~-r :'L

"

" . j~,

ELSEVIER

Geomorphology 11 (1995) 215-226

Landslides triggered by earthquakes and their relations with faults

and mountain slope geometry: an example from Ecuador
Alessandro Tibaldi, Luca Ferrari, Giorgio Pasquar~
Dipartimento di Scienze della Terra, Universit~ degli Studi di Milano, Milano, 20133, Italy
Received 31 January 1994; revised 10 June 1994; accepted 27 July 1994

Abstract
A large number of landslides occurred during two seismic events (respectively, 6.9 and 6.1 Msw) on 5 March 1987 in the
Ecuadorian Andes. These landslides have been mapped, digitized, and coregistered with topography at 1:50,000 scale. Geometry
of coseismic and Holocene faulting has been assessed integrating field and geophysical data. Landslide distribution and Holocene
tectonic features have been compared with earthquake foci, geological deposits, slope area, inclination and orientation, and
vegetation cover. The macro-seismic field deduced by landslide distribution is ellipse-shaped with the major axis striking NNE
and coinciding with the trend of the coseismic and Holocene faults. Landslide distribution also shows a correlation with respect
to the dip-direction of these faults and the orientation of mountain slopes. Slopes parallel to the coseismic fault planes, sloping
towards WNW, are almost unaffected by landsliding, while this increases on the slopes at high angle to the fault plane and lying
along its strike. If these results could be confirmed in other areas, a method of reconstruction of geometry of seismogenetic faults
from topographic effects appears feasible.

1. Introduction

Two seismic events with magnitude of surface waves

(Msw) 6.9 and 6.1 took place on 5 March 1987 in the
northeastern part of the Ecuadorian Andes, here known
as Cordillera Real (CR) (Fig. 1). During that day
hundreds of casualties were reported in the area, a number which is probably underestimated because the people are scattered in small rural villages and
communications are difficult throughout the region.
Many people died suddenly because of extraordinarily
widespread slope mass movements which affected a
very large area. A few hours after these events, several
other people lost their lives because of mudflows which
swept the main valley floors. These flows were induced
by the collapse of several temporary natural dams built
by the large and high-density landsliding.
0169-555X/95/$09.50 1995 Elsevier Science B.V. All rights reserved

SSDIOI69-555X(94)O0060-3

The large-scale slope denudation which occurred on

that day greatly facilitated the field survey and has
allowed us to recognise and map the geometry and
kinematics of the Holocene and coseismic main faults.
Because the study area is particularly inaccessible, with
steep mountains and dense rain forests, the field survey
was integrated with interpretation of aerial stereophotos and radar images. Stereophotos comprise 75 prints
at 1:80,000 scale, 352 at 1:40,000 scale and 12 at
1:20,000 scale, covering an area of more than 5000
km 2. Photos at 1:40,000 scale were taken before and
after the earthquakes. Radar images were obtained with
a Synthetic Aperture Radar of CLIRSEN (Centro de
Levantamientos Integrados de Recursos Naturales por
Sensores Remotos, Quito) with a Goodyear system.
The band we used is X ( 3.12 cm) and sensor resolution
is 10 meters. Topographical maps have been prepared

216

A. Tibaldi et aLI Geomotphology I I (1995) 215 226

/,

/-

~"

//

COLOMBIA

~\.,
(
/
I

<

'

i /
,/
/'

//
d

//
/
/"

,
!

-~

PERU

" \

,[

~. /

i/

.;

ited from the last main glacial and post-glacial phases.

These indicators comprise up-facing slope breaks,
marked uneroded scarps resembling "knife-cuts", dislocation of river channels, gullies, ridges, crests, subglacial erosion surfaces, or moraine deposits. Seven
main faults with such characteristics have been discovered in the eastern Ecuadorian Andes. The C a y a m b e Chingual Fault ( C C F in Fig. 2) can be followed for 76
km until it reaches the Colombian border. Along the
majority of this fault trace, dislocation of river channels
and gorges can be measured giving a mean value of
300 m. Taking into account that the dates of the last
glacial phases are ! 1,000+ 1000 yr BP for the areas
> 3800 m, and 27,500_+7500 yr BP for the areas
> 3000 m (Clapperton and Vera, 1986), and considcring that the majority of the CCF outcrops between
2700 and 3500 m, its horizontal rate of tectonic motion

