Roberts Et Al. (2013) - Paleointensidade

Quaternary Science Reviews 61 (2013) 1e16
Contents lists available at SciVerse ScienceDirect
Quaternary Science Reviews

journal homepage: www.elsevier.com/locate/quascirev
Invited review
Magnetic paleointensity stratigraphy and high-resolution Quaternary

geochronology: successes and future challenges
Andrew P. Roberts a, *, Lisa Tauxe b, David Heslop a
a
Research School of Earth Sciences, The Australian National University, 61 Mills Road, Canberra, ACT 0200, Australia
b
Scripps Institution of Oceanography, University of California San Diego, 9500 Gilman Drive, La Jolla, CA 92093, USA
a r t i c l e i n f o a b s t r a c t
Article history: Magnetic paleointensity stratigraphy is used to detect variations in the strength of Earth’s ancient
Received 12 September 2012 magnetic field. Paleointensity studies have demonstrated that a dominantly dipolar geomagnetic signal
Received in revised form can be recorded in a globally coherent manner in different types of sediments and in non-sedimentary
25 October 2012
archives, including ice core records and marine magnetic anomaly profiles. The dominantly dipolar
Accepted 29 October 2012
Available online 12 December 2012
nature of geomagnetic paleointensity variations provides a global geophysical signal that has come to be
widely used to date Quaternary sediments. Despite the many successful applications of paleointensity-
assisted chronology, the mechanisms by which sediments become magnetized remain poorly under-
Keywords:
Geomagnetic field
stood and there is no satisfactory theoretical foundation for paleointensity estimation. In this paper, we
Paleointensity outline past successes of sedimentary paleointensity analysis as well as remaining challenges that need
Quaternary to be addressed to place such work on a more secure theoretical and empirical foundation. We illustrate
Geochronology how common concepts for explaining sedimentary remanence acquisition can give rise to centennial to
Depositional remanent magnetization millennial offsets between paleomagnetic and other signals, which is a key limitation for using paleo-
Post-depositional remanent magnetization intensity signals for geochronology. Our approach is intended to help non-specialists to better under-
stand the legitimate uses and limitations of paleointensity stratigraphy, while pointing to outstanding
problems that require concerted specialist efforts to resolve.
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction global signal that can potentially be used to provide a high-

resolution (millennial scale) timescale for chronostratigraphy. This
Geomagnetic polarity reversals have been used widely in temporal resolution contrasts with geomagnetic reversals (Fig. 1),
Quaternary geochronology (Fig. 1) because they result from a virtu- which occurred only w4e5 times per million years over the last
ally synchronous global change in sign of the geomagnetic dipole. w31 Myr (Lowrie and Kent, 2004).
The geomagnetic polarity timescale (GPTS; e.g. Cande and Kent, Determining the magnetic polarity of a given geological unit is
1995) serves as the backbone for the geological timescale for the straightforward, whereas, as argued below, determining the
last 200 Myr, and is based on temporally calibrated records of Earth’s ancient geomagnetic field strength is not so simple. Given the
polarity history. Polarity reversals are accompanied by dramatic increasing use of paleointensity estimation in Quaternary
decreases in geomagnetic paleointensity (Fig. 1). Along with higher- geochronology, we provide an overview for a general audience of
frequency paleointensity variations within periods of stable polarity how geomagnetic paleointensities are estimated. We then
(Fig. 1), these major intensity changes can also provide a timescale summarize the strongest lines of evidence for why such estima-
that has come to be used widely in geochronology (e.g. Guyodo and tions appear to be robust, followed by discussion of some of the
Valet, 1996, 1999; Laj et al., 2000; Kiefer et al., 2001; Stoner et al., uses of geomagnetic paleointensity analysis in high-resolution
2002; Stott et al., 2002; Valet et al., 2005; Yamazaki and Oda, Quaternary geochronology. This treatment is representative of the
2005; Channell et al., 2009; Ziegler et al., 2011). The geomagnetic successes of paleointensity analyses of Quaternary sediments.
field is generated in Earth’s fluid outer core, and is dominated by the However, despite these outstanding successes, challenges remain.
dipole component, so that variations in field intensity have a strong The remainder of the paper is devoted to summarizing these
challenges. Our overall aim is to help Quaternary scientists to
understand better how sedimentary paleointensities are estimated,
* Corresponding author. Tel.: þ61 2 6125 3887. their potential chronostratigraphic value, and their limitations. We
E-mail address: andrew.roberts@anu.edu.au (A.P. Roberts). also point out problems that require concerted paleomagnetic
0277-3791/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.quascirev.2012.10.036
2 A.P. Roberts et al. / Quaternary Science Reviews 61 (2013) 1e16
measurements made by Gauss in 1835. Determination of geomag-
Brunhes precursor (797 ka)
Cobb Mt. (1190-1215 ka)

netic field intensities for time periods preceding the 19th Century,
Jaramillo (987-1068 ka)
Olduvai (1778-1945 ka)

Gardar (1465-1485 ka)
(Validated Excursions)
Iceland Basin (188 ka)
Pre-Olduvai (1977 ka)

Big Lost (560-580 ka)
Kamikatsura (883 ka)
Gilsa (1567-1575 ka)

Pringle Falls (211 ka)
Santa Rosa (932 ka)

therefore, requires paleomagnetic analysis of rocks or archaeolog-
Mono Lake (33 ka)
Punaruu (1115 ka)

Laschamp (41 ka)
Stage 17 (670 ka)

ical artefacts. Estimating the intensity of an ancient magnetic field
Bjorn (1255 ka)

Blake (120 ka)
is based on the assumption that the magnetization of a rock will be

Subchrons
related linearly to the geomagnetic field strength. It is, therefore,

expected that the natural remanent magnetization (NRM) of
a sample will be related to the ancient field intensity (Banc) as
follows:
Brunhes Matuyama
Magnetic
polarity NRMyaanc Banc ;
Nor.-Greenland Sea (70 ka)
CR0 (260 ka)
West Eiffel 4 (555 ka)
Intra-Jaramillo (1048 ka)

WE1 (722±38 ka)
CR2 (525 ka)
0.78 Ma
WE2 (626±24 ka)
Other possible
where aanc is a constant of proportionality. For certain igneous

excursions
rocks, archaeological artefacts or other materials that have cooled

from high temperatures, there is a robust physical theory and
10
experimental protocol that enables determination of the absolute
paleointensity from the recorded thermal remanent magnetization
(TRM) (Thellier, 1938; Néel, 1955; Thellier and Thellier, 1959).
8 Laboratory experiments are aimed at determining the proportion-
ality of the TRM intensity to the geomagnetic field strength
(Thellier and Thellier, 1959). After measuring the ancient TRM
(TRManc) at a given temperature, a TRM can be imparted in the
PADM (10 22 Am2 )
6
laboratory by heating the sample to the same temperature in
a known applied laboratory field (Blab). This enables determination
4 of the laboratory constant of proportionality (alab). Assuming that
alab is identical to aanc, which can be tested with carefully designed
experiments, the paleofield intensity can be determined from:
2
Banc ¼ ðTRManc =TRMlab ÞBlab :

0
While the laboratory normalization technique provides a theo-
retically grounded means of determining absolute ancient field
intensities, suitable materials with thermal remanences are neither
Interglacial
Warmer
3.5
temporally continuous nor are they globally available. Young
volcanic rocks are also notoriously difficult to date and only a small
fraction of available material yields useful paleointensity data.
δ18O 4.0 Sedimentary sequences are, therefore, an attractive target for
obtaining continuous records of geomagnetic paleointensity vari-
ations. However, identification of a robust procedure for laboratory
Colder
Glacial
4.5 normalization of a sedimentary NRM that is analogous to that for

