct.
Geol. 9, 797-808.
Crory, E. E., 1985. Long-term foundation studies of three bridges in the Fairbanks Area. U.S. Army Cold
Regions Res. And Engineering Lab. Technical Report.
Crouch, R.J., 1983. The role of tunnel erosion in gully head progression. J. Soil Conser., N.S.W. 39,
148-155.
Crouch, R.J., 1990. Rates and mechanisms of discontinuous gully erosion in a red-brown earth catchment,
New South Wales, Australia. Earth Surf. Processes Landforms 15, 277-282.
Crowley, T.J., North, G.R., 1991. Paleoclimatology. Oxford University Press, New York, p. 339.
Crozier, M.J., 1986. Landslides: Causes, Consequences and Environment. Croom Helm, London,
p. 252.
Currey, D.R., 1994. Hemiarid lake basins: hydrographic patterns. In: Abrahams, A.D., Parsons, A.J. (Eds),
Geomorphology of Desert Environments. Chapman and Hall, London, pp. 405-421.
References 709
Curtis, C.D., 1976. Chemistry of rock weathering: fundamental reactions and controls. In: Derbyshire, E.
(Ed.), Geomorphology and Climate. Wiley, New York, pp. 25-57.
Cutts, J.A., Blasius, K.R., Roberts, W.J., 1979. Evolution of Martian polar landscapes: interplay of long-
term variations in perennial ice cover and dust storm intensity. J. Geophys. Res. 84, 2975-2994.
Czeppe, Z., 1964. Exfoliation in a periglacial climate. Geographia Polonica 2, 5-10.
Czudek, T., 1964. Periglacial slope development in the area of the Bohemian Massif in Northern Moravia.
Biuletyn Peryglacjalny 14, 169-194.
Czudek, T., 1973. Zur Klimatischen Talasymmetrie des Westeiles der Tschechoslowakei. Z. Geomorphol.
17, 49-57.
Czudek, T., Demek, J., 1970a. Pleistocene cryopedimentation in Czechoslovakia. Acta Geographica
Lodziensia 24, 101 - 108.
Czudek, T., Demek, J., 1970b. Thermokarst in Siberia and its influence on the development of lowland
relief. Quaternary Res. 1, 103-120.
Czudek, T., Demek, J., 1971. Pleistocene cryoplanation in the Ceska vysocina highlands, Czechoslovakia.
Trans, Inst. Br. Geogr. 52, 95-112.
D'Almeida, G.A., 1989. Desert aerosol: characteristics and effects of climate. In: Leinen, M.,
Sarnthein, M. (Eds), Paleoclimatology and Palaeometereology: Modern and Past Patterns of Global
Atmospheric Transport. Kluwer Academic, Dordrecht, pp. 311-337.
D'Ercole, R., 1994. Natural hazards in the French West Indies. An overall view, Third European Intensive
Course on Applied Geomorphology. Tropical Regions. Universit6 Louis Pasteur, Strasbourg, p. 10.
Dahl, R., 1965. Plastically sculptured detail forms on rock surfaces in Northern Nordland, Norway. Geogr.
Ann. 47, 83 - 140.
Dahl, R., 1966. Blockfields and other weathering forms in the Narvik Mountains. Geogr. Ann. 48A,
55-85.
Danin, A., 1983. Weathering of limestone in Jerusalem by cyanobacteria. Z. Geomorphol. 27, 413-421.
Dansgaard, W., Johnsen, S.J., M611er, J., Langway, C.C. Jr., 1969. One thousand centuries of climatic
record from camp century on the Greenland ice sheet. Science 166, 377-381.
Dansgaard, W., Johnsen, S.J., Clausen, H.B., Dahl-Jensen, D., Gundestrup, N.S., Hammer, C.U.,
Hvidberg, C.S., Steffensen, J.P., Sveinbj6rnsdottir, A.E., Jouzel, J., Bond, G., 1993. Evidence for
general instability of past climate from a 250-ky ice-core record. Nature 364, 218-220.
Davies, J.L., 1969. Landforms of Cold Climates. The M.I.T. Press, Cambridge, p. 200.
Davis, L., 1992. Encyclopedia of Natural Disasters. Headline, London, p. 433.
Davis, W.M., 1899. The geographical cycle. Geogr. J. 14, 481-504.
Davis, W.M., 1905. The geographical cycle in arid climate. J. Geol. 13, 381-407.
Davis, W.M., 1906. The sculpture of mountains by glaciers. Scot. Geogr. Mag. 22, 76-89.
Davis, W.M., 1932. Piedmont benchlands and Prim~irrumpfe. Bull. Geol. Soc. Am. 43, 399-440.
Davis, W.M., 1954. Geographical Essays. Dover, New York, p. 777.
Davison, C., 1889. On the creeping on the soilcap through the action of frost. Geol. Mag. 6, 255-261.
Dawson, A.G., 1992. Ice age earth. Late Quaternary Geology and Climate. Routledge, London, p. 293.
Deacon, J., Lancaster, N., Scott, L., 1984. Evidence for Pate Quaternary climatic change in Southern
Africa. Late Cainozoic palaeoclimates of the Southern Hemisphere. In: Vogel, J. C., (Ed.) Balkema,
Rotterdam, pp. 21 - 34.
Dearman, W.R., Shibakova, V. S., 1989. The tropical environment. In: Dearman, W.R., Sergeev, E.M.,
Shibakoya, V.S. (Eds), Engineering Geology of the Earth. Nauka Publishers, Moscow, pp. 128-141.
Dearman, W.R., Baynes, F.J., Irfan, T.Y., 1978. Engineering grading of weathered granite. Eng. Geol. 12,
345-374.
De Chetelat, E., 1938. Le model~ lat6rique de l'ouest de la Guin~e francaise. Rev. G~ogr. Phys. G~ol. Dyn.
11, 5-12.
De Dapper, M., 1978. Couvertures limono-sableuses, stone-line, indurations ferrugineuses et action des
termites sur le plateau de La Manika (Kolwezi, Shaba, Zaire). G6o-Eco-Trop 2, 265-278.
De Dapper, M., 1989. Pedisediments and stone-line complexes in Peninsular Malaysia. In: Alexandre, J.,
Symoens, J.J. (Eds), Stone-Lines, Geo-Eco-Trop, Vol. 11. Acad6mie Royale des Sciences D'Outre-
Mer, Bruxelles, pp. 37-59.
De Gans, W., 1988. Pingo scars and their identification. In: Clark, M.J. (Ed.), Advances in Periglacial
Geomorphology. Wiley, Chichester, pp. 299-322.
DeGraff, J.V., 1978. Regional landslide evaluation: two Utah examples. Environ. Geol. 2, 203-214.
Demangeot, J., 1981. Les Milieux Natural D6sertique. SEDES, Paris, p. 261.
Demangeot, J., 1996. Certitudes et incertitudes du "global change". Bull. Assoc. G6ogr. Fr. 4,
278-282.
Demek, J., 1964. Castle koppies and tors in the Bohemian highland (Czechoslovakia). Biuletyn
Peryglacjalny 14, 195-216.
Demek, J., 1965. Slope development in granite areas of Bohemian massif (Czechoslovakia).
Demek, J., 1969. Cryogene processes and the development of cryoplanation terraces. Biuletyn
Dendy, F.E., Bolton, G.C., 1976. Sediment yield-runoff drainage area relationships in the United States.
J. Soil Water Conserv. 31,264-266.
Denny, C.S., 1967. Fans and pediments. Am. J. Sci. 265, 81-105.
De Ploey, J., 1964. Stone-lines and clayey-sandy mantles in Lower Congo: their formation and the effect
of termites. In: Bouillon, A. (Ed.), Etudes sur les Termites Africains. Universit6 de Louvanium,
Leopoldville, pp. 399-414.
De Ploey, J., 1965. Position g6omorphologique, g6n~se et chronologie de certains d6pots superficiels au
Congo occidentale. Quaternaria 7, 131 - 154.
De Ploey, J., 1980. Some field measurements and experimental data on wind-blown sands. In: Boodt,
M. de., Gabriels, D. (Eds), Assessment of Erosion. Wiley, New York, pp. 541-552.
De Ploey, J., Gabriels, D., 1980. Measuring soil loss and experimental studies. In: Kirkby, M.J.,
Morgan, R.P.C. (Eds), Soil Erosion. Wiley, New York, pp. 63-108.
De Ploey, J., Poesen, J., 1989. Experimental geomorphology and the interpretation of stone-lines.
In: Alexandre, J., Symoens, J.J. (Eds), Stone Lines, Vol. 11. Acad6mie Royale des Sciences D'Outre-
MerGeo, Geo-Eco-Trop, Bruxelles, pp. 75-82.
De Ploey, J., Savat, J., Moeyerons, D., 1976. The differential impact of some soil loss factors on flow,
runoff, creep, and rainwash. Earth Surf. Process. 1, 151-161.
De Ploey, J., Kirkby, M.J., Ahnert, F., 1991. Hillslope erosion by rainstorms - a magnitude - frequency
analysis. Earth Surf. Processes Landforms 16, 399-409.
Derbyshire, E., 1973. Introduction. In: Derbyshire, E. (Ed.), Climatic Geomorphology. Macmillan,
London, pp. 11 - 18.
Derbyshire, E., (Ed.), 1976a. Geomorphology and Climate. Wiley, London, p. 512.
Derbyshire, E., (Ed.), 1976b. Geomorphology and climate: background. Geomorphology and Climate.
Wiley, London, pp. 1-24.
Derbyshire, E., 1977. Periglacial environments. In: Hails, J.R. (Ed.), Applied Geomorphology. Elsevier,
Amsterdam, pp. 227- 276.
Derbyshire, E., 1983a. On the morphology, sediments and origin of the Loess Plateau of central China. In:
Gardner, R.A.M., Scoging, H. (Eds), Megageomorphology. Oxford University Press, Oxford, pp. 172-194.
Derbyshire, E., 1983b. Origin and characteristics of some Chinese loess at two locations in China. In:
Brookfield, M.E., Ahlbrandt, T.S. (Eds), Eolian Sediments and Processes. Elsevier, Amsterdam,
pp. 69-90.
Derbyshire, E., Evans, I.S., 1976. The climatic factor in cirque variation. In: Derbyshire, E. (Ed.),
Geomorphology and Climate. Wiley, New York, pp. 447-494.
Derbyshire, E., Love, M.A., 1986. Glacial environments. In: Fookes, P.G., Vaughan (Eds), A Handbook of
Engineering Geomorphology. Chapman & Hall, New York, pp. 66-81.
Derruau, M., 1965. Pr6cis de G6omorphologie. Masson, Paris.
DeWolf, Y., 1988. Stratified slope deposits. In: Clark, M.J. (Ed.), Advances in Periglacial Geomorphology.
Dias de Avila-Pires, F., 1974. Caracterizao zoogeogr~ifica de Provincia Amaz6nica. Ann. Acad. Cienc.
Brasil 46, 133-181.
Dickinson, R.E., 1986. Impact of human activities on climate - a framework. In: Clark, W.C., Munn, R.E.
(Eds), Sustainable Development of the Biosphere. Cambridge University Press, Cambridge,
pp. 252-289.
Dickinson, R.E., Meleshko, V., Randall, D., Sarashik, E., Silva-Dias, P., Slingo, A., 1996. Climate
Processes. In: Houghton, J.T. et al. (Eds), Climate Change. The Science of Climate Change. Cambridge
University Press, Cambridge, pp. 193-227.
Diester-Haas, L., 1976. Late Quaternary climatic variations in Northwest Africa deduced from east
Atlantic sediment cores. Quaternary Res. 6, 299-314.
Dietrich, R.V., 1977. Impact abrasion of harder by softer materials. J. Geol. 85, 242-246.
References 711
Dionne, J.P., 2004. Pot-hole. In: Goudie, A.S. (Ed.), Encyclopedia of Geomorphology. Routledge,
London, p. 807.
Dixon, J.C., 1994. Duricruts. In: Abrahams, A.D., Parson, A.J. (Eds), Geomorphology of Desert
Environments. Chapman & Hall, London, pp. 82-105.
Doelling, H.H., 1985. Geology of Arches National Park, Utah Geological and Mineral Survey, Map 74.
Dohrenwend, J.C., 1994. Pediments in arid environments. In: Abrahams, A.D., Parsons, A.J. (Eds),
Geomorphology of Desert Environments. Chapman & Hall, London, pp. 321-353.
Dokuchayev, V.V., 1883. The Russian Chernozem. St. Petersburg.
Donn, W.L., Farrand, W.R., Ewing, M., 1962. Pleistocene ice volumes and sea lowering. J. Geol. 70,
206-214.
Doornkamp, J.C., Ibrahim, H.A.M., 1990. Salt weathering. Prog. Phys. Geogr. 14, 335-348.
D'Orefice, M., Pecci, M., Smiraglia, C., Ventura, R., 2000. Retreat of Mediterranean Glaciers since the
Little Ice Age: Case Study of Ghiacciaio del Calderone, Central Apennines, Italy. Arct. Antarct. Alp.
Res. 32, 197-201.
Dorn, R.I., 1994a. Rock varnish as evidence of climatic change. In: Abrahams, A.D., Parsons, A.J. (Eds),
Dorn, R.I., 1994b. The role of climatic change in alluvial fan development. In: Abrahams, A.D., Parsons, A.J.
(Eds), Geomorphology of Desert Environments. Chapman & Hall, London, pp. 593-615.
Dorn, R.I., 1995. Alterations of ventifat surfaces at the glacier/desert interface. In: Tchakerian, V.P. (Ed.),
Desert Aeolian Processes. Chapman & Hall, London, pp. 199- 217.
Dorn, R.I., 1998. Rock Coatings. Elsevier, Amsterdam, p. 429.
Dorn, R.I., Oberlander, T.M., 1981. Rock varnish origin, characteristics, and usage. Z. Geomorphol. 25,
420-436.
Dorn, R.I., Oberlander, T.M., 1982. Rock varnish. Prog. Phys. Geogr. 6, 317-367.
Dorn, R.J., Tanner, D.L., Turrin, B.D., Dohrenwend, J.C., 1987. Cation-ratio dating of Quaternary
materials in the east-central Mojave Desert, California. Phys. Geogr. 8, 72-81.
Douglas, I., 1967. Man, vegetation and the sediment yield of rivers. Nature 215, 925-928.
Douglas, I., 1969. The efficiency of humid tropical denudation systems. Inst. Br. Geogr. Trans. 40,
1-16.
Douglas, I., 1977. Humid Landforms. The MIT Press, Cambridge, MA, p. 288.
Douglas, I., 1978. Tropical geomorphology: present problems and future prospects. In: Embleton, C.,
Brunsden, D., Jones, D.K.C. (Eds), Geomorphology: Present Problems and Future Prospects. Oxford
University Press, Oxford, pp. 162-184.
Douglas, I., 1986. Hot wetlands. In: Fookes, P.G., Vaughan, P.R. (Eds), A Handbook of Engineering
Geomorphology. Surrey University Press, Glasgow, pp. 137-149.
Douglas, I., Spencer, T., 1982. Introduction: applied geomorphology in the tropics. Z. Geomorphol.
Suppl.bd. 44, 1- 3.
Douglas, I., Spencer, T., 1985a. The history of geomorphology in low latitudes. In: Douglas, I.,
Spencer, T. (Eds), Environmental Change and Tropical Geomorphology. Allen and Unwin, London,
pp. 3-11.
Douglas, I., Spencer, T., 1985b. Present day processes as a key to the effects of environmental change.
In: Douglas, I., Spencer, T. (Eds), Environmental Change and Tropical Geomorphology. Allen and
Unwin, London, pp. 39-73.
Douglas, I., Spencer, T., (Eds), 1985c. Environmental Change and Tropical Geomorphology. Allen and
Unwin, London, p. 378.
Dov Nir, 1983. Man, a Geomorphological Agent. An Introduction to Anthropic Geomorphology. Keter
Publication, Jerusalem, p. 165.
Dowdeswell, J.A., Hamilton, G.S., Hagen, J.O., 1991. The duration of the active phase on surge-type
glaciers: contrasts between Svalbard and other regions. J. Glaciol. 37, 388-400.
Dragovich, D., 1969. The origin of cavernous surfaces (tafoni) in granitic rocks of Southern Australia.
Dregne, H.E., 1999. Pan. In: Mares, M.A. (Ed.), Encyclopaedia of Deserts. University of Oklahoma Press,
Norman, OK, pp. 499-500.
Dreimanis, A., 1953. Studies of friction cracks along shores of Cirrus Lake and Kasakokwog, Ontario.
Am. J. Sci. 251,769-783.
Dreimanis, A., 1988. Tills: their genetic terminology and classification. In: Goldthwait, R.P., Matsch, C.L.
(Eds), Classification of Glacigenic Deposits. Balkema, Rotterdam, pp. 17-83.
Dreimanis, A., Vagners, J.J., 1971. Bimodal distribution of rock and mineral fragments in basal tills.
In: Goldthwait, R.P. (Ed.), Till: A Symposium. Ohio State University Press, Columbus, pp. 237-250.
Dresch, J., (1938). Les surfaces de pi6mont dans les Djebilet et le Massif Central du Grand Atlas. Compte
Rendu Congress Internationale de G6ographie, Vol. 2, section 2a, p. 135. Amsterdam.
Dresch, J., 1957. P6diments et glacis d'erosion, pediplains et inselbergs. Inf. G6ogr. 21,183-196.
Dresch, J., 1970. Apropos du d6sert de Chihuahua. Ann. G6ogr. 79, 51-57.
Dresch, J., 1982. G6ographie des R6gions Arides. Presses Universitaires de France, Paris, p. 277.
Drewry, D., 1986. Glacial Geologic Processes. Edward Arnold, London, p. 276.
Duchaufour, P., 1977. P6dologie. 1. P6dog~nese et Classification. Masson, Paris, p. 477.
Dumas, B., 1967. Place et signification des glacis dan le Quaternaire. Bull. Assoc. Fr. Etude Quatern. 3,
223-244.
Dunne, T., 1980. Formation and control of channel networks. Prog. Phys. Geogr. 4, 211-239.
Dunne, T., Aubry, B.F., 1986. Evaluation of Horton's Theory of sheetwash and rill erosion on basis of field
experiments. In: Abrahams, A.D. (Ed.), Hillslope Processes. Allen and Unwin, Boston, pp. 31-53.
Dunne, T., Dietrich, W.E., 1980. Experimental study of Horton overland flow on tropical hillslopes. 1. Soil
conditions, infiltration and frequency of runnoff. Z. Geomorphol. Supplement Band 35, 40-59.
Durgin, P.B., 1977. Landslides and the weathering of granite rocks. In: Coates, D.R. (Ed.), Landslide
Perspectives, Reviews in Engineering Geology, Vol. III. Geological Society of America, Salt Lake
City, Utah, pp. 127-131.
Dury, G.H., 1969. Rational descriptive classification of duricrusts. Earth Sci. J. 3, 77-86.
Dury, G.H., 1971. Relict deep weathering and duricrusting in relation to the palaeoenvironments of middle
latitudes. Geogr. J. 137, 511-522.
Dury, G.H., 1978. The future of geomorphology. In: Embleton, C., Brunsden, D., Jones, D.K.C. (Eds),
Geomorphology: Present Problems and Future Prospects. Oxford University Press, Oxford,
pp. 263-274.
Dyke, A.S., Prest, V.K., 1987. Late Wisconsinan and Holocene history of the Laurentide Ice Sheet. G6ogr.
Phys. Quatern. 41,237- 264.
Eakin, H.M., 1916. The Yukon-Koyukuk region, Alaska. US Geol. Surv. Bull. 631, 88.
Ehlers, J., 1996. Quaternary and Glacial Geology. Wiley, Chichester, p. 578.
Eissmann, L., (1981). Periglaci~ire prozesse und permafroststrukturen aus sechs kaltzeiten des quart~is. Ein
beitrag zur periglazialgeologie aus des sicht des Saale-Elbe-Gebiets. Altenburger Naturwissenschaf-
tliche Forschunge, 1, p. 171.
Elson, J.A., 1989. Comment on glacitectonite, deformation till and comminution till. In: Goldthwait, R.P.,
Matsch, C.L. (Eds), Genetic Classification of Glacigenic Deposits. Balkema, Rotterdam, pp. 85-88.
Elvidge, C.D., Moore, C.B., 1979. A model for desert varnish formation. Geol. Soc. Am. Abstr. Prog. 11,
271.
Ellenberger, F., (1988). Histoire de la G~ologie. Tome I. Des Anciens ~t la Premiere Moiti~ du XVII e
Si~cle. Tecnique et Documentation-Lavoisier, p. 352 Paris.
Ellison, W.D., 1944. Studies of raindrop erosion. Agric. Eng. 25, 131-136, see also 181-182.
Embleton, C., 1979. Glacial Processes. In: Embleton, C., Thornes, J. (Eds), Process in Geomorphology.
Edward Arnold, London, pp. 272-306.
Embleton, C., 1989. Natural hazards and global change. ITC J. (1989), 169-178.
Embleton, C., King, C.A.M., 1975a. Glacial Geomorphology. Edward Arnold, London, p. 573.
Embleton, C., King, C.A.M., 1975b. Periglacial Geomorphology. Edward Arnold, London, p. 203.
Emery, K.O., 1960. Weathering of the Great Pyramide. J. Sediment. Petrol. 30, 140-143.
Emiliani, C., 1955. Pleistocene paleotemperatures. J. Geol. 63, 539-578.
Emmett, W.W., 1978. Overland flow. In: Kirkby, M.J. (Ed.), Hillslope Hydrology. Wiley, Chichester,
pp. 145-176.
Engle, C.E., Sharp, R.P., 1958. Chemical data on desert varnish. Bull. Geol. Soc. Am. 69, 487-518.
EPICA community members, 2004. Eight glacial cycles from an Antarctic ice core. Nature 429, 623-628.
Erhardt, H., 1951. Sur l'importance des ph6nom~nes biologiques dans la formation des cuirasses
ferrugineuses en zone tropicale. Compte Rendus de la Acad6mie des Sciences de Paris 223,
805-806.
Erhart, H., (1967). La Gen~se des Sols en tant que Ph6nom~ne G6ologique. Esquisse d'une Th6orie
G6ologique et G6ochimique. Biostasie et Rhexistasie. Masson, p. 177 Paris.
Eronen, M., 1983. Late Weichselian and Holowing shore displacement in Finland. In: Smith, D.E.,
Dawson, K.G., (Eds). Shorelines and Isostary. Academic Press, London, 183-208.
References 713
Eugster, H.P., Hardie, L.A., 1975. Sedimentation in an ancient playa-lake complex: the William Peak
Member of the Gran River Formation of Wyoming. Bull. Geol. Soc. Am. 86, 319-334.
Eugster, H.P., Kelts, K., 1983. Lacustrine chemical sediments. In: Goudie, A.S., Pye, K. (Eds), Chemical
Sediments and Geomorphology. Academic Press, London, pp. 321-367.
Euster, H.P., Hardie, L.A., 1978. Saline Lakes. In: Lerman, A. (Ed.), Lakes: Chemistry, Geology, Physics.
Springer, New York, pp. 237-293.
Evans, I.S., 1969. The geomorphology and morphometry of glacial and nival areas. In: Chorley, R.J. (Ed.),
Water, Earth and Man. Methuen, London, pp. 369-380.
Evans, D.J.A., 2004a. Surging glaciers. In: Goudie, A.S. (Ed.), Encyclopedia of Geomorphology.
Routledge, London, pp. 1028-1029.
Evans, D.J.A., 2004b. Fjord. In: Goudie, A.S. (Ed.), Encyclopedia of Geomorphology. Routledge, London,
pp. 374-376.
Evans, I.S., 2004c. Cirque, glacial. In: Goudie, A.S. (Ed.), Encyclopedia of Geomorphology. Routledge,
London, pp. 154-158.
Evans, I.S., 1969-1970. Salt crystallization and rock weathering: a review. Rev. G~omorphol. Dyn. 70,
153-177.
Evenari, M., Shanan, L., Tadmor, N.H., 1971a. The Negev: the Challenge of a Desert. Harvard University
Press, Cambridge, MA, p. 437.
Evenari, M.L., Shanan, N., Tadmor, N.H., 1971b. The Negev: the Challenge of a Desert. Harvard
University Press, Cambridge, MA.
Evenson, E.B., 1971. The relationship of macro and micro-fabric of till and the genesis of glacial
landforms in Jefferson County, Wisconsin. In: Goldthwait, R.P. (Ed.), Till: A Symposium. Ohio State
University Press, Columbus, OH, pp. 345-364.
Everard, C.E., 1963. Contrats in the form and evolution of hill-side slopes in central Cyprus. Trans. Inst.
Br. Geogr. 32, 31-47.
Eyles, N., (Ed.), 1983. Glacial Geology: An Introduction for Engineers and Earth Scientists. Pergamon
Press, Oxford, p. 409.
Eyles, N., 1993. Earth's glacial record and its tectonic setting. Earth-Sci. Rev. 35, 1-248.
Fahey, B.D., 1986. A comparative laboratory study of salt crystallisation and salt hydration as potential
weathering agents in deserts. Geogr. Ann. 68A, 107-111.
Fair, T.J.D., 1947. Slope form and development in the interior of Natal, South Africa. Trans. Geol. Soc.
S. Afr. 50, 105-120.
Fair, T.J.D., 1948. Hill-slopes and pediments of the semi-arid Karroo. S. Afr. Geogr. J. 30, 71-79.
Fairbanks, R.G., 1989. A 17,000-year glacio-eustatic sea level record: influence of glacial melting rates on
the Younger Dryas event and deep-ocean circulation. Nature 342, 637-642.
Fairbridge, R.W., 1968a. Fjord, fiord. In: Fairbridge, R.W. (Ed.), The Encyclopedia of Geomorphology.
Dowden, Hutchinson and Ross, Stroudsburg, pp. 358-359.
Fairbridge, R.W., 1968b. The Encyclopedia of Geomorphology. Dowden, Hutchinson and Ross,
Stroudsburg, p. 1295.
Faniran, A., Jeje, L.K., 1983. Humid Tropical Geomorphology. Longman, London, p. 414.
Farr, T.G., Adams, J.B., 1984. Rock coatings in Hawai. Bull. Geol. Soc. Am. 95, 1077-1083.
Fauque, L., Strecker, M.R., 1988. Large rock avalanche deposits. (Strurzstrome, sturzstroms) at Sierra
Aconquija, northern Sierras Pampeanas, Argentina. Eclogae Geol. Helv. 81,579-592.
Fayzullin, R.M., 1969. Lithological-facies characteristics and the gold contents of a series of
conglomerates of the Kamensk deposit. Chem. Abstr. 70, 108202b.
Feininger, T., 1971. Chemical weathering and glacial erosion of crystalline rocks and the origin of till. US
Geol. Surv. Prof. Paper 750, C65-C81.
Ferguson, R.I., 1973. Sinuosity of supraglacial streams. Bull. Geol. Soc. Am. 84, 251-256.
Fernfindez-Nieto, C., Fern~indez;, R., Guti~rrez Elorza, M., Arrese, F., 1981. Papel de la karstificaci6n
en la g~nesis de los yacimientos de Sierra Menera (Teruel y Guadalajara). Bol. Geol. Minero 92,
127-140.
Finkel, H.J., 1959. The barchans of southern Peru. J. Geol. 67, 614-647.
Fitzpatrick, E.A., 1963. Deeply weathered rock in northeast Scotland, its occurrence, age and contribution
to the soils. J. Soil Sci. 14, 33-42.
Fletcher, J.E., Carroll, P.H., 1948. Some properties of soils associated with piping in southeastern Arizona.
Soil Sci. Soc. Am. Proc. 13, 545-547.
Flint, R.F., 1971. Glacial and Quaternary Geology. Wiley, New York, p. 892.
F61ster, H., 1969. Slope development in SW Nigeria during the late Pleistocene and Holocene. G6ttinger
Bodenkundliche Berichte 10, 3 - 56.
Fookes, P.G., French, W.J., 1977. Soluble salt damage to surface roads in the Middle East. J. Inst. Highway
Eng. 24 (12), 10-20.
Fookes, P.G., Vaughan, P.R., 1986. A Handbook of Engineering Geomorphology. Surrey University Press,
Glasgow, p. 343.
Ford, D.C., Williams, D.W., 1989. Karst Geomorphology and Hydrology. Unwin and Hyman, London,
p. 601.
Fosberg, F.R., Gamier, B.J., Kuchler, A.W., 1961. Delimitation of the humid tropics. Geogr. Rev. 51,
333-347.
Foster, G.R., Meyer, L.D., 1975. Mathematical Simulation of Upland Erosion Using Fundamental Erosion
Mechanics, Present and Prospective Technology for Predicting Sediment Yields and Sources. US Depart-
ment of Agriculture, Mississippi, US Agricultural Research Service Report ARS-S-40, pp. 190-207.
Fournier, F., 1960. Climat et Erosion: La Relation entre l'l~rosion du Sol par l'Eau et les Pr6cipitations
Atmosph~riques. Presses Universitaires de France, Paris, p. 201.
Francis, C., 1994. Plants on desert hillslopes. In: Abrahams, A.D., Parsons, A.J. (Eds), Geomorphology of
Desert Enviroments. Chapman & Hall, London, pp. 243-254.
Francis, C.F., Thornes, J.B., 1990. Runoff hydrographs from three Mediterranean vegetation cover types.
In: Thornes, J.B. (Ed.), Vegetation and Erosion. Wiley, Chichester, pp. 363-384.
Francis, P., 1993. Volcanoes: A Planetary Perspective. Clarendon Press, Oxford, p. 443.
Francou, B., 1988. Eboulis stratifi6s dans les Hautes Andes Centrales du P6rou. Z. Geomorphol. 32,
47-76.
French, H.M., 1971. Slope asymmetry of the Beaufort Plain, northwest Banks Island, N.W.T, Canada.
Can. J. Earth Sci. 8, 717- 731.
French, H.M., 1987. Permafrost and ground ice. In: Gregory, K.J., Walling, D.E. (Eds), Human Activity
and Environmental Processes. Wiley, Chichester, pp. 237-269.
French, H.M., 1996. The Periglacial Environment. Longman, Essex, p. 341.
French, H.M., Egginton, P., 1973. Thermokarst development, Banks Island, Western Canadian Arctic,
Permafrost: North American Contribution, Second International Permafrost Conference, Yakutsk,
USSR. National Academy of Science, Washington, DC, Publication 2115, pp. 203-212.
French, H.M., Karte, J., 1988. A periglacial overview. In: Clark, M.J. (Ed.), Advances in Periglacial
French, H.M., Harris, S.A., Van Everdingen, R.O., 1983. The Klondike and Dawson. In: French, H.M.,
Heginbottom, J.A. (Eds), Guidebook to Permafrost and Related Features of the Northern Yukon
Territory and Mackenzie Delta, Canada, Fourth International Conference of Permafrost. Division of
Geological and Geophysical Surveys, State of Alaska, Fairbanks, pp. 35-63.
Friedman, J.D. et al., 1971. Observations on icelandic polygon and palsa areas. Photointerpretation and
field studies. Geogr. Ann. 53, 115-145.
Frostick, L.E., Reid, I., Layman, J.T., 1983. Changing size distribution of suspended sediment in arid-zone
flash floods. Spec. Publ. Int. Assoc. Sediment. 6, 97-106.
Frumkin, A., 1995. Morphology and development of salt caves. Bull. Natl Speleol. Soc. 56, 82-95.
Frumkin, A., Ford, D.C., 1995. Rapid entrenchment of stream profiles in the salt caves of Mount Sedom,
Israel. Earth Surf. Processes Landforms 20, 139-152.
Fryberger, S.G., Ahlbrandt, T.S., 1979. Mechanism for the formation of aeolian sand seas. Z. Geomorphol.
23,440-460.
Fryberger, S.G., Goudie, A.S., 1981. Arid geomorphology. Prog. Phys. Geogr. 5,420-428.
Frye, J.C., Willman, H.B., 1973. Wisconsinian climatic history interpreted from Lake Michigan lobe
deposits and soils. Geol. Soc. Am. Mem. 136, 135-152.
Furdada, G., 1996. Estudi de les allams al Pirineu occidental de Catalunya: predicaio espaniol I aplicanions
de la cartografia. Geoforma Ediciones, Logrono, p. 315.
Furrer, G., 1968. Unterswchungen an Strukturb6den in Ostspitbergen, ihre Bedentung ffir die Erforchung
rezenter und fossiler Frostmusterformen in den Alpen baw im Alpen vorland. Polarforchung 6, 202-206.
Furrer, G., 1972. Bewegungsmessungen auf solifluktionsdecken. Z. Geomorphol. Suppl.bd. 13,
87-101.
Gallart, F., Sold, A., Puigdeffibregas, J., Lfizaro, R., 2002. Badlands systems in the Mediterranean. In:
Bull, L.J., Kirby, M.J. (Eds), Drylands Rivers: Hydrology and Geomorphology of Semi-arid Channels.
Willey, New York, pp. 299-326.
References 715
Garcfa Salmer6n, J., 1967. Erosi6n E61ica. Ministerio de Agricultura, Madrid, p. 579.
Gardner, J.S., 1987. Evidence for headwall weathering zones, Boundary Glacier, Canadian Rocky
Mountains. J. Glaciol. 33, 60-67.
Gardner, L.R., 1972. Origin of the Mormon Mesa caliche, Clark County, Nevada. Bull. Geol. Soc. Am. 83,
143-156.
Gardner, R., 1983. Introduction. In: Gardner, R., y Scoging, H. (Eds), Mega-Geomorphology. Clarendon
Press, Oxford, pp. X-XIII.
Gaylord, D.R., Dawson, P.J., 1987. Airflow-terrain interactions through a mountain gap, with an example
of eolian activity beneath an atmospheric hydraulic jump. Geology 15, 789-792.
Gellert, J.F., 1970. Climatomorphology and palaeoclimates of the central European Tertiary. In: Pecsi, M.
(Ed.), Problems of Relief Planation. Akadtmiai Kaid6, Budapest, pp. 107-112.
Geological Society, 1990. Engineering group working party report: tropical residual soils. Q. J. Eng. Geol.
23, 1-101.
Georges, C., 2004. 20th Century Glacier Fluctuations in the Tropical Cordillera B lanca, Pert]. Arct.
Antarct. Alp. Res. 36, 100-107.
Gerits, J., Imeson, A.C., Verstraten, J.M., Bryan, R.B., 1987. Rill development and badland regolith
properties. Catena Suppl. 8, 141 - 160.
Gerrard, A.J., 1988. Rocks and Landforms. Unwin Hyman, London, p. 319.
Gerrard, A.J., 1990. Mountain Environments: An Examination of the Physical Geography of Mountains.
Belhaven Press, London, p. 317.
Gerson, R., 1982. Talus relicts in deserts: a key to major climatic fluctuations. Israel J. Earth Sci. 31,
123-132.
Gerson, R., Grossman, S., 1987. Geomorphic activity on escarpments and associated fluvial systems
in hot deserts. In: Rampino Sanders, J.E., Newman, W.S., Ktnigsson, L.K. (Eds), Climate, History,
Periodicity, Predictability. Van Nostrand Reinhold, New York, pp. 300-322.
Gerstenhauer, A., 1960. Der tropische kelgelkarst in Tabasco (Mexico). Z. Geomorphol. Suppl.bd. 2,
22-48.
Ghose, B., Kar, A., Hussain, Z., 1979. The lost courses of the Sarawasti river in the Great Indian Desert -
new evidence from Landsat imagery. Geogr. J. 145, 446-451.
Giardino, J.R., 1983. Movement of ice-cemented rock glaciers by hydrostatic pressure: an example from
Mt. Mestas, Colorado. Z. Geomorphol. 27, 297-310.
Giardino, J.R., Vick, S.G., 1987. Geologic engineering aspects of rock glaciers. In: Giardino, J.R.,
Schroeder, J.F. Jr., Vitek, J.D. (Eds), Rock Glaciers. Allen and Unwin, London, pp. 265-287.
Gibbs, R.J., 1967. The geochemistry of Amazon River System, Part I. The factors that control the salinity
and composition of the suspended solids. Bull. Geol. Soc. Am. 78, 203-1232.
Gilbert, G.K., 1875. Report on The Geology of Portions of Nevada, Utah, California and Arizona,
1871-1872. Report upon Geographical and Geological Explorations and Surveys West of the One
Hundredth Meridian 3, pp. 21-187.
Gilbert, G.K., 1877. Report on the geology of the Henry Mountains. United States Geographical and
Geological Survey of the Rocky Mountain Region. US Department of the Interior, Washington, DC.
Gilbert, G.K., 1890. Lake Bonneville. US Geol. Surv. Monogr. 1.
Gilbert, G.K., 1904. Domes and dome structure of the High Sierra. Bull. Geol. Soc. Am. 15, 29-36.
Gilbert, G.K., 1906. Crescentic gouges on glaciated surfaces. Bull. Geol. Soc. Am. 17, 303-313.
Gile, L., Hawley, J.W., 1966. Periodic sedimentation and soil formation on an alluvial-fan piedmont in
southern New Mexico. Soil Sci. Soc. Am. Proc. 30, 261-268.
Gile, L.H., 1975. Holocene soils and soil-geomorphic relations in an arid region of southern New Mexico.
Gile, L.H., Peterson, F.F., Grossman, R.B., 1966. Morphological and genetic sequences of carbonate
accumulation in desert soils. Soil Sci. 101,347-360.
Gilley, J.E., Kottwitz, E.R., Simanton, J.R., 1990. Hydraulic characteristics of rills. Trans. Am. Soc. Agr.
Eng. 33, 1900-1906.
Gjessing, J., 1965. On "plastic scouting" and "subglacial erosion". Norsk Geogr. Tidskr. 20, 1-37.
Glennie, K.W., 1970. Desert Sedimentary Environments. Elsevier, Amsterdam, p. 222.
Glennie, K.W., 1972. Permian Rotliegendes of northwest Europe interpreted in the light of modem desert
sedimentation studies. Bull. Am. Assoc. Petr. Geol. 56, 1048-1071.
Glennie, K.W., Pugh, J.M., Goodall, T.M., 1994. Late Quaternary Arabian desert models of Permian
Rotliegend reservoirs. Shell Explor. Bull. 274, 1-19.
Goldich, S.S., 1938. A study of rock weathering. J. Geol. 46, 17-58.