Fig. 1. Main morphologicalprovincesof Ecuador and location of the

study area. CP Coastal Plain, CO Cordillera Occidental, IV lnterandean Valley, CR Cordillera Real, SAZ Sub-AudeanZone, AP Amazonian Platform.Smallbox showslocationof Fig. 2, large box locates
Figs. 4 and 5.

for this study on the same scale. Geological data have

been derived from D G G M (1978), Balseca and Ferrafts (1987), INECEL (1988), and Tibaldi (1990),
and geophysical data derived from Barberi et al.
(1988).
Coregistration of these data sets permits a precise
comparison of landslide distribution with slope area,
inclination and orientation, rock lithology, epicentre
distribution, isoseismal patterns, and fault attitude.

/I

_..

"2. i
J
L

~/'
I' Cayambe%~ h, '//

._.p~/~ ~ J ~

fig 10

, L / %{R~v~.t~do~

2. Holocene faults
2. I. Structural field data

The outstanding feature of the Cordillera Real is a

marked and linear family of faults with an overall trend
of about N30 . Field surveys show that these faults are
near-vertical or have a dominant dip towards W N W ,
whatever their age (Pasquar~ et al., 1991; Tibaldi and
Ferrari, 1992). Only a few of these faults show Holocene tectonic activity.
In the high Andes the typical indicator of Holocene
faulting is the dislocation of geomorphic features inher-

....

~4t~13

8
I

25
km

50

Fig. 2. Main Holocene faults of the eastern EcuadorianAndes. CCF:

Cayambe-ChingualFault,LBF La BonitaFault.A, B, and C localities
discussed in the text. Boxes are locations of Figs. 6, 10, and 11.
Location given in Fig. 1.

A. Tibaldi et al. / Geomorphology 11 (1995) 215-226

may be estimated as 11.7 ___3.2 mm/yr. Using the offset

of at least 100 m (from aerial stereophotos) of the
Soche lava flow dated at 9670 yr BP (Hall and Beate,
1991) and considering a 10% error of offset and age
determination, the slip rate is approximately 10.5 ___2
mm/yr. The type of faulting is strike-slip right-lateral,
according to tectoglyphs measured on the fault plane
and the sense of fiver dislocation. The same features
were discovered along the La Bonita fault (LBF in Fig.
2). Here the Holocene rate of fault motion is more
uncertain because we found only two sites where dislocation could be measured.
Southwards, two main Holocene N - S faults were
recognised. At site A in Fig. 2, an escarpment facing
westward is outlined by dislocated glacial morphology
revealing a relative upward motion of the eastern block.
Topographic data indicate along this segment a total
amount of vertical dislocation of about 30 m (a rate of
2.7 + 0.2 m m / y r ) . The fault labelled B in Fig. 2 has an
opposite dip with a vertical component of motion of
the same order of magnitude.
A fault segment which probably moved during the
March 1987 earthquake (C in Fig. 2) has been recognised by an abrupt, fresh cut in the mountain slope
which creates an up-hill facing escarpment. The escarpment strikes NNE and the western block is up-raised.
These observations coincide with some possible precursory deformations which were observed some
months before the March 1987 earthquake. These
deformations consisted of a few centimetres net dislocation of small erosional grooves in bedrock stream
channels (Fig. 3). Fault traces strike NNE and show
reverse fight-lateral dislocations with up-raising of the
western block.
Another 20 minor Holocene faults have been discovered (Fig. 2). Field surveys confirmed that practically all the detected fault escarpments were due to
strike-slip or reverse faulting or a combination of the
two components of motions. Dip of minor fault planes
is also mostly towards W N W or W.
2.2. G e o p h y s i c a l d a t a

Epicentres of 93 shallow earthquakes extracted from

the CERESIS (1986), OAQ ( 1981 ) and National Oceanic Atmospheric Administration catalogues for the
years 1903-1987 AD were also coregistered, as well
as isoseismal pattems (Barberi et al., 1988). Wide-