a TRM has proved elusive. The problem is that there is no simple
means of determining aanc to calibrate the relationship between the
NRM of a sample and the strength of the magnetizing field. The
5.0
magnetization of sediments is affected by the strength of the
0 0.5 1.0 1.5 2.0
ambient magnetic field, the magnetic mineral that records the
paleomagnetic signal, the concentration of this magnetic mineral
Age (Ma)
fraction, its grain size and the mechanism by which the magneti-
Fig. 1. Geomagnetic polarity timescale for the last 2 million years, with geomagnetic zation was acquired. An empirical approach has been developed for
excursions, relative paleointensity variations and benthic d18O variations. Geomagnetic estimating paleointensities from sediments in which the NRM is
polarity is indicated at the top of the figure (black ¼ normal; white ¼ reversed polarity). normalized by an artificial laboratory-induced magnetization (Levi
Validated excursions (white) are indicated along with their respective ages above the
and Banerjee, 1976). The goal is to remove the influence of rock
polarity log in italics, with “possible” excursions (red) that have yet to be fully validated
are indicated in red below the polarity log (Laj and Channell, 2007; Roberts, 2008). Each magnetic variations with non-geomagnetic origins, and to validate
polarity reversal and excursion coincides with a paleointensity minimum. The paleo- the record by imposing strict rock magnetic selection criteria. These
magnetic axial dipole moment (PADM) model (Ziegler et al., 2011) is used to represent criteria traditionally require magnetite to be the only magnetic
paleointensity (blue). The climatic context of the geomagnetic variations is illustrated
mineral present and that it occurs within a narrow grain size and
using the global stacked benthic d18O record (red) of Lisiecki and Raymo (2005).
concentration range (King et al., 1983; Tauxe, 1993). While this
empirical approach appears to work (see discussion below), its
effort to resolve. Such efforts will be valuable for improving our theoretical underpinning is complicated (e.g. Tauxe et al., 2006).
understanding of the geodynamo as well as aiding geochronolog- The result is that, because we cannot determine absolute paleo-
ical applications. intensities from sediments, we seek to estimate relative paleo-
intensity variations by minimizing the number of variables that
2. How does paleointensity determination work? contribute to the magnetization of the sediment under
investigation.
Reliable contemporary measurement of the intensity of the Despite the lack of a first-principles theory for how sediments
geomagnetic field has only been possible since the first become magnetized, we can outline general principles by which
A.P. Roberts et al. / Quaternary Science Reviews 61 (2013) 1e16 3
normalized remanence records can be interpreted as having (SAPIS; Stoner et al., 2002), the global paleointensity stack (GLOPIS;
a geomagnetic origin. For relative paleointensity estimation to Laj et al., 2004), the North Atlantic ODP Site 983 record of Channell
work, the mechanism by which the sediment is magnetized must and Kleiven (2000) that spans the 700e1100 ka interval, and the
be constant throughout a studied sequence. This requirement paleointensity and stable isotope stack for the last 1.5 Myr (PISO-
cannot be tested and many questions remain about how sediments 1500; Channell et al., 2009). For the NAPIS, SAPIS and GLOPIS
acquire a magnetization (see discussion in Section 5). Furthermore, stacks, millennial-scale chronology is achieved by correlating
a key assumption in relative paleointensity studies is that there is sediment physical properties into a tight, internally consistent
a linear relationship between magnetization intensity and stratigraphy, and using oxygen isotope stratigraphies to correlate
geomagnetic field strength. There is incomplete empirical evidence millennial climatic events with those recorded in the Greenland
to support this assumption for most sedimentary environments, GISP2 ice core. For the PISO-1500 stack, simultaneous correlation of
including marine environments. For details, readers are referred to oxygen isotope and paleointensity records reduces the degree of
recent in-depth reviews (Valet, 2003; Tauxe and Yamazaki, 2007). freedom for correlation based on either parameter alone. Regional
We are, therefore, in the uncomfortable position of having an and global reproducibility of multiple records provides a powerful
inadequately grounded approach for relative paleointensity anal- argument for the robustness of relative geomagnetic paleointensity
ysis. In Section 3, we discuss the empirical evidence for why estimates from sediments.
normalized remanence records appear to be dominated by
geomagnetic signals, despite our inadequate understanding of 3.2. Cosmogenic radionuclides
paleointensity recording by sediments. This is followed by a brief
discussion of how paleointensity signals are used in Quaternary Cosmogenic radionuclides are produced by interaction of
geochronology (Section 4). We then discuss sedimentary rema- cosmic rays with Earth’s atmosphere. Production of cosmogenic
nence acquisition in Section 5 and issues related to relative pale- radioisotopes is modulated by variations in cosmic ray flux, solar
ointensity normalization in Section 6. activity and shielding by the geomagnetic field. Geomagnetic
dipole moment variations are the most important modulator of
3. Do sedimentary paleointensity estimates have production rate of cosmogenic radioisotopes, which varies
a geomagnetic origin? inversely with field strength. Variations in production of cosmo-
genic radionuclides, including 14C (half-life, T1/2 ¼ 5.73 kyr), 36Cl
Despite uncertainties about the details of how sediments (T1/2 ¼ 300 kyr), and 10Be (T1/2 ¼ 1500 kyr), from ice cores and
become magnetized, many credible relative paleointensity records sediments provide an independent measure of field intensity
have been recovered, which indicate that the geomagnetic field variations on a range of timescales (e.g. Frank et al., 1997; Raisbeck
intensity varied in a globally coherent manner on millennial and et al., 2006). These variations can be presented in terms of pre-
longer timescales (Fig. 1). Three principal “smoking guns” give dicted relative paleointensity by assuming that all of the cosmo-
confidence that sediments can provide robust relative geomagnetic genic radionuclide production stems from variations in the
paleointensity estimates. These include: 1. global reproducibility; 2. geomagnetic field and by transforming production rate into rela-
cosmogenic radionuclides; and 3. ocean crust magnetization. We tive paleointensity. Elsasser et al. (1956) used the simple formula
outline each of these lines of evidence below. ðQ =Q0 ÞfðM00:5 =MÞ, where Q is the radionuclide production rate at
a given dipole strength M, relative to initial values for both (Qo,
3.1. Global reproducibility Mo). Lal (1988) modified the relationship, particularly for low field
strengths; this modified relationship is normally used to convert
10
Collections of published relative paleointensity records have Be variations to predicted relative paleointensity variations (e.g.
been stacked to produce estimates of global field intensity varia- Frank et al., 1997). Good agreement between variations predicted
tions (e.g. Guyodo and Valet, 1996, 1999, 2006; Valet et al., 2005; for the paleomagnetic dipole from relative paleointensity data
Channell et al., 2009; Ziegler et al., 2011). Multiple records have (Ziegler et al., 2011) and a sedimentary record of 10Be production
been stacked in this way because of the overall global coherency of for the last 200 ka (Frank et al., 1997) (Fig. 2a) and Antarctic ice
the recorded signals (Fig. 1) despite differences in the sedimentary core 10Be flux data (Raisbeck et al., 2006) across the Matuyama/
environments from which the records were obtained. With devel- Brunhes boundary interval (Fig. 2b) provide strong evidence for
opment of increasing numbers of relative paleointensity records, a common signal.
such stacks have progressively worked back in time from the
present to 200 ka (SINT-200; Guyodo and Valet, 1996) to 800 ka 3.3. Ocean crust magnetization
(SINT-800; Guyodo and Valet, 1999) to 2.0 Ma, including SINT-2000
(Valet et al., 2005) and PADM2M (Ziegler et al., 2011), to 3 Ma Ocean crust provides a paleomagnetic record of geomagnetic
(EPAPIS-3 Ma; Yamazaki and Oda, 2005). In parallel with these polarity variations over the last w160 Ma (e.g. Cande and Kent,
developments, it has been recognized that stacking can affect the 1995; Gee and Kent, 2007). Data from deep-towed magnetom-
amplitude of paleointensity features when records have different eter surveys over fast-spreading ocean crust reveal coherent
chronological precision (e.g. Roberts et al., 1997; Guyodo and short-wavelength anomalies (Gee et al., 2000). Inversion of these
Channell, 2002) or where variable sedimentation rates and any anomaly profiles enables estimation of crustal magnetization
smoothing associated with paleomagnetic recording cause atten- (blue curve in Fig. 3), which compares well with paleomagnetic
uation of high-frequency features (Hartl and Tauxe, 1996; Guyodo dipole moments (red curve in Fig. 3) from the global paleo-
and Channell, 2002; Roberts and Winklhofer, 2004). Thus, in intensity stack of Ziegler et al. (2011). Minor offsets between the
addition to the above-cited paleointensity stacks, which tend to two records (e.g. at about 400 ka) result from imprecisions in the
have coarse age control on glacial/interglacial timescales (but respective age models, particularly the assumption of linear
sometimes better), an additional family of records or stacks of spreading rate for the magnetic anomaly stack. The overall
records has developed, usually from relatively rapidly deposited excellent agreement between these continuous records of ocean
sediments with millennial-scale chronological resolution. These crustal magnetization and sedimentary paleointensity provides
include the North Atlantic paleointensity stack for the last 75 ka a strong argument that they register a common geomagnetic
(NAPIS-75; Laj et al., 2000), the South Atlantic paleointensity stack signal.
2.0 4. Geomagnetic relative paleointensity in high-resolution

a Quaternary geochronology
1.5
Detailed relative paleointensity stacks now span the entire
Relative paleointensity
Quaternary (e.g. Valet et al., 2005; Yamazaki and Oda, 2005;

Channell et al., 2009; Ziegler et al., 2011). Recognition that precisely
dated, detailed paleointensity records have coherency on a global
1.0
scale has led to the notion that geomagnetic relative paleointensity
can be used to constrain the chronology of a sedimentary sequence.
This concept is often referred to as paleointensity-assisted chro-
0.5 nology (PAC). The key advantage of using paleointensity for high-
resolution dating is that dipolar geomagnetic variations are glob-
ally synchronous, although it is difficult to test for isochrony at
0.0 better than millennial scale. In particular, geomagnetic variations
0 50 100 150 200 are independent of climatic parameters, including seawater
2.0
chemistry, that are often used to synchronize marine sedimentary
b chronologies, and are therefore independent of the millennial-scale
leads and lags that can affect climatic parameters. Much paleo-
1.5 climate research is aimed at resolving such phase relationships and
independent methods that can constrain such problems are