Goldsmith, V., 1985. Coastal dunes. In: Davis, R.A. (Ed.), Coastal Sedimentary Environments. Springer,
New York, pp. 171-236.
Goldthwait, R.P., 1951. Development of end moraines in east-central Baffin Island. J. Geol. 59, 567-577.
G6mez Ortiz, A., 1996. E1Relleu d'Andorra. Morfologia glacial i periglacial. Govern d'Andorra. Ministeri
d'Educaci6, Joventut i Esports, p. 129, Andorra.
G6mez Ortiz, A., Salvador Franch, F., 1997. Procesos geom6rficos periglaciares en el Pandero del
Mulhac6n (Sierra Nevada). Experimentaci6n de campo sobre la efectividad de la geliturbaci6n y del
hielo en el suelo. Cuaternario y Geomorfologfa 11, 81-97.
G6mez Ortiz, A., Salvador Franch, F., 1998. Procesos periglaciares actuales en montafia mediterrfinea.
Ideas clave, trabajos de campo y resultados en Sierra Nevada. In: G6mez Ortiz, A., Salvador Franch, F.,
Schulte, L., Garcfa Navarro, A. (Eds), Procesos bioffsicos actuales en medios frios. Publicacions de la
Universitat de Barcelona, Barcelona, pp. 217-234.
G6mez Villar, A., 1996. Abanicos aluviales: aportaci6n te6rica a sus aspectos mils significativos.
Cuaternario y Geomorfologfa 10, 77-124.
Goodess, C.M., Palulikof, J.P., Davies, T.D., 1992. The Nature and Causes of Climate Change. Assessing
the Long Term Future. Belhaven Press, London, p. 248.
Goodsell, B., Hambrey, M.J., Glasser, N.F., 2002. Formation of band ogives and associated structures at
Bas Glacier d'Arolla, Valais, Switzerland. J. Glaciol. 48, 207-300.
Goosens, D., Offer, Z.I., 1990. A wind-tunnel simulation and field verification of desert dust deposition.
Gordon, J.E., 1981. Ice-scoured topography and its relationships to bedrock structure and ice movement in
parts of northern Scotland and west Greenland. Geogr. Ann. 63A, 55-65.
Gordon, J.E., Whalley, W.B., Gellatly, A.F., Vere, D.M., 1992. The formation of glacial flutes: assessment
of models with evidence from Lyngsdalen, north Norway. Quaternary Sci. Rev. 11,709-731.
Gossmann, H., 1976. L'importance des processus se d6roulant fi la ligne de partage locale des eaux por
l'6volution des versants sous la dominance du ruissellement pluvial (fi l'aide de formules
math6matiques 616mentaires). Actes du Symposium sur les versants en pays m6diterran6es, Aix en
Provence, 28-30 Avril 1975. Aix en Provence: Centre d'Etudes G6ographiques et de Recherches
M6diterran6ennes Vol. V, 139-143.
Goudie, A.S., 1972a. On the definition of calcrete deposits. Z. Geomorphol. 16, 464-468.
Goudie, A.S., 1972b. The chemistry of world calcrete deposits. J. Geol. 80, 449-463.
Goudie, A.S., 1973. Duricrust in Tropical and Subtropical Landscapes. Clarendon Press, Oxford, p. 174.
Goudie, A.S., 1974. Further experimental investigation of rock weathering by salt and other mechanical
processes. Z. Geomorphol. 21, 1- 12.
Goudie, A.S., 1977. Sodium sulphate weathering and the disintegration of Mohenjo-Daro, Pakistan. Earth
Surf. Processes 2, 75-86.
Goudie, A.S., 1978. Dust storms and their geomorphological implications. J. Arid Environ. 1, 291-310.
Goudie, A.S., 198 l a. Geomorphological Techniques. Allen and Unwin, London, p. 395.
Goudie, A.S., 198 lb. The Human Impact. Man's Role in Environmental Change. B lackwell, Oxford, p. 326.
Goudie, A.S., 1983a. The arid earth. In: Gardner, R., Scoging, H. (Eds), Megageomorphology. Clarendon
Press, Oxford, pp. 152-171.
Goudie, A.S., 1983b. Calcrete. In: Goudie, A.S., Pye, K. (Eds), Chemical Sediments and
Geomorphology: Precipitates and Residua in the Near Surface Environment. Academic Press,
Goudie, A.S., 1986. The Human Impact on the Natural Environment. Blackwell, Oxford, p. 338.
Goudie, A.S., 1988. The geomorphological role of termites and earthworms in the tropics. In: Viles, H.
(Ed.), Biogeomorphology. Blackwell, Oxford, p. 365.
Goudie, A.S., 1989a. Weathering processes. In: Thomas, D.S.G. (Ed.), Arid Zone Geomorphology.
Belhaven Press, London, pp. 11-24.
Goudie, A.S., 1989b. Wind erosion in deserts. Proc. Geologist. Assoc. 100, 83-92.
Goudie, A.S., 1990a. Desert degradation. In: Goudie, A.S. (Ed.), Techniques for Desert Reclamation.
Goudie, A.S., 1990b. The global geomorphological future. Z. Geomorphol. Suppl.bd. 79, 51-62.
Goudie, A.S., (Ed.), 1990c. Techniques for Desert Reclamation. Wiley, Chichester, p. 271.
Goudie, A.S., 1991. Pans. Prog. Phys. Geog. 15, 221-237.
Goudie, A.S., 1992. Environmental Change. Clarendon Press, Oxford, p. 329.
References 717
Goudie, A.S., 1994a. Salt attack on buildings and other structures in arid lands. In: Fookes, P.G.,
Parry, R.H.G. (Eds), Engineering Characteristics of Arid Soils. Balkema, Rotterdam, pp. 15-28.
Goudie, A.S., 1994b. Deserts in a warmer world. In: Millington, A.C., Pye, K. (Eds), Environmental
Change in Drylands: Biogeographical and Biogeomorphological Perspectives. Wiley, Chichester,
pp. 1-24.
Goudie, A.S., 1995. The Changing Earth. Rates of Geomorphological Processes. Blackwell, Oxford,
p. 302.
Goudie, A.S., 1998. The salt weathering hazards in deserts. In: Kalvoda, J., Rosenfeld, C.L. (Eds),
Geomorphological Hazards in High Mountain Areas. Kluwer, Dordrecht, pp. 107-120.
Goudie, A.S., 2004a. Glacial erosion. In: Goudie, A.S. (Ed.), Encyclopedia of Geomorphology. Routledge,
pp. 447-448.
Goudie, A.S., 2004b. Pan. In: Goudie, A.S. (Ed.), Encyclopaedia of Geomorphology. Routledge, London,
758-759.
Goudie, A.S., Pye, K., (Eds), 1983. Chemical Sediments and Geomorphology. Academic Press, London,
p. 439.
Goudie, A.S., Thomas, D.S.G., 1985. Pans in southern Africa with particular reference to South Africa and
Zimbabwe. Z. Geomorphol. NF 29, 1-19.
Goudie, A.S., Viles, H., 1995. The nature and pattern of debris liberation by salt weathering: a laboratory
study. Earth Surf. Processes Landforms 20, 437-449.
Goudie, A.S., Viles, H., 1997. Salt Weathering Hazards. Wiley, Chichester, p. 241.
Goudie, A.S., Wells, G.L., 1995. The nature, distribution and formation of pans in arid zones. Earth Sci.
Rev. 38, 1-69.
Goudie, A.S., Wilkinson, J., 1977. The Warm Desert Environment. Cambridge University Press,
Cambridge, MA, p. 88.
Goudie, A.S., Cooke, R.U., Evans, I., 1970. Experimental investigation of rock weathering by salts. Area,
42-48.
Goudie, A.S., Warren, A., Jones, D.K.C., Cooke, R.U., 1987. The character and possible origins of the
aeolian sediments of the Wahiba Sand Sea, Oman. Geogr. J. 153, 231-256.
Gow, A.J., Epstein, S., Stteeby, W., 1979. On the origin of stratified debris in ice cores from the bottom of
the Antarctic ice sheet. J. Glaciol. 23, 185-192.
Gracia, F.J., 1985. Geomorfologfa de las Bardenas Orientales, Tesis de Licenciatura. Facultad de Ciencias,
Zaragoza, p. 172.
Gracia, F.J., 1995. Shoreline forms and deposits in Gallocanta Lake (NE Spain). Geomorphology 11,
323-335.
Graf, W.L., 1979. The development of montane arroyos and gullies. Earth Surf. Processes 4, 1-14.
Graf, W.L., 1980. The effect of dam closure on downstream rapids. Water Resour. Res. 16, 129-136.
Graf, W.L., 1988. Fluvial Processes in Dryland Rivers. Springer, Berlin, p. 346.
Grant, W.H., 1969. Abrasion pH, and index of weathering. Clays Clay Miner. 17, 151-155.
Gravis, G.F., 1969. Fossil slope deposits in the northern Arctic asymmetrical valleys. Biuletyn
Gray, J.M., 1981. P-forms from the Isle of Mull. Scot. J. Geol. 17, 39-47.
Gray, J.M., 1991. Glaciofluvial landforms. In: Ehlers, J., Gibbard, P.L., Rose, J. (Eds), Glacial Deposits in
Britain and Ireland. Balkema, Rotterdam, pp. 443-453.
Greeley, R., Iversen, J.D., 1985. Wind as a Geological Process on Earth, Mars, Venus and Titan.
Cambridge University Press, Cambridge, MA, p. 333.
Greenway, D.R., 1987. Vegetation and slope stability. In: Anderson, M.G., Richards, K.S. (Eds), Slope
Stability. Wiley, Chichester, pp. 187-230.
Gregory, K.J., 1976. Drainage networks and climate. In: Derbyshire, E. (Ed.), Geomorphology and
Climate. Wiley, London, pp. 289-315.
Gribbin, J., Lamb, H.H., 1978. Climatic change in historical times. In: Gribbin, J. (Ed.), Climatic Change.
Cambridge University Press, Cambridge, MA, pp. 68-82.
Griggs, D.T., 1936. The factor of fatigue in rock exfoliation. J. Geol. 44, 783-796.
Grimm, E.C., Jacobson, G.L., Watts, W.A., Hansen, B.C.S., Maasch, K., 1993. A 50,000 year record of
climate oscillations from Florida and its temporal correlation with the Heinrich events. Science 261,
198-200.
Grolier, M.J., McCauley, J.F., Breed, C.S., Embabi, N.S., 1980. Yardangs of the Western desert. Geogr. J.
135, 191-212.
Grossman, S., Gerson, R., 1987. Fluviatile deposits and morphology of alluvial surfaces as indicators
of Quaternary environmental changes in the southern Negev, Israel. In: Frostick, L., Reid, Y. (Eds),
Desert Sediments: Ancient and Modern, Geological Society of London Special Publication,
pp. 17-29.
Grosso, S.A., Corte, A.E., 1989. Pleistocene ice-wedge coasts at 34~ eastern Andes piedmont, south-west
of South America. Geogr. Ann. 71A, 125-136.
Grove, A.T., 1958. The ancient ergs of Hausaland and similar formations of the south side of the Sahara.
Geogr. J. 124, 528-533.
Grove, A.T., 1977. The geography of semi-arid lands. Philos. Trans. R. Soc. London Ser. B 278,
457-475.
Grove, A.T., 1985. The physical evolution of the river basins. In: Grove, A.T. (Ed.). The Niger and its
neighbours. Balkema, Rotterdam, pp. 2-60.
Grove, A.T., Warren, A., 1968. Quaternary landforms and climate on the south side of the Sahara. Geogr.
J. 134, 194-208.
Grove, J.M., 1979. The glacial history of the Holocene. Prog. Phys. Geog. 3, 1-54.
Grove, J.M., 1988. The Little Ice Age. Routledge, London, p. 498.
Grund, A., 1914. Der geographische zyclus im karst. Z. Gesselchaft Erdkund 52, 621-740.
Grtinert, J., Busche, D., 1980. Large-scale fossil landslides at the Msak Mallat and Hamadat Manghini
escarpement. In: Salem, M.J., Busrewil, M.T. (Eds), Geology of Libya. Academic Press, London,
pp. 849-860.
Guill6n, J., Diaz, J.I., Palanques, A., 1992. Cuantificaci6n y evoluci6n durante el Siglo XX de los aportes
de sedimentos transportados como carga de fondo por el rio Ebro al medio marino. Revista de la
Sociedad Geol6gica de Espafia 5, 27-37.
Guillien, Y., 1951. Les gr6zes lit6es de Charente. Revue Geographique de Pyr6n6es et de Sud-Ouest 22,
154-162.
Guillien, Y., Lautridou, J.P., 1970. Recherches de g61ifraccion experimentale du Centre de G6omor-
phologic. I. Calcaires de Charentes. Bull. Cent. G6omorphol. Caen 5, 45, C.N.R.S.
Giinster, A., Hadley, N.F., Mitchell, D., Nobel, P.S., Seely, M., 1993. Form and function. In: Seely, M.
(Ed.), Deserts. Weldon Ower Limited, Australia, pp. 66-81.
Gupta, A., 1984. Urban hydrology and sedimentation in the humid tropics. In: Costa, J.E., Fleisher, P.J.
(Eds), Developments and Applications of Geomorphology. Springer, Berlfn, pp. 240-267.
Gupta, A., 1988. Large floods as geomorphic events in the humid tropics. In: Baker, V.R., Craig Kockel, R.,
Patton, P.C. (Eds), Flood Geomorphology. Wiley, New York, pp. 301-315.
Gupta, A., 1993. The changing geomorphology of the humid tropics. Geomorphology 7, 165-186.
Gupta, A., Dutt, A., 1989. The Auranga: description of a tropical monsoon river. Z. Geomorphol. 33,
73-92.
Gupta, A., Rahman, A., Poh Poh, W., Pitts, J., 1987. The old alluvium of Singapore and the extinct
drainage system of the South China Sea. Earth Surf. Processes Landforms 12, 259-275.
Guti6rrez, F., 1995. Synsedimentary and postsedimentary subsidence due to gypsum karstification
and derived hazards (Calatayud Graben, Iberian Range, Spain). In: Barends, F.B.J., Brouwer, F.J.J.,
Schr6der, F.H. (Eds), Land Subsidence, Natural Causes Measuring Techniques and Groningen
Gasfields. Balkema, The Hague, pp. 47-57.
Guti6rrez, F., 1996. Gypsum karstification induced subsidence: effects on alluvial systems and derived
geohazards, Calatayud Graben, Iberian Range, Spain. Geomorphology 16, 277-293.
Guti&rez, F., 1998. Fen6menos de subsidencia por disoluci6n de formaciones evaporfticas en las
fosas ne6genas de Teruel y Calatayud (Cordillera Ib6rica). Tesis Doctoral, Facultad de Ciencias de
Zaragoza.
Guti4rrez, F., Guti6rrez, M., Sancho, C., 1998. Geomorphological and sedimentological analysis of
a catastrophic flash flood in the Arfis drainage basin (Central Pyrenees, Spain). Geomorphology 22,
265-283.
Guti6rrez, M., 1986. Some remarks on the problem of climatic changes and geomorphological processes in
arid zones. In: L6pez-Vera, F. (Ed.), Quaternary Climate in Western Mediterranean, pp. 3-29, Madrid.
Guti6rrez, M., 1990. Historia de la Geomorfologfa, Historia de la Geologfa. Real Academia de Ciencias
Exactas, Ffsicas y Naturales, Madrid, pp. 115-131.
Guti6rrez, M., 2001. Geomorfologia Climatica Omega, Barcelona, p. 642.
Guti6rrez, M., Ibfifiez, M.J., 1979. Las "gnammas" de la regi6n de Alcafiiz. Estudios Geol6gicos 35,
193-198.
References 719
Gutirrrez, M., Gracia, F.J., 1997. Environmental interpretation and evolution of the Tertiary erosion
surfaces in the Iberian Range (Spain). In: Widdowson, M. (Ed.), Paleosurfaces: Recognition,
Reconstruction and Palaeoenvironmental Interpretation, Vol. 120. The Geological Society Special
Publication, London, pp. 147-158.
Guti~rrez, M., Gutirrrez, F., 1998. Geomorphology of the Tertiary gypsum formations in the Ebro
Depression. Geoderma, 1-29.
Gutirrrez, M., Martinez, V.H., 2001. Multiple talus flatirons, variations of scarp retreat rates and the
evolution of the slopes in Almaz~in Basin (semi-arid central Spain). Geomorphology 38, 19-29.
Gutirrrez, M., Pefia, J.L., 1976. Glacis y terrazas en el curso medio del rio Alfambra. Boletfn Geolrgico y
Minero 87, 561-570, Madrid.
Gutirrrez, M., Pefia, J.L., 1977. Las acumulaciones periglaciares del Macizo del Tremedal (Sierra de
Albarracfn). Boletfn Geol6gico y Minero 92, 101-110.
Gutirrrez, M., Pefia, J.L., 1981. Los glaciares rocosos y el modelado acompafiante en el firea de la
Bonaigua (Pirineo de Lrrida). Boletfn Geolrgico y Minero 92, 101-110.
Gutirrrez, M., Pefia, J.L., 1989. Upper Holocene climatic change and geomorphological processes on
slopes and infilled valleys from archaeological dating (NE-Spain), European Conference on Landscape
Ecological Impact of Climate Change Lunteren, Netherlands, 3-7 December 1989. Dutch Ministry of
the Environment, Lunteren, p. 21.
Gutirrrez, M., Pefia, J.L., 1992. Evolucirn clim~itica y geomorfolrgica de Holoceno Superior (Cordillera
Ibrrica, Depresirn del Ebro y Pre-Pirineo). In: Cearreta, A., Ugarte, F.M. (Eds), The Late Quaternary
in the Western Pyrenean Region. Servicio Editorial de la Universidad del Pals Vasco, Bilbao,
pp. 109-124.
Gutirrrez, M, Pefia, J.L., 1994. Depresirn del Ebro. In: Gutirrrez, M. (Coord.), Geomorfolog/a de Espafia.
Rueda, Madrid, pp. 305-349.
Gutirrrez, M., Pefia, J.L., 1998. Geomorphology and Upper Holocene climatic change in northeastern
Spain. Geomorphology 23, 205- 217.
Gutirrrez, M., Rodrfguez, J., 1978. Consideraciones sobre la morfogrnesis del Sistema Central. Boletfn
Geolrgico y Minero 89, 109-113.
Gutirrrez, M., Sancho, C., 1993. Applied geomorphology in arid and semiarid regions, Second Intensive
Course of Applied Geomorphology: Arid Regions. Universidad de Zaragoza-Erasmus, Zaragoza,
pp. 3-31.
Guti~rrez, M., Ib~ifiez,M.J., Pefia, J.L., Rodrfguez, J., Soriano, M.A., 1985. Quelques exemples de karst sur
gypse dans la D~pression de l'Ebre. Karstologfa 6, 29-36.
Gutirrrez, M., Benito, G., Rodriguez, J., 1988. Piping in badlands areas of the middle Ebro basin, Spain.
Catena Suppl. 13, 49-60.
Gutirrrez, M., Sancho, C., Arauzo, T., 1998a. Scarp retreat in semiarid environments from talus flatirons
(Ebro Basin). Geomorphology 25, 111 - 121.
Guti~rrez, M., Sancho, C., Arauzo, T., Pefia, J.L., 1998b. Evolution and paleoclimatic meaning of the talus
flatirons in the Ebro Basin (NE of Spain). In: Alsharhan, A.S., Glennie, K.W., Whittle, G.L. (Eds),
Quaternary Deserts and Climatic Change. Balkema, Rotterdam, pp. 593-599.
Gutirrrez, M., Sancho, C., Desir, G., Sirvent, J., Benito, G., Calvo, A., 1995. Erosirn hfdrica en terrenos
arcillosos y yesfferos de la Depresirn del Ebro, Proyecto Lucdeme. Ministerio de Agricultura, Pesca y
Alimentacirn, Zaragoza, p. 389.
Gutirrrez, M., Sancho, C., Benito, G., Sirvent, J., Desir, G., 1997. Quantitative study of piping processes in
badland areas of the Ebro Basin. Geomorphology 20, 237-253.
Gutirrrez, M., Desir, G., Gutirrrez, F., 2002. Yardangs in the semiarid central sector of the Ebro
Depression (NE Spain). Geomorphology 44, 155-170.
Guti~rrez, M., Gutirrrez, F., Desir, G., 2005. Considerations on the chronological and causal relationships
between tapus flatiron and palaeoclimatic changes in central and northeastern Spain. Geomorphology
(accepted).
Guti~rrez, M., Desir G., Gutirrrez, F., Marin C., 2005. Origin and evolution of playas and blowouts in the
semi arid zone of Tierra de Pinares (Duero Basin, Spain). Geomorphology (accepted).
Hack, J.T., 1960. Interpretation of erosional topography in humid temperate regions. Am. J. Sci. 258A,
80-97.
Hack, J.T., 1975. Dynamic equilibrium and landscape evolution. In: Melhorn, W.C., Flemal, R.C. (Eds),
Theories of Landform Development. George Allen and Unwin, London, pp. 87-102.
Hackman, R.J., 1965. Interpretation of Alaskan post-earthquake photographs. Photogramm., 31,604-620.
Hadley, R.F., Schumm, S.A., 1961. Sediment sources and drainage basin characteristics in upper
Cheyenne River basin. US Geol. Surv. Water Supply Paper 1531, 137-198.
Hagedorn, H., 1968. Uber ~iolischeAbtragung und Formung in der Stidest-Sahara. Erdekunde 22, 257-269.
Haigh, M.J., 1978. Microrills and dessication cracks: some observations. Z. Geomorphol. 12, 457-461.
Halimov, M., Fezer, F., 1989. Eight yardang types in central Asia. Z. Geomorphol. NF 33, 205-217.
Hall, A.M., 1986. Deep weathering patterns in north-east Scotland and their geomorphological signi-
ficance. Z. Geomorphol. 30, 407-422.
Hall, A.M., 1988. The characteristics and significance of deep weathering in the Gaick area, Grampian
Highlands, Scotland. Geogr. Ann. 70A, 309-314.
Hall, A.M., Thomas, M.F., Thorp, M.B., 1985. Late Quaternary alluvial placer development in the humid
tropics: the case of the Birim Diamond Placer, Ghana. J. Geol. Soc. 142, 777-787.
Hall, A.M., Mellor, A.M., Wilson, M.J., 1989. The clay mineralogy and age of deeply weathered rock in
north-east Scotland. Z. Geomorphol. Suppl.bd. 72, 97-108.
Hall, K., Thorn, C.E., Matsuoka, N., Prick, A., 2002. Weathering in cold regions: some thoughts and
perspectives. Prog. Phys. Geog. 26, 577-603.
Hallet, B., 1979. Subglacial regelation water film. J. Glaciol. 23, 321-334.
Hallsworth, EG., Robertson, G.K., Gibbons, F.R., 1955. Studies of pedogenesis in New South Wales, VII:
the "gilgai" soils. J. Soil Sci. 6, 1-31.
Hambrey, M.J., 1977. Foliation, minor folds and strain in glacier ice. Tectonophysics 39, 397-416.
Hambrey, M., 1994. Glacial Environments. UCL Press, London, p. 296.
Hambrey, M., Alean, J.C., 1992. Glaciers. Cambridge University Press, Cambridge, MA, p. 207.
Hamilton, A., 1976. The significance of patterns of distribution shown by forest plants and animals
in tropical Africa for the reconstruction of Upper Pleistocene palaeoenvironments: a review. In: Van
Zinderen Baker, E.M. (Ed.), Palaeoecology of Africa, Vol. 9. Balkema, Rotterdam, pp. 63-97.
Hansen, W.R., 1965. Effects of the earthquake of March 27, 1964, at Anchorage, Alaska. US Geol. Surv.
Prof. Paper 542-A, 64.
Hardie, L.A., Smoot, J.P., Eugster, H.P., 1978. Saline lakes and their deposits: a sedimentological
approach. In: Matter, A., Tucker, M. (Eds), Modern and Ancient Lake Sediments, International
Association of Sedimentologists, Special Publication, Vol. 2, pp. 7-41.
Harding, A.F., 1982. Introduction: climate change and archaeology. In: Harding, A.F. (Ed.), Climatic
change in Later Prehistory. Edinburgh University Press, Edinburgh, pp. 1-10.
Harris, C., Davies, M.C.R., 2000. Gelifluction: observations from large-scale laboratory simulations. Arct.
Harris, C., Davies, M.C.R., Rea, B.R., 2003. Gelifluction: viscous flow or plastic creep? Earth Surf.
Processes Landforms 28, 1289-1301.
Harris, S.A., 1968. Gilgai. In: Fairbridge, R.W. (Ed.), The Encyclopedia of Geomorphology. Dowden,
Hutchinson and Ross, Stroudsburg, PA, pp. 425-426.
Harris, S.A., 1981. Climatic Relationships of Permafrost Zones in areas of low winter snow-cover.
Biuletyn Periyglacjalny, 28, 227-240.
Harris, S.A., 1986. The Permafrost Environment. Croom Helm, London, p. 276.
Harris, S.A., 1988. The alpine periglacial zone. In: Clark, M.J. (Ed.), Advances in Periglacial Geo-
morphology. Wiley, Chichester, pp. 369-413.
Harris, S.A., 1998. A genetic classification of the paisa-like mounds in Western Canada. Biuletyn
Harris, S.E., 1943. Friction cracks and the direction of glacial movement. J. Geol. 51,244-258.
Harry, D.G., 1988. Ground ice and permafrost. In: Clark, M.J. (Ed.), Advances in Periglacial
Hart, M.G., 1986. Geomorphology Pure and Applied. Allen and Unwin, London, p. 228.
Harvey, A.M., 1978. Dissected alluvial fans in southeast Spain. Catena 5, 177-211.
Harvey, A.M., 1984a. Debris flows and fluvial deposits in Spanish Quaternary alluvial fans: implications
for fan geomorphology. In: Koster, E.H., Steel, R.J. (Eds), Sedimentology of Gravels and Conglo-
merates, Canadian Society Petroleum Geologist, Memoir, Vol. 10, pp. 123-132.
Harvey, A.M., 1984b. Aggradation and dissection sequences on spanish alluvial fans: influence on
morphological development. Catena 11,289-304.
Harvey, A.M., 1987a. Alluvial fans dissection: relationship between morphology and sedimentation.
In: Frostik, L., Reid, I. (Eds), Desert Sediments, Ancient and Modern, Geological Society of London,
35, pp. 87-103.
References 721
Harvey, A.M., 1987b. Patterns of Quaternary aggradational and dissectional landform development in
the Almeria region, southeast Spain: a dry region, tectonically active landscape. Die Erde 118,
193-215.
Harvey, A.M., 1990. Factors influencing Quaternary alluvial fans development in southeast Spain.
In: Rachocki, A., Church, M. (Eds), Alluvial fans. A Field Approach. Wiley, Chichester, pp. 247-269.
Harvey, A.M., 1996. The role of alluvial fans in the mountain fluvial systems of southeast Spain:
implications of climatic change. Earth Surf. Processes Landforms 21,543-553.
Harvey, A.M., 1997. The role of alluvial fans in arid zone fluvial systems. In: Thomas, D.G.S. (Ed.), Arid
Zone Geomorphology. Process, Form and Change in Drylands. Wiley, Chichester, pp. 231-259.
Haugen, R.K., Brown, J., 1971. Natural and man-induced disturbances of permafrost terrane. In: Coates,
D.R. (Ed.), Environmental Geomorphology. Geom. State University, Binghamton, Nueva York,
pp. 139-149.
Hay, R.L., Wiggins, B., 1980. Pellets, ooids, sepiolite and silica in three calcretes of the southwest United
States. Sedimentology 27, 559-576.
Hayden, B.P., 1988. Flood climates. In: Baker, V.R., Craig Kockel, R., Patton, P.C. (Eds), Flood
Hays, J., 1967. Land surfaces and laterites of the Northern Territory. In: Jennings, J.N., Mabbutt, J.A.
(Eds), Landform Studies from Australasia and New Guinea. Cambridge University Press, Cambridge,
MA, pp. 182-210.
Heathcote, R.L., 1983. The Arid Lands: Their Use and Abuse. Longman, London.
Heede, B.H., 1971. Characteristics and processes of soil piping in gullies. United States Department of
Agriculture Forest Service, Rocky Mountain Forest and Range Experiment Station Research Paper
RM-68, p. 15.
Heginbottom, J.A., 2002. Permafrost mapping: a review. Prog. Phys. Geog. 26, 623-642.
Heine, K., Walter, R., 1996. Gypcretes of the central Namib Desert, Namibia. In: Heine, K. (Ed.),
Palaecology of Africa and the Surroundings Islands, Vol. 24. Balkema, Rotterdam, pp. 173-201.
Heinrich, H., 1988. Origin and consequences of cyclic ice rafting in the Northeast Atlantic ocean during
the past 130,000 years. Quaternary Res. 29, 142-152.
Henderson, E.P., 1956. Large nivation hollows near Knob Lake, Quebec. J. Geol. 64, 607-616.
Henderson-Sellers, A., 1992. Continental cloudiness changes this century. GeoJournal 27, 255-262.
Hereford, R., 1984. Climate and ephemeral-stream processes: twentieth-century geomorphology in
alluvial stratigraphy of the Little Colorado River, Arizona. Bull. Geol. Soc. Am. 95, 654-668.
Herrero, J., Aragii~s, R., 1988. Suelos afectados por salinidad en Arag6n. Surcos de Arag6n 9, 5-10.
Herrero, J., Aragii6s, R., Amezketa, E., 1993. Salt-affected soils and agriculture in the Ebro Basin. In:
Guti6rrez, M., Sancho, C., Benito, G. (Eds), Second Intensive Course of Applied Geomorphology: Arid
Regions. Universidad de Zaragoza-Erasmus, Zaragoza, pp. 139-150.
Herron, S., Langway, C.C., 1979. The debris-laden ice at the bottom of the Greenland ice sheet. J. Glaciol.
23, 193- 207.
Hestners, E., 1985. A contribution to the prediction of slush avalanches. Ann. Glaciol. 6, 1-4.
Hewitt, K.J., 1967. Ice-front deposition and the seasonal affect, a Himalayan example. Trans. Inst. Br.
Geogr. 42, 93-106.
Hickson, C.J., Peterson, D.W., (Eds), 1990. Special Symposium Commemorating the 10th Anniversary
of the Eruption of Mount St Helens, May 18, 1980, Geosci. Can., 17, pp. 125-187.
Higgins, C.G., 1956. Formation of small ventifacts. J. Geol. 64, 506-516.
Higgins, C.G., 1975. Theories of landscape development: a perspective. In: Melhorn, W.N., Flemal, R.C.
(Eds), Theories of Landform Development. Allen and Unwin, Boston, pp. 1-28.
Higgins, C.G., 1984. Piping and sapping: development of landforms by groundwater flow. In: LaFleur,
R.G. (Ed.), Groundwater as a Geomorphic Agent. Allen and Unwin, Boston, pp. 18-58.
Higgins, C.G., Osterkamp, W.R., 1990. Seepage-induced cliff recession and regional denudation. In:
Higgins, C.G., Coates, D.R. (Eds), Groundwater Geomorphology, Geological Society of America
Special Paper, Vol. 252, pp. 291-317.
Hills, E.S., 1940. The lunette, a new landform of aeolian origin. Aust. Geogr. 3, 15-21.
Hill, I.D., Rackham, I.J., 1978. Indications of mass movement on the Jos Plateau, Nigeria. Z. Geomorphol.
22, 258- 274.
Hodges, W.K., 1982. Hydrologic characteristics of a badland pseudopediment slope system during
simulated rainstorm experiments. In: Bryan, R.B., Yair, A. (Eds), Badland Geomorphology and Piping.
Geobooks, Norwich, pp. 127-152.
Hodges, W.K., Bryan, R., 1982. The influence of material behaviour on runoff initiation in the Dinosaur
Badlands, Canada. In: Bryan, R., Yair, A. (Eds), Badland Geomorphology and Piping. GeoBooks,
Norwich, pp. 13-46.
Holmes, A., 1944. Principles of Physical Geology. Thomas Nelson, London.
Holmes, A., 1965. Principles of Physical Geology. Nelson, London, p. 1288.
Holmes, G.W., Hopkins, D.M., Foster, H.L., 1968. Pingos in central Alaska. Geogr. Surv. Bull. 1341 H,
40, Washington.
Holzer, T.L., 1984. Ground failure induced by ground-water withdrawal from unconsolidated sediment.
In: Holzer, T.L. (Ed.), Man-Induced Land Subsidence, The Geological Society of America. Reviews in
Engineering Geology, Vol. 6. Boulder, Colorado, pp. 67-105.
Holzer, T.L., Davis, S.N., Lofgren, B.E., 1979. Faulting caused by groundwater extraction in south-central
Arizona. J. Geophys. Res. 84, 603-612.
H611ermann, P., 1983. Blockgletscher als mesoformen der periglazialstufe. Bonner Geogr. Abh. 67, 1-73.
Hooke, R., Le, B., 1967. Processes on arid-region alluvial fans. J. Geol. 75, 438-460.
Hooke, R., Le, B., 1972. Geomorphic evidence for late-Wisconsin and Holocene tectonic deformation,
Death Valley, California. Bull. Geol. Soc. Am. 83, 2073-2098.
Hooke, R., Le, B., 1987. Mass movement in semi-arid environments and the morphology of alluvial fans.
In: Anderson, M.G., Richards, K.S. (Eds), Slope Stability. Wiley, Chichester, pp. 505-529.
Hooke, R., Le, B., Rohrer, W.L., 1979. Geometry of alluvial fans: effect of discharge and sediment size.
Earth Surf. Processes 4, 147-166.
Hooke, R., Le, B., Miller, S.B., Kohler, J., 1988. Character of englacial and subglacial drainage system in
the upper part of the ablation area of Storglaci~iren, Sweden. J. Glaciol. 34, 228-231.
Hoppe, G., 1959. Glacial morphology and inland ice recession in northern Sweden. Geogr. Ann. 41,
193-212.
Horton, R.E., 1933. The role of infiltration in the hydrologic cycle. Trans. Am. Geophys. Union 14,
446-460.
Horton, R.E., 1945. Erosional development of streams and their drainage basins: hydrophysical approach
to quantitative morphology. Bull. Geol. Soc. Am. 56, 275-370.
Houghton, J.T., Jenkins, G.C., Ephraums, J.J., 1990. Climatic Change, The IPCC Scientific Assessment.
Cambridge University Press, Cambridge, MA, p. 356.
Houghton, J.T., Meira Filho, L.G., Callander, B.A., Harris, N., Kattenberg, A., Maskell, K., 1996. Climate
Change 1995, The Science of Climatic Change. Published for the IPCC. Cambridge University Press,
Howard, A.D., 1942. Pediment passes and the pediment problem. J. Geomorphol. 5, 3-32, 95-136.
Howard, A.D., Dolan, R., 1981. Geomorphology of the Colorado River in the Grand Canyon. J. Geol. 89,
269-298.
Howard, A.D., Selby, M.J., 1994. Rock slopes. In: Abrahams, A.D., Parsons, A.J. (Eds), Geomorphology
of Desert Environments. Chapman and Hall, London, pp. 123-172.
Howard, A.D., Morton, J.B., Gad-el-Hak, M., Pierce, D.B., 1978. Sand transport model of barchan dune
equilibrium. Sedimentology 25, 307-338.
Hubble, G.D., Isbell, R.F., Northcote, K.H., 1983. Features of Australian soils, Soils: an Australian
viewpoint. Academic Press, Melbourne, pp. 17-47.
Hudson, N., 1981. Soil Conservation. Batsford, London, p. 324.
Hughes, T.J., 1992. Abrupt climatic change related to unstable ice-sheet dynamics: toward a new
paradigm. Palaeogeogr. Palaeoclimatol. Palaeoecol. 97, 203-234.
Humlum, O., 1982. Rock glacier types on Disko, Central West Greenland. Jesgr. Tidgskr., 82, 59-66.
Hunt, C.B., y Mabey, D.R., 1966. Stratigraphy and Structure. Death Valley, California. US Geol. Surv.
Prof. Paper 494A, 162.
Hunt, C.B., Washburn, A.L., 1960. Salt features that simulate ground patterns formed in cold climates.
US Geol. Surv. Prof. Paper 494-B, 104-133.
Hunt, C.B., Washburn, A.L., 1966. Patterned ground. In: Hunt, C.B. et al. (Eds), Hydrologic basin Death
Valley, California, US Geological Survey Professional Paper, 494B, pp. 104-133.
Hunter, R.E., 1985. A kinematic model for the structure of lee-side deposits. Sedimentology 32, 409-422.
Hunter, R.E., Richmond, B.M., Alpha, T.R., 1983. Storm-controlled oblique dunes of the Oregon coast.
Bull. Geol. Soc. Am. 94, 1450-1465.
Hutchinson, G.E., 1957. A Treatise on Limnology: Geography, Physics and Chemistry, Vol. 1. Wiley,
New York, p. 1015.
References 723
Hutchinson, J.N., 1968. Mass movement. In: Fairbridge, R.W. (Ed.), The Encyclopedia of
Geomorphology. Reinhold, New York, pp. 688-695.
Hutchinson, J.N., 1988. General report: morphological and geotechnical parameters of landslides in
relation to geology and hydrogeology, Proceedings Fifth International Symposium on Landslides
(Laussanne). Balkema, Rotterdam, pp. 3-35.
Hutton, J.T., Dixon, J.C., 1981. The chemistry and mineralogy of some South Australian calcretes and
associated soft carbonates and their dolomitization. J. Geol. Soc. Aust. 28, 71-79.
Idso, S.B., 1974. Tornado or dust devil: enigma of desert whirlwinds. Am. Sci. 62, 530-541.
Imeson, A.C., Kwaad, F.J.P.M., Verstratten, J.M., 1982. The relationship of soil physical and chemical
properties to the development of badlands in Morocco. In: Bryan, R.B., Yair, A. (Eds), Badland
Geomorphology and Piping. Geobooks, Norwich, England, pp. 47-70.
Inbar, M., 1972. A geomorphic analysis of a catastrophic flood in a Mediterranean basaltic watershed.
University of Haifa, 22 International Geographical Congress Publications.
Ingles, O.G., Aitchinson, G.D., 1970. Soil water disequilibrium as a cause of subsidence in natural soils
and earth embankments, International Symposium on Land Subsidence, International Association of
Hydrological Sciences, Publication 89, pp. 342-353.
Innes, J., 1983. Debris flows. Prog. Phys. Geogr. 7, 469-501.
Isarin, R.F.B., 1997. Permafrost distribution and temperatures in Europe during the Younger Dryas.
Permafrost Periglac. Processes 8, 313-333.
Ishihara, K., 1989. Liquefaction-induced landslide and debris flow in Tajikistan, URSS. Landslide News 3, 6-7.
International Society of Soil Mechanics and Foundation Engineering, 1985. Progress Report: peculiarities
of Geotechnical behaviour of tropical lateritic and saprolitic soils, First International Conference and
Saprolitic Soils. Committee on Tropical Soils of International Society of Soil Mechanics and
Foundation Engineering, Sao Paulo, p. 449.
Iverson, R.N., 1980. Processes of accelerated pluvial erosion on desert hillslopes modified by vehicular
activity. Earth Surf. Processes 5, 369-388.
Iverson, N.R., 1995. Processes of erosion. In: Menzies, J. (Ed.), Modern Glacial Environments: Processes,
Dynamics and Sedimets. Butterworth-Heineman, Oxford, pp. 241-260.
Iversen, J.D., Greeley, R., Marshall, J.R., Pollack, J., 1987. Aeolian saltation threshold: effect of density
ratio. Sedimentology 34, 699-706.
Jahn, A., 1972. Tundra polygons in the Mackenzie Delta. G6ttinger Geogr. Abh. 60, 285-292.
Jahn, A., 1975. Problems of the Periglacial Zone. National Technical Information Service. Panstowe
Wydawnistwo Naukowe, Warsaw, Poland, p. 223.
Jahn, A., 1976. Geomorphological modeling and nature protection in Arctic and Subarctic environments.
Geoforum 7, 121-137.
Jahns, R.H., 1943. Sheet structure in granites: its origin and use as a measure of glacial erosion in New
England. J. Geol. 60, 7-98.
Jakcus, L., 1977. Morphogenetics of Karst Regions. Adam Hilger, Bristol, p. 284.
Jausen, J.H.F., Van Iperen, J.M., 1991. A 220,000 years climatic record for the east equatorial Atlantic
Ocean and equatorial Africa: evidence from diatoms and opal phytoliths in the Zaire (Congo) deep sea
fan. Palaeoceanography 6, 573-591.
Jauzein, A., 1974. Les donnres sur le syst~me CaSO4"H20 et leurs implications grologiques. Rev. G~ogr.
Phys. Grol. Dyn. 16, 151 - 159.
Jeje, L.K., 1973. Inselberg evolution in a humid tropical environment: the example of south-western
Nigeria. Z. Geomorphol. 17, 194-225.
Jennings, J.N., 1971. Karst. The MIT Press, Cambridge, MA, p. 252.
Jennings, J.N., 1985. Karst Geomorphology. Blackwell, Oxford, p. 293.
Jenny, H., 1941. Factors of Soil Formation. McGraw-Hill, New York, p. 281.
Jenny, H., 1950. Origin of soils. In: Trask, P.D. (Ed.), Applied Sedimentation. Wiley, New York, pp. 41-61.
Jiskoot, H., Murray, T., Boyle, P., 2000. Controls on the distribution of surge-type glaciers in Svalbard.
J. Glaciol. 46, 412-422.
Johannessen, C.L., Feireisen, J.J., Wells, A.K., 1982. Weathering of ocean cliffs by salt expansion in mid-
latitude coastal environment. Shore Beach 50, 26-34.
Johnson, A.M., 1970. Physical Processes in Geology. Freeman Cooper and Co., San Francisco, p. 577.
Johnson, A.M., Rodine, J.R., 1984. Debris flow. In: Brunsden, D., Prior, D.B. (Eds), Slope Instability.
Johnson, D.W., 1932a. Rock planes of arid regions. Geogr. Rev. 22, 656-665.
Johnson, D.W., 1932b. Rock fans of arid regions. Am. J. Sci. 23, 389-416.
Johnson, G., 1960. Cryoturbartion at Zaragoza, Northern Spain. Z. Geomorphol. 4, 74-80.
Johnson, W.H., Menzies, J., 1996. Pleistocene supraglacial and ice-marginal deposits and landforms. In:
Menzies, J. (Ed.), Past Glacial Environments. Sediments, Forms and Techniques. Butterworth-
Heinemann, Oxford, pp. 137-160.
Joly, F., 1950. P~diments et glacis d'6rosion dans le Sud-Est du Maroc. Congres Internationale de
G6ographie 2, 110-125, Lisboa.
Jones, D.E., Holtz, W.G., 1973. Expansive soils - the hidden disaster: civil engineering. Am. Soc. Civil
Eng. 43 (8), 49-51.
Jones, D.K.C., Brunsden, D., Goudie, A.S., 1983. A preliminary geomorphological assessment of the
Karakoram Highway. Q. J. Eng. Geol. 16, 331-355.
Jones, F.O., 1973. Landslides of Rio de Janeiro and the Serra das Araras Scarpment, Brazil. US Geol. Surv.
Prof. Paper 697, p. 42.
Jones, J.A.A., 1981. The Nature of Soil Piping - - A Review of Research. Geobooks, Norwich, p. 301.
Jones, J.A.A., 1990. Piping effects in humid lands. Geol. Soc. Am. Spec. Paper 252, 111-138.
Jouzel, J., Lorius, C., 1999. Evolution du climat: du pass6 r6cent vers le futur. C. R. Acad. Sci. Paris 328,
229-239.
Jouzel, J., Lorius, C., Petit, J.R., Barkev, N.I., Kotlyakov, V.M., Petrov, V.M., 1987. Vostok ice core.
A continuous isotopic temperature record over the last climatic cycle (160,000 years). Nature 329,
403 -408.
Jungerius, P.D., 1984. A simulation model of blowout development. Earth Surf. Processes Landforms 9,
509-512.
Jungerius, P.D., van der Meulen, F., 1989. The development of dune blowouts, as measured with erosion
pins and sequential air photos. Catena 16, 369-376.
K~i~ib, A., Haeberli, W., 2001. Evolution of a high-mountain thermokarst lake in the Swiss Alps. Arct.
Kamb, B., LaChapelle, E.R., 1964. Direct observations of the mechanism of glacier sliding over bedrock.
J. Glaciol. 6, 159-172.
Kapitsa, A.P., Ridley, J.K., Robin, G.de Q., Siegert, M.J., Zotikov, I.A., 1996. A large deep freshwater lake
beneath the ice of central East Antarctica. Nature 381,684-686.
Kaplar, C.W., 1965. Stone migration by freezing of soils. Science 149, 1520-1521.
Kasse, C., 1997. Cold-climate sand sheet formation in North-Western Europe (c.14-12.4 ka); a response
to permafrost degradation and increased aridity. Permafrost Periglac. Processes 8, 295-311.
Kasse, C., 2002. Sandy aeolian deposits and environments and their relation to climate during the Last
Glacial Maximum and Late glacial in northwest and central Europe. Prog. Phys. Geog. 26,
507-532.
Kassler, P., 1973. The structural and geomorphic evolution of the Persian Gulf. In: Purser, B.H. (Ed.), The
Persian Gulf. Springer, New York, pp. 11-32.
Kaye, C.A., 1967. Kaolinisation of bedrock of the Boston, Massachusetts area. US Geol. Surv. Prof. Paper
575-C, C165-C172.
Keller, C., 1946. E1 Departamento de Arica. Ministerio de Economfa y Comercio, Santiago de Chile.
Keller, W.D., 1957. The Principles of Chemical Weathering. Lucas Brothers, Columbia, MO, p. 111.
Kemmis, T.L., 1989. Importance of the vegetation process to certain properties of basal tills deposited by
the Laurentide ice sheet in Iowa and Illinois, USA. Ann. Glaciol. 2, 147-152.
Kerr, R.C., Nigra, J.O., 1952. Eolian sand control. Bull. Am. Assoc. Petrol. Geol. 36, 1541-1573.
Kershaw, G.P., Gill, D., 1979. Growth and decay of palsas and peat plateaux in the Macmillan Pass -
Tsichu River area, Northwest Territories, Canada. Can. J. Earth Sci. 16, 1362-1367.
Kesel, R.H., 1973. Inselberg landform elements: definition and synthesis. Rev. G6omorphol. Dyn. 22,
97-108.
Kesel, R.H., 1976. The use of the refraction-seismic techniques in geomorphology. Catena 3, 91-98.
Kesel, R.H., 1977a. Slope runoff and denudation in the Rupununi savanna, Guyana. J. Trop. Geogr. 44,
33-42.
Kesel, R.H., 1977b. Some aspects of the geomorphology of inselbergs in central Arizona, USA.
Kesseli, J.E., Beaty, C.B., 1959. Desert flat conditions in the White mountains of California and Nevada.
US Army Quartermaster Research and Engineering Center. Technical Report, EP-108.
Keylock, C., 1997. Snow avalanches. Prog. Phys. Geogr. 21,481-500.
References 725
Khalil, G.M., 1990. Floods in Bangladesh: a question of disciplining the rivers. Nat. Hazards 3,
379-401.
Kiewietdejonge, C.J., 1984. Bfidel's geomorphology II. Prog. Phys. Geogr. 8, 365-397.
Killigrew, L.P., Gilkes, R.J., 1974. Development of playa lakes in south western Australia. Nature 247,
454-455.
King, C.A.M., 1976a. Introduction. In: King, C.A.M. (Ed.), Periglacial Processes. Dowden, Hutchinson
and Ross, Stroudsburg, pp. 1-6.
King, C.A.M., 1976b. Landforms and Geomorphology. Concepts and History. Dowden, Hutchinson and
Ross, Stroudsburg, p. 404.
King, L.C., 1942. South African Scenery. A Textbook of Geomorphology. Oliver and Boyd, Edinburgh,
p. 308.
King, L.C., 1948. A theory of bornhardts. Geogr. J. 112, 83-87.
King, L.C., 1949. The pediment problem: some current problems. Geol. Mag. 86, 245-250.
King, L.C., 1953. Canons of landscape evolution. Bull. Geol. Soc. Am. 64, 721-752.
King, L.C., 1957a. Landscape study in southern Africa. Proc. Geol. Soc. S. Afr. 50, 22-52.
King, L.C., 1957b. The uniformitarian nature of hillslopes. Trans. Edin. Geol. Soc. 17, 81-102.
King, L.C., 1962. The Morphology of the Earth. A Study and Synthesis of World Scenery. Oliver and
Boyd, Edinburgh, p. 726.
King, L.C., 1975. Bornhardt landforms and what they teach. Z. Geomorphol. 19, 299-318.
King, L.C., 1976. Planation remnants upon high lands. Z. Geomorphol. N.F. 20, 133-148.
King, L.C., 1983. Wandering Continents and Spreading Sea Floors on an Expanding Earth. Wiley,
Chichester, p. 232.
King, P.B., Schumm, S.A., 1980. The Physical Geography (Geomorphology) of William Morris Davis.
Geo Books, Norwich, p. 217.
Kirkbride, M.P., 1995. Processes of transportation. In: Menzies, J. (Ed.), Modern Glacial Environments:
Processes, Dynamics and Sediments. Butterworth-Heineman, Oxford, pp. 261-292.
Kirkby, A.T.V., Kirkby, M.J., 1974. Surface wash at the semiarid break in slope. Z. Geomorphol.
Suppl.bd. 21, 151 - 171.
Kirkby, M.J., 1994. Thresholds and instability in stream head hollow: a model of magnitude and frequency
for wash processes. In: Kirkby, M.J. (Ed.), Process, Models and Theoretical Geomorphology. Wiley,
Chichester, pp. 215- 314.
Klammer, C., 1982. Die Palaeowfiste des Pantanal von Mato Grosso und die Pleistozane Klimageschichte
der Brazilianischen Randtropen. Z. Geomorphol. 26, 393-416.
Klappa, C.F., 1979. Lichen stromatolites: criterion for subaerial exposure and a mechanism for the
formation of laminar calcretes (caliche). J. Sediment. Petrol. 49, 387-400.
Klein, M., 1984. Weathering rates of limestones tombstones measured in Haifa, Israel. Z. Geomorphol. 28,
105-111.
Klimchouk, A., Andrejchuk, V., 1996. Environmental problems in gypsum karst terrains. Int. J. Speleol.
25, 145-156.
Klimchouk, A., Forti, P., Cooper, A., 1996. Gypsum karst of the world: a brief overview. Int. J. Speleol. 25,
159-181.
Knight, M.J., 1980. Structural analysis and mechanical origins of gilgai at Boorook, Victoria, Australia.
Geoderma 23, 245-283.
Knight, P.G., Waller, R.I., Patterson, C.J., Jones, A.P., Robinson, Z.P., 2002. Discharge of debris from ice
at the margin of the Greenland ice sheet. J. Glaciol. 48, 192-200.
Knighton, D., 1998. Fluvial Form and Processes. A New Perspective. Arnold, London, p. 383.
Knox, J.C., 1972. Valley alluviation in southwestern Wisconsin. Ann. Assoc. Am. Geogr. 62,
401-410.
Knox, J.C., 1983. Responses of river systems to Holocene climates. In: Wright, H.E. Jr. (Ed.), Late
Quaternary Environments of the United States, Holocene, Vol. 2. University of Minnesota Press,
Minneapolis, pp. 26-41.
Knox, J.C., 1984. Fluvial responses to small scale climate changes. In: Costa, J.E., Fleisher, P.J. (Eds),
Developments and Applications of Geomorphology. Springer, Berlin, pp. 318-342.
Knox, J.C., 1993. Large increases in flood magnitude in response to modest changes in climate. Nature
361,430-432.
Knox, J.C., 1995. Fluvial systems since 20,000 years BP. In: Gregory, K.J., Starkel, L., Baker, V.R. (Eds),
Global Continental Palaeohydrology. Wiley, Chichester, pp. 87-108.
Kocurek, G., 1988. First-order and super bounding surfaces in eolian sequences. Bounding surfaces
revisited. In: Kocurek, G. (Ed.), Late Paleozoic and Mesozoic eolian deposits of the western interior of
the United States, Sediment. Geol., 56, pp. 193-206.
Kocurek, G., Nielson, J., 1986. Conditions favourable for the formation of warm-climate aeolian sand
sheets. Sedimentology 33, 795-816.
Kocurek, G., Harholm, K.G., Deynoux, M., Blakey, R.C., 1991. Amalgamated accumulations resulting
from climatic and eustatic changes, Akchar Erg, Mauritania. Sedimentology 38, 751-772.
Kocurek, G., Townsley, M., Yeh, E., Sweet, M., Havholm, K., 1992. Dune and dune-field development
stages on Padre Island, Texas: effects of lee airflow and sand saturation levels and implications for
interdune deposition. J. Sediment. Petrol. 62, 622-635.
Kolla, V., Biscaye, P.E., Hanley, A.F., 1979. Distribution of quartz in Late Quaternary Atlantic sediments
in relation to climate. Quaternary Res. 11,261-277.
Koons, D., 1955. Cliff retreat in the southwestern United States. Am. J. Sci. 253, 44-52.
K6ppen, W., 1936. Das geographische system der klimate, Handbuch der Klimatologie, Vol. 1. Gerbruder
Borntraeger, Berlin.
Kor, P.S.G., Shaw, J., Sharpe, D.R., 1991. Erosion of bedrock by subglacial meltwater, Georgian Bay,
Ontario: a regional review. Can. J. Earth Sci. 28, 623-642.
Koster, E.A., Dijkmans, J.W.A., 1988. Niveo-aeolian deposits and denivation forms, with special
reference to the Great Kobuk Sand Dunes, Northwestern Alaska. Earth Surf. Processes Landforms 13,
153-170.
Krahn, J., Johnson, R.F., Fredlund, D.G., Clifton, A.W., 1979. A highway failure in Cretaceous sediments
at Maymount, Saskatchewan. Can. Geotech. J. 16, 703-715.
Krainer, K., Mostler, W., 2002. Hydrology of active rock glaciers: examples from the Austrian Alps. Arct.
Krauskopf, K.B., 1956. Dissolution and precipitation of silica at low temperatures. Acta Geochim.
Cosmochim. 10, 1- 26.
Krinsley, D.B., 1970. A geomorphological and paleoclimatological study of the playas of Iran. United
States Geological Survey, Washington, 2 Volumes.
Krinsley, D., Wellendorf, W., 1980. Wind velocities determined from the surface textures of sand grains.
Nature 283, 372-373.
Krumbein, W.E., Jens, K., 1981. Biogenic rock varnishes of the Negev Desert (Israel): an ecological study
of iron and manganese transformation by cyano-bacteria and fungi. Oecologia 50, 25-28.
Kuenen, Ph., 1960. Experimental abrasion, 4: eolian action. J. Geol. 68, 427-449.
Ktihnel, R.A., 1987. The role of cationic and anionic scavengers in laterites. Chem. Geol. 60, 31-40.
Lachenbruch, A., 1962. Mechanics of thermal contraction cracks and ice-wedge polygons in permafrost.
Geol. Soc. Am. Spec. Paper 70, p. 69.
Lachenbruch, A.H., 1970. Thermal considerations in permafrost. In: Adkison, W.L., Borsge, M.M. (Eds),
Geological Seminar on the North Slope of Alaska, Proc. Am. Assoc. Petr. Geol., J1-2 and discussion
J2-5. Los Angeles.
Laity, J.E., 1994. Landforms of aeolian erosion. In: Abrahams, A.D., Parsons, A.J. (Eds), Geomorphology
of Desert Environments. Chapman & Hall, London, pp. 506-537.
Laity, J.E., 1995. Wind abrasion and ventifact formation in California. In: Tchakerian, V.P. (Ed.), Desert
Aeolian Processes. Chapman & Hall, London, pp. 295-321.
Laity, J.E., Malin, M.C., 1985. Sapping processes and the development of theater-headed valley networks
in the Colorado Plateau. Bull. Geol. Soc. Am. 96, 203-217.
Lal, R., 1986. Deforestation and soil erosion. In: Lal, R., Sfinchez, P.A., Cummings, R.W. (Eds), Land
Cleating and Development in the Tropics. Balkema, Rotterdam, pp. 299-315.
Lal, R., 1990. Water erosion and conservation: an assessment of the water erosion problem and the
techniques available for soil conservation. In: Goudie, A. (Ed.), Techniques for Desert Reclamation.
Wiley, Chichester, pp. 161 - 198.
Lal, R., Lawson, T.L., Anastase, A.H., 1981. Erosivity of tropical rains. In: De Broodt, M., Gabriels, D.
(Eds), Assessment of Erosion. Wiley, Chichester, pp. 143-151.
Lamb, H.H., 1972. Climate: Past, Present and Future, Vol. 1. Methuen, London, p. 613.
Lamb, H.H., 1977. Climates: Present, Past and Future, Vol. II. Methuen, London, p. 835.
Lancaster, I.N., 1978. The pans of the Southern Kalahari, Botswana. Geogr. J. 144, 80-98.
Lancaster, N., 1981. Palaeoenvironmental implications of fixed dune systems in southern Africa.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 33, 327-346.
References 727
Lancaster, N., 1982. Dunes on the Skeleton Coast, Namibia (South West Africa): geomorphology and
grain size relationships. Earth Surf. Processes Landforms 7, 575-587.
Lancaster, N., 1983. Linear dunes of the Namib sand sea. Z. Geomorphol. Suppl.bd. 45, 27-49.
Lancaster, N., 1985. Variations in wind velocity and sand transport on the windward flanks of desert sand
dunes. Sedimentology 32, 581 - 593.
Lancaster, N., 1987. Variations in wind velocity and sand transport on the windward flanks of desert sand
dunes-reply. Sedimentology 43, 516-520.
Lancaster, N., 1988. Development of linear dunes in the southwestern Kalahari, southern Africa. J. Arid
Environ. 14, 233-244.
Lancaster, N., 1989a. The Namib Sand Sea: Dune Forms, Processes and Sediments. Balkema, Rotterdam,
p. 180.
Lancaster, N., 1989b. The dynamics of star dunes: an example from the Gran Desierto. Mexico.
Lancaster, N., 1989c. Star dunes. Prog. Phys. Geogr., 67-91.
Lancaster, N., 1994. Dune morphology and dynamics. In: Abrahams, A.D., Parsons, A.J. (Eds),
Lancaster, N., 1995. Geomorphology of Desert Dunes. Routledge, London, p. 290.
Lancaster, N., Ollier, C.D., 1983. Sources of sand for the Namib Sand Sea. Z. Geomorphol. Suppl.bd. 45,
71-83.
Land, L.S., 1970. Phreatic versus vadose meteoric diagenesis of limestones: evidence from a fossil water
table. Sedimentology 14, 175-185.
Langbein, W.B., 1949. Annual runoff in the United States. US Geol. Surv. Circ. 52, p. 14.
Langbein, W.B., 1961. Salinity and hydrology of closed lakes. US Geol. Surv. Prof. Paper 412, 1-20.
Langbein, W.B., Schumm, S.A., 1958. Yield of sediment in relation to mean annual precipitation. Trans.
Am. Geophys. Union 39, 1076-1084.
Lautridou, J.P., 1988. Recent advances in cryogenic weathering. In: Clark, M.J. (Ed.), Advances in
Periglacial Geomorphology. Wiley, Chichester, pp. 33-47.
Lawson, A.C., 1915. The epigene profiles of the desert. Univ. Calif. Dep. Geol. Bull. 9, 23-48.
Lawson, D.E., 1979. A comparison of the pebble orientations in ice and deposits of the Matanuska glacier,
Alaska. J. Geol. 87, 629-645.
Lawson, M.P., Thomas, D.S.G., 2002. Late Quaternary lunette dune sedimentation in the southwestern
Kalahari Desert, South Africa: luminescence based chronologies of aeolian activity. Quaternary Sci.
Rev. 21,825-836.
Lecce, S.A., 1990. The alluvial fan problem. In: Rachocki, A.H., Church, M. (Eds), Alluvial Fans. A Field
Approach. Wiley, Chichester, pp. 3-24.
Lee, J.A., Wigner, K.A., Gregory, J.M., 1993. Drought, wind and blowing dust on the southern High Plains
of the United States. Phys. Geogr. 14, 56-67.
Lee, S.W., 1981. Landslides in Taiwan. Proceedings South East Asian regional symposium on Problems of
soil erosion and sedimentation, Bangkok, pp. 195-206.
Legget, R.F., 1974. Glacial landforms and civil engineering. In: Coates, D.R. (Ed.), Glacial Geomor-
phology. State University of New York, Binghamton, pp. 351-374.
Lehmann, H., 1936. Morphologische Studien auf Java. Geogr. Abhandl III Heft 9, Stuttgart, p. 114.
Lehmann, H., 1955. Der tropische kegelkarst in westindien. Tagungsbericht Wiss. Abhandl., 126-131,
Wiesbaden.
Lehmann, H., Roglic, J., Rathjens, C., Lasserre, G., Harrassowitz, H., Corbel, J., Birot, P., 1954.
Karstph~inomen in den verschiedenen klimazonen. Erdkunde 8, 112-122.
Lelong, F., 1966. R6gimes des nappes phreatiques contenues dans las formations d'alt6ration tropicale.
Cons6quences pour la p6dogen~se. Sciences de la Terre 11,203-244.
Leopold, L.B., 1968. Hydrology for urban planning - a guidebook on the hydrologic effects of urban land
use. US Geol. Surv. Circ. 554, p. 18.
Leopold, L.B., 1976. Reversal of erosion cycle and climatic change. Quaternary Res. 6, 557-562.
Leopold, L.B., 1994. A View of the River. Harvard University Press, Cambridge, MA, p. 298.
Leopold, L.B., Maddock, T., 1954. The Flood Control Controversy. Ronald Press, New York, p. 178.
Leopold, L.B., Wolman, M.G., Miller, G.P., 1964. Fluvial Processes in Geomorphology. W.H. Freeman,
San Francisco, p. 522.
Leopold, L.B., Emmet, W.W., Myrick, R.M., 1966. Channel and hillslope processes in a semiarid area of
New Mexico, US Geol. Surv. Prof. Paper, 352G.
L6tolle, R., Mainguet, M., 1993. Aral. Springer, Paris, p. 357.