217

Fig. 3. Possible precursorydeformations which were observedsome

months before the 5 March 1987 earthquakes consisting in a few
centimeters net dislocations of small erosional grooves in bedrock
stream channels. Fault traces strike NNE and show reverse rightlateral dislocations with up-raising of the western block. North is on
the upper side of the picture ( Photo by G. Pasquar~).

spread seismic activity comprising 7 large earthquakes

(estimated intensity of Mercalli scale > IX) occurred
inside and near the study area mostly along a belt running NNE (Fig. 4, OAQ, 1959). The focus of the two
large events of 5 March 1987 (b and c in Fig. 4) lie at
a d e p t h o f 14km (Msw = 6 . 1 ) and 10km (Msw = 6.9),
while their epicentres show a N N E - S S W alignment,
coincident with the orientation of the Holocene faults
recognised in the field. A similar alignment is shown
by the microseismicity recorded during the periods January-April 1981 (1 in Fig. 4) and December 1986March 1987 (2 and 3 in Fig. 4) (Barberi et al., 1988).
Also the isoseismal pattern of the 1987 earthquake
shows an elliptical shape with a major axis striking
NNE (Fig. 5). The focal mechanisms of some of these
events are shown in Fig. 4. The events of 5 March, 1987

218

A. Tibaldi et al. / Geomorphology 11 f 1995) 215-226

/
/,
/

,)
/S

iv

CO

f
)
t

"''/f~'
/ /"

OR /

/
/

l
j/

/ J"

\,

,,i /

QUITO

SAZ

/
/

X o

/~

X
/
/'

/'z

"1i

"f

'ltt:-

:<,<]

ol

79

774

77

Fig. 4. Epicentre distribution map of period 1903-1987. Numbers I, 2 and 3 refer to microseismicity analyses discussed in the text. Letters refer

to focal mechanisms: b and c after Barberi et al. ( 19881, e after Woodward and Clyde ( 19811, Schmidt projection, lower hemisphere; black
arrows show the horizontal direction of the P-axis. CO Cordillera Occidental, 1V lnterandean Valley, CR Cordillera Real, SAZ Sub-Andean
Zone. Box shows location of Fig. 8. Location given in Fig. 1.

(USGS, 1987) show a dominant thrust motion with a

right-lateral strike-slip component (b and c in Fig. 4)
along a W dipping fault. Centroid Moment Tensor solutions for the same events (Dziewonski et al., 1988)
show nearly pure thrust motions but very similar P and

T axes trending E - W and vertical respectively. The

fault dip is westward. The event of 5 August, 1949
(Woodward and Clyde, 1981 ) (e in Fig. 4) shows an
almost pure transcurrent motion along a NE striking
plane with a northwestward high angle dip.

A, Tibaldi et al. / Geomorphology I I (1995) 215-226

219

/
/

/ /

~-___~,~-~,_~,,-.-.-...-...~.,

/,

i/

/
/

//

,,o

--

o
l

lo

2o

30Kin

Fig. 5. Isoseismic map of the 5 March, 1987 earthquakes (after Barberi et al., 1988). The degree of intensity is indicated for each line. Black
boxes represent the two main shock epicenters with magnitude (Msw). The smaller sinuous lines are three isolines encircling areas of landslide
different density ( 1-39%, 40-59% and 60-100%). Location given in Fig. 1.
3. L a n d s l i d e s a n d t o p o g r a p h y
In the study area several small landslides of rock fall
and soil slide type are located along already k n o w n

major active faults and are strictly linked to the fault

traces (e.g. in Fig. 6). This distribution is consistent
with the presence of active tectonic motions along the

220

A. Tibaldi et al. / Geon oq~holog v 11 (1995) 215-226

,,,..

--~

..

......

17 ."Tv

5 Z;/i,~

.....