extremely valuable in paleoclimate studies. Paleointensity varia-
1.0 tions have been used to resolve phasing issues associated with
paleoclimate signals through marine isotope stage (MIS) 3, even in
records in which there is clear manifestation of millennial-scale
climate variability (e.g. Laj et al., 2000; Stoner et al., 2002).
0.5
However, in other records in which this northern hemisphere
climate variability is not evident, Channell et al. (2000) argued that
paleointensity can provide a basis for inter-hemispheric correlation
0.0 of marine sediments at a resolution that is difficult to achieve with
680 700 720 740 760 780 800
d18O stratigraphy alone (Fig. 1). Paleointensity variations have,
Age (ka) therefore, been widely used either as one parameter among others
Fig. 2. Comparison of relative paleointensity and cosmogenic radionuclide production
to constrain a chronology (e.g. Channell et al., 2000; Stoner et al.,
records for two time intervals. (a) Relative paleointensity (blue) predicted from 2000, 2002; Kiefer et al., 2001; Stott et al., 2002; Evans et al.,
normalized 10Be stacks (from Frank et al., 1997) and relative paleointensity (red) from the 2007), or, in lake or marine environments that lack suitable
PADM2M model (Ziegler et al., 2011) for the last 200 ka. (b) Same but with predicted material for radiocarbon dating or for d18O stratigraphy, to develop
relative paleointensity from 10Be flux in the Antarctic EPICA Dome C ice core (blue) across
an independent chronology that is based largely or entirely on
the MatuyamaeBrunhes boundary interval (from Raisbeck et al., 2006), normalized in
same way as in (a). The PADM2M model (red) has relatively low resolution compared to paleointensity variations (e.g. Sagnotti et al., 2001; Brachfeld et al.,
the ice core records, but the same large-scale paleointensity features are evident in both 2003; Macrí et al., 2005, 2006; Willmott et al., 2006; Zimmerman
the paleointensity and cosmogenic radionuclide production records. The data repre- et al., 2006; Lisé-Provonost et al., 2009). The success of the PAC
sented by the blue curves were recalculated so that both curves are directly comparable approach is indicated by the fact that the scientific objectives of
with the PADM2M dipole moment record (as discussed briefly in the main text).
research cruises of the Integrated Ocean Drilling Program (IODP),
which represents the largest international program in Earth and
ocean science, have been based on the use of PAC to assess ice
sheeteoceaneatmosphere interactions on millennial timescales
1.5
(e.g. Shipboard Scientific Party, 2005).
As demonstrated above, the geomagnetic field varies in a globally
1.0 coherent manner on millennial timescales (e.g. Channell et al., 2000;
Laj et al., 2000, 2004; Stoner et al., 2002; Channell et al., 2009).
Sedimentation rates in most marine environments are not high