Levson, V.M., Morison, S.R., 1996. Geology of placer deposits in glaciated environments. In: Menzies, J.
(Ed.), Past Glacial Environments: Sediments, Forms and Techniques. Butterworth-Heinemann,
Oxford, pp. 441-478.
Linell, K.A., Johnston, G.H., 1973. Engineering design and construction in permafrost regions,
Proceedings of Permafrost 2st International Conference. Natural Research Council, Yakutsk, URSS,
pp. 553-575.
Linell, K.A., Lobacz, E.F., 1978. Some experiences with tunnel entrances in permafrost, Proceedings
Third International Conference of Permafrost, pp. 813-819. Ottawa.
Linton, D.L., 1955. The problem of tors. Geogr. J. 121,470-487.
Linton, D.L., 1963. The forms of glacial erosion. Trans. Inst. Br. Geogr. 33, 1-28.
Lipman, P.W., Mullineaux, D.R., (Eds), 1981. The 1980 Eruptions of Mount St. Helens, Washington,
US Geol. Surv. Prof. Paper, 1250.
Lister, L.A., Secrest, C.D., 1985. Giant desiccation cracks and differential surface subsidence, Red Lake
Playa, Mohave County, Arizona. Bull. Assoc. Eng. Geol. 22, 299-314.
Livingstone, I., Warren, A., 1996. Aeolian Geomorphology. An Introduction. Longman, Essex, p. 211.
Llamas, M.R., 1962. Estudio geol6gico-t6cnico de los terrenos yesfferos de la Cuenca del Ebro y de los
problemas que plantean en los canales. Ministerio de Obras P6blicas, Boletfn 12, Madrid, p. 192.
Lliboutry, L., 1965. Trait4 de Glaciologie. Masson, Paris, 2 Volumes, p. 1040.
Lliboutry, L., Briat, M., Cresoveur, M., Fourchet, M., 1976. 15 m deep temperatures in the glaciers
of Mont Blanc (French Alps). J. Glaciol. 16, 197-203.
Lloyd, J.W., 1983. Hydrogeology in glaciated terrains. In: Eyles, N. (Ed.), Glacial Geology.
An Introduction for Engineers and Earth Scientists. Pergamon Press, Oxford, pp. 349-368.
Lobacz, E.F., Quinn, W.F., 1966. Thermal regime beneath buildings constructed on permafrost,
Proceedings First International Conference of Permafrost. National Academy of Sciences, Washington,
Publication 1287, pp. 447-491.
L6ffler, E., 1974. Piping and pseudokarst features in the tropical lowlands of New Guinea. Erdkunde 28,
13-18.
L6pez-Bermfdez, F., 1973. La Vega Alta del Segura. Departamento de Geograffa. Universidad de Murcia,
Murcia, p. 288.
L6pez Bermfdez, F., 1986. Evaluaci6n de la erosi6n hfdrica en las fireas receptoras de los embalses de la
Cuenca del Segura. Aplicaci6n de la USLE. In: L6pez Bermfdez, F., Thornes, J. (Eds), Estudios sobre
Geomorfologfa del Sur de Espafia. Universidad de Murcia, Murcia, pp. 93-101.
L6pez Bermfdez, F., 1988. Desertificaci6n: magnitud del problema y estado actual de las investigaciones.
In: Guti6rrez, M., Pefia, J.L. (Eds), Perspectivas en Geomorfologfa, Vol. 2. Monograffa de la Sociedad
Espafiola de Geomorfologfa, Zaragoza, pp. 155-170.
L6pez-Bermfdez, F., Conesa-Garcfa, C., Alonso-Sarrifi, F., 2002. Floods: magnitude and frequency in
ephemeral streams of the Spanish Mediterranean region. In: Bull, L.J., Kirby, M.J. (Eds),
Drylands Rivers: Hydrology and Geomorphology of Semi-arid Channels. Wiley, New York, pp.
329-350.
L6pez-Martfnez, J., 1988. E1 riesgo debido a los aludes. In: Ayala, F.J., Durfin, J.J., v Peinado, T. (Eds),
Riesgos geol6gicos. Instituto Geol6gico y Minero de Espafia, Madrid, pp. 215-225.
Lorius, C., Barkov, N.I., Jouzel, J., Korotkevich, Y.S., Kotlyakov, V.M., Raynaud, D., 1988. Antarctic ice
core: CO2 and change over the last climatic cycle. EOS 69, 681-684.
Loughnan, F.C., 1969. Chemical Weathering of the Silicate Minerals. Elsevier, New York, p. 154.
Loughnan, F.C., Bayliss, P., 1961. The mineralogy of the bauxitic deposits near Weipa, Queensland.
Am. Mineral. 46, 209- 217.
Loughnan, F.C., Grim, R.W., Vernet, J., 1962. Weathering of some Triassic shales in the Sydney area.
J. Geol. Soc. Aust. 8, 245-257.
Lovegrove, G.W., Fookes, P.G., 1972. The planning and implementation of a site investigation for
a highway in tropical conditions in Fiji. Q. J. Eng. Geol. 5, 43-68.
Lozinski, W.Von., 1909. Ober die mechanische Verwitterung der Sandsteine im gem/issigten Klima.
Acad. Sci. Cracovie Bull. Int. C1. Sci. Math. Nat. 1, 1-25.
Lucchitta, I., 1975. Application of ERTS images and image processing to regional geologic problems and
geologic mapping in northern Arizona - Part IV B, the Shivwits Plateau. National Aeronautics and
Space Administration Technical Report 32-1597, 41-72.
Lumb, P., 1962. The properties of decomposed granite. G6otechnique 12, 226-243.
References 729
Lumb, P., 1983. Engineering properties of fresh and decomposed igneous rocks from Hong-Kong.
Eng. Geol. 27, 287-298.
Lundquist, G., 1969. Earth and ice mounds: a terminological discussion. In: Prwr, T.L. (Ed.),
The Periglacial Environment. McGill-Queen's University Press, Montreal, pp. 203-215.
Lundqvist, J., 1989. Rogen (ribbed) moraine: identification and possible origin. Sediment. Geol. 62,
281-292.
Lvovitch, M.I., 1967. Water resources of the world and their future. General Assembly of Bern. Int. Assoc.
Sci. Hydrol., 317- 322.
Lyford, F.P., Qashu, H.K., 1969. Infiltration rates as affected by desert vegetation. Water Resour. Res. 5,
1373-1376.
Mabbutt, J.A., 1961a. Basal surface or weathering front. Proc. Geol. Assoc. 72, 357-358.
Mabbutt, J.A., 1961b. A stripped landsurface in Western Australia. Trans. Inst. Br. Geogr. 29,
101-114.
Mabbutt, J.A., 1965a. Stone distribution in a Stony Table-land soil. Aust. J. Soil Res. 3, 131-142.
Mabbutt, J.A., 1965b. The weathered landsurface of central Australia. Z. Geomorphol. 9, 82-114.
Mabbutt, J.A., 1966. Mantle-controlled planation of pediments. Am. J. Sci. 264, 78-91.
Mabbutt, J.A., 1977. Desert landforms. The MIT Press, Cambridge, MA, p. 340.
MacAyeal, D.R., 1993. A low-order model of the Heinrich event cycle. Paleoceanography 8, 767-773.
Mackay, J.R., 1973a. Problems in the origin of glassif Icy Beds, Western Arctic, Canada. North American
Contribution. Permafrost. 2nd Int. Conf., Yakutsk, USSR. Nat. Acad. Sciences, Washington, 10,
223-228.
Mackay, J.R., 1973b. The growth of pingos, western Arctic coast, Canada. Can. J. Earth Sci. 10,
979-1004.
Mackay, J.R., 1979. Pingos of the Tuktoyaktuk Peninsula area, Northwest Territories. G~ogr. Phys.
Quatern. 33, 3-61.
M/ickel, R., 1974. Dambos. A study in morphodynamic activity on the plateau regions of Zambia. Catena
1,327- 365.
M~ickel, R., 1985. Dambos and related landforms in Africa - an example for the ecological approach to
tropical geomorphology. Z. Geomorphol. Suppl.bd. 52, 1-23.
Machette, M.N., 1985. Calcic soils of the southwestern United States. In: Weide, D. (Ed.), Soils and
Quaternary geology of the Southwestern United States, Geological Society of America Special Paper,
Vol. 203, pp. 1- 21.
Magarerz, M., Enzel, Y., 1990. Standing-water deposits as indicators of Late Quaternary dune migration in
the northwestern Neger, Israel. Climatic Change, 336, 367-318.
Maignien, R., 1966. Review of Research on Laterites, Vol. 4. UNESCO Natural Resources Research,
Paris.
Mainguet, M., 1972. Le Model~ des Gr~s. Problemes Grn~raux. Institut Grographique National, Paris, p. 657.
Mainguet, M., 1975. Etude comparre des ergs ~ l'rchelle continentale (Sahara et d~serts d'Australie). Bull.
Assoc. Grogr. Franqaises 52, 135-140.
Mainguet, M., 1978. The influence of trade winds, local air masses and topographic obstacles on the
aeolian movement of sand particles and the origin and distribution of ergs in the Sahara and Australia.
Geoforum 9, 17-28.
Mainguet, M., 1983. Dunes vives, dunes fixres, dunes v~tues: une classification selon le bilan
d'alimentation, le r~gime ~olien et la dynamique des 6difices sableux. Z. Geomorphol. Suppl.bd. 45,
265-285.
Mainguet, M., 1984. A classification of dunes based on aeolian dynamics and the sand budget. In: E1 Baz,
F. (Ed.), Deserts and Arid Lands. Martinus Nijhof, The Hague, pp. 59-77.
Mainguet, M., 1991. Desertification. Natural Background and Human Mismanagement. Springer, Berlin,
p. 306.
Mainguet, M., Canon, N., Chemin, M., 1980. Le Sahara: g~omorphologie et palrogromorphologie
6oliennes. In: Williams, M.A.J., Faure, H. (Eds), The Sahara and the Nile. Balkema, Rotterdam,
pp. 17-35.
Maizels, J.K., 1981. Freeze/thaw experiments in the simulation of sediment cracking patterns. Bedford
College London, Papers in Geography 13.
Maizels, J.K., 1987. Plio-Pleistocene raised channel systems of the western Sharqiya (Wahiba), Oman.
In: Frostick, L.E., Reid, I. (Eds), Desert Sediments: Ancient and Modern. Blackwell Scientific, Oxford,
pp. 31-50.
Maizels, J., 1995. Sediments and landforms of modern proglacial terrestrial environments. In: Menzies, J.
(Ed.), Modern Glacial Environments. Processes, Dynamics and Sediments. Butterworth-Heinemann,
Maizels, J., 1997. J6kulhlaup deposits in proglacial areas. Quaternary Sci. Rev. 16, 783-819.
Mammerickx, J., 1964. Quantitative observations on pediments in the Mojave and Sonoran deserts
(Southwestern United States). Am. J. Sci. 262, 417-435.
Mann, A.W., 1984. Mobility of gold and silver in lateritic weathering profiles: some observations from
western Australia. Econ. Geol. 79, 38-49.
Mannion, A.M., 1997. Global Environmental Change. Longman, Essex, p. 387.
Marko, J.R., Fissel, D.B., Miller, J.D., 1988. Iceberg movement prediction off the Canadian east coast.
In: E1-Sabh, M.I., Murty, T.S. (Eds), Natural and Man-Made Hazards. D. Reidel Publishers, Dordrecht,
pp. 435-462.
Marsden, S.S., Davis, S.N., 1967. Geological subsidence. Sci. Am. 216, 93-100.
Marshall, A.F., Workman, J.P., 1977. Identification of dispersive clays in the Texas Gulf Coast Area.
In: Sherard, J.L., Decker, R.S. (Eds), Dispersive Clays, Related Piping, and Erosion in Geotechnical
Projects, pp. 274-286.
Martfnez, J., Garz6n, M.G., Arche, A., 1987. Avenidas e inundaciones. Ministerio de Obras Pfiblicas y
Urbanismo, Madrid, p. 67.
Martini, J.P., 1978. Tafoni weathering, with examples from Toscany, Italy. Z. Geomorphol. 22, 44-67.
Martfn-Serrano, A., 1988. E1 relieve de la regi6n occidental zamorana. La evoluci6n geomorfol6gica de
un borde del Macizo Hesprrico. Instituto de Estudios Zamoranos Flori~in de Ocampo, Zamora, p. 311.
Martonne, E. de, Aufr~re, L., 1928. L'extension des rrgions privres d'rcoulement vers l'ocean. Ann. de
Grogr., 38, 1-24.
Mason, B., 1958. Principles of Geochemistry. Wiley, New York, p. 310.
Matthes, F.E., 1900. Glacial sculpture of the Bighorn Mountains, Wyoming, United States Geological
Survey 21 st Annual Report 1899-1900, pp. 167-190.
Matthes, F.E., 1930. Geologic history of the Yosemite valley. US Geol. Surv. Prof. Paper 160, 54-103.
Maxson, J.H., 1940. Fluting and faceting of rock fragments. J. Geol. 48, 717-751.
Mayer, L., Gerson, R., Bull, W.B., 1984. Alluvial gravel production and deposition - a useful indicator
of Quaternary climatic changes in deserts (a case study in south-western Arizona). Catena Suppl. 5,
137-151.
McBean, G.A., Liss, P.S., Schneider, S.H., 1996. Advancing our understanding. In: Houghton, J.T. (Ed.),
Climate Change 1995. The Science of Climatic Change. Cambridge University Press, Cambridge, MA,
pp. 517-531.
McCall, J.G., 1960. The flow characteristics of a cirque glacier and their effect on cirque formation. In:
Lewis, W.V. (Ed.), Investigations on Norvegian Cirque Glaciers, Royal Geographers Society,
Research Series, Vol. 4, pp. 39-62.
McCauley, J.F., Grolier, M.J., Breed, C.S., 1977. Yardangs. In: Doehring, D.O. (Ed.), Geomorphology
in Arid Regions. Allen and Unwin, London, pp. 233-269.
McCauley, J.F., Breed, C.S., E1-Baz, F., Whitney, M.I., Grolier, M.J., Ward, A.W., 1979. Pitted and fluted
rocks in the Western Desert of Egypt-Viking comparisons. J. Geophys. Res. 84, 8222-8232.
McCauley, J.F. et al., 1982. Subsurface valleys and geoarchaeology of the eastern Sahara revealed
by Shuttle Radar. Science 218, 1004-1020.
McClung, D.M., 2003. Magnitude and frequency of avalanches in relation to terrain and forest cover. Arct.
Antarc. Alp. Res. 35, 82-90.
McClung, D., Schaerer, P., 1993. The Avalanche Handbook, The Mountaineers, p. 272.
McClure, H.A., 1978. ArRub' Alkhali. In: A1-Sayari, S.S., Z6tl, J.G. (Eds), Quaternary Period in
Saudi Arabia. Springer, Viena, pp. 252-263.
McFarlane, M.J., 1976. Laterite and Landscape. Academic Press, London, p. 151.
McFarlane, M.J., 1983. Laterites. In: Goudie, A.S., Pye, K. (Eds), Chemical Sediments and Geomor-
phology. Academic Press, London, pp. 7-58.
McFarlane, M.J., 1987. The key role of microorganisms in the process of bauxitisation. Mod. Geol. 11,
325-344.
McFarlane, M.J., 1989. Dambos - their characteristics and geomorphological evolution in parts of Malawi
and Zimbabwe, with particular reference to their role in the hydrogeological regime of surviving areas
of African surface. Proceedings of the groundwater exploration. Harare, Zimbabwe, 15-24 June 1987.
Commonwealth Science Council, Vol. 1,254-308.
References 731
McFarlane, M.J., Pollard, S., 1989. Some aspects of stone-lines and dissolution fronts associated with
regolith and dambo profiles in parts of Malawi and Zimbabwe. In: Alexandre, J., Symoens, J.J. (Eds),
Stone-Lines, Geo-Eco-Trop, Vol. 11. Acadrmie Royale des Sciences D'Outre-Mer, Bruxelles,
pp. 23-35.
McFarlane, M.J., Whitlow, R., 1990. Key factors affecting the initiation and progress of gullying in
dambos in parts of Zimbabwe and Malawi. Land Degrad. Rehabil. 2, 215-235.
McGee, W.J., 1897. Sheetflood erosion. Bull. Geol. Soc. Am. 8, 87-112.
McGowen, J.H., 1979. Alluvial fan systems. In: Galloway, W.E., Kreitler, C.V., McGowen, J.H. (Eds),
Depositional and Ground Water Flow Systems in the Exploration for Uranium, Texas Bureau of
Economic Geology Research Colloquium, pp. 43-79.
McGreevy, J.P., 1982. Frost and salt weathering: further experimental results. Earth Surf. Processes
Landforms 7, 475-488.
McGregor, G.R., Nieuwolt, S., 1998. Tropical climatology, An Introduction to the Climates of the Low
Latitudes. Wiley, Chichester, p. 339.
Mclntyre, D.S., 1979. Exchangeable sodium, subplasticity, and hydraulic conductivity of some Australian
soils. Aust. J. Soil Res. 17, 115-120.
McKee, E.D., 1979a. A study of global sand seas. US Geol. Soc. Prof. Paper 1052, Washington,
p. 429.
McKee, E.D., 1979b. Introduction to a study of global sand seas. In: McKee, E.D. (Ed.), A Study of
Global Sand Seas, US Geological Survey Professional Paper, Vol. 1052, pp. 1-19, Washington.
McKee, E.D., Bigarella, J.J., 1979. Sedimentary structures in dunes. In: McKee, E.D. (Ed.), A Study
of Global Sand Seas, US Geol. Surv. Prof. Paper, 1052, pp. 83-134, Washington.
McManus, J.F., 2004. A great grand-daddy of ice cores. Nature 429, 611-612.
McTainsh, G.H., 1987. Desert loess in northern Nigeria. Z. Geomorphol. NF 31, 145-165.
Meadows, M.E., 1985. Dambos and environmental change in Malawi, central Africa. Z. Geomorphol.
Suppl.bd. 52, 147-169.
Mehta, P.K., 1983. Mechanism of sulphate attack on Portland cement concrete-another look. Cem. Concr.
Res. 13, 401-406.
Meigs, P., 1953. The world distribution of arid and semiarid homoclimates, Reviews of Research on Arid
Zone Hydrology. UNESCO, Paris, pp. 203-209.
Meir, M.F., Post, A., 1969. What are glacier surges? Can. J. Earth Sci. 6, 807-817.
Mellet, J.S., 1990. Ground-penetrating radar enhances knowledge of Earth's surface layer. Geotimes 35,
12-14.
Mensching, H., 1958. Glacis-Fussfl~iche-Pediment. Z. Geomorphol. N.F. 2, 165-186.
Menzies, J., 1979. A review of the literature on the formation and location of drumlins. Earth Sci. Rev. 14,
315-359.
Menzies, J., 1995a. Glaciers and ice sheets. In: Menzies, J. (Ed.), Modem Glacial Environments.
Processes, Dynamics and Sediments. Butterworth-Heinemann, Oxford, pp. 101-138.
Menzies, J., (Ed.), 1995b. Modern Glacial Environments: Processes, Dynamics and Sediments.
Butterwoth-Heinemann, Oxford, p. 621.
Menzies, J., 1995c. Hydrology of glaciers. In: Menzies, J. (Ed.), Modem Glacial Environments. Processes,
Dynamics and Sediments. Butterworth-Heinemann, Oxford, pp. 197-239.
Menzies, J., (Ed.), 1996. Past GlacialEnvironments: Sediments, Forms and Techniques. Butterwoth-
Heinemann, Oxford, p. 598.
Menzies, J., Shilts, W.W., 1996. Subglacial environments. In: Menzies, J. (Ed.), Past Glacial
Environments. Sediments, Forms and Techniques. Butterworth-Heinemann, Oxford, pp. 15-136.
Meybeck, M., 1979. Concentrations des eaux fluviales en 616ments majeurs et apports en solution aux
ocrans. Rev. Grol. Dyn. Grogr. Phys. 21,215-246.
Meyer, L.D., 1986. Erosion processes and sediment properties for agricultural cropland. In: Abrahams,
A.D. (Ed.), Hillslope Processes. Allen and Unwin, Boston, pp. 55-76.
Miall, A.D., 1978. Lithofacies types and vertical profile models in braided river deposits: a summary.
In: Miall, A.D. (Ed.), Fluvial Sedimentology, Canadian Society Petroleum Geologists Memoir, 5,
pp. 597-604, Calgary.
Michel, P., 1973. Les basins des fleuves Srnrgal et Gambie. Etude gromorphologique. Memoirs
ORSTROM 63, (3 Volumes).
Middleton, N., 1997. Desert dust. In: Thomas, D.S.G. (Ed.), Arid Zone Geomorphology. Process, Form in
Change in Drylands. Wiley, Chichester, pp. 413-436.
Middleton, N., Thomas, D., (Eds), 1997. World Atlas of Desertification. Arnold, London, p. 182.
Middleton, N.J., 1990. Wind erosion and dust-storm control. In: Goudie, A. (Ed.), Techniques for Desert
Reclamation. Wiley, Chichester, pp. 87-108.
Middleton, N.J., Goudie, A.S., Wells, G.L., 1986. The frequency and source areas of dust storms. In:
Nickling, W.G. (Ed.), Aeolian Geomorphology. Allen and Unwin, Boston, pp. 237-268.
Milnes, A.R., Thiry, M., Wright, V.P., 1991. Silica accumulations in saprolite and soils in South Australia.
In: Nettleton, W.D. (Ed.), Occurrence, characteristics, and genesis of carbonate, gypsum and silica
accumulation in soils, Soil Science Society of America, Special Publication, 26, pp. 121-149.
Milliman, J.D., Meade, R.H., 1983. World-wide delivery of river sediment to the oceans. J. Geol. 91,
1-21.
Millington, A.C., Drake, N.A., White, K., Bryant, R.G., 1995. Salt ramps: wind-induced depositional
features on Tunisian playas. Earth Surf. Processes Landforms 20, 105-113.
Millot, G., 1964. G6ologie des Argiles. Masson, Paris, p. 499.
Mills, H.H., 1990. Thickness and character of regolith on mountain slopes in the vicinity of Mountain
Lake, Virginia, as indicated by seismic refraction, and implications for hillslope evolution. Geomor-
phology 3, 143-157.
Moeyersons, J., 1989. The concentration of stones into a stone-line, as a result from subsurface movements
in fine and loose soils in the tropics. In: Alexandre, J., Symoens, J.J. (Eds), Stone Lines, Acad6mie
Royale des Sciences D'Outre-Mer, Geo-Eco-Trop, Vol. 11, pp. 11-22.
Mol, J., Vandenberghe, J., Kasse, C., 2000. River response of variations of periglacial climate. Geomor-
phology 33, 131 - 148.
Molina, E., Blanco, J.A., 1980. Quelques pr6cisions sur l'alt6ration du Massif Hercynien espagnol.
C. R. Acad. Sci. Paris 280, 1293-1296.
Molina, E., Pellitero, E., 1982. Formas periglaciares actuales en la Paramera de Avila. Hip6tesis sobre
su origen. Botelfn de la Real Sociedad Espafiola de Historia Natural 80, 49-56.
Monroe, W.H., 1964. Lithological control in the formation of the towers of Puerto Rico, 20th International
Geographical Conference Karst Symposium.
Monroe, W.H., 1966. Formation of tropical karst topography by limestone solution and reprecipitation.
Caribb. J. Sci. 6, 1-7.
Mook, W., Woillard, G., 1982. Carbon-14 dates at Gran Pile. Correlation of land and sea chronologies.
Science 215, 159-161.
Ministerio de Obras Pfblicas y Urbanismo, 1987. Medio Ambiente en Espafia, 1986, Direcci6n General de
Medio Ambiente. Ministerio de Obras P6blicas y Urbanismo, Madrid, p. 368.
Moran, S., 1971. Glaciotectonic structures in drift. In: Goldthwait, R.P. (Ed.), Till: A Symposium. Ohio
State University Press, Columbus, pp. 127-148.
Morgan, R.P.C., 1986. Soil Erosion and Conservation. Longman, Harlow, p. 298.
Morgan, R.P.C., Finney, H.J., Lavee, H., Merrit, E., Noble, C.A., 1986. Plant cover effects on hillslope
runoff and erosion: evidence from two laboratory experiments. In: Abrahams, A.D. (Ed.), Hillslope
Processes, Binghampton Symposia in Geomorphology. Unwin, London, pp. 77-90.
Morin, J., Benyami, Y., 1977. Rainfall infiltration into bare soils. Water Resour. Res. 13, 813-817.
M6rner, N.A., 1980. The Fennoscandian uplift: geological data and their geodynamical implication.
In: M6rner, N.A. (Ed.), Earth Rheology, Isostasy and Eustasy. Wiley, Chichester, pp. 251-284.
Morrison, R.B., 1965. Quaternary geology of the Great Basin. In: Wright, H.E., Frey, D.G. (Eds), The
Quaternary of the United States. Princeton University Press, Princeton, NJ, pp. 265-285.
Moss, R.P., 1965. Slope development and soil morphology in a part of south west Nigeria. J. Soil Sci. 16,
192-209.
Motts, W.S., 1970 (Ed.). Geology and Hydrology of Selected Playas in Western United States. Geol. Dept.
Unit of Massachussets, Amherst.
Muller, S.W., 1945. Permafrost or permanently frozen ground and related engineering problems. US Geol.
Surv. Spec. Paper 62, p. 231.
Mtiller, F., 1959. Beobachtung uber pingos. Meddeleser om Gronland 153, p. 127.
Mfiller, F., 1962. Zonation in the accumulation area of the glaciers of Axel Heiberg Island, N.W.T.,
Canada. J. Glaciol. 4, 302-318.
Murchinson, S.B., 1989. Fluctuation history of Great Salt Lake, Utah, during the last 13,000 years.
Ph. Dissertation, University of Utah, Salt Lake City.
Murty, T.S., E1-Sabh, M.I., 1988. Edge waves and ice ride-up on shorelines. In: E1-Sabh, M.I., Murty, T.S.
(Eds), Natural and Man-Made Hazards. D. Reidel Publication, The Netherlands, pp. 429-434.
References 733
Mustoe, G.E., 1982. Origin of honeycomb weathering. Bull. Geol. Soc. Am. 93, 108-115.
Mustoe, G.E., 1983. Cavernous weathering in the Capitol Reef Desert, Utah. Earth Surf. Processes
Landforms 8, 517- 526.
Muxart, T., Birot, P., 1977. L'Alteration Meteorique des Roches, Publications du D~partement de
G6ographie de l'Universit~ de Paris-Sorbonne, p. 279.
Nadler, C.T., Schumm, S.A., 1981. Metamorphosis of South Platte and Arkansas rivers, eastern Colorado.
Phys. Geogr. 2, 95-115.
Nagel, R.H., 1962. Geology of the Serra do Navio manganese district. Econ. Geol. 57, 481-498.
Nahon, D., 1986. Evolution of iron crusts in tropical landscapes. In: Colman, S.M., Dethier, D.P.
(Eds), Rates of Chemical Weathering of Rocks and Minerals. Academic Press, New York, pp.
169-191.
Nahon, D., Beauvales, A., Trescases, J.J., 1985. Manganese concentration through chemical weathering
of metamorphic rocks under lateritic conditions. In: Drever, J.I. (Ed.), The Chemistry of Weathering.
Reidel, Boston, pp. 277-291.
Nahon, D.B., 1991. Introduction to the Petrology of Soils and Chemical Weathering. Wiley, New York,
p. 313.
Nalpanis, P., 1985. Saltating and suspended particles over flat and sloping surfaces. II. Experiments
and numerical simulations. In: Barndorff-Nielsen, O.E., Moller, J.T., Romer-Rasmussen, K., Willets,
B.B. (Eds), Proceedings of the International Workshop on the Physics of Blown Sand, 8. Institute of
Mathematics, University of Aarhus, Denmark, pp. 37-66.
Nfivar, J., Bryan, R., 1990. Interception loss and rainfall redestribution by three semiarid growing shrubs
in northeastern Mexico. J. Hydrol. 115, 51-63.
Neal, J.T., 1965. Environmental setting and general surface characteristics of playas. In: Neal, J.T. (Ed.),
Geology, Mineralogy and Hydrology of United States Playas, Air Force Cambridge Research
Laboratory, Environmental Research Paper, Vol. 96, pp. 1-29.
Neal, J.T., (Ed.), 1975. Playas and Dried Lakes. Dowden, Hutchinson and Ross, Stroudsburg, p. 411.
Netterberg, F., 1969. Ages of calcretes in southern Africa. S. Afr. Archeol. Bull. 24, 117-122.
Newman, J.C., Philips, J.R.H., 1957. Tunnel erosion in the Riverina. J. Soil Conserv. Serv. NSW 13,
159-169.
Nickling, W.G., 1983. Grain size characteristics of sediment transported during dust storms. J. Sediment.
Petrol. 53, 1011 - 1024.
Nickling, W.G., 1988. The initiation of particle movement by wind. Sedimentology 35, 499-511.
Nickling, W.G., Gillies, J.A., 1993. Dust emission and transport in Mali, West Africa. Sedimentology 40,
859-868.
Nicholas, R.M., Dixon, J.C., 1986. Sandstone scarp form and retreat in the Land of Standing Rocks,
Canyonlands National Park, Utah. Z. Geomorphol. 30, 167-187.
Nicholson, S.E., 1981. Saharan climates in historical times. In: Allan, J.A. (Ed.), The Sahara: Ecological
Change and Early Economic History. Menas Press, London, pp. 35-59.
Nicholson, S.E., 1993. What is a desert? In: Seely, M. (Ed.), Deserts. Weldon Ower Limited, Australia,
pp. 14-25.
Nicholson, S.E., Flohn, H., 1980. African environmental and climatic changes and the general atmospheric
circulation in Late Pleistocene and Holocene. Climatic Change 2, 313-348.
Nicod, J., 1992. Formes d'aplanissement et de regularisation des versants dans les roches carbonat~es:
position des probl~mes et elements de discussion. Tfibinger Geogr. Stud. 109, 1-22, Ttibingen.
Nicoletti, P.G., Sorriso-Valvo, M., 1991. Geomorphic controls of the shape and mobility of rock
avalanches. Bull. Geol. Soc. Am. 103, 1365-1373.
Nielson, J., Kocurek, G., 1987. Surface processes, deposits, and development of star dunes, Dumont dune
field, California. Bull. Geol. Soc. Am. 99, 177-186.
Nilsen, T.H., 1993. What is an alluvial fan? Int. Work. Group Fan Deltas Newslett. 1, 3-4.
Nilsen, T.H., Moore, T.E., 1984. Bibliography of Alluvial Fan Deposits. Geo Books, Norwich,
p. 96.
Nortcliff, S., Ross, S.M., Thornes, J.B., 1990. Soil moisture, runoff and sediment yield from differentially
cleared tropical rainforest plots. In: Thornes, J.B. (Ed.), Vegetation and Erosion. Wiley, Chichester,
pp. 420-436.
Norton, S.A., 1973. Laterite and bauxite formation. Econ. Geol. 68, 353-361.
Nyamweru, C., 1980. Rifts and Volcanoes, A Study of the East African Rift System. Nelson Africa,
Nairobi, Kenya, p. 128.
Nye, J.F., 1952. The mechanics of glacier flow. J. Glaciol. 2, 82-93.