/e,

'- ......

faults which create dislocations in rock walls. Exhumation of steep surfaces coinciding with the fault plane
creates the conditions for the occurrence of rock falls.
For example, the rock fall deposit A in Fig. 6 is due to
right-lateral strike-slip sllifting of the southern lava
block which is no longer supported by the former
adjoining northern lava block. Different degrees of vegetation colonisation on the various landslide deposits
mark various collapse episodes repeated through time.
In order to distinguish between these landslides and
those produced by the 5 March 1987 seismic events,
all the area was studied on pre- and post-earthquake
photo series. The March landslides are also easily identified on the photos because of the high difference in
reflectance between the exhumed rock and the surrounding jungle.
Several thousands of landslides affected an area of
about 2000 km 2, depicting a catastrophic scenario. The
presence of such a large number of collapses may be
referred to a series of contemporaneous causes. First of
all, a particularly intense rainfall regime affected the
area in the four weeks preceding the earthquakes. The
precipitation during February 1987 was heavy, with a
value of 600 mm, which probably induced soil saturation. Such a value is about twice the average monthly
precipitation recorded in the area during the years
1975-1984 (Fig. 7). Other causes which could have
contributed to the exceptional phenomenon are: the
relatively large seismic magnitude of the two events
which occurred within a short time interval, the abun]
6004

~OO-(4:~ !,;38Z~

1975 84

j ~
~-- 2004

Fig. 6. Landslides, dragging effects on tributaries, and geomorphic

and geological indicators of Holocene motions along the CayambeChingual Fault. Note that distortion of the hydrographic pattern is
consistent with the right-lateral strike-slip motions detected on the
fault plane tectoglyphs. 1 Latest Pleistocene-Holocene lava flow: 2
fault trace; 3 triangular facets; 4 landslide scarp; 5 major landslide
scree-tongue. "A" is a site discussed in the text. Location given in
Fig. 2.

--

100~

oi

. . . . .

Fig. 7. Average monthly precipitation recorded at two stations within

the study area (San Raphael for 1975-1984, full triangles, and Rio
Salado for 1977-1984, open squares). The unusual precipitation of

February 1987 is indicated.

Ct

Fig. 8. Isolines of the percentage of slid area measured with grid cells of 1 sq. km (0, 1, 4, 7, 10, 20, 40, 60, 80, 100%). Areas encircled by dashed lines and marked C L are cloud
covered. The black dot is the summit vent of Reventador volcano. Location given in Fig. 4.

~f

CI.

to
to

to
o~

to

t~

e~

222

,4. Tibaldi et al / G e o m o r p h o l o g y

dance of pyroclastic deposits, the high degree of rock

fracturation and the high average slope steepness.
Sliding usually started from the uppermost part of
the mountain slope reaching the portion of the valley
floor occupied by the river channel. The denuded area
is strongly elongate with an aspect ratio ( length/width:
l/w) mostly between 3 and 5. The average volume of
these landslides is about 60 X 103 ms. Larger landslides
were also able to cross the widest valley floors reaching
the opposite valley side. These landslides show an average l/w = 2.5, and an average volume of about 2 x l ()"
m 3"

Three main types of landslide deposits have been

recognised. One is composed of a fine matrix, with
granulometric fractions of clay, silt and sand, supporting angular clasts and boulders up to several metres in
diameter. The collapse surface is planar or gently
curved. This type corresponds to the collapse of soil
covering, talus, and bed-rock and should coincide with
the rock slide type of Keefer (1984). A second type is
made of clay, silt and sand, coming from the collapse
of soil sheets one to a few metres thick. These features
correspond to disrupted soil slides and present basal
shear surfaces mostly formed at soil-bedrock contact.
A third type is only made up of angular clasts and
boulders corresponding to rock fall. Disrupted soil
slides are the most represented type, having been favoured by the widespread presence of pyroclastic fall
deposits resting on bedrock or on paleosols, and
induced by several phases of explosive volcanism in
the surrounding volcanic centres (Fig. 2). This situa-

I0
F-

"7
4 ,7

z<

0 0 40

80

120 160 200 240

SLOPEORIENTATION

280 320 360

Fig. 9. H i s t o g r a m representing the percentage of landslided area ( left

y-axis) for each slope orientation (x-axis expressed as slope dip
azimuth ). The segmented line gives the percentage of area ( right yaxis) for a given slope orientation in respect to the total study area.