0.5 enough to enable temporal resolution at better than millennial
timescales. However, centennial-scale paleointensity variations have
been documented in exceptionally high-resolution datasets (e.g.
0.0
Willmott et al., 2006; Lisé-Provonost et al., 2009; Barletta et al., 2010).
Our rapid summary of the development and use of PAC suggests
-0.5 that it has real promise for high-resolution geochronology. Despite
the optimism that might be assumed from our treatment above,
many difficulties remain with use of continuous relative geomag-
-1.0 netic paleointensity records. Signal variability in individual records
0 100 200 300 400 500 600 700 800 900
Age (ka) often does not correlate well among records (or with stacks). To
illustrate this, we plot for the last 800 ka the SINT-2000 stack (Valet
Fig. 3. Comparison of predicted relative paleointensity from the magnetization et al., 2005) alongside the PISO-1500 stack of Channell et al. (2009)
inverted from a high-resolution marine magnetic anomaly stack (blue; data from Gee in Fig. 4a (renormalized to a mean of unity over the interval shown).
et al., 2000) and dipole moments from the PADM2M paleointensity stack (red; Ziegler
et al., 2011). Ages for the magnetic anomaly record were rescaled to a common age for
Although there are similarities between these two stacks, as
the MatuyamaeBrunhes boundary (see discussion in the text concerning age offsets of emphasized above, there are intervals, as shown in the boxes, where
some paleointensity features). agreement is less impressive. These differences far exceed
2.5
a
2.0
b
1.5
1.0 c
0.5
SINT2000 PISO1500
0.0
0 100 200 300 400 500 600 700 800
180
d e Geomagia
ODP984 (60.4°N) 160
|Lat.| < 30°
ODP983 (60°N) 140 |Lat.| > 50°
VADM (ZAm 2 )
MD95-2024 120
(50°N) 100
80
ODP769 (9°N)
60
ODP768B (8°N)
40
ODP768A (8°N)
20
0 1 2 3 4 5 6 7
0 20 40 Age (ka)
Age (ka)
Fig. 4. Comparison of relative paleointensity stacks from different latitudes. (a) Comparison of SINT-2000 (Valet et al., 2005) with PISO-1500 (Channell et al., 2009) for the last
800 ka. Boxes denote periods with poor agreement between the two stacks. (b) Locations of cores in the PISO-1500 and (c) SINT-2000 stacks. (d) Examples of representative
individual records from high (top three) and low (bottom three) latitudes (data are as compiled by Tauxe and Yamazaki (2007) and references therein). (e) Subset of the Geomagia50
database (Korhonen et al., 2008) with data from latitudes greater than 50 away from the equator and those from no more than 30 away. Lack of systematic latitudinal differences
in the archaeointensity data for the last 50,000 years indicates that the observed differences in the paleointensity stacks are not due to non-dipolar effects.
measurement errors and any errors associated with stacking of limitations in relative geomagnetic paleointensity determinations
multiple records. Why do such discrepancies occur and what do they in order to better understand the measured signals and to find
mean? Do they result from imprecisions in chronological resolution, solutions to the questions posed. The authors of this paper represent
loss of signal amplitude due to stacking, or could they result from part of the spectrum of views in relation to the issues at stake and do
non-dipolar field behaviour (i.e. regional versus global signals, not agree on all aspects. In the following treatment, we attempt to
where the PISO-1500 stack is dominated by relatively high latitude represent the spectrum of views in order to articulate the problems
records compared to the more global distribution of sites in the that need to be addressed through concerted research effort.
SINT-2000 stack (Fig. 4b, c))? One of the biggest differences among
these records occurs for the last 40 ka (Fig. 4d). By comparing 5. Sedimentary remanence acquisition
absolute paleointensities from the Geomagia50 database (Korhonen
et al., 2008), we observe no profound difference between high A major uncertainty concerning relative paleointensity studies
(>50 ) and low (<30 ) latitude data (Fig. 4e). This suggests that relates to sedimentary remanence acquisition. While sedimentary
a non-dipole explanation for the differences between the stacks is remanence acquisition has been subjected to detailed experimental,
unlikely and that other explanations are necessary. Are they, then, theoretical and numerical investigation for over 60 years (e.g.
due to variability in remanence acquisition mechanisms within the Johnson et al., 1948; King, 1955; Nagata, 1961; Irving and Major, 1964;
studied sediments? Do they result from imperfect normalization? Kent, 1973; Verosub, 1977; deMenocal et al., 1990; Tauxe et al., 1996,
Are there grain size variations that are not adequately normalized by 2006; Carter-Stiglitz et al., 2006; Heslop, 2007a; Tauxe, 2006; Liu
standard approaches? What influence does diagenesis have on et al., 2008; Mitra and Tauxe, 2009; Roberts et al., 2012), much
recording of paleointensity signals? Are there other problems? The remains unknown. Uncertainties surround the numerous variables
temptation to throw away relative paleointensity analysis because that remain difficult to constrain fully. These relate to sedimentology,
of the existence of such problems should be avoided. The case for flocculation and pelletization of sediment particles, salinity, bio-
why relative paleointensity works has been made above. It should turbation, compaction, diagenesis, and differential contributions
be remembered that marine d18O records can also be complicated from biogenic and detrital magnetic minerals. We now describe
due to variable freshwater inputs (e.g. ice melting, strong evapora- known issues in relation to these variables and how they influence
tion in marginal basins, monsoon variations and ocean stratification, remanence acquisition and relative paleointensity studies.
etc.), and that synchronization of paleoclimate records can be
blighted by many problems. Tuning and correlation remain 5.1. Classical depositional and post-depositional remanent
a problem with age model construction in paleoclimate studies and magnetizations
standard tuning procedures gloss over attempts to independently
determine phase relationships among climate signals of interest. It The journey of a magnetic particle begins as it settles through
remains important to resolve the reasons that lie behind the the water column and ends when it is locked into position in the
sediment (Fig. 5). Magnetic particles suspended in water are spun between the two vectors. The magnetic moment will rotate to bring
about by turbulence, which acts to randomize the magnetic m into alignment with B, but this motion will be opposed by the
moments. Depending on the chemistry of the water (salinity and/or viscosity of the aqueous fluid. Nagata (1961) solved the equation of
pH), magnetic particles will adhere to clays and clay particles will motion and demonstrated that the time constant of alignment (s) of
coalesce to form larger, less magnetic flocs (Shcherbakov and a particle with the ambient field can be approximated by:
Shcherbakova, 1983; Lu et al., 1990; Katari and Tauxe, 2000;
Tauxe et al., 2006) or fecal pellets. These flocs and pellets are l 6h
s¼ ¼ ;
exported to the bottom of the water column. During their descent mB MB
they are subject to magnetic and hydrodynamic torques. The action
of any hydrodynamic torque will dominate the magnetic torque for where l is the viscosity coefficient opposing the motion of the
non-spherical flocs during settling, thereby contributing to an particle through the fluid (defined as the surface area of the particle
inefficient magnetization of the resultant sediment (Heslop, times the viscosity of the fluid h), and M is the magnetization
2007a). The sediment/water interface is usually not well defined. (moment normalized by volume). Choosing reasonable values for
The lowermost part of the water column is a zone in which the magnetic minerals, fields and fluid viscosities, a magnetic particle
concentration of suspended sediment increases dramatically will align fully and almost instantaneously with the magnetic field.
(Fig. 5). This zone is referred to as the nepheloid layer or the benthic Simple DRM theory, therefore, predicts that for sediments
boundary layer, in which sediment remains in suspension due to composed of isolated magnetic particles, the particles should have
friction associated with movement of bottom waters over the magnetic moments that are fully aligned with the geomagnetic
sediment substrate. At the base of the nepheloid layer, there is field. As a result, a DRM should be insensitive to changing field
a higher density of fluffy floc-rich suspended matter, below which strengths. However, as we have demonstrated above, sedimentary
occurs the uppermost part of the sediment column. The sediment is remanences record a strong signal due to field strength variations.
water rich and is actively mixed by organisms (bioturbation) who The magnetic torque on isolated particles can, therefore, only be
repeatedly ingest, excrete and re-suspend the material until it one part of a complex story.
finally joins the more consolidated sediment layer below, in which Other factors are also important for DRM and pDRM acquisition.
the probability of re-suspension drops dramatically. As the sedi- When a magnetic particle falls through a fluid (air or water), it can
ment dewaters and compacts, magnetic minerals may undergo roll when it impacts the sedimentary substrate. The net effect is
further rotation. Finally, magnetic minerals may undergo chemical that the recorded paleomagnetic inclination will be systematically
change(s) during diagenesis, either growing from non-magnetic shallow (e.g. Johnson et al., 1948; King, 1955; Tauxe and Kent,
precursory phases, changing from one magnetic phase to another, 2004). Inclination shallowing is, therefore, often associated with
or dissolving altogether. Stages of remanence acquisition have a DRM. In contrast, many sediments provide superb recording of