Nye, J.F., 1958. Surges in glaciers. Nature 181, 1450-1451.
O'Hara, S.L., 1997. Human impacts on drylands geomorphic processes. In: Thomas, D.S.G. (Ed.), Arid
Zone Geomorphology. Process, Form and Change in Drylands. Wiley, Chichester, pp. 639-658.
Oberlander, T.M., 1972. Morphogenesis of granitic boulder slopes in the Mojave Desert California. J. Geol.
80, 1-20.
Oberlander, T.M., 1977a. Origin of segmented cliffs in massive sandstones of southeastern Utah. In: Doehring,
D.O. (Ed.), Geomorphology in Arid Regions. George Allen and Unwin, London, pp. 79-114.
Oberlander, T.M., 1997b. Slope and pediment systems. In: Thomas, D.S.G. (Ed.), Arid Zone
Oberlander, T.M., 1994. Rock varnish in deserts. In: Abrahams, A.D., Parsons, A.J. (Eds), Geomorphology
of Desert Environments. Chapman & Hall, London, pp. 106-119.
Oerlemans, J., 1989a. On the response of valley glaciers to climatic change. In: Oerlemans, J. (Ed.),
Glacier Fluctuations and Climatic Change. Kluwer, Dordrecht, pp. 353-371.
Oerlemans, J., (Ed.), 1989b. Glacier Fluctuations and Climatic Change. Kluwer, Dordrecht, p. 417.
Ogunkoya, O.O., Jeje, L.K., 1987. Sediment yield from some third order basins of the basement complex
rocks in southwestern Nigeria. Catena 14, 383-397.
Ollier, C., 1977. Applications of weathering studies. In: Hails, J.R. (Ed.), Applied Geomorphology.
Elsevier, Amsterdam, pp. 9-50.
Ollier, C., 1984. Weathering. Longman, London, p. 304.
Ollier, C.D., 1960. The inselbergs of Uganda. Z. Geomorphol. 4, 43-52.
Ollier, C.D., 1965. Some features of granite weathering in Australia. Z. Geomorphol. 9, 285-304.
Ollier, C.D., 1966. Desert gilgai. Nature 212, 581-583.
Ollier, C.D., 1978. Inselbergs of the Namib Desert, processes and history. Z. Geomorphol. Suppl.bd. 31,
161-176.
Ollier, C.D., 1981. Tectonics and Landforms. Longman, London, p. 324.
Ollier, C.D., 1984. Weathering. Longman, London, p. 270.
Ollier, C.D., 1988a. Deep weathering, groundwater and climate. Geogr. Ann. 70A, 285-290.
Ollier, C.D., 1988b. The regolith in Australia. Earth Sci. Rev. 25, 355-362.
Ollier, C.D., 1991. Ancient Landforms. Belhaven Press, London, p. 233.
Ollier, C.D., Ash, J.E., 1983. Fire and rock breakdown. Z. Geomorphol. 27, 363-374.
Ollier, C.D., Pain, C., 1996. Regolith, Soils and Landforms. Wiley, Chichester, p. 316.
Oostwoud, D.J.W., Bryan, R.R., 1994. Gully head cuts as sediment sources on the Njemps Flats and initial
low-cost gully control measures. In: Bryan, R.R. (Ed.), Soil erosion, land degradation and social
transition, Advances in GeoEcology, Vol. 27, pp. 205-229.
Ortf, F., 1989. Evaporitas marinas. In: Arche, A. (Coor.), Sedimentologfa, Vol. II. Consejo Superior de
Investigaciones Cientfficas, Madrid, pp. 89-177.
Ortiz, A., 1993. Hydrology information and voice communication automatic system in the Ebro basin.
In: Guti&rez, M., Sancho, C., Benito, G. (Eds), Second Intensive Course on Applied Geomorphology:
Arid Regions, Zaragoza, pp. 89-105.
Osterkamp, W.R., Wood, W.W., 1987. Playa-lake basins on the Southern High Plains and New Mexico:
Part 1 - hydrologic, geomorphic and geologic evidence for their development. Bull. Geol. Soc. Am.
99, 215-223.
Ostrem, G., Haakensen, N., Melander, O., 1973. Atlas over Breer i Nord-Skandinavia. Norges Veerdrags
of Elektrisitetsvesen on Stockholm Univ., 315 p.
Oviatt, C.G., 1997. Lake Bonneville fluctuations and global climate change. Geology 25, 155-158.
Owens, I., 2004. Avalanche, snow. In: Goudie, A.S. (Ed.), Encyclopedia of Geomorphology, Routledge,
pp. 41-44.
Owens, L.B., Watson, J.P., 1979. Landscape reduction by weathering in small Rhodesian watersheds.
Geology 7, 281 - 284.
Pachur, H.J., R6per, H.P., Kroplin, S., Goschin, M., 1987. Late Quarternary hydrography of the eastern
Sahara. Berliner Geowissenschaften Abhandlungen, 75, 331-384.
Pain, C.F., Ollier, C.D., 1981. Geomorphology of a Pliocene granite in Papua New Guinea. Z. Geomorphol.
25, 249-258.
Palmer, R.S., 1963. The influence of a thin water layer on water drop impact forces. Int. Assoc. Sci.
Hydrol. 65, 141 - 148.
Pallister, J.W., 1956. Slope development in Buganda. Geogr. J. 122, 80-87.
References 735
Parker, G.G., 1963. Piping, a geomorphic agent in landform development of the drylands. Int. Assoc. Sci.
Hydrol. Publ. 65, 103-113.
Parker, G.G., Jenne, E.A., 1967. Structural failure of Western US highways caused by piping, United
States Geological Survey. Water Resources Division, p. 27.
Parker, G.G., Higgins, G.C., Parker, G.G., Wood, W.W., 1990. Piping and pseudokarst in drylands.
In: Higgins, G.C., Coates, D.R. (Eds), Groundwater Geomorphology: The Role of Subsurface Water
in Earth-surface Processes and Landforms. Geological Society of America, Boulder, pp. 77-110,
Special Paper 252.
Parmenter, C., Folger, D.W., 1974. Eolian biogenic detritus in deep sea sediments: a possible index of
equatorial Ice Age aridity. Science 185, 695-698.
Parra, E., Capeda, H., 1990. Volcanic hazards maps of the Nevado del Ruiz volcano, Columbia.
J. Volcanol. Geotherm. Res. 42, 117-127.
Passarge, S., 1904. Die Kalahari: Versuch einer physisch-geographischen darstellung der sandfelden des
Siidafrikanischen beckens. Reiner, Berlin.
Passarge, S., 1931. Geomorfologfa. Ed. Labor, Barcelona, p. 189.
Patr-Cornell, E., 1996. Uncertainties in global climate change estimates. An editorial essay. Climatic
Change 33, 145-149.
Paterson, W.S.B., 1994. The Physics of Glaciers. Pergamon, New York, p. 480.
Pavich, M.J., 1985. Appalachian piedmont morphogenesis: weathering, erosion, and Cenozoic uplift. In:
Morisawa, M., Hack, J.T. (Eds), Tectonic Geomorphology. George Allen and Unwin, London, pp. 27- 51.
Pavlov, A.V., 1994. Current changes of climate and permafrost in the Arctic and Sub-Arctic of Russia.
Permafrost Periglac. Processes 5, 101 - 110.
Pearce, F., 1994. Not warming, but cooling. New Sci. 9, 37-41, July.
Prcsi, M., 1968. Loess. In: Fairbridge, R.W. (Ed.), The Encyclopedia of Geomorphology. Reinold,
Prcsi, M., Sailfird, J., 1970. Planated surfaces: principal problems of research and terminology. In:
Prcsi, M. (Ed.), Problems of Relief Planation. Akadrmiai Kiad6, Budapest, pp. 13-27.
Pedraza, J., Sanz, M.A., Martin, A., 1989. Formas Granfticas de La Pedriza. Agencia de Medio Ambiente,
Madrid, p. 205.
Pedro, G., 1966. Essai sur le caractrrisation grochimique des diffrrents processus zonaux rrsultant de
l'altrration superficielle. C. R. Acad. Sci. Paris 262, 1828-1831.
Pedro, G., 1968. Distributions des principaux types d'altrration chimique h la surface du globe. Rev.
Grogr. Phys. Grol. Dyn. 10, 457-470.
Pedro, G., Jamague, M., Bejon, J.C., 1969. Mineral interactions and transformations in relation to
pedogenesis during the Quaternary. Soil Sci. 107, 462-469.
Peel, R.F., 1974. Insolation weathering: some measurements of diurnal temperature changes in exposed
rocks in the Tibesti region, central Sahara. Z. Geomorph. - Suppl., 21, 19-28.
Peltier, L.C., 1950. The geographic cycle in periglacial regions as it is related to climatic geomorphology.
Ann. Assoc. Am. Geogr. 40, 214-236.
Peltier, W.R., Andrews, J.T., 1983. Glacial geology and glacial isostasy of the Hudson Bay region. In:
Smith, D.E., Dawson, A.G. (Eds), Shorelines and Isostasy. Academic Press, London, pp. 285-320.
Pefia, J.L., Chueca, J., Julifin, A., 1998. Los derrubios estratificados del sector central pirenaico: cronologfa
y lfmites altitudinales. In: G6mez Ortiz, A., Salvador Franch, F., Schulte, L., Garcfa Navarro, A. (Eds),
Procesos biof/sicos actuales en medios frfos. Publicaciones de la Universitat de Barcelona, Barcelona,
pp. 205- 216.
Pefia, J.L., Echeverrfa, M.T., Petit-Maire, N., Lafont, R., 1993. Cronolog/a e interpretacirn de las
acumulaciones holocenas de la Val de las Lenas (Depresirn del Ebro, Zaragoza). Geographicalia,
Zaragoza, 30, pp. 321-332.
Penck, A., 1905. Glacial features in the surface of the Alps. J. Geol. 13, 1-19.
Penck, W., 1924. Die Morphologische Analyse. Geogr. Abh. 2, Stuttgart, 1-283.
Penck, W., 1953. Morphological Analysis of Landforms. Macmillan, London, p. 429.
Prrez del Campo, P., 1989. Nuevos datos sobre colapsos kfirsticos en la Depresi6n del Ebro. Direccirn
e incidencia en la nueva lfnea de alta velocidad Madrid-Barcelona. II Reuni6n del Cuaternario Ibrrico,
Madrid, pp. 905-911.
Persons, B.S., 1970. Laterite: Genesis, Location and Use. Plenum Press, New York, p. 103.
Persons, B.S., 1984. Laterite, Engineering Geology. In: Finkl, C.W. (Ed.), The Encyclopedia of Applied
Geology. Van Nostrand Reinhold, New York, pp. 320-326.
Perthuisot, J.P., Jauzein, A., 1975. Sebkha et dunes d' argile: 1' enclave endoreique de Pont du Fars, Tunisie.
Rev. Ge6gr. Phys. G6ol. Dyn. 17, 295-306.
Petit-Maire, N., 1999. Variabilit6 naturelle des environments terrestres: les deux dernieres extremes
climatiques (18,000 + 1000 ans BP). C. R. Acad. Sci. Paris 328, 273-279.
Petri, S., Ftilgaro, V.F., 1983. Geolog/a do Brasil. Editora da Universidade de Sao Paulo, Sao Paulo, p. 631.
Petts, G., Foster, I., 1985. Rivers and Landscape. Arnold, London, p. 274.
P6w6, T.L., 1955. Origin of the upland silt near Fairbanks, Alaska. Bull. Geol. Soc. Am. 66, 699-724.
P6w6, T.L., 1959. Sand wedge polygons (tesselations) in the McMurdo Sound region, Antarctica.
Am. J. Sci. 257, 545-552.
P6w6, T.L., 1966. Paleoclimatic significance of fossil wedges. Biuletyn Peryglacjalny 15, 65-73.
P6w6, T.L., 1981. Desert dust: an overview. In: Pew6, T.L. (Ed.), Desert Dust: Origin, Characteristics
and Effect on Man. Geological Society of America, Boulder, pp. 1-10, Special Paper 186.
P6w6, T.L., 1983. The periglacial environment in North America during Wisconsin time. In: Porter, S.C.
(Ed.), Late Quaternary of the United States, The Late Pleistocene, Vol. I. Longman, London, pp. 157-189.
P6w~, T.L., Tungsheng, L., Slatt, R.M., Bingyuan, L., 1995. Origin and Character of Loesslike Silt in
the Southern Qinghai-Xizang (Tibet) Plateau, China, United States Geological Survey Professional
Paper, Vol. 1549, p. 55.
Pfeffer, W.-T., 1992. Stress-induced foliation in the terminus of variegated glacier, Alaska, USA, formed
during the 1982-1983 surge. J. Glaciol. 38, 213-222.
Philbrick, S.S., 1970. Horizontal configuration and the rate of erosion of Niagara Falls. Bull. Geol. Soc.
Am. 81, 3723-3732.
Pias, J., 1970. Les Formations S6dimentaires Tertiaires et Quaternaires de la Cuvette Tchadienne et les
Sols qui en D6rivent. ORSTOM Mem. 43.
Pichler, E., 1958. Aspectos geol6gicos dos escorregamentos de Santos. Noticia Geomorfol6gica 2,
40 -44.
Pierson, T.C., 1983. Soil pipes and slope stability. Q. J. Eng. Geol. 16, 1-11, London.
Pilgrim, D., Cordery, I., Baron, B., 1982. Effects of catchment size on runoff relationships. J. Hydrol. 58,
205-221.
Pissart, A., 1965. Les pingos des Hautes Fagnes: le probl~me de leur gen~se. Ann. Soc. G6ol. Bel. 89,
377-402.
Pissart, A., 1970. Les ph6nom6nes physiques essentiels li6s au gel; les structures p6riglaciaires qui en
r~sultent et leur signification climatique. Ann. Soc. G6ol. Belg. 93, 7-49.
Pissart, A., 1988. Pingos: an overview of the present state of knowledge. In: Clark, M.J. (Ed.), Advances
in Periglacial Geomorphology. Wiley, Chichester, pp. 279-297.
Pissart, A., 1990. Advances in periglacial geomorphology. Z. Geomorphol. Suppl.bd. 79, 119-131.
Pissart, A., 2000. Remnants of lithalsas of the Hautes Fagnes, Belgium: a summary of present-day
knowledge. Permafrost Periglac. Processes 11,327-355.
Pissart, A., 2004. Thermokarst. In: Goudie, A.S. (Ed.), Encyclopedia of Geomorphology. Routledge,
London, pp. 1049-1051.
Plafker, G., Ericksen, G.E., 1978. Nevados Huascar~in avalanches, Peril. In: Voight, B. (Ed.), Rockslides
and Avalanches, Vol. 1. Elsevier, Amsterdam, pp. 277-314.
Ploey, J. de., 1980. Some field measurements and experimental data on wind-blown sands. In: Boodt, M. de
y Gabriels, D., (Eds), Assessment of erosion. John Wiley, New York, pp. 541-552.
Poessen, J., 1986. Surface sealing as influenced by slope angle and position of simulated stones in the top
layer of loose sediments. Earth Surf. Processes Landforms 11, 1-10.
Poessen, J., Lavee, H., 1991. Effects of size and incorporation of synthetic mulch on runoff and sediment
yield from interrills in a laboratory study with simulated rainfall. Soil Till. Res. 21,209-223.
Poessen, J., Ingelmo-S~inchez, F., Mucher, H., 1990. The hydrological response of soil surfaces to rainfall
as affected by cover and position of rock fragments in the top layer. Earth Surf. Processes Landforms
15, 653-672.
Pohjola, V.A., 1994. TV-video observations of bed and basal sliding on Storglaci~iren, Sweden. J. Glaciol.
39, 111-118.
Pokras, E.M., Mix, A.C., 1985. Eolian evidence for spatial variability of Late Quaternary climates in
tropical Africa. Quaternary Res. 24, 137-149.
Poland, J.F., Davis, G.H., 1969. Land subsidence due to withdrawal of fluids. In: Varnes, D.J., Kiersch, G.
(Eds), Reviews in Engineering Geology, Geological Society of America, Vol. 2, pp. 187-269.
Polynov, B., 1937. The Cycle of Weathering. Murby, London, p. 220.
References 737
Pollard, W.E., 1988. Seasonal frost mounds. In: Clark, M.J. (Ed.), Advances in Periglacial
Ponti, D.J., 1985. The Quaternary alluvial sequence of the Antelope Valley, California. Geol. Soc. Am.
Spec. Paper 203, 79-96.
Porath, A., Shick, A.P., 1974. The use of remote sensing systems in monitoring desert floods. Int. Assoc.
Hydrol. Sci. Publ. 112, 133-138.
Porter, S.C., Zhisheng, An., 1995. Correlation between climate events in the North Atlantic and China
during the last glaciation. Nature 375, 305-308.
Posamentier, H.W., 1978. Thoughts on ogive formation. J. Glaciol. 20, 218-220.
Poser, H., 1948. Boden-und klimaverh~iltnisse in Mittel und Westeuropa w~ihrend der Wtirmmeiszeit.
Erdkunde 2, 53-68.
Poser, H., 1950. Zur rekonstruktion der sp~itglazialen luftdruckverh~iltnisse in Mittel-und Westeuropa auf
grund der vorzeitlichen binnendfinen. Erdkunde 4, 81-88.
Potter, N., 1972. Ice-cored rock glacier, Galena Creek, Northern Absaroka Mountains, Wyoming. Bull.
Geol. Soc. Am. 83, 3025-3057.
Potter, N., Moss, J.H., 1968. Origin of the Blue Rocks block field and adjacent deposits, Berks County,
Pennsylvania. Bull. Geol. Soc. Am. 79, 255-262.
Powell, R.D., 1981. A model for sedimentation by tidewater glaciers. Ann. Glaciol. 2, 129-134.
Powell, R.D., 2003. Subaquatic landsystems: fjords. In: Evans, D.J.A. (Ed.), Glacial Landsystems. Arnold,
London, pp. 313 - 347.
Prest, N.K., 1968. Nomenclature of moraines and ice-flow features as applied to the glacial map of Canada.
Geol. Surv. Can., Paper 66-57.
Price, R.J., 1966. Eskers near the Casement, Alaska glacier. Geogr. Ann. 48, 111-125.
Price, R.J., 1969. Moraines, sandar, kames and eskers near Breidamerkurj6kull, Iceland. Trans. Inst. Br.
Geogr. 46, 17-43.
Price, R.J., 1970. Moraines at Fjallsj6kull, Iceland. Arctic Alpine Res. 2, 27-42.
Price, R.J., 1973. Glacial and Fluvioglacial Landforms. Longman, Edinburgh, p. 242.
Price, R.J., 1980. Rates of geomorphological changes in proglacial areas. In: Cullinford, R.A., Davidson,
D.A., Lewin, J. (Eds), Timescales in Geomorphology. Wiley, Chichester, pp. 79-93.
Price, W.A., 1963. Physico-chemical and environmental factors in clay-dune genesis. J. Sediment. Petrol.
33, 766-778.
Price, W.A., 1968. Carolina bays. In: Fairbridge, R.W. (Ed.), The Encyclopaedia of Geomorphology.
Reinhold, New York, pp. 102-109.
Priesnitz, K., 1988. Cryoplanation. In: Clark, M.J. (Ed.), Advances in Periglacial Geomorphology. Wiley,
Prospero, J.M., Bonatti, E., Schubert, C., Carlson, T.N., 1970. Dust in the Caribbean atmosphere traced to
an African dust storm. Earth Planet. Sci. Lett. 9, 287-293.
Prouty, W.F., 1952. Carolina bays and their origin. Bull. Geol. Soc. Am. 63, 167-224.
Pueyo, J.J., 1978-1979. La precipitaci6n evaporftica actual en las lagunas del firea: Bujaraloz, Sfistago,
Caspe, Alcafiiz y Calanda (Provs. de Zaragoza y Teruel). Revista del Instituto de Investigaciones
Geol6gicas 33, 5-56, Barcelona.
Pugh, J.C., 1956. Fringing pediments and marginal depressions in the inselberg landscape of Nigeria.
Trans. Inst. Br. Geogr. 22, 15-31.
Pugh, J.C., 1966. Landforms in low latitudes. In: Dury, G.H. (Ed.), Essays in Geomorphology. American
Elsevier, New York, pp. 121-138.
Pulido-Bosch, A., 1986. Le karst dans les gypses de Sorbas (Almerfa). Aspects morphologiques et
hydrog6ologiques. Karstologie, 27-35, Memoire 1.
Punkari, M., 1997. Subglacial processes of the Scandinavian Ice Sheet in Fennoscandia inferred from flow-
parallel features and lithostratigraphy. Sediment. Geol. 111,263-283.
Pye, K., 1982. Morphological development of coastal dunes in a humid tropical environment, Cape
Bedford and Cape Flattery, North Queensland. Geogr. Ann. A64, 212-227.
Pye, K., 1987. Aeolian Dust and Dust Deposits. Academic Press, London, p. 334.
Pye, K., 1995. The nature, origin and accumulation of loess. Quaternary Sci. Rev. 14, 653-667.
Pye, K., Tsoar, H., 1987. The mechanics and geological implications of dust transport and deposition in
deserts, with particular reference to loess formation and dune sand diagenesis in the northern Negev,
Israel. In: Frostick, L.E., Reid, I. (Eds), Desert Sediments, Ancient and Modem, Special Publication 35,
Geological Society of London. Blackwell, Oxford, pp. 139-156.
Pye, K., Tsoar, H., 1990. Aeolian Sand and Sand Deposits. Unwin Hyman, London, p. 396.
Quirantes, J., 1965. Notas sobre las lagunas de Bujalaroz-S~istago. Geographica 12, 30-34.
Rachocki, A., 1981. Alluvial Fans. An Attempt at an Empirical Approach. Wiley, Chichester, p. 161.
Rachocki, A.H., Church, M., (Eds), 1990. Alluvial Fans. A Field Approach. Wiley, Chichester,
p. 391.
Ramfrez, E., Francou, B., Ribstein, P., Descloitres, M., Gu4rin, R., Mendoza, J., Gallaire, R., Poyaud, B.,
Jordan, E., 2001. Small glacier disappearing in the tropical Andes: a case-study in Bolivia: Glaciar
Chacaltaya (16~ J. Glaciol. 47 (157), 187-194.
Rantz, S.E., 1970. Urban sprawl and flooding in Southern California. US Geol. Surv. Circ. 60lB.
Rapp, A., 1957. Studien fiber Schutthalden in Lappland und auf Spitzbergen. Z. Geomorphol. 1,
179-200.
Rapp, A., 1984. Are terra rossa soils in Europe eolian deposits from Africa? Geologiska F6reningens I
Stocklom F6rhendlingar 105, 161 - 168.
Rapp, A., 1986. Slope processes in high latitude mountains. Prog. Phys. Geogr. 10, 53-68.
Rasid, H., Pramanik, M.A.H., 1993. Areal extent of the 1988 flood in Bangladesh: how much did the
satellite imagery show? Nat. Hazards 8, 189-200.
Raso, J.M., Garcfa Loureiro, M.C., 1998. Oscillacions termombtiques al'entorn de 08 C a Andorra. In:
G6mez Ortiz, A., Salvador Franch, F., Schulte, L., Garcfa Navarro, A. (Eds), Procesos bioffsicos
actuales en medios frios. Publicacions de la Universitat de Barcelona, Barcelona, pp. 217-234.
Raunet, M., 1985. Les bas-fonds en Afrique et fi Madagascar. Z. Geomorphol. Suppl.bd. 52, 25-62.
Raymo, M.E., Ruddiman, W.F., Shackleton, N.J., Oppo, D.W., 1990. Evolution of Atlantic-Pacific dl3C
gradients over the last 2.5 m.y. Earth Planet. Sci. Lett. 97, 353-368.
Rea, B.R., Evans, D.J.A., 1996. Landscapes of areal scouring in NW Scotland. Scot. Geogr. Mag. 112,
47-50.
Rea, B.R., Evans, D.J.A., Dixon, T.S., Whalley, W.B., 2000. Contemporaneous, localized, basal ice-flow
variations: implications for bedrock erosion and the origin of p-forms. J. Glaciol. 46, 470-476.
Reading, A.J., Thompson, R.D., Millington, A.C., 1995. Humid Tropical Environments. Blackwell,
Oxford, p. 429.
Reeves, C.C., 1964. Gas rings from Terry County, Texas. J. Sediment. Petrol. 34, 190-193.
Reeves, C.C., 1968. Introduction to Paleolimnology. Elsevier, Amsterdam, p. 228.
Reeves, C.C., 1976. Caliche: Origin, Classification, Morphology and Uses. Estacado Books, Lubbock, TX,
p. 233.
Regiies, D., Pardini, G., Pini, R., 1992. Estudio del comportamiento de regolitos arcillosos en zonas
abarrancadas frente a variaciones de la temperatura y la humedad. In: L6pez, F., Conesa, C., Romero,
M.A. (Eds), Estudios de Geomorfologfa en Espafia, pp. 171-182, Murcia.
Reiche, P., 1950. A Survey of Weathering Processes and Products, Publications in Geology n ~ 3. New
Mexico University, p. 95.
Reid, I., 1994. River landforms and sediments: evidence of climatic change. In: Abrahams, A.D., Parsons,
A.J. (Eds), Geomorphology of Desert Environments. Chapman & Hall, London, pp. 571-592.
Renard, K.G., 1969. Evaporation from an ephemeral stream bed: discussion. J. Hydr. Div. Proc. Am. Soc.
Civ. Eng. 95, 2200-2204.
Renard, K.G., 1970. The hydrology of semiarid rangeland watersheds. US Dept. Agr. Res. Ser. Pub.,
41-162.
Renard, K.G., Keppel, R.V., 1966. Hydrographs of ephemeral streams in the Southwest. Proc. Am. Soc.
Civ. Eng. J. Hydrol. Div. 92, 33-52.
Rettig, S.L., Jones, B.F., Risacher, F., 1980. Geochemical evolution of brines in the Salar of Uyuni,
Bolivia. Chem. Geol. 30, 57-79.
Rhoades, J.D., 1990. Soil salinity-causes and controls. In: Goudie, A.S. (Ed.), Techniques for Desert
Reclamation. Wiley, Chichester, pp. 109-134.
Rhoads, B.L., 1986. Flood hazard assessment for land-use planning near desert mountains. Environ.
Manage. 10, 97-106.
Riba, O., Llamas, M.R., 1962. Canales de La Violada, Monegros y Flumen, Libro-Gufa del Viaje de
Estudio n~ 5. I Coloquio Internacional sobre las Obras P6blicas en Terrenos Yesfferos. Servicio
Geol6gico de Obras Pfblicas, Madrid, p. 28.
Richardson, C., Holmund, P., 1996. Glacial cirque formation in northern Scandinavia. Ann. Glaciol. 22,
102-106.
Ritter, D.F., 1978. Process Geomorphology. WM. C. Brown, Dubuque, Iowa, p. 603.
References 739
Roberts, C.R., Mitchell, C.W., 1987. Spring mounds in southern Tunisia. In: Frostick, L.E., Reid, I. (Eds),
Desert Sediments: Ancient and Modern, Geological Society of London, Special Publication 35,
pp. 321-336.
Robin, G., De, Q., Drewry, D.J., Meldrum, D.T., 1977. International studies of ice sheet and bedrock.
Philos. Trans. R. Soc. London 279, 185-196.
Robinson, D.A., Williams, R.B.G., 1992. Sandstone weathering in the High Atlas, Morocco.
Rodrfguez, J., Navascu6s, L., 1982. La tafonizaci6n de las areniscas miocenas en los alrededores de
Huesca. Tecniterrae 19, 7-12.
Rogers, R.D., Schumm, S.A., 1991. The effect of sparse vegetative cover on erosion and sediment yield.
J. Hydrol. 123, 19-24.
Rognon, P., 1980. Pluvial and arid phases in the Sahara: the role of nonclimatic factors. In: Sarnthein, M.,
Seibold, E., Rognon, P. (Eds), Palaeoecology of Africa. Sahara and Surrounding Seas. Balkema,
Rotterdam, pp. 45-62.
Rognon, P., 1982. Pluvial and arid phases in the Sahara: the role of nonclimatic factors. Palaeoecology of
Africa, 12, 45-62.
Rognon, P., 1987. Late Quarternary climatic reconstruction for the Maghreb (North Africa).
Palaeogeography, Palaeoclimatology, Palaeoecology, 58, 11-34.
Rogowski, A.S., Khanbilvardi, R.M., Deangelis, R.J., 1985. Estimating erosion on plot, field and
watershed scales. In: E1-Swaifi, S.A., Molldenhaner, W.C., Andrew, L. (Eds), Soil Erosion and
Conservation. Soil Conservation Society of America, pp. 149-166.
Rojo, L., Sfinchez, M.C., 1996. Red de Estaciones Experimentales de Seguimiento y Evaluaci6n de la
Erosi6n y Desertificaci6n (RESEL), Catfilogo de Estaciones, Proyecto Lucdeme. Direcci6n General de
Conservaci6n de la Naturaleza. Ministerio de Medio Ambiente, Madrid, p. 121.
Roose, E.J., Lelong, F., 1976. Les facteurs de l'4rosion hydrique en Afrique tropicale. Etudes sur petites
parcelles exp~rimentales du sol. Rev. G4ogr. Phys. G6ol. Dyn. 18, 365-374.
Roque, C., Pallf, L., 1998. Geologfa de l'Ardenya i Formes Granftiques Associades. Publicacions de
l'Institut d'Estudis del Baix Empordfi 17, 5-42.
Rosen, M.R., 1994. The importance of groundwater in playas: a review of playa classifications and the
sedimentology and hydrology of playas. In: Rosen, M.R. (Ed.), Paleoclimate and basin evolution of
playa systems, Geological Society of America, Special Paper 289, pp. 1-18.
Rust, B.R., 1979. Facies models, coarse alluvial deposits. In: Walker, R.G. (Ed.), Facies models,
Geoscience Canada Reprint Series, Vol. 1, pp. 9-21.
Rutte, E., 1958. Kalkruste in Spanien. Neues Jahrb. Geol. Pal~ieontol. 106, 52-138.
Rutter, N.W., 1965. Foliation pattern of Gulkana glacier, Alaska Range, Alaska. J. Glaciol. 5,
711-718.
Ruxton, B.P., 1958. Weathering and sub-surface erosion in granites at the piedmont angle, Balos, Sudan.
Geol. Mag. 95, 353-377.
Ruxton, B.P., 1967. Slope wash under nature primary rainforest in northern Papua. In: Jennings, J.N.,
Mabbutt, J.A. (Eds), Landform Studies from Australia and New Guinea. Cambridge University Press,
London, pp. 85-94.
Ruxton, B.P., Berry, L., 1957. Weathering of granite and associated erosional features in Hong Kong. Bull.
Geol. Soc. 68, 1263-1292.
Ruxton, B.P., Berry, L., 1961. Weathering profiles and geomorphic position on granite in two tropical
regions. Rev. G4omorphol. Dyn. 12, 16- 31.
Sack, D., 1994. Geomorphic evidence of climate change from desert-basin paleolakes. In: Abrahams,
A.D., Parsons, A.J. (Eds), Geomorphology of Desert Environments. Chapman & Hall, London,
pp. 616-630.
Sakamoto-Arnold, C.M., 1981. Eolian features produced by the December, 1977 windstorm, southern San
Joaquin Valley, California. J. Geol. 89, 129-137.
Salinas, J.L., 1988. Riesgos ligados a arcillas expansivas. In: Ayala, F.J., Durfin, J.J., Peinado, T. (Eds),
Riesgos Geol6gicos. Instituto Geol6gico y Minero de Espafia, Madrid, pp. 295-304.
Sanches Furtado, A.F.A., 1968. Alt6ration des granites dans les r6gions intertropicales sous diff4rents
climats. Ninth International Congress of Soil Science, Adelaide 4, 403-409.
Sfinchez, J.A., Martfnez, F.J., Florfa, E., Schumann, S., 1993. Hydrogeological characterization of the
Monegros Lakes. In: Guti6rrez, M., Sancho, C., Benito, G. (Eds), Second Intensive Course on Applied
Geomorphology: Arid Regions. Universidad de Zaragoza, Zaragoza, pp. 245-252.
740 Climatic G e o m o r p h o l o g y
Sancho, C., Benito, G., 1990. Factors controlling tafoni weathering in the Ebro Basin (NE Spain).
Zeitscrift ffir Geomorphol. 34, 165-177.
Sancho, C., Guti6rrez, M., 1990. Anfilisis morfom&rico de la tafonizaci6n de la arenisca de Peraltilla
(Anticlinal de Barbastro. Depresi6n del Ebro)" Influencia de los factores mineral6gicos-texturales.
Cuatemario y Geomorfologfa 4, 131-145.
Sancho, C., Guti6rrez, M., 1993. Geomorphological features of the Bujalaroz salt lakes, Second Intensive
Course of Applied Geomorphology: Arid Regions. Universidad de Zaragoza, Zaragoza, pp. 241-243.
Sancho, C., Mel6ndez, A., 1992. G6nesis y significado ambiental de los caliches pleistocenos de la regi6n
del Cinca (Depresi6n del Ebro). Revista de la Sociedad Geol6gica de Espafia 5, 81-93.
Sancho, C., Guti4rrez, M., Pefia, J.L., Burillo, F., 1988. A quantitative approach to scarp retreat starting
from triangular slope facets (Central Ebro Basin Spain). Catena Suppl. 13, 139-146.
Sancho, C., Mel6ndez, A., Signes, M., Bastida, J., 1992. Chemical and mineralogical characteristics of
Pleistocene caliche deposits from the Central Ebro basin. Clay Miner. 27, 293-308.
Sanjaume, E., 1985. Las Costas Valencianas. Sedimentologfa y Geomorfologfa. Secci6n de Geograffa.
Universidad de Valencia, Valencia, p. 505.
Santer, B.D., Wigley, T.M.L., Barnett, T.P., Anyamba, B., 1996. Detection of climatic change and
attribution of causes. In: Houghton, J.T et al. (Eds), Climate Change 1995. The Science of Climatic
Change. Cambridge University Press, Cambridge, MA, pp. 406-443.
Satin, M.M., Krishnaswami, S., Dilli, K., Somayajulu, B.L.K., Moore, W.S., 1989. Major ion chemistry
of the Ganga-Brahmaputra river system: weathering processes and fluxes to the Bay of Bengal. Acta
Geochim. Cosmochim. 53,997-1009.
Sarnthein, M., 1978. Sand deserts during the last glacial maximum and climatic optimum. Nature 272,
43-46.
Sarnthein, M., Walger, K., 1974. Der /iolische Sandstrom aus der W-Sahara zur Atlantikktiste.
Geologische Rundschau 63, 1065-1087.
Saunders, I., Young, A., 1983. Rate of surface processes on slopes, slope retreat and denudation. Earth
Surf. Processes Landforms 8, 473-501.
Savat, J., Poesen, J., 1981. Detachment and transportation of loose sediments by raindrop splash.
Part I. The calculation of absolute data on detachability and transportability. Catena 8, 1-17.
Savigear, R.A.G., 1952. Some observations on slope development in South Wales. Trans. Inst. Br. Geogr.
18, 31-51.
Savigear, R.A.G., 1960. Slopes and hills in West Africa. Z. Geomorphol. Suppl.bd. 1,156-171.
Scheidegger, A.E., Schumm, S.A., Fairbridge, R.W., 1968. Badlands. In: Fairbridge, R.W. (Ed.), The
Encyclopedia of Geomorphology. Hutchinson and Ross, Stroudsburg, PA, pp. 43-48.
Schellmann, W., 1977. The formation of lateritic silicate bauxites and criteria for their exploration and
assessment of deposits. Nat. Resour. Dev. 5, 119-134.
Schellmann, W., 1981. Considerations on the definition and classification of laterites. Proceedings
International Seminar on Lateritisation Processes, Trivandrum. Lateritisation Processes. Balkema,
Rotterdam, pp. 1 - 10.
Schick, A.P., Lekach, J., Hassan, M.A., 1987. Vertical exchange of coarse bedload in desert streams.
In: Frostick, L.E., Reid, I. (Eds), Desert Sediments: Ancient and Modem, Geological Society of
London, Special Publication 35. Blackwell, Oxford, pp. 7-16.
Schipull, K., 1980. Die Cedar Mesa - Schichtstufe aud dem Colorado Plateau - ein Beispiel ftir die
Morphodynamik arider Schichtstufen. Z. Geomorphol. 24, 318-331.
Schmidt, K.H., 1980. Eine neue Metode zur Ermittlung von Stufenruckwanderungsrate, dargestellt am
Beispiel der Black Mesa Schichtstufen, Colorado Plateau, USA. Z. Geomorphol. 24, 180-191.
Schmidt, K.H., 1987. Factors influencing structural landform dynamics on the Colorado Plateau about
the necessity of calibrating theoretical models by empirical data. Catena Suppl. 10, 51-66.
Schmidt, K.H., 1988. Rates of scarp retreat: a means of dating neotectonic activity. In: Jacobshagen, V.H.
(Ed.), The Atlas System of Morocco Studies on its Geodynamic Evolution, Lecture Notes in Earth
Science 15, Berlin, pp. 445-462.
Schmidt, K.H., 1989a. Talus and pediment flatirons - erosional and depositional features of dryland cuesta
scarps. Catena Suppl. 14, 107-118.
Schmidt, K.H., 1989b. The significance of scarp retreat for Cenozoic landform evolution on the Colorado
Plateau, USA. Earth Surf. Processes Landforms 14, 93-105.
Schmidt, K.H., 1994. Hillslopes as evidence of climatic change. In: Abrahams, A.D., Parsons, J. (Eds),
Geomorphology of Deserts Environments. Chapman & Hall, London, pp. 553-570.
References 741
Schmidt, K.H., 1996. Talus and pediment flatirons-indicators of climatic change on scarp slope on the
Colorado Plateau, USA. Z. Geom6rphol. NF Suppl.bd. 103, 135-158.
Schumm, S.A., 1956a. Evolution of drainage systems and slopes in badlands at Perth Amboy, New Jersey.
Bull. Geol. Soc. Am. 67, 597-646.
Schumm, S.A., 1956b. The role of creep and rainwash on the retreat of badland slopes. Am. J. Sci. 254,
693-706.
Schumm, S.A., 1962. Erosion on miniature pediments in Badlands National Monument, South Dakota.
Bull. Geol. Soc. Am. 73, 719-724.
Schumm, S.A., 1964. Seasonal variations of erosion rates and processes on hillslopes in western Colorado.
Z. Geomorphol. Suppl.bd. 5, 215- 238.
Schumm, S.A., 1965. Quaternary paleohydrology. In: Wright, H.E., Frey, D.G. (Eds), The Quaternary of
the United States. Princeton University Press, Princeton, NJ, pp. 783-794.
Schumm, S.A., 1968. River adjustment to altered hydrologic regime. Murrumbidgee River and
paleochannels. US Geol. Surv. Prof. Paper 598.
Schumm, S.A., 1969. Geomorphic implications of climatic change. In: Chorley, R.J. (Ed.), Water, Earth
and Man. Methuen, London, pp. 525-534.
Schumm, S.A., 1973. Geomorphic thresholds and complex response of drainage systems. In: Morisawa,
M. (Ed.), Fluvial Geomorphology. George Allen and Unwin, London, pp. 299-310.
Schumm, S.A., 1977. The Fluvial System. Wiley, New York, p. 338.
Schumm, S.A., 1979. Geomorphic thresholds: the concept and its applications. Inst. Br. Geogr. Trans. 4,
485-515.
Schumm, S.A., 1981. Evolution and response of the fluvial system, sedimentological implications. Soc.
Econ. Paleontol. Mineral. 31, 18-29, Special Publication.
Schumm, S.A., Chorley, R.J., 1966. Talus weathering and scarp recession in the Colorado Plateau.
Z. Geomorphol. 10, 11 - 36.
Schumm, S.A., Hadley, R.F., 1957. Arroyos and the semiarid cycle of erosion. Am. J. Sci. 255,
161-174.
Schumm, S.A., Lichty, R.W., 1963. Channel widening and flood plain construction along Cimarron fiver
in south-western Kansas. US Geol. Surv. Prof. Paper 352D, 71-88.
Schumm, S.A., Lichty, R.W., 1965. Time, space and causality in geomorphology. Am. J. Sci. 263,
110-119.
Schumm, S.A., Lusby, G.C., 1963. Seasonal variation of infiltration capacity and runoff on hillslopes in
western Colorado. J. Geophys. Res. 68, 3655-3666.
Schumm, S.A., Harvey, M.D., Watson, C.C., 1984. Incised Channels: Morphology, Dynamics and
Control. Water Resources Publications, Littleton, CO, p. 200.
Schumm, S.A., Mosley, M.P., Weaver, W.E., 1987. Experimental Fluvial Geomorphology. Wiley,
Chichester, p. 413.
Schunke, E., 1974. Formungsvorg~iuge an schneeflecken im isl~indischen Hochland. In: Poser, H. (Ed.),
Geomorphologische Prozesse und Prozesskombinationen in der Gegenwart unter Verschiedenen
Klimabedingungen, Akademie Wissenschaften zu G6ttingen Abh. Math.-Phys., 29, pp. 274-286.
Schunke, E., 1975. Die periglazialerscheinungen Islands in Abh~ingigkeit von Klima und substrat.
Akademie Wissenchaften zu G6ttingen Math.-Phys., K1. Folge 3 (30), 273.
Schunke, E., Zoltai, S.C., 1988. Earth hummocks (thufur). In: Clark, M.J. (Ed.), Advances in Periglacial
Schwarzbach, M., 1967. Islandische wasserfalle und eire genetische systematik des wasserfalles
fiberhaupt. Z. Geomorphol. 11,377-417.
Scoging, H., 1989. Run-off generation and sediment mobilization by water. In: Thomas, D.S.G. (Ed.), Arid
Zone Geomorphology. Belhaven Press, London, pp. 87-116.
Segura, F.S., 1990. Las Ramblas Valencianas. Departamento de Geograffa, Universidad de Valencia, p. 229.
Selby, M.J., 1972. Antarctic tors. Z. Geomorphol. Suppl.bd. 13, 73-86.
Selby, M.J., 1977. Bornhardts of the Namib Desert. Z. Geomorphol. 21, 1-13.
Selby, M.J., 1985. Earth's Changing Surface. Clarendon Press, Oxford, p. 607.
Selby, M.J., 1993. Hillslope Materials and Processes. Oxford University Press, Oxford, p. 451.
Selby, M.J., Hendy, C.H., Seely, M.K., 1979. A late Quaternary lake in the central Namib Desert, southern
Africa, and some implications. Palaeogeogr. Palaeoclimatol. Palaeoecol. 26, 37-41.
Sepp~l~i, M., 1988. Palsas and related forms. In: Clark, M.J. (Ed.), Advances in Periglacial
Sepp~il~i, M., 2004. Wind as a Geomorphic Agent in Cold Climates. Cambridge University Press,
Sepp~il~i, M., Lind& K., 1978. Wind tunnel studies of ripple formation. Geografiska Annaler 60A, 29-40.
Serrano, E., Agudo, C., 1998. Glaciares rocosos activos de los Pirineos. In: G6mez Ortiz, A., Salvador
Franch, L., Schulte, A., Garcfa Navarro, A. (Eds), Procesos bioffsicos actuales en medios flffOS.
Publicaciones de la Universitat de Barcelona, Barcelona, pp. 133-154.
Serrat, D., 1979. Rock glacier morainic deposits in the eastern Pyrenees. In: Schltichter, Ch. (Ed.),
Moraines and Varves. Origin, Genesis, Classification. Balkema, Rotterdam, pp. 93-100.
Servant, M., Servant-Vildary, S., 1980. L'environnement Quaternaire du bassin du Tchad. In: Williams,
M.A.J., Faure, H. (Eds), The Sahara and the Nile. Balkema, Rotterdam, pp. 133-162.
Shackleton, N.J., Kennett, J.P., 1975. Paleotemperature history of the Cenozoic and initiation of Antarctic
glaciation: oxygen and carbon isotope analyses in DSDP Sites 277, 279 and 281. In: Kennett, J.P.,
Houtz, R.E., Andrews, P.B., Edwards, A.R. et al. (Eds), Initial Reports of the Deep Sea Drilling Project
No. 29. US Government Printing Office, Washington, DC.
Shackleton, N.J., Opdyke, N.D., 1973. Oxygen isotope and paleomagnetic stratigraphy of equatorial
Pacific core V28-238: oxygen isotope temperatures and ice volumes on a 105 and 10 6 year scale.
Shackleton, N.J., Opdyke, N.D., 1976. Oxygen isotope and paleomagnetic stratigraphy of equatorial
Pacific core V28-239, late Pliocene to latest Pleistocene. In: Cline, R.M., Hays, R.D. (Eds), Investi-
gation of Late Quaternary Paleoceanography and Paleoclimatology, Memoirs of the Geological
Society of America, Vol. 145, pp. 449-464.
Sharon, D., 1962. On the nature of hamadas in Israel. Z. Geomorphol. 6, 129-147.
Sharp, M.J., 1982. Modification of clasts in lodgement tills by glacial erosion. J. Glaciol. 28, 475-481.
Sharp, R.P., 1940. Geomorphology of the Ruby-East Humboldt Range, Nevada. Bull. Geol. Soc. Am. 51,
337-372.
Sharp, R.P., 1942a. Soil structures in the St Elias Range, Yukon Territory. J. Geomorphol. 5, 274-304.
Sharp, R.P., 1942b. Periglacial involutions in northeastern Illinois. J. Geol. 50, 113-133.
Sharp, R.P., 1949. Pleistocene ventifacts east of the Bighorn Mountains, Wyoming. J. Geol. 57,
175-195.
Sharp, R.P., 1958. Malaspina Glacier, Alaska. Bull. Geol. Soc. Am. 69, 617-646.
Sharp, R.P., 1960. Glaciers. Condon Lecture Publishing, University Oregon Press, Eugene, p. 78.
Sharp, R.P., 1963. Wind tipples. J. Geol. 71, 617-636.
Sharp, R.P., 1964. Wind-driven sand in the Coachella valley, California. Bull. Geol. Soc. Am. 75,
785-804.
Sharp, R.P., 1980. Wind-driven sand in Coachella Valley, California: further data. Bull. Geol. Soc. Am.
91,724-730.
Sharp, R.P., 1988. Living Ice. Cambridge University Press, Cambridge, MA, p. 225.
Sharpe, C.F.S., 1938. Landslides and Related Phenomena. Columbia University Press, New York, p. 137.
Sharpe, D.R., 1967. The stratified nature of drumlins from Victoria Island and southern Ontario.
In: Menzies, I n :
References 743
Sheridan, D., 1981. Problems of Desertification of the United States, Council Environmental Quality,
p. 142.
Sherman, G.D., 1952. The genesis and morphology of the alumina-rich laterite clays, Problems of Clay
and Laterite Genesis. American Institute of Mining and Metallurgical, New York, pp. 154-161.
Shroder, J.E., 1973. Erosional residual in Malawi near the edge of the African Rift. Z. Geomorphol.
Suppl.bd 18, 121-143.
Simanton, J.R., Rawitz, E., Shirley, E., 1984. Efects of rock fragments on erosion of semiarid rangeland
soils. Soil Sci. Soc. Am. Spec. Publ. 13, 65-72.
Simonett, D.S., 1967. Landslide distribution and earth-quakes in the Bewanni and Toricelli Mountains,
New Guinea, a statistical analysis. In: Jennings, J.N., Mabbutt, J.A. (Eds), Landform Studies from
Australia and New Guinea. Cambridge University Press, London, pp. 64-84.
Simons, F.S., Eriksen, G.E., 1953. Some desert features of northwest central Peru. Boletin de la Sociedad
Geol6gica del Peril 26, 229-245.
Singer, A., 1979/1980. The paleoclimatic interpretation of clay minerals in soils and weathering profiles.
Earth-Sci. Rev. 15, 303- 326.
Singh, G., 1971. The Indus valley culture seen in context of post-glacial climate and ecological studies in
northwest India. Archaeol. Anthropol. Oceania 6, 177-189.
Sirvent, J., Desir, G., Guti6rrez, M., Sancho, C., Benito, G., 1997. Erosion rates in badland areas recorded
by collectors, erosion pins and profilometer techniques (Ebro Basin, NE-Spain). Geomorphology 18,
61-75.
Sivarajasingham, S., Alexander, L.T., Cady, J.G., Cline, M.G., 1962. Laterite. Adv. Agron. 14, 1-60.
Small, R.J., 1970. The Study of Landforms. A Textbook of Geomorphology. Cambridge University Press,
Small, R.J., 1987. Englacial and supraglacial sediment: transport and deposition. In: Gurnell, A.M.,
Clark, M.J. (Eds), Glacio-Fluvial Sediment Transfer-an Alpine Perspective. Wiley, Chichester,
pp. 111-145.
Smalley, I.J., 1972. The interaction of great rivers and large deposits of primary loess. Trans. NY Acad.
Sci. 34, 534-542.
Smalley, I.J., 1990. Possible formation mechanisms for the modal coarse-silt quartz particles in loess.
Quaternary Int. 7/8, 23-27.
Smalley, I.J., 1995. Maxing the material: the formation of silt-sized primary mineral particles for loess
deposits. Quaternary Sci. Rev. 14, 645-651.
Smalley, I.J., Unwin, D.J., 1968. The formation and shape of drumlins and their distribution and
orientation in drumlin fields. J. Glaciol. 7, 377-390.
Smith, B., Warke, P., 1997. Controls and uncertainties in the weathering environment. In: Thomas, D.S.G.
(Ed.), Arid Zone Geomorphology. Wiley, Chichester, pp. 41-54.
Smith, B.J., 1978. The origin and geomorphic implications of cliff foot recesses and tafoni on limestone
hamadas in the northwest Sahara. Zeitscrift ffir Geomorphol. 22, 21-43.
Smith, B.J., 1994. Weathering processes and forms. In: Abrahams, A.D., Parson, J.P. (Eds),
Smith, B.J., de Sfinchez, B.A., 1992. Erosion hazards in a Brazilian suburb. Geogr. Rev. 6, 37-41.
Smith, G.I., Streat-Perrott, F.A., 1983. Pluvial lakes of the Western United States. In: Porter, S.C. (Ed.),
Late-Quaternary Environments of the United States. University of Minnesota Press, Minneapolis,
pp. 190-212.
Smith, H.T.U., 1948. Giant glacial groves in Northwest Canada. Am. J. Sci. 246, 503-514.
Smith, H.T.U., 1953. The Hickory Run boulderfield, Carbon County, Pennsylvania. Am. J. Sci. 25,
625-642.
Smith, H.T.U., Smith, A.P., 1945. Periglacial rock streams in the Blue Ridge area. Bull. Geol. Soc. Am. 56,
1198.
Smith, K., Ward, R., 1998. Floods: Physical Processes and Human Impacts. Wiley, Chichester, p. 382.
Smith, M.W., 1993. Climatic change and permafrost. In: French, H.M., Slaymaker, O. (Eds), Canada's
Cold Environments. McGill-Queen's University Press, Montreal, pp. 291-311.
Smith, N.D., Minter, W.E.C., 1980. Sedimentologic central of gold and uranium and two Witwatersrand
palaeoplacers. Econ. Geol. 75, 1-14.
Smith, N.D., Vendl, M.A., Kennedy, S.K., 1982. Comparison of sedimentation regimes in four glacier-
fed lakes of western Alberta. In: Davidson-Arnott, R., Nickling, W., Fahey, B.D. (Eds), Research in
Glacial, Glaciofluvial and Glaciolacustrine Systems. Geo Books, Norwich, pp. 203-238.
So, C.L., 1971. Mass movements associated with the rainstorm of June 1966 in Hong Kong. Trans. Inst.
Br. Geogr. 53, 55-66.
Srderman, G., 1985. Planation and weathering in eastern Fennoscandia. Fennia 163, 347-352.
Soloviev, P.A., 1973. Thermokarst phenomena and landforms due to frost heaving in Central Yakutia.
Biuletyn Peryglacjalny 23, 135-155.
Soriano, M.A., Simrn, J.L., 1995. Alluvial dolines in the central Ebro basin, Spain: a spatial and
developmental hazard analysis. Geomorphology 11,295-309.
Soriano, M.A., Simrn, J.L., Gracia, J., Salvador, T., 1994. Alluvial sinkholes over gypsum in the Ebro
basin (Spain): genesis and environmental impact. Hydrol. Sci. 39, 257-268.
Sorriso-Valvo, M., 1988. Landslide-related fans in Calabria. Catena Suppl. 13, 109-121.
Souchez, P., 1966. Grlivation et 6volution des versants en bordure de l'Islandsis d'Antartides orientale.
In: Macar, P. (Ed.), L'Evolution des Versants. Les Congr~s et Colloques de L'Univesit6 de Lirge, Vol.
40, pp. 291-298.
Spaulding, W.G., 1990. Vegetation and climatic development of the Mojave Desert: The last glacial
maximum to the present. In: Betancourt, J.L., Van Devender, T.R., Martin, P.S. (Eds), Packrat
Middens, the Last 40,000 Years of Biotic Change. University of Arizona Press, Tucson, pp. 166-199.
Sperling, C.H.B., Cooke, R.U., 1985. Laboratory simulation of rock weathering by salt crystallization and
hydration processes in hot, arid environments. Earth Surf. Processes Landforms 10, 541-555.
Srivastava, P., Juyal, N., Singhve, A.K., Wasson, R.J., Bateman, M.D., 2001. Luminiscence chronology of
river-adjustment and incision of Quaternary sediments in the alluvial plain, of the Sabarmati River,
north Gujarat, India. Geomorphology 36, 217-229.
Stager, J.K., 1956. Progress report on the analysis on the characteristics and distribution of pingos east of
the Mackenzie Delta. Can. Geogr. 7, 13-20.
Stanistreet, I.G., McCarthy, T.S., 1993. The Okavango Fan and the classification of subaerial fan systems.
Sediment. Geol. 85, 115-133.
Starkel, L., 1976. The role of extreme (catastrophic) meteorological events in contemporary evolution of
slopes. In: Derbyshire, E. (Ed.), Geomorphology and Climate. Wiley, London, pp. 203-246.
Steijn, H.van, Boelhouwers, J., Harris, S., Hrtu, B., 2002. Recent research on the nature, origin and
climatic relations of blocky and stratified slope deposits. Prog. Phys. Geogr. 26, 551-575.
Stenborg, T., 1970. Delay of run-off from a glacier basin. Geogr. Ann. 52, 1-30.
Stocking, M., 1976. Tunnel erosion. Rhod. Agric. J. 73, 35-39.
Stockton, C.W., Fritts, H.C., 1971. Conditional probability of occurrence for variations in climate based on
width of annual tree-rings in Arizona. Tree-Ring Bull. 31, 3-24.
Stoddart, D.R., 1969a. Climatic geomorphology: review and re-assessment. Prog. Geogr. 1, 159-222.
Stoddart, D.R., 1969b. Climatic geomorphology. In: Chorley, R.J. (Ed.), Introduction to Fluvial Processes.
Methuen, Suffolk, pp. 189- 201.
Stoddart, D.R., 1969c. Climatic geomorphology: review and reassessment. In: Board, C. et al. (Eds),
Progress in Geography. Arnold, London, pp. 160-222.
Stokes, S., 1997. Dating of desert sequences. In: Thomas, D.S.G. (Ed.), Arid Zone Geomorphology. Wiley,
Chichester.
Strachan, A.D., Dearman, W.R., 1983. Engineering geological mapping in glaciated terrains. In: Eyles, N.
(Ed.), Glacial Geology: An Introduction for Engineerings and Earth Scientists. Pergamon Press,
Strahler, A.N., 1952. Dynamic basis of geomorphology. Bull. Geol. Soc. Am. 66, 923-938.
Strahler, A.N., 1965. Introduction to Physical Geography. Wiley, New York, p. 643.
Strakhov, N.M., 1967. Principles of Lithogenesis, Vol. 1. Oliver and Boyd, Edinburgh, p. 245.
Street, G.J., Anderson, A., 1993. Airborne electromagnetic surveys of the regolith. Explor. Geophys. 24,
795 -800.
Stret-Perrott, F.A., Roberto, N., Metcalfe, S.L., 1985. Geomorphic implications of late Quaternary
hydrological and climatic changes in the Northern Hemisphere tropics. In: Douglas, I., Spencer, T.
(Eds), Environmental Change and Tropical Geomorphology. Allen and Unwin, London,
pp. 165-183.
Stroeven, A., Wal, R.van de, Oerlemans, J., 1989. Historic front variations of the Rhone Glacier:
simulation with an ice flow model. In: Oerlemans, J. (Ed.), Glacier Fluctuations and Climatic Change.
Kluwer, Dordrecht, pp. 391-405.
Sueoka, T., 1988. Identification and classification of granitic residual soils using chemical weathering
index, Geomechanics in Tropical Soils, Vol. 1. Balkema, Rotterdam, pp. 55-61.
References 745
Sugden, D.E., 1974. Landscapes of glacial erosion in Greenland and their relationship to ice, topographic
and bedrock conditions. In: Brown, E.H., Waters, R.S. (Eds), Progress in Geomorphology, Vol. 7,
pp. 177-195, Institute British Geographers.
Sugden, D.E., 1977. Reconstruction of the morphology, dynamics and thermal characteristics of the
Laurentide ice sheet at its maximum. Arctic Alpine Res. 9, 27-47.
Sugden, D.E., 1978. Glacial erosion by the Laurentide ice sheet. J. Glaciol. 20, 367-391.
Sugden, D.E., 1982. Arctic and Antarctic. A Modern Geographical Systems. Blackwell, Oxford, p. 472.
Sugden, D.E., John, B.S., 1976. Glaciers and Landscape. Edward Arnold, London, p. 376.
Sugden, D.E., Glasser, N.F., Clapperton, C.M., 1992. Evolution of large roches moutonnres. Geogr. Ann.
74A, 253-264.
Summerfield, M.A., 1983. Silcrete. In: Goudie, A.S., Pye, K. (Eds), Chemical Sediments and Geomor-
phology: Precipitates and Residua in the Near Surface Environment. Academic Press, New York,
pp. 59-91.
Sundquist, E.T., 1993. The global carbon dioxide budget. Science 259, 934-941.
Sung-Chiao, C., Xing, J., 1982. Origin and development of the Shamo gand deserts and the Goby stony
deserts of China. In: Smiley, T. L. (Ed.). The geological history of the world's deserts. University
of Uppsala Press, Uppsala, 79-91.
Sutherland, D.G., 1984. Geomorphology and mineral exploration: some samples for exploration for
diamondiferous placer deposits. Z. Geomorphol. Suppl.bd. 51, 95-108.
Sutherland, D.G., 1985. Geomorphological controls on the distribution of placer deposits. J. Geol. Soc.
Lond. 142, 727-737.
Swan, S.B.StC., 1972. Land surface evolution and related problems with reference to a humid tropical
region: Johor, Malaya. Z. Geomorphol. 16, 160-181.
Sweet, M.L., Kocurek, G., 1990. An empirical model of aeolian dune lee-face airflow. Sedimentology 37,
1023-1038.
Sweeting, M.M., 1958. The karstlands of Jamaica. Geogr. J. 124, 184-199.
Sweeting, M.M., 1966. The weathering of limestones with particular reference to the Carboniferous
Limestones of northern England. In: Dury, G.H. (Ed.), Essays in Geomorphology. Elsevier,
Amsterdam, pp. 177- 210.
Sweeting, M.M., 1972. Karst Landforms. Macmillan, London, p. 362.
Sweeting, M.M., 1995. Karst in China. Its Geomorphology and Environment. Springer, Berlin, p. 265.
Swinzow, G.K., 1969. Certain aspects of engineering geology in permafrost. Eng. Geol. 3, 177-215.
Swoboda-Colberg, N.G., Drever, J., 1993. Mineral dissolution rates in plot-scale field and laboratory
experiments. Chem. Geol. 105, 51-69.
Syvitski, J.P.M., Burrell, D.C., Skei, J.M., 1987. Fjords: Processes and Products. Springer, New York, p. 379.
Taber, S., 1929. Frost heaving. J. Geol. 37, 428-461.
Takahashi, T., 1991. Debris Flow. A.A. Balkema, Rotterdam, p. 165.
Takei, A., Kobashi, S., Fukushima, Y., 1981. Erosion and sediment transport measurement in a weathered
granite mountain areas. Symposium on Erosion and Sediment Transport Measurement, Florence 133,
493-502, International Association Hydrological Sciences.
Talbot, M.R., 1980. Environmental responses to climatic change in the West African Sahel over the
past 20,000 years. In: Williams, M.A.J., Faure, H. (Eds), The Sahara and the Nile. Balkema, Rotterdam,
pp. 37-62.
Talbot, M.R., Williams, M.A.J., 1978. Erosion of fixed dunes in the Sahel, Central Niger. Earth Surf.
Processess 3, 107-113.
Tanner, W.F., 1961. An alternate approach to morphogenetic climates. Southeastern Geol. 2, 251-257.
Tardy, Y., Roquin, C., 1992. Geochemistry and evolution of lateritic landscapes. In: Martini, I.P.,
Chesworth, W. (Eds), Weathering, Soils and Paleosols. Elsevier, Amsterdam, pp. 407-443.
Tator, B.A., 1952. Pediment characteristics and terminology (Part I). Ann. Assoc. Am. Geogr. 42,
295-317.
Tator, B.A., 1953. Pediment characteristics and terminology (Part II). Ann. Assoc. Am. Geogr. 43, 47-53.
Taylor, G.R., Eggleton, R.A., 2001. Regolith Geology and Geomorphology. Wiley, Chichester, p. 375.
Taylor, G.R., Eggleton, R.A., Holzhauer, C.C., Maconachie, L.A., Gordon, M., Brown, M.C., McQueen,
K.G., 1992. Cool climate lateritic and bauxitic weathering. J. Geol. 100, 669-677.
Taylor, K.C., Lamorey, G.W., Doyle, G.A., Alley, R.B., Grootes, P.M., Mayewskill, P.A., White, J.W.C.,
Barlow, L.K., 1993. The "flickering switch" of the Late Pleistocene climatic change. Nature, 361,
432-436.
Tchakerian, V.P., 1994. Palaeoclimatic interpretations from desert dunes and sediments. In: Abrahams,
A.D., Parsons, A.J. (Eds), Geomorphology of Desert Environments. Chapman & Hall, London,
pp. 631-643.
Tchakerian, V.P., 1999a. Playa. In: Mares, M.A. (Ed.), Encyclopaedia of Deserts. University of Oklahoma
Press, Norman, pp. 443-444.
Tchakerian, V.P., 1999b. Sebkha. In: Mares, M.A. (Ed.), Encyclopaedia of Deserts. University of
Oklahoma Press, Norman, p. 485.
Temple, P.H., Rapp, A., 1972. Landslides in the Mgeta area, western Uluguru Mountains, Tanzania.
Geogr. Ann. 54A, 157-193.
Thomas, D.S., 1988. The biogeomorphology of arid and semi-arid environments. In: Viles, H. (Ed.),
Biogeomorphology. Blackwell, Oxford, pp. 193-221.
Thomas, D.S.G., 1989. The nature of arid environments. In: Thomas, D.S.G. (Ed.), Arid Zone Geomor-
phology. Belhaven Press, London, pp. 1-8.
Thomas, D.S.G., (Ed.), 1989. Arid Zone Geomorphology. Belhaven Press, London, p. 372.
Thomas, D.S.G., 1992. Desert dune activity: concepts and significance. J. Arid Environ. 22, 31-38.
Thomas, D.S.G., (Ed.), 1997a. Arid Zone Geomorphology. Process, Form and Change in Drylands. Wiley,
Chichester, p. 713.
Thomas, D.S.G., 1997b. Arid environments: their nature and extent. In: Thomas, D.S.G. (Ed.), Arid Zone
Geomorphology. Process, Form and Change in Drylands. Wiley, Chichester, pp. 3-12.
Thomas, D.S.G., 1997c. Reconstructing ancient arid environments. In: Thomas, D.S.G. (Ed.), Arid Zone
Thomas, D.S.G., 1997d. Sand seas and aeolian bedforms. In: Thomas, D.S.G. (Ed.), Arid Zone
Geomorphology: Process, Form and Change in Drylands. Wiley, Chichester, pp. 373-412.
Thomas, D.S.G., Goudie, A.S., 1984. Ancient ergs of the Southern Hemisphere. In: Vogel, J.C. (Ed.), Late
Cenozoic Palaeoclimates of the Southern Hemisphere. Balkema, Rotterdam, pp. 407-418.
Thomas, D.S.G., Middleton, N.J., 1994. Desertification: Exploding the Myth. Wiley, Chichester, p. 194.
Thomas, D.S.G., Shaw, P.A., 1991. The Kalahari Environment. Cambridge University Press, Cambridge,
MA, p. 284.
Thomas, M.F., 1965. Some aspects of the geomorphology of domes and tors in Nigeria. Z. Geomorphol. 9,
63-81.
Thomas, M.F., 1966. Some geomorphological implications of deep weathering patterns in crystalline
rocks in Nigeria. Trans. Inst. Br. Geogr. 40, 173-193.
Thomas, M.F., 1974. Tropical Geomorphology. A Study of Weathering and Landform Development in
Warm Climates. Macmillan, London, p. 329.
Thomas, M.F., 1976. Criteria for the recognition of climatically induced variations in granite landforms.
In: Derbyshire, E. (Ed.), Geomorphology and Climate. Wiley, London, pp. 411-445.
Thomas, M.F., 1978. The study of inselbergs. Z. Geomorphol. Suppl.bd. 31, 1-41.
Thomas, M.F., 1983. Contemporary denudation systems and the effects of climatic change in the humid
tropics - some problems from Sierra Leone. In: Briggs, D.J., Waters, R.S. (Eds), Studies in Quaternary
Geomorphology. Geo Books, Norwich, pp. 195-214.
Thomas, M.F., 1986. Savanna. In: Fookes, P.G., Vaughan, P.R. (Eds), A Handbook of Engineering
Geology. Survey University Press, Glasgow, pp. 125-136.
Thomas, M.F., 1989. The role of etch processes in landform development: I. Etching concepts and
the formation of relief. Z. Geomorphol. 33, 129-142.
Thomas, M.F., 1994. Geomorphology in the Tropics. A Study of Weathering and Denudation
in Low Latitudes. Wiley, Chichester, p. 460.
Thomas, M.F., Goudie, A.S., (Eds), 1985. Dambos: small channelless valleys in the tropics. Character-
istics, formation, utilisation. Z. Geomorphol. Suppl.bd., 52, p. 228.
Thomas, M.F., Thorp, M.B., 1985. Environmental change and episodic explanation in the humid tropics
of Sierra Leona: the Koidn etchplain. In: Douglas, I., Spencer, T. (Eds), Environmental Change and
Tropical Geomorphology. Allen and Unwin, London, pp. 239-267.
Thomas, M.F., Thorp, M.B., 1992. Landscape dynamics and surface deposits arising from late
Quaternary fluctuations in the forest-savanna boundary. In: Furley, P.A., Proctor, J., Ratter, J.A.
(Eds), Nature and Dynamics of Forest-Savanna Boundaries. Chapman & Hall, London, pp.
215-253.
Thomas, M.F., Thorp, M.B., 1993. The geomorphology of some Quaternary placer deposits.
References 747
Thomas, M.F., Thorp, M.B., 1995. Geomorphic response to rapid climatic and hydrologic change during
the Late Pleistocene and Early Holecene in the humid and sub-humid tropics. Quaternary Sci. Rev. 14,
193-207.
Thorarinsson, S., 1953. Some new aspects of the Grimsv6tn problem. J. Glaciol. 2, 267-274.
Thorbecke, F., 1927. Morphologie der klimazonen. DUsseldorfer Geogr. Vortr/~ge Er6rterungen, 1-100,
Part 3.
Thorn, C.E., 1988a. Introduction to Theoretical Geomorphology. Unwin Hyman, Boston, p. 247.
Thorn, C.E., 1988b. Nivation: a geomorphic chimera. In: Clark, M.J. (Ed.), Advances in Periglacial
Thom, C.E., 1992. Periglacial geomorphology: what, where, when? In: Dixon, J.C., Abrahams, A.D.
(Eds), Periglacial Geomorphology. Wiley, Chichester, pp. 1-30.
Thom, C.E., Hall, K., 1980. Nivation: an arctic-alpine comparison and reappraisal. J. Glaciol. 25, 109-124.
Thorn, C.E., Hall, K., 2002. Nivation and cryoplanation: the case for scrutiny and integration. Prog. Phys.
Geogr. 26, 533-550.
Thombury, W.D., 1954. Principles of Geomorphology. Wiley, New York, p. 618.
Thomes, J., 1976. Semiarid Erosional Systems. In: London School of Economic, Geographical Papers 7,
London, p. 79.
Thomes, J.B., 1994. Catchment and channel hydrology. In: Abrahams, A.P., Parsons, A.J. (Eds),
Thomthwaite, C.W., 1948. An approach towards a rational classification of climate. Geogr. Rev. 38,
55-94.
Torri, D., Sfalanga, M., Del Sette, M., 1987. Splash detachment: runoff depth and soil cohesion. Catena 14,
149-155.
Toy, T.J., 1977. Hillslope form and climate. Geol. Soc. Am. Bull. 88, 16-22.
Trendall, A.F., 1962. The formation of "apparent peneplains" by a process of combined lateritisation and
surface wash. Z. Geomorphol. 6, 183-197.
Trescasses, J. J., 1973. L'~volution g~ochimique superg~ne des roches ultiabasiques en zone tropicale et la
formation des egisements nichrlifere, de Novella Calrdonia. Thrse. Strasbourg. p. 347.
Tricart, J., 1952. Cours de Gromorphologie. Gromorphologie Climatique. CDU, Paris, p. 193.
Tricart, J., 1956. Geomorphologic dynamique du Delta du Senegal. Revue de Gromorphologie
Dynamique 7, 65-86.
Tricart, J., 1965. Principes et M&hodes de la Gromorphologie. Masson, Paris, p. 496.
Tricart, J., 1967. Le Model6 des Rrgions Prriglaciaires. SEDES, Paris, p. 512.
Tricart, J., 1969. Le Model6 des Rrgions Srches. SEDES, Paris, p. 472.
Tricart, J., 1974a. Existence de prriodes s~ches au Quaternaire en Amazonie et dans les rrgions voisines.
Revue de Gromorphologic Dynamique 23, 145-158.
Tricart, J., 1974b. Le Model6 des R~gions Chaudes: For&s et Savanes. SEDES, Paris, p. 345.
Tricart, J., 1975. Influence des oscillations climatiques rrcents sur le model6 en Amazonie Orientale
(rrgion de Santardm) d'aprrs des images de radar latrral. Z. Geomorphol. 19, 140-163.
Tricart, J., 1979. E1 concepto de "pluvial". Actas de la III Reuni6n Nacional del Grupo Espafiol de Trabajo
del Cuaternario, Madrid, pp. 7-20.
Tricart, J., 1982. E1 Pantanal: Un ejemplo del impacto geomorfol6gico sobre el ambiente, Informaciones
Geogrfificas, Vol. 29. Universidad de Chile, Santiago, pp. 81-97.
Tricart, J., 1985. Evidence of Upper Pleistocene dry climates in northern South America. In: Douglas, I.,
Spencer, T. (Eds), Environmental Change and Tropical Geomorphology. Allen and Unwin, London,
pp. 197-217.
Tricart, J., Alfonsi, P., 1981. Actions 6oliennes rrcents aux abords du delta de l'Orrnoque. Bull. AGG 476,
75-82.
Tricart, J., Cailleux, A., 1962. Le Model~ Glaciaire et Nival. SEDES, Paris, p. 508.
Tricart, J., Cailleux, A., 1965. Introduction ?ala Gromorphologie Climatique. SEDES, Paris, p. 306.
Tricart, J., Raynal, R., Besancon, J., 1972. C6nes rocheux, pediments, glacis. Ann. Grogr. 43, 1-24.
Troeh, F.R., 1965. Landform equations fitted to topographic maps. Am. J. Sci. 263, 616-627.
Troll, C., 1944. Strukturb6den, solifluktion und frostklimate der Erde. Geologische Rundschau 34,
545-694.
Trombe, F., 1952. Trait~ de Sp~lrologie. Payot, Paris, p. 376.
Trudgill, S.T., 1976. Rock weathering in climate: quantitative and experimental aspects. In: Derbyshire, E.
(Ed.), Geomorphology and Climate. Wiley, London, pp. 59-99.
Trudgill, S., 1985. Limestone Geomorphology. Longman, London, p. 196.