/ I ~ 199,5) 2 1 5 - 2 2 6

tion is similar to the 1960 Chile earthquake which was

accompanied by thousands of disrupted soil slides,
though they were restricted to deforested slopes
( Wright and Mella, 1963 ).
The limit of each landslide was digitized and the
affected area was calculated, using a grid cell of 1 km 2
( Fig. 8). The isolines, expressing percentage slid area
in respect to the cell area, are incomplete in the northwestern part of Fig. 8 because post-earthquake photos
were not available there. Analysis of Landsat images
of late 1987 shows that landslide frequency rapidly
decreases in that direction (Fig. 5). The resulting isoline map defines an elliptic-shaped zone with a NNE
trending major axis. The coseismic landslides are
observable along the NNE axis for about 70 km. Such
a length is comparable with the fault rupture length
expected from the magnitude of the events according
to the curves by Slemmons (1977). Also the size of
the area affected by landslides (2 000 km 2) is consistent with the earthquake magnitude, when compared
with data by Keefer (1984), which report a range from
0 km 2 in a M = 4 earthquake to 500,000 km 2 in a
M = 9.2 earthquake.
In more detail, the limit of the area unaffected by
landslides is asymmetric, with a rectilinear SE boundary and a more complicated opposite side. The shape
of the latter, which can be better appreciated in Fig. 5,
is given by a segmented line trending N-S in the western zone and NE in the northwestern zone. The area
affected by a landsliding percentage lower than 10
shows a regular pattern following the 0 isoline (Fig.
8). The areas included within the isoline 10 tend to
widen symmetrically in correspondence with the central zone depicting a NNE elongated symmetric ellipse.
With the exception of the largest landslides involving bed-rock, which are less sensitive to surface characteristics, landslide distribution data have been
compared with topography and geology in order to
ascertain what really determines the isoline shape. The
results indicate that the isoline shape is principally the
expression of the seismic shake from the following
arguments: (a) in the study area, variations of slope
above an inclination of 20-30 did not influence the
landslide distribution, (b) the average slope steepness
of the collapsed area is almost equal to the conterminous zones and higher than 20 except towards the east
where there is a gradual transition to the Amazon Platform, (c) the majority of the isolines cross the geolog-

A. Tibaldi et al. / Geomorphology 11 (1995) 215-226

223

Fig. 10. Aerial photo showing different spatial frequency of landslides on mountain slopes with different orientations. Large side of picture is
about 8 km (photogram N, 24099 R-128, authorization n. 930634-IGM-d 3459). Location given in Fig. 2.

224

A. Tibaldi et al. / Geomorphology I l t1995 ) 215-226

Fig. 11. Aerial photo showing a lava flow of the Cayambe volcano affected by maximum landslide density along the slope facing toward ESE
and minimum density along the slope facing in the opposite direction. Large side of picture is about 8 km (photogram N. 24206 R-129.
authorization n. 930634-IGM-d 3459). Location given in Fig. 2.

A. Tibaldi et al. /Geomorphology 11 (1995) 215-226

ical boundaries showing them to be independent of the

rheological characteristics of rocks (compare Fig. 8
with DGGM, 1978, or with fig. 2 of Pasquar~ et al.,
1991 ). Only the eastern part of the central area affected
by the maximum percentage of landsliding would have
been influenced by the local presence of particularly
thick (tens of meters) pyroclastic deposits lying on the
flanks of the Reventador volcano (Fig. 8). The landslide density rises here by about 30-M0%. The presence
of the Reventador volcano produced a local anomaly
but did not affect the general tendency of increase of
landslide density in this central zone; this increase is
already present all around the volcano base where pyroclastic deposit thickness is usually a few metres. A local
increase (by about 20%) in landslide density has been
noted at the passage between intrusive and metamorphic bedrock with no pyroclastic cover.
The rectilinear SE boundary given by isolines 0, 1,
and 4 which trend NE, corresponds both to a NE striking rectilinear fault and to the highest average slope
decrease of the study area. The zone located to the SE
of this boundary, structurally known as Napo Uplift, is
mostly flat with slopes ranging between 0 and 6 . Some
valleys with various orientations dissect this territory
and are characterised by slope inclinations between 15
and 80. A few small landslides have been reported
along these valley flanks.
Another type of comparison of landslide distribution
with topography, in the sense of slope orientation and
inclination, shows that the landslide density is maximum along the slopes facing towards ESE (25-35%
Fig. 9; e.g. in Fig. 10), is lower along slopes facing in
the other directions, and reaches a minimum value
along mountain slopes facing towards WNW. This
result was weighted by computing the percentage of
area for the various slope orientations showing a nonsignificant result only for the orientation sector 100120 (Fig. 9). Vegetation influence should be minimal
because field observations do not suggest variation in
density or type of vegetation on slopes with different
orientations for a given altimetric zone. Also, geological influence should be minimal because the same preferential landslide distribution characterizes zones of
different lithology and rock geometry. For example,
the area shown in Fig. 10 is on sedimentary rock, while
the area in Fig. 11 is on intrusive and extrusive igneous
rocks and on metamorphic rocks. In particular in Fig.
11 two slopes facing in opposite directions towards