traditionally been described as a depositional remanent magneti- the expected time-averaged geomagnetic field for a geocentric
zation (DRM), where magnetic particles rotate freely in aqueous axial dipole (e.g. Opdyke and Henry, 1969). Such sediments are
solution, a post-depositional remanent magnetization (pDRM), often bioturbated, which suggests that bioturbation lessens the
where magnetic particles respond to magnetic torques in the density of particle packing, which then allows the geomagnetic
consolidating zone, and a chemical remanent magnetization (CRM), field to exert a torque on magnetic particles to realign them with
which is acquired when magnetic particles grow through a critical the field after the last mixing event, thereby giving rise to a pDRM
blocking volume during diagenesis. (Irving and Major, 1964; Kent, 1973). Subsequent compaction of
When considering DRM and pDRM acquisition, it is important to clay-rich sediments can give rise to an additional type of inclination
consider how magnetic particles can realign with the ambient shallowing (e.g. Anson and Kodama, 1987; Arason and Levi, 1990).
geomagnetic field. The torque on the magnetic moment m of Both types of inclination shallowing can be corrected for (Tauxe and
a particle by the field B is: m B, or mB sin a, where a is the angle Kent, 2004). Regardless, compaction-induced inclination shallow-
ing is not normally observed above depths of 60e85 m below the
sediment/water interface (Arason and Levi, 1990) or even at depths
of hundreds of metres (Anson and Kodama, 1987). This phenom-
Particle
concentration (%) enon is, therefore, not normally important for Quaternary settings
0 100 of relevance here. Overall, the poorer recording fidelity expected for
Input of magnetic minerals ( ), clays ( )
a DRM is less desirable than the recording expected for a pDRM.
Turbulent Water
flocculation/ Regardless, other complexities need to be considered before
pellitization accepting the legitimacy of either the classically defined DRM or
pDRM concepts (Fig. 5).
magnetic/
export to bottom hydrodynamic
torques 5.2. Salinity and flocculation
benthic dilute
boundary suspension The observed field dependence of DRM implies a much longer
layer nepheloid resuspension/ time constant of alignment than that predicted by Nagata (1961) for
layer “fluffy” layer reflocculation/ settling of isolated magnetic particles. There are several ways to
pellitization
surface mixed layer bioturbation accomplish this from a theoretical point of view (see Tauxe et al.
consolidation (2006) for a review). The most promising approach, however, is
consolidating “historical” layer compaction that of Shcherbakov and Shcherbakova (1983) who recognized that
lock-in
diagenesis in natural waters, particles adhere to each other (which results in
“coagulation” or “flocculation”). Qualitatively, particles are drawn
Fig. 5. Schematic illustration of the journey of magnetic particles (black ellipsoids) together by van der Waals forces. In deionized water, clay particles
from the water column to burial with the processes that are likely to contribute to
eventual recording of the paleomagnetic field by the particles (modified from Tauxe
are surrounded by a double layer of ions that impart an electrical
and Yamazaki (2007)). See discussion in the text for a more detailed explanation of charge to the particles. The charges repulse one another, which
the processes depicted. keeps the clays apart in a stable colloid. Addition of salt (or other
electrolytes) interferes with the double layer, thinning it, and particles are treated as collections of “plates and spheres”, but
allowing van der Waals forces to become important, thus making differs substantially in the processes involved. Instead of assuming
the particles more likely to adhere to one another. Flocs of such two distinct magnetic grain shape populations, Mitra and Tauxe
particles are composites of magnetic and non-magnetic minerals (2009) divided a continuous distribution of floc sizes into two
and have a much lower net magnetization than isolated magnetic groups: one small enough to respond mainly to magnetic torques
particles. The lower resulting M in the equation of Nagata (1961) (group M) and one large enough to be governed by hydrodynamic
increases the theoretical time constant of alignment with the torques (group H). The net magnetic moment of group M is
ambient field. essentially parallel to the applied field while group H flocs are more
The effect of embedding magnetic particles within flocs is influenced by hydrodynamic torques. The flocs attain hydrody-
illustrated in Fig. 6a. Numerical simulations predict that small flocs namic stability while settling (with long axes on average hori-
(5 mm) will be essentially fully aligned with the applied field (i.e. the zontal); the magnetic moments then attempt to align with the field.
curves are saturated), but the magnetic remanence of larger flocs When the flocs are too large to maintain equilibrium with the field,
(25 mm) will be far from saturation and will have the expected their magnetization is essentially randomized. Therefore, the net
quasi-linear dependence of DRM with respect to B (Tauxe et al., magnetic declination of group H flocs tracks the field azimuth, but
2006). This simple floc model, which has a single magnetic the net inclination is near zero (Fig. 8). The net magnetic moments
moment within each floc (similar to the model of Katari and of both groups of flocs contribute to the observed DRM (Fig. 8). In
Bloxham, 2001), needed to be modified to explain the results of reality, a floc is not expected to behave according to the simple
laboratory re-deposition experiments (Tauxe et al., 2006). In the scheme of particles aligning through action of a magnetic torque, as
model of Tauxe et al. (2006), flocs coalesce to form compound flocs envisaged in the model, but it would instead follow a complicated
(Fig. 6b inset) so that saturation is never achieved in weak Earth- trajectory under the simultaneous influence of magnetic and
like fields. hydrodynamic torques (Heslop, 2007a). However, the average
Van Vreumingen (1993) demonstrated that, in addition to behaviour of an ensemble of flocs can be approximated by this
a strong influence of salinity on DRM intensity (Fig. 7), the degree of simple conceptual model. Mitra and Tauxe (2009) used this model
paleomagnetic inclination shallowing also depends strongly on to successfully simulate the laboratory results shown in Fig. 7.
salinity. The flocculation model of Tauxe et al. (2006), with spher- A key assumption in sedimentary paleointensity studies is that
ical composite flocs, cannot account for this phenomenon. Floccu- DRM is linearly related to the applied field, yet laboratory re-
lated particles are hydrodynamically different from isolated deposition in known fields sometimes produces a non-linear
particles. They are porous, loose and fragile. Although flocs have DRM-applied field strength relationship (e.g. Fig. 6b; Tauxe et al.,
irregular shapes and are not rigid, a useful first step is to model 2006). From theory, larger flocs will respond in a more linear
them as rigid ellipsoids, which undergo complex motion when fashion with applied field than smaller flocs (Fig. 6a), which reach
settling in a fluid (Belmonte et al., 1998; Galdi and Vaidya, 2001). saturation rapidly (although larger particle sizes are essentially
With this approximation, Heslop (2007a) demonstrated that with randomly oriented). Given the importance of the linearity
increasing floc size, the hydrodynamic torque increases at a much assumption to relative paleointensity studies, we note that labo-
higher rate than the magnetic torque. For prolate ellipsoids with ratory re-deposition experiments suggest that these non-linear
a particular aspect ratio, flocs become dominated by hydrodynamic DRMs, expressed as a fraction of the saturation isothermal rema-
torques at a critical size. nent magnetization (sIRM; i.e. the strongest remanence that can be
Mitra and Tauxe (2009) incorporated the approach of Heslop acquired by a sample), are much larger than natural sedimentary
(2007a) into the flocculation model of Tauxe et al. (2006). The DRM/sIRM ratios (which are usually a few % at most). Such DRMs
new model is conceptually similar to that of King (1955), where will likely have a quasi-linear relationship with applied field.
a 1.0
b 40
1 ppt
5
10 2.2 ± 1.6
0.8
30
15 2.3 ± 1.6
DRM/sIRM (%)
5 ppt
0.6
DRM/sIRM
2.8 ± 1.6
20 20
0.4 2.9 ± 1.6
25 μm 10
0.2
0.0
0
0 10 20 30 40 50 0 10 20 30 40
B (μ T) B (μT)
Fig. 6. Calculated results for magnetizations associated with flocs consisting of magnetic and non-magnetic particles. (a) Results of a simple numerical simulation whereby
a magnetic moment is embedded in flocs of increasing radius (in microns) and solving for the net DRM after settling through 0.2 m of water. The DRM is expressed as a ratio of the
saturation isothermal remanent magnetization (sIRM), which is unity when the magnetic moment of the floc is fully aligned with the ambient field. In this calculation, the floc has
a single magnetic moment. (b) Results of laboratory re-deposition experiments as a function of field in a flocculating environment (red dots) for compound flocs (see schematic in
inset). Salinities (in parts per thousand, ppt) are 1 and 5 for two different sets of experiments. Solid lines are from numerical simulations using distributions of compound flocs with
assumed log-normal size distributions (mean and standard deviations are indicated in microns). The inset is an example of a compound floc with several smaller flocs (magnetic
moments are red arrows) that coalesce to form a compound floc whose magnetization is the sum of the three individual moments. In contrast to (a), the more realistic scenario in
(b), with compound flocs, never achieves saturation in the weak Earth-like ambient fields used for the experiment (modified from Tauxe et al. (2006)).
70 70 that only small decreases in water content are necessary to lock-in