Tseo, G., 1993. Two types of longitudinal dune fields and possible mechanisms for their development.
Tsoar, H., 1983a. Dynamic processes acting on a longitudinal (seif) dune. Sedimentology 30,
567-578.
Tsoar, H., 1983b. Wind tunnel modelling of echo and climbing dunes. In: Brookfield, M.E., Ahlbrandt,
T.S. (Eds), Eolian Sediments and Processes. Elsevier, Amsterdam, pp. 247-259.
Tsoar, H., Moiler, J.T., 1986. The role of vegetation in the formation of linear dunes. In: Nickling, W.G.
(Ed.), Aeolian Geomorphology. Allen and Unwin, Boston, pp. 75-95.
Tsoar, H., Pye, K., 1987. Dust transport and the question of desert loess formation. Sedimentology 34,
139-154.
Tsukamoto, Y., Kusakabe, D., 1984. Vegetative influences on debris slides occurrences on steep slopes in
Japan. Proc. Symp. On Effects Forest Land Use on Erosion and Slope Stability. Environment and
Policy Institute. Honolulu, Hawaii.
Tucker, M.E., 1978. Gypsum crust (gypcrete) and patterned ground from northern Iraq. Z. Geomorphol.
22, 89-100.
Tufnell, L., 1972. Ploughing blocks with special reference to north-west England. Biuletyn Peryglacjalny
21,237-270.
Tufnell, L., 1984. Glacier Hazards. Longman, London, p. 97.
Twidale, C.R., 1962. Steepened margins of inselbergs from north-western Eyre Peninsula, South Australia.
Z. Geomorphol. NF 6, 51-69.
Twidale, C.R., 1964. A contribution to the general theory of domed inselbergs. Trans. Inst. Br. Geogr. 34,
91-113.
Twidale, C.R., 1967. Hillslopes and pediments in the Flinders Ranges, South Australia. In: Jennings, J.N.,
Mabbutt, J.A. (Eds), Landform Studies from Australia and New Guinea. Cambridge University Press,
Cambridge, MA, pp. 95-117.
Twidale, C.R., 1968a. Geomorphology with Special Reference to Australia. Nelson, Melbourne.
Twidale, C.R., 1968b. Inselberg. In: Fairbridge, R.W. (Ed.), Encyclopedia of Geomorphology. Reinhold,
Twidale, C.R., 1973. On the origin of the sheet jointing. Rock Mech. 5, 163-187.
Twidale, C.R., 1976a. Analysis of landforms. Wiley, Sydney, p. 572.
Twidale, C.R., 1976b. On the survival of palaeoforms. Am. J. Sci. 276, 77-94.
Twidale, C.R., 1978. On the origin of Ayers Rock, Central Australia. Z. Geomorphol. Suppl.bd. 31, 177-206.
Twidale, C.R., 1981. Granite Inselbergs. Geogr. J. 147, 54-71.
Twidale, C.R., 1982a. The evolution of bornhardts. Am. Sci. 70, 268-276.
Twidale, C.R., 1982b. Granite Landforms. Elsevier, Amsterdam, p. 372.
Twidale, C.R., 1983. Pediments, peneplains and ultiplains. Revue de Gromorphologie Dynamique 32,
1-35.
Twidale, C.R., 1990. The origin and implications of some erosional landforms. J. Geol. 98, 343-364.
Twidale, C.R., 1992. King of the plains: Lester King' s contribution to Geomorphology. Geomorphology 5,
491-509.
Twidale, C.R., Bourne, J.A., 1975. Episodic exposure of inselbergs. Bull. Geol. Soc. Am. 86, 1473-1481.
Twidale, C.R., Bourne, J.A., 1978. Bornhardts. Z. Geomorphol. Suppl.bd. 31, 111-137.
Twidale, C.R., Corbin, E.M., 1963. Gnammas. Revue de Gromorphologie Dynamique 14, 1-20.
Twidale, C.R., Lageat, Y., 1994. Climatic geomorphology: a critique. Prog. Phys. Geogr. 18, 319-334.
UNEP, 1992. World Atlas of Desertification. Arnold, London, p. 69.
UNESCO, 1981. Avalanche Atlas. UNESCO, Paris, p. 265.
UNESCO-FAO, 1977. Desertification; An Overview, Desertification. Its Causes and Consequences.
Pergamon Press, Oxford, pp. 1-61.
USGS, 1983. Permafrost. In: Tank, R.W. (Ed.), Environmental Geology. Oxford University Press, Oxford,
pp. 163-170.
Valeton, I., 1972. Bauxites. Elsevier, Amsterdam, p. 226.
Valeton, I., 1994. Element concentration and formation of ore deposits by weathering. Catena 21,
99-129.
Vandenberghe, J., Thorn, C.E., 2002. Progress in periglacial research. Prog. Phys. Geogr. 26, 475-477.
Vandenberghe, J., Woo, M., 2002. Modern and ancient periglacial river types. Prog. Phys. Geogr. 26,
479-506.
References 749
Van der Wateren, D.F.M., 1987. Structural geology and sedimentology of the Dammer Berge push
moraine, FRG. In: Van der Meer, J.J.M. (Ed.), Tills and Glaciotectonics. Balkema, Rotterdam,
pp. 157-182.
Van der Wateren, D.F.M., 1992. Structural geology and sedimentology of push moraines. Processes of
soft sediment deformation in a glacial environment and the distribution of glaciotectonic styles.
Dissertation, Amsterdam, p. 230.
Van der Wateren, F.M., 1995. Processes of glaciotectonism. In: Menzies, J. (Ed.), Modern Glacial
Environments. Processes, Dynamics and Sediments. Butterworth-Heinemann, Oxford, pp. 307-335.
Vaniman, D.T., Chipera, S.J., Bish, D.L., 1994. Pedogenesis of siliceous calcretes at Yucca Mountains,
Nevada. Geoderma 63, 1-17.
Varnes, D.J., 1958. Landslide types and processes. In: Eckel, E.B. (Ed.), Landslides and Engineering
Practice, Highway Research Board. National Academy of Sciences, Washington, DC, pp. 20-47,
Special Report, 29.
Varnes, D.J., 1978. Slope movement types and processes. In: Schuster, R.L., Krizek, R.J. (Eds), Landslide
Analysis and Control. National Academy of Sciences, Washington, DC, pp. 11-33, Special Report No.
176.
Veder, C., 1981. Landslides and their Stabilization. Springer, New York, p. 247.
Verger, F., 1964. Mottereaux et gilgais. Ann. Grogr. 73,413-430.
Verstappen, H.T., 1964. Karst morphology of the Star Mountains (Central New Guinea) and its relation
to lithology and climate. Z. Geomorphol. 8, 40-49.
Verstappen, H.Th., 1968. On the origin of longitudinal (seif) dunes. Z. Geomorphol. NF 12, 200-220.
Verstappen, H.Th., 1970. Aeolian geomorphology of the Thar Desert and palaeo-climates. Z. Geomorphol.
Suppl.bd. 10, 104-120.
Veste, M., 1995. Structures of geomorphological and ecological units and ecosystems processes in the
linear dune system near Nizzana/Negev. Bielefelder 0kologische Beitrage 8, 85-96.
Vidal Romani, J.R., Twidale, C.R., 1998. Formas y Paisajes Granfticos, Universidade da Corufia. Servicio
de Publications, A Corufia, p. 411.
Vidal Romanf, J.R., Twidale, C.R., Campbell, E.M., Centeno, J.D., 1995. Pruebas morfol6gicas y
estructurales sobre el origen de las fracturas por descamaci6n. Cadernos Laboratorio Xeol6xico de
Laxe 20, 307-346.
Vilaplana, J.M., 1987. Guia dels Paisatges Granftics dels Pa'fsos Catalans. Kapel SA, Barcelona, p. 182.
Viles, H.A., 1984. Biokarst: review and prospect. Prog. Phys. Geogr. 8, 523-542.
Viles, H.A., 1995. Ecological perspectives on rock surface weathering: towards a conceptual model.
Geomorphology 13, 21-35.
Vischer, S.S., 1945. Climatic maps of geological interest. Bull. Geol. Soc. Am. 56, 713-736.
Vivian, R., 1979. Les Glaciers sont Vivants. Ed. Deno~ll, Paris, p. 240.
Vogt, J., 1959. Aspects de l'rvolution morphologique r4cente de l'Ouest Africain. Ann. Grogr. 68,
193-206.
Vogt, J., 1966. Le complexe de la stone-line. Mise au point. Bulletin du Bureau de Recherches
Grologiques et Minirres 4, 3-51.
Vogt, T., 1984. Crofites Calcaires: types et Genrse. Universit6 Louis Pasteur. Institut de G4ographie,
Strasbourg, p. 239.
Voight, B., 1990. The 1985 Nevado del Ruiz Volcano catastrophe: anatomy and retrospection. J. Volcanol.
Geotherm. Res. 42, 151-188.
Voight, B., Pariseau, G., 1978. Rockslides and avalanches: an introduction. In: Voight, B. (Ed.),
Rockslides and Avalanches, 1. Natural Phenomena. Elsevier, Amsterdam, pp. 1-67.
V61kel, J., Grunert, J., 1990. To the problem of dune formation and dune weathering during the Late
Pleistocene and Holocene in the southern Sahara and Sahel. Z. Geomorphol. 34, 117.
Wahrhaftig, C., Cox, A., 1959. Rock glaciers in the Alaska Range. Bull. Geol. Soc. Am. 70, 383-436.
Walcott, R.I., 1970. Isostatic response to loading of the crust in Canada. Can. J. Earth Sci. 7, 716-726.
Walder, J.S., 1982. Stability of sheet flow of water beneath temperate glaciers and implications for glacier
surging. J. Glaciol. 28, 273-293.
Walker, G., 2004. Frozen time. Nature 429, 596-597.
Walker, H., 1973. The morphology of the North Slope. In: Britton, M. (Ed.), Alaska Arctic Tundra. AINA,
Washington, DC, pp. 49-92.
Walker, H.J., 1986. Periglacial environments. In: Fookes, P.G., Vaughan, P.R. (Eds), A Handbook of
Engineering Geomorphology. Surrey University Press, London, pp. 82-96.
Walker, J.G.C., Sloan, L.C., 1992. Something is wrong with climate theory. Geotimes, 16-18, June.
Walling, D.E., 1984. The sediment yields of African rivers. In: Walling, D.E., Foster, S.S.D., Wurzel, P.
(Eds), Challenges in African Hydrology and Water Resources, International Association of Hydrolo
gical Sciences, No. 144, pp. 265-283.
Walling, D.E., Webb, B.W., 1986. Solutes in fiver systems. In: Trudgill, S.T. (Ed.), Solute Processes.
Waltham, A.C., 1989. Ground Subsidence. Blackie, Glasgow, p. 202.
Walther, J., 1915. Laterite in Westaustralien. Zeitschrift der Deustschen Geologischen Gessellschaft 62
(1-7), 46-53.
Ward, A.W., Greeley, R., 1984. Evolution of yardangs at Rogers Lake, California. Bull. Geol. Soc. Am.
95, 829-837.
Ward, R., 1978. Floods: A Geographical Perspective. Macmillan, London.
Warren, A., Knott, P., 1983. Desert dunes: a short review of needs in desert dune research and a recent
study of micrometeorological dune-initiation mechanisms. In: Brookfield, M.E., Ahlbrandt, T.S. (Eds),
Eolian Sediments and Processes. Elsevier, Amsterdam, pp. 343-352.
Washburn, A.L., 1956. Classification of patterned ground and review of suggested origins. Bull. Geol. Soc.
Am. 67, 823-865.
Washburn, A.L., 1967. Instrumental observations on mass wasting in the Mesters Vig District, Northeast
Greenland. Meddelelser om Gronland 166, p. 318.
Washburn, A.L., 1969. Weathering, frost action and patterned ground in the Mesters Vig District,
Northeast Greenland. Meddelelser om Gronland 176, 303.
Washburn, A.L., 1979. Geocryology. A Survey of Periglacial Processes and Environments. Arnold,
London, p. 406.
Wasson, R.J., 1977. Last-glacial alluvial fan sedimentation in the lower Derwent Valley, Tasmania.
Wasson, R.J., 1983. Dune sediment types, sand colour, sediment provenance and hydrology in the
Strzelecki-Simpson dunefield, Australia. In: Brookfield, M.E., Ahlbrandt, T.S. (Eds), Eolian Sediments
and Processes. Elsevier, Amsterdam, pp. 165-195.
Wasson, R.J., 1984. Late Quaternary palaeo-environments in the desert dunnefields of Australia.
In: Vogel, J.C. (Ed.), Late Cenozoic Palaeoenvironmental of the Southern Hemisphere. Balkema,
Rotterdam, pp. 419-432.
Wasson, R.J., Rajaguru, S.N., Misra, V.N., Agarwal, D.P., Dhir, R.P., Singhvi, A.K., Kameswara Rao, K.,
1983. Geomorphology, Late Quaternary stratigraphy and palaeoclimatology of the Thar dune field.
Watson, A., 1979. Gypsum crusts in deserts. J. Arid Environ. 2, 3-20.
Watson, A., 1983. Gypsum crusts. In: Goudie, A.S., Pye, K. (Eds), Chemical Sediments and
Geomorphology: Precipitates and Residua in the Near Surface Environment. Academic Press,
Watson, A., 1985. Structure, chemistry and origins of gypsum crusts in southern Tunisia and the central
Namib Desert. Sedimentology 32, 855-875.
Watson, A., 1989a. Desert crusts and varnishes. In: Thomas, D.S.G. (Ed.), Arid Zone Geomorphology.
Hallstead Press, New York, pp. 25-55.
Watson, A., 1989b. Windflow characteristics and aeolian entrainment. In: Thomas, D.S.G. (Ed.),
Arid Zone Geomorphology. Belhaven Press, London, pp. 209-231.
Watson, A., 1990. The control of blowing sand and mobile desert dunes. In: Goudie, A. (Ed.), Techniques
for Desert Reclamation. Wiley, New York, pp. 35-85.
Watson, A., Price-Williams, D., Goudie, A.S., 1983. Palaeoenvironmental interpretation of colluvial
sediments and palaeosols of the Late Pleistocene hypothermal in southern Africa. Palaeogeogr.
Palaeoclim. Palaeoecol. 45, 225-250.
Watts, S.H., 1980. Quaternary pedogenic calcretes from the Kalahari (southern Africa): mineralogy,
genesis and diagenesis. Sedimentology 27, 661-686.
Wayland, E.J., 1933. Peneplains and some other erosional platforms, Annual Report and Bulletin,
Protectorate of Uganda Geological Survey, Department of Mines, Note 1, pp. 77-79.
Weaver, A.J., 2003. The science of climatic change. Geosci. Can. 30, 91-109.
Webb, R.H., Wilshire, H.G., (Eds), 1983. Environmental Effects of Off-Road Vehicles. Springer, New York.
Weertman, J., 1957. On the sliding of glaciers. J. Glaciol. 3, 33-38.
Weertman, J., 1983. Creep deformation of ice. Annu. Rev. Earth Planet. Sci. 11,215-240.
References 751
Wells, S.G., Harvey, A.M., 1987. Sedimentologic and geomorphic variations in storm generated alluvial
fans, Howgill Fells, Norhwest England. Bull. Geol. Soc. Am. 98, 182-198.
Wells, S.G., Metadden, L.D., Dohrenwend, J.C., 1987. Influence of late Quaternary climatic changes on
geomorphic and pedogenic processes on a desert piedmont, eastern Mojave Desert, California.
Wesson, R.L., 1983. The eruption of Mount St Helens - entering the era of real-time geology. In: Tank,
R.W. (Ed.), Environmental Geology. Oxford University Press, New York, pp. 50-62.
Whitaker, C.R., 1973. A Biblography of Pediments. GeoAbstracts, Norwich, p. 95.
Whitaker, C.R., 1979. The use of the term 'pediment' and related terminology. Z. Geomorphol. 23,427-439.
White, W.A., 1972. Deep erosion by continental ice sheets. Bull. Geol. Soc. Am. 83, 1037-1056.
Whiteman, C.A., 1995. Processes of terrestrial deposition. In: Menzies, J. (Ed.), Modern Glacial
Environments: Processes, Dynamics and Sediments. Butterworth-Heineman, Oxford, pp. 293-308.
Whitney, M.I., 1979. Electron micrography of mineral surfaces subject to wind-blast erosion. Bull. Geol.
Soc. Am. 90, 917-934.
Whitney, M.I., 1985. Yardangs. J. Geol. Educ. 33, 93-96.
Whitney, M.I., Dietrich, R.V., 1973. Ventifact sculpture by windblown dust. Bull. Geol. Soc. Am. 84,
2561-2581.
Wickham, S.S., Johnson, W.H., 1981. The Tiskilwa till, a regional view of this origin and depositional
processes. Ann. Glaciol. 2, 176-182.
Wilford, G.E., Wall, J.A.D., 1965. Karst topography in Sarawak. J. Trop. Geogr. 21, 44-70.
Wilhelmy, H., 1958. Klimamorphologie der Massengesteine. Georg Westerman Verlag, Braunschweig,
p. 238.
Wilson, A.F., 1984. Origin of quartz-free gold nuggets and supergene gold found in laterites and soils -
a review and some new observations. Aust. J. Earth Sci. 31,303-316.
Wilson, I.G., 1971. Desert sandflow basins and a model for the development of ergs. Geogr. J. 137,
180-199.
Wilson, I.G., 1972. Aeolian bedforms - their development and origins. Sedimentology 19, 173-210.
Wilson, I.G., 1973. Ergs. Sediment. Geol. 10, 77-106.
Wilson, L., 1968. Morphogenetic classification. In: Fairbridge, R.W. (Ed.), The Encyclopedia
of Geomorphology. Dowden, Hutchinson and Ross, Stroudsburg, PA, pp. 717-729.
Wilson, L., 1969. Les relations entre les processus gromorphologiques et le climat moderne comme
mrthode de paleoclimatologie. Revue de Grographie Physique et de Grologie Dynamique 11, 301 - 314.
Wilson, L., 1973. Variations in mean annual sediment yield as a function of mean annual precipitation.
Am. J. Sci. 273, 335-349.
Willets, B., 1983. Transport by wind of granular materials of different grain shapes and densities.
Williams, G.E., 1973. Late Quaternary piedmont sedimentation, soil formation and paleoclimates in arid
south Australia. Z. Geomorphol. 17, 102-125.
Williams, J.J., Butterfield, G.R., Clark, D.G., 1994. Aerodynamic entrainment thresholds: effect
of boundary layer flow conditions. Sedimentology 41,309-328.
Williams, M.A.J., 1968. Termites and soil development near Brocks Creek, Northern Territory. Aust. J. Sci.
31, 135-154.
Williams, M.A.J., 1978. Termites, soils and landscape equilibrium in the Northern Territory of Australia.
In: Davis, J.L., Williams, M.A.J. (Eds), Landform Evolution in Australasia. Australian National
University Press, Canberra, pp. 128-141.
Williams, M.A.J., 1994. Cenozoic climatic changes in deserts: a synthesis. In: Abrahams, A.D., Parsons,
A.J. (Eds), Geomorphology of Desert Environments. Chapman & Hall, London, pp. 644-670.
Williams, P.J., Smith, M.W., 1989. The Frozen Earth. Fundamentals of Geocryology. Cambridge
University Press, Cambridge, p. 306.
Williams, P.W., 1969. The geomorphic effects of ground water. In: Chorley, R.J. (Ed.), Water, Earth and
Man. Methuen, London, pp. 269-284.
Williams, P.W., 1972. Morphometric analysis of polygonal karst in New Guinea. Bull. Geol. Soc. Am. 83,
761-796.
Williams, R.B.G., Robinson, D.A., 1981. Weathering of sandstone by the combined action of frost and salt.
Willis, B., 1936. East African Plateaus and Rift Valleys - Studies in Comparative Seismology. Carnegie
Institute, Publication, Washington, DC, p. 470.
Winkler, E.M., Singer, P.C., 1972. Crystallisation pressure of salts in stone and concrete. Bull. Geol. Soc.
Am. 83, 3509-3514.
Wintle, A.G., 1993. Luminescence dating of aeolian sands - an overview. In: Pye, K. (Ed.), Dynamics and
Environmental Context of Aeolian Sedimentary Systems. Geological Society of London, London,
pp. 49-58, Special Paper 72.
Wirthmann, A., 1964. Die Landformen der Edge-Inse6 in Stidost-Spitzbergen. Ergebnisse der Stanforland-
Expedition 2, 53.
Wirthmann, A., 2000. Geomorphology of the Tropics. Springer, Berlin, p. 314.
Wolman, M.G., Gerson, R., 1978. Relative scales of time and effectiveness of climate in watershed
geomorphology. Earth Surf. Processes 3, 189-208.
Wood, A., 1942. The development of hillside slopes. Proc. Geol. Assoc. 53, 128-140.
Woodward, J.C., 1995. Pattern of erosion and suspended yield in Mediterranean fiver basins. In: Foster,
I.D.L., Gurnell, A.W., Webb, B.W. (Eds), Sediment and Water Quality in River Catchments. Wiley,
Wopfner, H., 1978. Silcretes of northern South Australia and adjacent regions. In: Langford Smith, T.
(Ed.), Department of Geography, University of New England, pp. 93-141.
World Glacier Monitoring Service, 1989. World Glacier Inventory, IAHS, (ICSI). UNEP-UNESCO.
Worsley, P., 1974. Recent "annual" moraine ridges at Austre Okstindebreen, North Norway. J. Glaciol. 13,
265-277.
Wright, V.P., Tucker, M.E., (Eds), 1991. Calcretes, Reprint Series of the International Association
of Sedimentologits, Vol. 2. Blackwell, Oxford, p. 352.
Yair, A., Gerson, R., 1974. Mode and rate of escarpment retreat in an extremely arid environment (Sharm
el Sheikh, southern Sinai Peninsula). Z. Geomorphol. 21,106-121.
Yair, A., Lavee, H., Bryan, R.B., Adar, E., 1980. Runoff and erosion processes and rates in the Zian Valley
badlands, northern Negev, Israel. Earth Surf. Processes 5, 205-225.
Yazawa, D., Toya, H., Kaizuka, S., 1971. Alluvial Fans. Kokon Shoin, Tokyo, p. 318.
Young, A., 1964. Slope profile analysis. Z. Geomorphol. Suppl.bd. 5, 17-27.
Young, A., 1970. Slope form in part of the Mato Grosso, Brasil. Geogr. J. 136, 383-392.
Young, A., 1972. Slopes. Oliver and Boyd, Edingurgh, p. 278.
Young, A., 1976. Tropical Soils and Soil Survey. Cambridge University Press, London, p. 468.
Young, R.A., 1985. Geomorphic evolution of the Colorado Plateau margin in West-Central Arizona:
a tectonic model to distinguish between the causes of rapid, symmetrical scarp retreat and scarp
dissection. In: Morisawa, M., Hack, J.T. (Eds), Tectonic Geomorphology. Allen and Unwin, Boston,
pp. 261-278.
Young, R.M., 1987. Salt as an agent in the development of cavernous weathering. Geology 15,962-966.
Yuan, D., 1987. New observations on tower karst. In: Gardiner, V. (Ed.), International Geomorphology
1986, Proceedings of First International Conference on Geomorphology, Part II. Wiley, Chichester,
pp. 1109-1123.
Zachar, D., 1982. Soil Erosion. Elsevier, Amsterdam, p. 547.
Zakaria, A.S., 1977. Controls upon the mineral outputs from three small catchments in New England.
Unpublished Ph.D. thesis, University of England, Australia.
Zaruba, Q., Mencl, V., 1969. Landslides and their Control. Elsevier, Amsterdam, p. 213.
Zazo, C., Dabrio, C.J., Borja, F., Goy, J.L., Lezine, A.M., Lario, J., Polo, M.D., Hoyos, M., Boersma, J.R.,
1999. Pleistocene and Holocene aeolian facies along the Huelva Coast (southern Spain): climatic and
neotectonic implications, Geologie en Mijabouw (en prensa).
Zekri, S., Albisu, L.M., Aragties, R., Herrero, J., 1990. Impacto econ6mico de la salinidad de los suelos en
la agricultura de Bftrdenas I. Comunicaciones Instituto Nacional de Investigaciones Agrarias, Madrid,
Serie Economia 36, p. 129.
Zhende, Z., 1984. In: E1-Baz, F. (Ed.), Aeolian landforms in the Taklimakan Desert. Martinus Nijhoff,
The Hague, pp. 133-144.
Zoltai, S.C., 1971. Southern limit of permafrost features in peat landforms, Manitoba and Saskatchewan.
Geol. Assoc. Can. Spec. Paper 9, 305-310.
Zuidam, R.A.van, 1976a. Geomorphological Development of the Zaragoza Region, Spain. International
Institute for Aerial Survey and Earth Sciences (ITC), Enschede, p. 211.
Zuidam, R.A.van, 1976b. Periglacial-like features in the Zaragoza region, Spain. Z. Geomorphol. 20,
227-234.
Subject Index
abrasion, 62 bench, 558