225

ESE and WNW, but with the same inclination and

lithological characteristics (a lava flow from the Cayambe volcano), display maximum and minimum landslide density respectively.

4. Conclusions

The total field of landslides produced during the 5

March 1987 earthquakes has been mapped and shows
an elliptical shape with a strong NNE elongation. This
elongation is largely independent of geological limits
and vegetation cover, while the abrupt rectilinear SE
boundary of landslide density isolines could have been
influenced by the high slope gradient of this area.
The elliptical zone has a good correlation with the
location and strike of the fault segments which moved
during the earthquake and with the general geometry
of the Holocene faults. The zone of maximum landslide
density is very close to the epicentres of the two seismic
events.
At a closer look, landslides preferentially occurred
on mountain slopes facing towards ESE. Landslide
density decreases along the slopes facing in the other
directions to the minimum value on slopes facing
towards WNW. This result was weighted by computation of the areas of slopes with different orientations
and was compared with lithology and geological structure showing a clear independence from these factors.
WNW dipping slopes are parallel or sub-parallel to the
dominant fault planes of this area of the Ecuadorian
Andes. Having established the contemporaneity
between these landslides and the two seismic events of
5 March 1987, we would expect a correlation between
earthquake ground motions, geometry of slope failures
and seismogenetic faults. When seismic energy propagated from the shallow focus of the two events, seismic waves crossed the surface with different incidence
angles mostly depending on topography. Evidently
there was some amplification of the ground response
when the slope is perpendicular or sub-perpendicular
to the seismogenetic fault plane and lies along its strike,
whereas maximum attenuation occurred on slopes parallel or sub-parallel to the seismic fault plane.
These intriguing results, deduced from landslide distribution and field studies, could represent a topic of
further multidisciplinary investigations and for microzonation purposes. To our knowledge, a preferred dis-

226

A. 7ibaldi et al. / Geomorphology 11 (1995) 215-226

t r i b u t i o n o f e a r t h q u a k e - i n d u c e d landslides with respect

to hillslope o r i e n t a t i o n h a s b e e n r e p o r t e d only by
P e a r c e a n d O ' L o u g h l i n ( 1 9 8 5 ) for the m a s s m o v e m e n t s w h i c h o c c u r r e d d u r i n g the 1929 M = 7.7 earthq u a k e in n o r t h w e s t e r n S o u t h Island, N e w Z e a l a n d .
C o s e i s m i c fault g e o m e t r y ( H e n d e r s o n , 1937) does not
s u g g e s t the s a m e linking w i t h l a n d s l i d e d i s t r i b u t i o n as
in the p r e s e n t paper. F u r t h e r studies in o t h e r areas
s h o u l d b e e n c o u r a g e d , in o r d e r to assess the potential
o f r e c o n s t r u c t i o n o f g e o m e t r y o f s e i s m o g e n e t i c faults
f r o m the t o p o g r a p h i c effects o f e a r t h q u a k e s .

Acknowledgements
W e are i n d e b t e d to A. C a v a l l i n for c o - o p e r a t i o n in
s o m e o f the c o m p u t e r elaborations. W e a c k n o w l e d g e
m a n y s u g g e s t i o n s o n an early v e r s i o n o f the m a n u s c r i p t
b y A. C a n c e l l i a n d t w o a n o n y m o u s referees, and
i m p r o v e m e n t s to the E n g l i s h style b y A . M . H a r v e y . All
the authors are grateful to E L C - E l e c t r o c o n s u l t for
logistical support. A.T. benefitted f r o m a Ph.D. grant
by M i n i s t e r o Italiano della P u b b l i c a I s t r u z i o n e a n d a
f e l l o w s h i p b y Istituto N a z i o n a l e di G e o f i s i c a - G r u p p o
N a z i o n a l e p e r la V u i c a n o l o g i a .