the pDRM, which suggests that shallow lock-in with minimal delays
60 in paleomagnetic recording might be expected to be common
60 (Kent, 1973). The observed linear relationship between pDRM and
50 applied field in such experiments (Kent, 1973; Verosub et al., 1979;
Intensity (mA/m)
Inclination (°)
Barton et al., 1980; Tucker, 1980) would also be expected to make
40 50 such sediments suitable for relative paleointensity studies. Never-
theless, as indicated above, the pDRM/sIRM ratio in natural sedi-
30
ments is always much lower than in laboratory re-deposition
40 experiments, which means that these simple and elegant experi-
20
ments do not capture the full complexity of sedimentary magne-
10 tizations. Katari et al. (2000) argued that, because these re-
30
deposition experiments were carried out with deionized water or
0 with added anti-coagulants, which inhibited flocculation, they do
0 4 8 not provide good analogues for understanding remanence acqui-
Salinity (ppt) sition in marine sediments.
Despite the fact that a pDRM is considered to be an important
Fig. 7. Illustration of the effect of variations in water salinity on DRM intensity and remanence acquisition mechanism, verification of the reality of
paleomagnetic inclination. The relationship between DRM intensity and salinity for
pDRM and quantification of lock-in depth have proved elusive.
synthetic sediment composed of a mixture of kaolinite and maghemite is indicated by
the solid black line, while inclination versus salinity is indicated by the dashed red line. Studies that have been widely cited as providing evidence for pDRM
The heavy red line represents the inclination of the applied field (67 ) during the re- acquisition (e.g. Lund and Keigwin, 1994; Kent and Schneider, 1995)
deposition experiment (data are from Van Vreumingen (1993)). were critically assessed by Tauxe et al. (2006). In the case of Lund
and Keigwin (1994), Tauxe et al. (2006) argued that paleomag-
5.3. Critical assessment of the classical pDRM concept netic records from marine sediments at Bermuda Rise and lake
sediments from Minnesota could be correlated without invoking
In Section 5.1, we briefly described classical concepts concerning smoothing due to pDRM acquisition. In the case of Kent and
DRM and pDRM acquisition in sediments, and we introduced the Schneider (1995), removal of one data set with a poor chronology
important complications that arise from particle flocculation in removed the basis for invoking paleomagnetic smoothing associ-
Section 5.2. We now provide a more detailed, and critical, assess- ated with pDRM acquisition.
ment of the evidence for pDRM acquisition in sediments. The pDRM Several attempts have been made to quantify the pDRM lock-in
concept is based on laboratory experiments in which wet sedi- depth for sediments. The pDRM lock-in depth is usually estimated
ments are re-deposited in a controlled ambient field, and by plotting the depth difference between two stratigraphic markers
progressively dried with or without stirring to simulate post- of known age versus sedimentation rate for a number of locations.
depositional sedimentary remanence acquisition processes (e.g. deMenocal et al. (1990) compiled the depth differences between
Irving and Major, 1964; Kent, 1973). The resultant magnetizations the MatuyamaeBrunhes boundary and MIS 19 (n ¼ 9) and
record no inclination error, which is consistent with paleomagnetic concluded that the lock-in depth in marine sediments is, on
observations from bioturbated sediments. Additionally, it is argued average, about 16 cm. Tauxe et al. (1996) re-explored this
a b c
H H
ΔI H
N = 21 N = 12 N =7
ΔI ΔI
ΔI
M
N = 13 M N = 17
N =9
M
M 1400 M M
2000 1 1 2000 1
Floc Count
1000 1500
1500
H H H
1000 1000
600
500 0 0 500 0
200
0 20 40 60 80 100 0 20 40 60 80 100 120 0 200 400 600 800