abrasion process, 63 bergschrund, 59
abrasion rate, 63 biological, 526
active dunes, 675 biological modifications, 687
active layer, 148 block-fields, 185
active rock glaciers, 190 block-slopes, 185
actualism, 5 block-streams, 185
aeolian action, 169 blowouts, 447
aeolian action, 696 bombardment, 412
aeolian dominion, 236 bornhardts, 318, 566
aeolian erosion, 418 boulder hamada, 259
aeolian erosion landforms, 418 boulder inselbergs, 566
alas, 207
alas valleys, 207 calcrete, 276
albedo effect, 233 caliches, 276
alluvial fan, 357, 478, 549 cambering, 561
alluvial-fan development, 364 cascades, 545
alluvial fan morphology, 361 castle kopje, 319, 567
alpine permafrost, 146 cataracts, 545
alterite, 516 catastrophic flooding, 595
alveoles, 255 catastrophism, 5
ancient pingos, 636 causes of aridity, 232
animal activity, 393 causes of the paleoclimate fluctuations, 685
annual moraines, 104 caves, 255
Antarctic, 36 cavetto forms, 72
anthropic action, 557 changes in channel type, 475
areic drainage, 290 changes in volume, 467
ar~tes, 78 channel flows, 374
arid period, 643 channel metamorphosis, 659
arroyos, 310, 651 chelation, 510
asymmetric valleys, 202 chemical weathering, 159
atmospheric dust, 612 chiflones, 417
aufeis, 148 circles, 172
avalanches, 162 circular and stepped gilgai, 269
azonal, 25 cirque glaciers, 47, 626
cirques, 75
badland, 333 classification of dunes, 437
barchan, 438 cleaved boulder, 243
basal slipping, 50 cliff-top dunes, 452
basins, 83 climate, 526
754 Subject Index
climate changes, 349, 601,658 debris flows, 162, 372, 478