References
Balseca, W. and Ferraris, M., 1987. lnfornles geologico sobre el
volchn Reventador. Tech. Repts. Coca-Codo Sinclair Proj.. INECEL, Quito, Ecuador, 85 pp.
Barberi, F., Belloni, L., Ferrari, L., Pasquar~, G., Previtali, F.. Tibaldi.
A. and Zonno, G., 1988. Riesgo Sismico. Tech. Rep. Coca-Codo
Sinclair Project, Quito, Ecuador, 294 pp.
CERESIS, 1986. Catalogo de terremotos para America del Sur. Datos
de hipocentros y intensidades.
Clapperton, C.M. and Vera, R., 1986. The Quaternary glacial
sequence in Ecuador: a reinterpretation of the work of Walter
Sauer. J. Quat. Sci., 1: 45-56.
DGGM (Direccion General de Geologia y Minas), 1978. Mapas
Geologicas de Ecuador, escala 1: 100,000. N. 68-82, Quito, Ecuador.

Dziewonski. A.M., EkstrOm, G., Woodhouse, J.H. and Zwart, G.,

1988, Centroid moment tensor solutions for January-March,
1987. Phys. Earth Planet. Inter., 50:116-126.
Hall, LM. and Beate, B., 1991. El volcanismo plio-quatemario en
las Andes del Ecuador. Est. Geogr. Quito, 4: 5-17.
Henderson, J., 1937. The West Nelson earthquakes of 1929. N. Z. J.
Sci. Technol., XIX(2): 65-144.
INECEL (lnsfituto Ecuadoriano de Electrification), 1988. Estudio
Vulcanologico de '~EI Reventador". Ministerio de Energia y
Minas, Quito, Ecuador, 117 pp.
Keefer, D.K., 1984. Landslides caused by earthquakes. Geol. Soc.
Am. Bull., 95: 406~421.
OAQ (Observatorio Astronomico de Quito), 1959. Breve historia
de los principales terremotos de la Republica del Ecuador. Publicacion del Comit~ del Ano Geofisico lnternacional de Ecuador,
Quito, Ecuador.
OAQ I Observatorio Astronomico de Quito 1, 198 I. Catalogo de Sismos del Ecuador, 1900-1980. Quito, Ecuador.
Pasquarb, G., Tibaldi, A. and Ferrari, L., 1991. Relationships
between plate convergence and tectonic evolution of the Ecuadorian active Thrust Belt. In: S.S. Augustfiitis (Editor), Critical
Aspects of Plate Tectonic Theory. Theophrastus Publications,
Athens, pp. 365-388.
Pearce, A.J. and O'Loughlin, C.L., 1985. Landsliding during a M
7.7 earthquake: Influence of geology and topography. Geology,
13: 855-858.
Slemmons, D.B., 1977. Faults and earthquake magnitude. U.S. Army
Congr. of Engineers, Waterways Experimental Stations. Miscellaneous Papers S 73-1, Reprint 6:1-129.
Tibaldi, A., 1990. Transcurrent tectonics and anomalous volcanic
arcs: Analysis of some key areas in Ecuador and Mexico. Ph.D.
Thesis, Universitb.degli Studi di Milano, Italia, 195 pp. ( in Italian
with an English abstr. I.
Tibaldi, A. and Ferrari, L., 1992. Latest Pleistocene-Holocene tectonics of the Ecuadorian Andes. Tectonophysics, 205: 109-125.
USGS ( United States Geological Survey), 1987. Preliminary deterruination of epicenters (PDE). Monthly listing, March 1987.
U.S. Dept. of the Interior.
Woodward and Clyde Consultant, San Francisco, Cal., USA, 1981.
lnvestigaciones para los estudios del riesgo sismico del sitio de
la presa Agoyan. Technical Report for the Agoyan Dam Project,
lnstituto Ecuadoriano de Electrification (INECEL), Quito, Ecuador.
Wright, C. and Mella, A., 1963. Modifications to the soil pattern of
south-central Chile resulting from seismic and associated phenomena during the period May to August 1960. Seismol. Soc.
Am. Bull., 53(6): 1367-1402.