Floc Size (μm) Floc Size (μm) Floc Size (μm)
Fig. 8. Numerical simulation of how paleomagnetic intensity and inclination in sediments are affected by floc size. Lower: the solid blue and red dashed lines represent floc
distributions in simulated M and H groups, respectively, where M represents flocs that respond to magnetic torques and H represents flocs that are dominated by hydrodynamic
torques. A northward-directed applied field with 45 inclination and 45 mT intensity was used for the simulations. In equal area projections of floc moments for the M (upper) and H
(lower) groups, no distinction is made between hemispheres; the lower plots are normalized by the maximum concentration of flocs. High and low concentrations are indicated by
darker and lighter shading, respectively, and homogeneity (or lack of alignment) of moment direction is indicated by mid tone shading over the entire stereographic projection (as
in c). For the M group, small dark areas indicate that the majority of moments are well aligned with the applied field. Upper insets: schematic representations of three cases. Left:
blue dots and red squares in stereographic projections represent individual floc directions from the M and H groups, respectively. The orange cross is the applied field direction.
Right: blue and red arrows represent the recorded paleomagnetic inclinations for the M and H groups. The black arrow is the resultant, and DI is the inclination flattening. (a) For
small floc sizes, most flocs are in group M; few flocs are in group H, and DI is small. (b) For larger floc sizes, more flocs are in group H and DI increases. (c) For the largest floc sizes,
most flocs are in group H and are so large that they are oriented randomly with respect to the field. The small number of group M flocs is sufficient for DI to become small. The net
magnetic moment decreases from (a) to (c) because of the less efficient alignment of increasingly larger flocs (modified from Mitra and Tauxe (2009)).
relationship with a larger data set (n ¼ 19) and concluded that the forward numerical simulations and inverse parameter estimations
lock-in depth in marine sediments is negligible: 2.2 cm on average and determined that the best-fit pDRM filter function, surprisingly,
for all data, and only 1.0 cm if the most problematical data are has a Gaussian form. Conventionally, pDRM lock-in has been
removed from the analysis. Liu et al. (2008) subsequently demon- assumed to develop with progressive sediment compaction and
strated that such analyses depend crucially on construction of dewatering, which is expected to proceed in an exponential
precise millennial chronologies. The foraminiferal d18O records manner during early burial (Lambe and Whitman, 1969). The
used for such analyses depend on the oceanic water masses in Gaussian filter function estimated by Suganuma et al. (2011) is the
which the foraminifera dwelt. The potential presence of different only empirically determined filter function reported to date, and
water masses at different water depths, in different ocean basins, suggests that compaction and dewatering may not be the most
and water-mass dependent differences in d18O signals in benthic important early processes associated with pDRM acquisition,
and planktic foraminifera mean that such analyses can be affected although they must be important in the lower part of the lock-in
by age offsets of the order of several kyr (e.g. Skinner and zone. This result also demonstrates that we know too little about
Shackleton, 2005). Liu et al. (2008) argued that such problems pDRM acquisition.
can be overcome by restricting inter-core comparisons to planktice If a pDRM is an important remanence acquisition mechanism, it
planktic or benthicebenthic d18O correlations for sediments would be surprising if the lock-in depth was to have a fixed value
deposited beneath the same water masses. This substantially throughout a given sedimentary sequence. Sagnotti et al. (2005)
decreases the available global data set for such comparisons. No provide a rare glimpse of this likelihood. The maximum depth of
suitably large global data set has been subsequently constructed. bioturbation in the surface mixed layer is expected to be a key
Any meaningful assessment of the pDRM lock-in depth requires determinant of the lock-in depth. Burrowers focus their activity on
analysis of paired local datasets. Liu et al. (2008) analysed one such surface sediments where there is an ample supply of high-grade
pair of sediment cores with modern high-quality paleomagnetic organic matter (Gage and Tyler, 1991). Trauth et al. (1997)
and benthicebenthic d18O results from the North Atlantic Ocean demonstrated that it is the carbon flux that determines the bio-
(Venz et al., 1999; Channell and Kleiven, 2000), and concluded that turbational mixing depth, and that time-variable mixing due to
the recorded pDRM has a lock-in depth of 23 6 cm. Liu et al. changes in productivity and organic carbon flux to the seafloor are
(2008) concluded that most of the depth offset associated with important for obtaining high-resolution chronologies from sedi-
the pDRM is associated with the depth of the surface mixed layer ments. This is likely to be particularly important for high-resolution
(due to bioturbation), which in this region is estimated at 10e20 cm paleomagnetic studies of sediments, and remains poorly con-
(Thomson et al., 2000) and that the additional depth offset due to strained. The potential influence of time-variable lock-in, as well as
pDRM acquisition was only about 5 cm. This conclusion is consis- the influence of sedimentation rate variations (e.g. Roberts and
tent with the earlier analysis of Channell et al. (2004) and Channell Winklhofer, 2004), ought to be borne in mind when interpreting
and Guyodo (2004) from the same sequences, but based on apparently continuous paleomagnetic records from sediments.
systematic offsets in recording of the MatuyamaeBrunhes
boundary and of multiple paleomagnetic reversals through the 5.4. Effects of pDRM acquisition on paleointensity signals: model
Matuyama Chron. Tauxe et al. (2006) critiqued the approach of results
Channell et al. (2004) by arguing that the millennial-scale age
offsets documented in the studied North Atlantic drift deposits are Based on the above evidence, which is not as extensive as one
due to reworking of foraminifera. However, as argued by Liu et al. would expect given that a pDRM has been postulated as a realistic
(2008), the d18O signals in question are carried by benthic forami- remanence acquisition mechanism for 50 years, we now provide
nifera, which live within the sediment and are unlikely to have results of numerical simulations to illustrate the potential effects of
been reworked. We, therefore, consider the evidence from these pDRM acquisition and lock-in depth variations. Any such treatment
North Atlantic sites to be indicative of pDRM acquisition. involves significant assumptions because much remains unknown
Other evidence for pDRM acquisition is outlined as follows. about pDRM acquisition. Most pDRM models assume that rema-
Sagnotti et al. (2005) analysed two sediment cores from the Gulf of nence lock-in can only begin once substantial surface mixing has
Salerno, and correlated the cores using high-resolution environ- ceased. Bioturbation in the surface mixed layer will cause a delay
mental magnetic measurements. Within this tight stratigraphic between sediment deposition and remanence recording. The
framework, they observed that well-defined paleomagnetic direc- extent of the delay will depend on the sedimentation rate and the
tional features are variably offset with respect to correlative thickness of the surface mixed layer, which has been argued to have
susceptibility features between the two cores. No clear pattern was an average of w10 cm for marine sediments, with a standard
observed, with paleomagnetic recording in one core leading the deviation of 4.5 cm and minimum and maximum values of 2 and
signal recorded in the other core in some stratigraphic intervals, 30 cm, respectively (e.g. Boudreau, 1994, 1998). Various lock-in
and lagging it in others. The maximum amplitude of these offsets functions have been used in pDRM acquisition models (e.g. Løvlie,
was between 12 cm and þ15 cm. Sagnotti et al. (2005) concluded 1976; Hamano, 1980; Otofuji and Sasajima, 1981; Kent and
that such stratigraphically variable lock-in depths cause centennial- Schneider, 1995; Meynadier and Valet, 1996; Bleil and von
scale uncertainty in age models for such rapidly deposited sedi- Dobeneck, 1999; Spassov et al., 2003; Channell and Guyodo,
ments. The possibility of such age offsets needs to be taken into 2004; Roberts and Winklhofer, 2004; Suganuma et al., 2011).
account when using paleomagnetic records for dating sedimentary Remanence lock-in is assumed to result from progressive
sequences. Using more slowly deposited sediments, Suganuma compaction and dewatering, with expulsion of interstitial water
et al. (2010, 2011) observed offsets between the relative paleo- increasing inter-particle friction to overcome the realigning
intensity minimum associated with the MatuyamaeBrunhes geomagnetic torque on a magnetic particle and fixing the magne-
boundary and its precursor (e.g. Hartl and Tauxe, 1996) and tization. An exponential lock-in function (Fig. 9) has been used
a maximum in 10Be flux in marine sediment cores. They reported most commonly for pDRM modelling (e.g. Løvlie, 1974; Hamano,
a pDRM lock-in depth of 17 cm. Suganuma et al. (2010) argued that 1980; Otofuji and Sasajima, 1981; Kent and Schneider, 1995;
15e17 cm offsets associated with pDRM lock-in can cause age Meynadier and Valet, 1996; Roberts and Winklhofer, 2004) because
differences of more than 10 kyr in these slowly deposited (1e2 cm/ sediment consolidation is expected to proceed in an exponential
kyr) carbonate sediments. Suganuma et al. (2011) carried out manner in early stages (Lambe and Whitman, 1969). However, the
PDRM locked in (0−100%) This is a much more efficient function; rapid lock-in produces much
less temporal delay and less signal distortion or smoothing. The
100 80 60 40 20 0
cubic function also produces complete lock in within the lock-in
zone, whereas an exponential function approaches saturation
asymptotically, but never reaches it, which is not physically
realistic.
Surface mixed layer: Increasing temporal offsets for progressively slower sedimen-
no lock-in tation rates and deeper lock-in depths (Fig. 10) illustrate the con-
trasting recording efficiency of Gaussian and cubic lock-in
functions. These model results also illustrate the time offset that is
introduced into paleomagnetic records by pDRM acquisition. When
sedimentary paleomagnetic records are only used to construct
Depth below surface
magnetic polarity records, such offsets are usually unimportant.

However, in the modern era in which continuous high-resolution
paleointensity records are prized for providing chronological
constraints to assess leads and lags between millennial scale
climatic variations, such offsets become a crucial limitation and
must be reckoned with as part of the age uncertainty of such
Lock-in depth
studies. Our ignorance of details of the remanence acquisition
mechanism, and its possible variability with time (e.g. Sagnotti
et al., 2005), is a key limiting factor for paleointensity-assisted
chronology. This ignorance illustrates the crucial importance of
constant making concerted efforts to understand better the processes by
linear which sediments become magnetized. Improved understanding of
cubic pDRM lock-in could then lead to the use of modelling to provide
Gaussian probabilistic assessment of uncertainties in high-resolution sedi-
exponential ment age models. These uncertainties are reduced in environments
with the highest sedimentation rates, but it is precisely such
environments that are targetted to resolve details of climate forcing
Fig. 9. Illustration of lock-in filter functions used for modelling pDRM acquisition in
and response. Probabilistic assessment of uncertainties is applied
sediments. No pDRM lock-in is assumed within the surface mixed layer. The curves in
the lock-in zone illustrate the cumulative lock-in for constant, linear, cubic, Gaussian routinely when constructing high-resolution radiocarbon chro-
and exponential functions. The pDRM is completely locked in at the base of the lock-in nologies (e.g. Ramsey, 2008), and could be readily adapted to bring
zone, except for the Gaussian and exponential functions, which never reach full lock-in greater rigour to uncertainty assessment in paleointensity-based
(they are, therefore, truncated at þ3.5 standard deviations and 99.9% at the base of the
age models.
lock-in zone, respectively). Numerical pDRM calculations are made by dividing the
lock-in zone into discrete depth slices, convolving the geomagnetic input signal with
the lock-in function and summing the magnetizations at each depth interval (for 5.5. The effects of sediment type on paleointensity signal recording
details, see Roberts and Winklhofer (2004)).
Sedimentological factors are potentially important for paleo-
only empirically determined lock-in function is the Gaussian (Fig. 9) intensity signal recording. For example, do clay-rich sediments
function of Suganuma et al. (2011). For the purposes of illustration, have different recording characteristics than carbonate-rich sedi-
we present model results for Gaussian and cubic lock-in functions ments? Few studies have directly addressed this question (e.g.
in Fig. 10. There is no empirical or theoretical support for a cubic Carter-Stiglitz et al., 2006). In re-deposition experiments in known
function. It has been used (e.g. Roberts and Winklhofer, 2004) magnetic fields with variable water salinities, with and without
simply because the pDRM locks in more rapidly and at shallower deflocculants, Spassov and Valet (2012) tested for differences in
depths than with other commonly used functions (Fig. 9). It, magnetic recording quality for carbonate and clay-rich sediments.
therefore, constrains best-case recording scenarios, which can be They found that flocculation effects are dominant in clay-rich
used to understand and illustrate the likely minimum effects of sediments, while carbonates are less affected by flocculation and
signal filtering associated with pDRM acquisition. have more efficient magnetizations with a linear relationship
We use the same approach as Roberts and Winklhofer (2004) to between magnetizing field and magnetization. Carbonates, there-
model pDRM acquisition and do not repeat the details of the fore, should be ideal paleointensity recorders. Nevertheless,
procedure here. We use the high-frequency relative paleointensity generally slow sedimentation rates mean that carbonate paleo-
signal from ODP Site 983 (Channell, 1999) as an input geomagnetic intensity records are normally more smoothed and have lower
signal. The paleomagnetic signal at Site 983 is interpreted to be resolution (e.g. Guyodo and Channell, 2002; and references
a pDRM (e.g. Channell and Guyodo, 2004; Channell et al., 2004). If therein). Spassov and Valet (2012) confirmed the importance of
so, it will be a filtered representation of the geomagnetic field. It is flocculation, and suggested that lithological factors are important
the highest resolution record available, so we use it as an input for paleointensity signal recording. The likely importance of litho-
paleointensity signal to simulate scenarios with lower sedimenta- logical effects means that further experimental work is needed to
tion rates than those at Site 983. In Fig. 10, we illustrate how the understand better any differences between paleomagnetic
input paleointensity signal is recorded after filtering by Gaussian recording in carbonates and clay-rich sediments and the impor-
(Fig. 10a) and cubic (Fig. 10b) lock-in functions at different sedi- tance of flocculation in clay-rich sediments.
mentation rates for lock-in zones with 10 and 20 cm thicknesses,
respectively. The Gaussian function is symmetric with 50% of the 5.6. Diagenesis
remanence recorded at half the lock-in depth (Fig. 9). In contrast,
50% of the remanence is recorded at w16% of the lock-in depth for Burial of organic matter can have a controlling impact on the
the cubic function and w90% is recorded at half the lock-in depth. magnetic properties of sediments. Assessing the influence and
a Paleointensity (µT) 10 cm Gaussian lock-in 20 cm Gaussian lock-in