climate fluctuations in periglacial regions, 638 debris slides, 162
climatic factor, 366 deflation, 261, 418
climatic interpretation, 650 deflation basins, 423
climatic parameters, 19 deforestation, 498
climatogenic geomorphology, 26 degradation of soil, 492
climatomorphological zones, 27 deltas, 123
climbing dunes, 451 depositional landforms, 624
closed depressions, 561 desert dust, 452
coalescent cirques, 78 desert glaze, 273
coastal dunes, 450 desert lakes, 484
cold desert, 233 desert pavements, 259
cold ice, 42 desert rivers, 659
cold ocean currents, 232 desert varnish, 270
collapsed pingos, 207 desertification, 492
complex response, 606 desiccation cracks, 263
complex slopes, 321 different morphoclimatic areas, 26
composite slopes, 321 diffluence, 85
compressive flow, 52 discharge, 106
concentration zones, 247 discontinuous permafrost, 146
cone karst, 577, 578 disintegration moraines, 104
constitution and structure, 524 dissolution, 463
continental ice sheets, 43 dissolution of carbonates, 511
continentality, 232 dolocretes, 278
continuous permafrost, 146 dome dunes, 439
convergence, 403 domed inselbergs, 566
convergence of forms, 261,263 domes, 398
costra calcfirea, 276 dominant processes, 433
covered pediment, 345 dormant dunes, 675
crag and tail, 87 double planation surfaces, 562
craters, 393 drainage basin, 361
creep, 413, 540, 542 dreikanter, 418
creeping, 412 dropstones, 43
crests, 435 drumlins, 100
crevasses, 55 dune formation, 434
critical stream power threshold, 364 dune systems, 670
cross-sectional profile, 82 dunes anchored by vegetation, 447
cryokarst, 202 duricrusts, 275
cryopediments, 181 dust devils, 454
cryoplanation terraces, 181 dust storms, 236
cryostatic pressure, 156 dynamic geomorphology, 13
cryoturbations, 156
crystal growth, 249 earthquakes, 557
echo dunes, 451
dambos, 547 Eh, 509
dammed lakes, 130 einkanter, 418
dams, 477 endogenous drainage, 289
Davis, 8 endorheic, 289
dayas, 389 engineering geology, 583
De Geer moraines, 104 engineering problems in permafrost areas, 215
Subject Index 755
englacial debris, 91 fluvial systems, 644, 692

environmental change, 601 flyggberg, 73
episodically active dunes, 675 folds, 54
equifinality, 171,263 foliation, 54
equilibrium line, 39 forbes bands, 55
erg, 429 fossil rock glaciers, 192
erosion rates, 338 fossil wedges, 636
erosive basins, 389 fracturing, 64
erosive morphologies, 624 fragment mobilization, 66
erratic blocks, 94 free face retreat velocity, 330
eskers, 118 friction cracks, 70
etch surface, 350 frost action, 151
etched plains, 350, 561 frost cracking, 157
etching, 419 frost-creep, 160
evaporation, 288 frost heaving, 153
exhumation, 573 frost-pull hypothesis, 154
exhumed plains, 562 frost-push hypothesis, 154
exogenous drainage, 289 frost sorting, 158
exoreic drainage, 291 fussflS.che, 345
exposed glaciomarine bottom muds, 123
extending flow, 53 gassi, 441
extrinsic factors, 369 gelifluction, 160
gelifluction benches, 180
fabric, 95 gelifluction lobes, 180
faced spurs, 83 gelifluction sheets, 180
facets, 418 gelifluction slopes, 180
factors, 552 gelifluction streams, 180
factors affecting generation, 525 geocryology, 145
falling dunes, 452 geological environmental change, 601
fan apex, 361 geological structure, 349
fan area, 364 geomorphic threshold, 605
fan gradient, 364 ghourds, 443
fan shape, 364 giant desiccation fissures, 265
fans of arid zones, 360 glacial ice, 148
fans of temperate zones, 360 glacial karstic topography, 105
faults, 54 glacial polishing, 70
feeder channel, 361 glacial tables, 38
firths, 89 glacial valleys, 78
fixed dunes, 676 glacioeustacy, 630
fjord, 87 glaciolacustrine beaches, 123
flash floods, 374 glaciological factors, 63
flat-bottomed valleys, 202 glaciotectonic, 116
flatirons, 326 glaciotectonic activity, 137
floods, 295, 480 glaciotectonite, 95
fluid threshold, 413 glacis, 343
fluted moraines, 93, 100 gnammas, 256
flutes, 418 gossan, 588
fluvial activity, 167 grain size distribution, 95
fluvial processes, 374 grazes lit6es, 193
fluvial sedimentary landforms, 546 grooves, 69, 418
756 Subject Index
ground moraines, 100 intersection point, 361

gypsum crusts, 283 interstadials, 617
intraglacial channels, 105
intrazonal, 25
haboob, 236, 456
involutions, 156
half-oranges, 537
ion exchange, 510
haloclasty, 462
isobases, 632
hamada, 259
hanging valleys, 83
hazards, 592 j6kulhlaups, 106, 132, 136, 622
headward-eroding gullies, 361
helicoid forms, 419 kame topography, 121
hill and lake topography, 86 kame complex, 121
horns, 78 kame delta, 120
humans, 601 kame terraces, 121
human-induced subsidence, 486 kames, 120
hummocks, 173 karstification, 589
hummocky moraine, 104 kegelkarst, 577
hydration, 250, 510 kettle holes, 121
hydrographs, 294 kettle topography, 121
hydrology of lakes, 393 kettled sandur, 121
King, 12
knick-points, 346, 543
ice avalanches, 127
ice caps, 43
ice-dammed lakes, 130 labile landforms, 607
ice dumped moraines, 102 lacustrine phase, 643
ice lenses, 148 lahars, 135
ice needles, 155 lake expansions, 664
ice sheet of Greenland, 37 land use, 655
ice sheets, 36 landforms, 558
ice shelves, 43 landslides, 372
ice veins, 149 lateral moraines, 97
ice wedges, 157 lateral planation, 351
icebergs, 125 lateral sorting, 158
icings, 148 laterite, 523, 528, 588
~mpact craters, 393 laterite genesis, 529
impact threshold, 413 laterite profile, 528
inactive dunes, 676 lee dunes, 452
reactive rock glaciers, 190 lee-side slopes, 436
incised channels, 361,375 lineal infrastructures, 220
inclination, 556 linear dunes, 440
infiltration capacity, 296 lithologic factor, 366
influence of climate, 246 lithology, 349
inselbergs, 537, 561,562, 565, 691 littoral sebkhas, 386, 403
msolation weathering, 243 loess, 170, 457
internal deformation, 50 longitudinal profile, 83
lnterpluvial, 643 long-term climate changes, 612
rater-rill areas, 297 long-term suspension, 415
inter-fill erosion, 300 lunette, 389, 424, 452
Subject Index 757
mantled pediment, 345 origin of the drumlins, 100

marginal channels, 105 origin of trenching, 654
marginal discharge, 94 origins, 247, 388
mass balance in a glacier, 37 orographic effect, 232
mass displacement, 156 ouadis, 343
mass movements, 160, 214, 551 ouads, 652
maximal extension during the Pleistocene, 37 outlet glaciers, 47
mechanical behaviour of regolith, 552 outwash, 121
mechanical sorting, 158 overgrazing, 498
medial moraines, 97 overland flows, 375
megaripples, 416
mesas, 558 pack ice, 43
microforms, 397 paleoclimatic information, 679
models, 417 paleolakes, 661
modified saltation, 414 pallid zone, 528
mollisol, 148 palsas, 176
morphoclimatic regions, 28 pan, 383
morphologic classifications, 43 parabolic dunes, 449
morphometry, 76 patterned ground, 171,262
mottled clay, 528 pedestals tables, 38
moulins, 105 pediment morphology, 345
Mount St. Helens volcano, 135 pediment origin, 350
mountain and basin deserts, 236 pediments, 345, 480
movement of glaciers, 50 pediplains, 356
mud curls, 264 pediplanation cycle, 356
pedogenic caliches, 281
mud flows, 162, 373
Penck, 12
mudslides, 162
pergelisol, 146
periglacial, 145
nebkhas, 447
periglacial breaking of the bottom, 65
neotectonic, 349
periglacial environments, 634
net balance, 39 periglacial processes, 151
nets, 173 P-forms, 72
network dunes, 444 pH, 509
Nevado del Ruiz volcano, 135 physical characteristics, 410
niches, 165 physical-chemical processes, 249
nlval processes, 162 piedmont glacier, 47
nlvation, 78, 162, 165 piedmonttreppen of pediplains, 356
nlvation cirques, 165 plngos, 150, 177
nlvation niches, 196 pinnacles, 402
nlvation ridges, 197 piping, 303, 393, 473
nonpedogenic caliches, 282 plpkrake, 155
nubbins, 155 pits, 418
pitted outwash, 121
ogives, 55 placer deposits, 141
oil and gas pipelines, 222 plains, 82
on laterites, 558 plains of glaciolacustrine origin, 123
oriented lakes, 204 plateau basalts, 317
origin of particles, 411 platforms, 558
origin of p-forms, 112 playa, 383
758 Subject Index
plots, 535 sabkha, 383

ploughing blocks, 161 saline blisters, 400
pluvial lakes, 667 saline polygons, 400
pluvial period, 643 saline ramps, 400
polar permafrost, 146 salinization, 494
polygons, 173 salt source, 396
position inselberg, 346 salt weathering, 245
potholes, 112 siltation, 414
powder avalanches, 164 salts, 247
precipitation, 285 sand sea, 429
proglacial channels, 106 sand sheets, 429, 444
protalus ramparts, 197 sand strips, 444
pure siltation, 414 sand wedges, 157
push moraines, 94, 104 sandur, 111, 121
pyramidal dunes, 443 sandur plains, 121
sandur valleys, 121
saprolite, 516
quaternary ice sheets, 620 savanna, 502
quick clays, 140 scarp retreat hypothesis, 573
screes, 160, 637
sea level zones, 634
rain forest, 502
sea lochs, 89
rainfall, 554
sebkha, 383
rainsplash, 300
sediment supply processes, 371
raised beaches, 123
sedimentary depressions, 393
ramblas, 651
segmented fans, 369
rapids, 543
sensitivity, 606
regolith, 516, 688
sequence, 648
regolith characteristics, 583 s6racs, 56
relict, 192 shadow dunes, 447
relict dunes, 676 sheet floods, 351,374
relict weathering profiles, 688 sheeting, 65, 317, 505
relief inversion, 561 shelters, 255
reptation, 413 shield and platform deserts, 236
resistant inselberg, 346 short-term climate changes, 612
reversing dunes, 440 short-term suspension, 415
reworking processes, 375 shoulders, 82
rill erosion, 300 sichelwannen, 72
rills, 297 sieve deposits, 375
ripples, 415 sif, 441
roches moutonn6es, 73 silcrete, 283
rock avalanches, 372 silica glaze, 273
rock basins, 74 simple slopes, 321
rock drumlins, 74, 86 sinuous channels, 543
rock-falls, 160 skjaergard, 89
rock glaciers, 189, 636 slab avalanches, 165
rock varnish, 270 slide, 552
rocky hamada, 259 slope failures, 592
rogen moraines, 102 slopes, 538
run-off, 296 slopes in arid zones, 317
Subject Index 759
slopes on crystalline rocks, 317 thaw lakes, 204

slopes on stratified rocks, 319 theories on general circulation, 685
sluggish landforms, 608 thermal classification, 42
small basins, 536 thermal expansion, 250
snow avalanches, 210 thermocirques, 196
sodification, 494 thermo-erosion niches, 199
soil-loss quantification, 535 thermokarst, 202
solifluction, 160 thresholds, 83
solubility of the gypsum, 465 till, 94
solution water, 510 till sheets, 100
spaces around clasts, 156 topographic factors, 364
speed of the snow-ice transformation, 41 topography, 526
spheroidal weathering, 518 tors, 181,566
sporadic permafrost, 148 tower karst, 577, 580
spring mounds, 400 transfluence, 85
star dunes, 443 transforms into glacier ice, 41
stepped cirques, 78 transition zone, 528
stepped levels, 652 transpiration, 288
steps, 174 transverse dunes, 438
stone lines, 539 trap, 317
straie, 68 triangular slope facets, 199, 326, 647, 648
straight channels, 542 tripartite slopes, 199, 326, 648
strandflats, 89 tropical karst, 576
stratification, 54 tropical rivers, 536
streamlined spurs, 74 tumuli, 398
striations, 68 tunnelling, 303
stripes, 174 turmkarst, 577
structures, 525 types, 247
subglacial channels, 105
subglacial debris, 91 uplifting, 412
subglacial melting, 92, 93 U-shaped valleys, 202
submarine permafrost, 146
sulphate aerosols, 612 valley glaciers, 626
supraglacial channels, 105 valley glaciers of the alpine type, 47
supraglacial debris, 91 valley-transverse moraines, 102
surging glaciers, 52 varves, 123
suspended load, 536 vegetal cover, 367, 655
vegetation, 287, 556
tafonis, 255 ventifacts, 418
talik, 148 vertical sorting, 158
talus, 160, 637 volcanic activity, 612
talus cones, 183 volcanism and glaciers, 135
talus flatirons, 326, 648
talus slopes, 181 warm deserts, 233
tectonic basins, 388 warm ice, 42
tectonic factor, 367 water capillary, 246
terminal moraines, 97 water erosion, 261, 471
termitaria, 507 weathering, 377
termites, 540 weathering basal surface, 516
thaw collapses, 207 weathering degrees, 512
760 Subject Index
weathering depth, 516 whalebacks, 74

weathering front, 516 wind action, 377, 488
weathering pits, 256 windward slopes, 435
weathering processes, 243
weathering products, 513 yardangs, 420
weathering profiles, 65, 516
weathering rates, 523 zanj6n, 578
weathering zones, 518 zibars, 446
wedges, 149 zonal, 25
wetting and drying, 462 zonal factor, 232
wetting and drying weathering, 243 zweikanter, 418

Climatic Geomorphology PDF

Uploaded by

Copyright:

Available Formats

Climatic Geomorphology PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Climatic Geomorphology PDF

Uploaded by

Copyright:

Available Formats

ELSEVIE R 8.\'. ELSEVIER Inc.

ELSEVIER Ltd ELSEVIER ltd

Working together to grow

4 Glacial transport and sedimentation 91

5 Fluvioglacial ero ion and sedimentation 105

6 Geomorphology applied to glacial regions 125

7 The periglacial environment 145

8 Peri glacial landfom1s 17 1

9 Some aspects on applied geomorphology in perigl acial regions 209

l 0 T he arid zone 229

II Weathering proce. ses and resulting fonn · 241

13 T he action of water in arid regions 285

14 Slopes in arid zones 3 17

15 The arid region piedmonts: glacis and alluvial fans 343

I6 Desert Jakes: playa Jakes and sebkhas 383

17 Aeolian processes and erosive landforms 409

18 Aeolian accumulations 429

19 Applied geomorpho logy and arid regions 46 1

20 The humid tropics: weathering and laterites 50 1

21 Tropical land forms 533

22 Applied geomorphology in tropical regions 583

23 Environmental change 601

24 Climate change in g lacial and periglacial regions 6 15

25 Climate change in arid and tropical humid regions 643

Subject Index 753

by the discipline of geotectonics after micro-structural analysis was developed, which

2. History of climatic geomorphology

2.1. Geomorphology before the second half of the 19th century

2.2. The dawning of climatic geomorphology

2.3. The impact of landscape evolution models on climatic geomorphology

Figure 1.2. Karsticcycle (Grund, 1914).

2.4. Process geomorphology

i~AS~ L E V ~ A i LE~yFLO0~. ~ / ,~,~"~~"~-~ ,

PENCK ACCORDING TO VON ENGELN

~T Valley deepening to obtain

LTI~UDE of MAIN V,~.LLEYs . ~ . "' / _

2.5. Structure of climatic geomorphology

On 22-23rd September 1926, in Dtisseldorf, Germany, researchers held a meeting on

however, ideas of climatic geomorphology lead directly to climatogenic geomorphology,

2.6. The development of climatic geomorphology

on experimental basins. Studies on fluvial transport were undertaken by Douglas, Spencer

2.7. Applications and future trends in climatic geomorphology

3. Climatic geomorphology: processes and morphoclimatic zonation

Chemical weathering Frost action

IO~ -12~ IOOF -12oC

Weathering regions Pluvial erosion

Mass movement Wind action

Worldwide scale approaches to climate and vegetation controls on weathering intensity

203 178 152 127 102 76 50 25 cm

. ~.-.Z. Climatic r6gimes and

/I>2000 [~IIII1M \ ;~YJ e ~ "

Figure 1.9. World distribution of fluvial erosion (Fournier, 1960).

\ @a"o o ,~, LATiTUD E'~"X,~/~ TROPICAL

(a) First-order morphogenetic regions, considered as non-seasonal by Beckinsale and

4. The zonal concept in climatic geomorphology

From a planetary perspective, the variations occurring at the lithosphere-atmosphere

5. The main morphoclimatic zones

(1) Cold Zone.

Subtropical zone of pediment