0 40 80 120 10 cm/kyr 3 cm/kyr 1 cm/kyr 10 cm/kyr 3 cm/kyr 1 cm/kyr
200
250
300
Age (ka)
350
400
450
500
Offset 1.5 kyr 5.0 kyr 14.9 kyr 2.0 kyr 6.6 kyr 19.9 kyr
b Paleointensity (µT) 10 cm cubic lock-in 20 cm cubic lock-in

0 40 80 120 10 cm/kyr 3 cm/kyr 1 cm/kyr 10 cm/kyr 3 cm/kyr 1 cm/kyr
200
250
300
Age (ka)
350
400
450
500
Offset 1.2 kyr 3.8 kyr 11.2 kyr 1.3 kyr 4.2 kyr 12.8 kyr
Fig. 10. Illustration of the delays and smoothing of paleointensity signals associated with pDRM recording. In each case, the high-resolution paleointensity record from ODP Site 983
(Channell, 1999) is used as the geomagnetic input signal. This signal is attenuated with delayed recording, as illustrated for a range of sedimentation rates for: (a) a Gaussian lock-in
function (see Suganuma et al., 2011), and (b) a cubic lock-in function (see Roberts and Winklhofer, 2004). In each case, results are illustrated for lock-in depths (see Fig. 9) of 10 and
20 cm, with sedimentation rates of 1 cm/kyr (typical of pelagic carbonates), 3 cm/kyr, and 10 cm/kyr. See text for discussion.
extent of diagenesis is, therefore, a routine part of sedimentary rates in oxic sedimentary environments make them an infrequent
paleomagnetic studies. When organic carbon flux is low, oxygen target for paleointensity investigations. In settings with moderate
can diffuse into the sediment and organisms, including microbes, organic carbon fluxes, molecular oxygen is progressively consumed
will consume or oxidize the organic matter so that organic matter through microbial degradation of organic matter until virtually no
diagenesis does not progress far. In such settings, oxygen diffusion oxygen remains (i.e. the suboxic diagenetic zone). Organic carbon
into the sediment can partially oxidize detrital magnetite to form degradation then proceeds though microbial use and eventual
maghemite (e.g. Cui et al., 1994) or titanomaghemite (e.g. Xuan and depletion or consumption of a sequence of the most efficient
Channell, 2010) shells on the surface of (titano)magnetite grains. remaining oxidants. Nitrate and labile manganese oxides are
The paleomagnetic complexities that result from oxidation can progressively reduced, followed by iron in reactive iron oxides
make oxic sediments unsuitable for paleointensity studies (e.g. (these reactions all occur under suboxic conditions; see Froelich
Xuan and Channell, 2010). Regardless, generally low sedimentation et al., 1979), followed by sulphate reduction and methanogenesis
(under anoxic conditions). The depths at which these reactions paleointensity studies discuss diagenetic complications much less
occur depend principally on organic carbon flux and sedimentation often than other sedimentary paleomagnetic studies.
rate, which vary from setting to setting.
Iron-bearing minerals are strongly affected by organic matter 5.7. Summary
diagenesis (Berner, 1981). Magnetite and other iron oxides can
undergo dissolution (e.g. Karlin and Levi, 1983, 1985; Canfield and Given the uncertainties described above, it might be surprising
Berner, 1987), while iron sulphides, including pyrite and greigite that relative paleointensity signals have so much apparently robust
(e.g. Berner, 1984; Roberts and Turner, 1993), can grow authigeni- structure that can be globally or regionally correlated (see Sections
cally at their expense. Dissolution of highly reactive iron-bearing 3 and 4). The inescapable conclusion from Section 5 is that much
minerals (i.e. ferric hydrous oxide, lepidocrocite and ferrihydrite) remains to be learned about how paleomagnetic, including paleo-
occurs in the iron reduction zone (Poulton et al., 2004), whereas intensity, signals are recorded by sediments. We urgently need to
magnetite dissolution is ubiquitous in the sulphate reduction zone develop a better theoretical, numerical, and experimental under-
(Canfield and Berner, 1987) and can lead to near-total dissolution of standing of the physics of sedimentary remanence acquisition to
the paleomagnetically useful detrital magnetic mineral assem- underpin high-resolution paleomagnetic studies of sediments.
blage. Paleointensity analyses usually target sediments deposited at
rates above 1 cm/kyr, and preferably tens of cm/kyr. At these sedi- 6. Relative paleointensity normalization
mentation rates, diagenetic iron reduction and sulphate reduction
can become paleomagnetically important. Many attempted paleo- As discussed in Section 2, relative paleointensities are estimated
intensity studies have been compromised by magnetite dissolution by normalizing the NRM of a sediment by an artificial laboratory-
in sulphate reducing environments. It is impossible to obtain pale- induced magnetization to cancel the influence of non-
ointensity results when no detrital magnetite population remains, geomagnetic mineralogical variations on the NRM. Readers are
so sulphate reducing diagenetic environments can be ruled out as referred to Levi and Banerjee (1976), King et al. (1983), Tauxe
useful targets for such studies. Of greater interest, therefore, is the (1993), and Tauxe and Yamazaki (2007) for details of the ratio-
effect of iron reduction on paleointensity signals. nale behind relative paleointensity normalization. It is well recog-
Few studies have explicitly addressed the effects of iron reduc- nized that paleointensity normalization is a “brute force” method
tion on paleointensity records. Tarduno et al. (1998) made detailed that will have complexities associated with various factors. In this
rock magnetic analyses across the iron reduction front in carbonate section, we provide an up-to-date perspective on some of these
sediments from the Ontong-Java Plateau. They reported a shift in complicating factors.
the distribution of hysteresis properties and a slight shift in the
peak coercivity below the Fe redox front. They attributed this 6.1. Flocculation and linear versus non-linear magnetic recording
change to the occurrence of magnetite-producing magnetotactic regimes
bacteria within the sediment at this depth. If true, this observation
will have fundamentally important implications for the timing of A fundamental assumption in relative paleointensity studies is
acquisition of the paleomagnetic signal, which Tarduno et al. (1998) that the recorded magnetization (DRM or pDRM) is related linearly
referred to as a biogeochemical remanent magnetization. The small to the strength of the magnetizing field. As stated in Section 2, we
change in anhysteretic remanent magnetization (ARM) coercivity can only assume this and cannot test whether the assumed linear
that they reported across this boundary, however, is unlikely to relationship holds. In Section 5.2, we described the likely impor-
significantly affect relative paleointensity normalizations. tance of sediment flocculation for sedimentary remanence acqui-
Yamazaki and Solheid (2011) reported a similar shift in hysteresis sition and presented model results (Fig. 6) for how flocculated
properties to that reported by Tarduno et al. (1998) across the Fe composite aggregations of clay and magnetite particles respond to
redox front, which they attributed to dissolution of a surficial Earth-like magnetic fields. Tauxe et al. (2006) argued that such flocs
maghemite skin on magnetite particles. By comparing the paleo- respond in a significantly non-linear manner to the magnetizing
intensity record from the studied core with other paleointensity field. They also suggested that current methods for normalizing
records, Yamazaki and Solheid (2011) concluded that maghemite sedimentary remanence records do not take into account changes
reduction at the Fe redox boundary does not significantly affect in floc size and can, therefore, only be partially effective in isolating
their relative paleointensity estimation. They also interpreted the the geomagnetic contribution to the NRM. Tauxe et al. (2006)
lack of discordance between their paleointensity record and others attributed the unexplained scatter seen in relative paleointensity
as evidence for lack of a significant biogeochemical remanence as records (e.g. Fig. 4) to unaccounted for changes in floc size within
suggested by Tarduno et al. (1998). The possible influence of the analysed sediments. This conclusion indicates that we urgently
a biogeochemical remanent magnetization (Tarduno et al., 1998; require methods to assess the presence and size of flocs, particu-
Abrajevitch and Kodama, 2009; Roberts et al., 2011, 2012) is best larly in clay-rich sediments, and a stronger empirical and numerical
gauged by direct observations, or by comparison with other pale- understanding of the effects of flocs on sedimentary paleointensity
ointensity records. The proposed lack of influence of iron reduction records. Such knowledge is needed to assess whether a particular
on relative paleointensity estimation (Yamazaki and Solheid, 2011) floc size distribution will place a paleomagnetic record within the
needs to be tested at other locations because a coercivity shift for non-linear recording regime. Any such work is likely to require
a magnetic mineral assemblage might normally be expected to a level of sedimentological knowledge that is currently unusual in
influence relative paleointensity normalizations. paleointensity studies and a fundamental shift in the way that
Overall, although diagenesis will affect all sedimentary paleo- paleointensity investigations are undertaken.
magnetic records to some extent, it can be argued that successful
paleointensity studies are the least likely to be strongly affected by 6.2. Effects of mixed biogenic and detrital magnetite populations
diagenesis. The requirement for excellent paleomagnetic stability
and the need to pass strict rock magnetic selection criteria mean The rock magnetic selection criteria that have been developed
that environments that have been strongly affected by diagenetic for relative paleointensity studies (see Section 2) require that
magnetic mineral alteration will not normally yield robust paleo- magnetite is the only magnetic mineral present and that it must
intensity results. This probably also explains why relative occur within a narrow grain size and concentration range (King
et al., 1983; Tauxe, 1993). Most rock magnetic methods that have 2007b; Xuan and Channell, 2008). There is yet to be a convincing
been used to assess grain size variations in relative paleointensity demonstration that orbital periodicities recorded in normalized
studies are indicative of bulk magnetic properties rather than remanence records reflect orbital energization of the geomagnetic
providing a detailed view of the presence and grain size distribu- field. Nevertheless, the fact that orbital periodicities have
tions of different magnetic mineral components within the sedi- contaminated paleointensity signals indicates that imperfections in
ment. Stratigraphic variation in the magnetic contribution of normalization methods need to be better understood (see detailed
different components can potentially bias paleointensity normali- discussion by Xuan and Channell (2008)).
zation. For example, with the advent and routine application of
magnetic techniques that enable discrimination of different 6.4. Magnetostatic interactions
magnetic mineral components within a sediment (Roberts et al.,
2000; Egli, 2004; Weiss et al., 2004; Kopp et al., 2006; Egli et al., A key empirical criterion for selecting sediments for relative
2010), magnetite magnetofossils (i.e. the inorganic remains of paleointensity investigations requires that the magnetite concen-
magnetite-producing magnetotactic bacteria) have been routinely tration in a sediment should not vary by more than a factor of 10
identified in addition to a detrital magnetite fraction. Such (e.g. King et al., 1983; Tauxe, 1993). The purpose of this criterion is
magnetic mineral assemblages are particularly common in pelagic to avoid the effects of magnetostatic interactions among particles,
carbonate sediments (e.g. Abrajevitch and Kodama, 2009; Roberts which increase with larger magnetite concentrations, and can
et al., 2011, 2012; Larrasoaña et al., 2012). As pointed out by significantly affect paleomagnetic recording fidelity (e.g. Sugiura,
Roberts et al. (2012), this could have a significant impact on relative 1979; Muxworthy et al., 2003; Heslop et al., 2006). By using first-
paleointensity estimations for several reasons. First, the potential order reversal curve (FORC) diagrams (Pike et al., 1999; Roberts
presence of a stratigraphically variable biogeochemical remanent et al., 2000), Yamazaki (2008) argued that non-interacting
magnetization could cause variable recording fidelity with respect biogenic magnetite occurs commonly in pelagic carbonates, along
to any DRM or pDRM associated with the detrital magnetic mineral with a more strongly interacting detrital magnetite component.
component. Second, the empirical framework for relative paleo- Observed glacialeinterglacial variations in the concentration of
intensity determination applies to magnetite in the 1e15 mm size these components will have a stratigraphically variable effect on
range (King et al., 1983; Tauxe, 1993). Magnetite magnetofossils ARM acquisition, which could then affect ARM normalizations used
occur in the 30e100 nm size range, for which there is no empirical for relative paleointensity estimations. This could also give rise to
evidence for linearity of NRM with respect to magnetizing field or coherence between the normalized remanence and the normal-
with respect to the NRM fraction recorded by magnetite particles in izing parameter, which is known to contaminate some paleo-
the 1e15 mm size range. The fact that magnetofossils occur in the intensity records (see Section 6.3). While it is useful to assess
magnetically ideal single domain (SD) size range means that they whether magnetostatic interactions influence relative paleointen-
will contribute more strongly than larger pseudo-single-domain sity normalizations, we argue that the approach of Yamazaki (2008)
(PSD) particles to the ARM, which is a widely used laboratory- is probably not valid. The vertical spread in FORC diagrams that is
induced magnetization for paleointensity normalization. If there used to assess the importance of magnetostatic interactions (Pike
are two different grain size fractions (detrital and biogenic) in et al., 1999; Roberts et al., 2000) has a different origin in SD
a sediment, different, and potentially stratigraphically variable, compared to multi-domain (MD) magnetic materials. Magneto-
concentrations of these components means that standard methods static interactions among SD particles result from the close prox-
of paleointensity normalization will be complicated in ways for imity of particles whose magnetic moments physically interact
which there are currently no correction methods. Roberts et al. with each other. In contrast, PSD and MD materials give rise to an
(2012) suggested that this might explain some difficulties inherent vertical spread in FORC diagrams due to processes internal
encountered in paleointensity studies, particularly in pelagic to a particle, such as vortex structures, domain wall interactions,
carbonates. and nucleation, annihilation and pinning of domain walls (e.g. Pike
et al., 2001), rather than the particleeparticle interactions assumed
6.3. Contamination of normalized remanence records by climatic by Yamazaki (2008). We, therefore, argue that Yamazaki (2008) did
variability not document different magnetostatic interaction regimes, but
rather the inherent characteristics of two magnetite grain size
With increased publication of relative paleointensity records distributions (biogenic and detrital). The issues that arise in relation
over the last 20 years, orbital (or near-orbital) periodicities have to paleointensity signal recording then revert to the type discussed
sometimes been documented within paleointensity signals (e.g. in Section 6.2. If our reasoning is correct, the effects of magneto-
Channell et al., 1998; Yamazaki, 1999; Yamazaki and Oda, 2002), static interactions remain largely un-assessed in relation to their
while others have shown that near-orbital periodicities are not potential importance in relative paleointensity investigations.
coherent with climatic records (e.g. Tauxe and Shackleton, 1994).
This has led to revival of the debate about whether the geomagnetic 6.5. Summary
field is powered not only by an internal dynamo mechanism, due to
thermal and compositional convection in Earth’s electrically con- As outlined above, several phenomena can complicate relative
ducting liquid outer core, but also by an external orbital component paleointensity estimation. Paleointensity practitioners need
(e.g. Malkus, 1968). Critical analysis of such claims has led to improved information of several types to isolate these effects to
identification of orbital contamination of relative paleointensity avoid, or correct for, these effects. For example, routine sedi-
signals due to failure of normalization procedures to remove subtle mentological analysis of clay-rich sediments is needed to assess
expressions of climatically forced lithological variations from the the presence and size distribution of flocs. Improved rock
paleointensity signal (e.g. Guyodo et al., 2000). In other cases, magnetic methods are needed to detect and screen for non-linear
claims of orbital control on the geodynamo have been based on recording regimes associated with large flocs. Magnetic methods
spectral analyses (e.g. Yamazaki and Oda, 2002) that are not that enable determination of the grain size distribution of
statistically significant and that have highly variable phase magnetic mineral components in a sediment need to be routinely
compared to the consistent orbital modulation expected for used rather than bulk magnetic methods that cannot discriminate
a genuinely orbitally forced signal (Roberts et al., 2003; Heslop, between different magnetite populations or their stratigraphic
variations. Placing paleointensity analysis on a sound theoretical Anson, G.L., Kodama, K.P., 1987. Compaction-induced inclination shallowing of the
post-depositional remanent magnetization in a synthetic sediment. Geophys.
and experimental footing requires development of methods
J. R. Astron. Soc. 88, 673e692.
that can identify and correct for effects that cause non-ideal Arason, P., Levi, S., 1990. Compaction and inclination shallowing in deep-sea sedi-
normalization. ments from the Pacific Ocean. J. Geophys. Res. 95, 4501e4510.
Barletta, F., St-Onge, G., Stoner, J.S., Lajeunesse, P., Locat, J., 2010. A high-resolution
Holocene paleomagnetic secular variation and relative palaeointensity stack
7. Conclusions and future directions from eastern Canada. Earth Planet. Sci. Lett. 298, 162e174.
Barton, C.E., McElhinny, M.W., Edwards, D.J., 1980. Laboratory studies of deposi-
tional DRM. Geophys. J. R. Astron. Soc. 61, 355e377.
Developments in relative paleointensity studies over the last Belmonte, A., Eisenberg, H., Moses, E., 1998. From flutter to tumble: inertial drag
two decades provide much new information about dynamic geo- and Froude similarity in falling paper. Phys. Rev. Lett. 81, 345e348.
magnetic field behaviour and provide a reference signal that has Berner, R.A., 1981. A new geochemical classification of sedimentary environments.
J. Sediment. Petrol. 51, 359e365.
become widely used for dating sediments, often at millennial or Berner, R.A., 1984. Sedimentary pyrite formation: an update. Geochim. Cosmochim.
higher resolution. The continuity of data provided by sedimentary Acta 48, 605e615.
relative paleointensity studies provides a level of detail and Bleil, U., von Dobeneck, T., 1999. Geomagnetic events and relative paleointensity
records e clues to high-resolution paleomagnetic chronostratigraphies of Late
temporal continuity that is unavailable from absolute paleointen-
Quaternary marine sediments? In: Fischer, G., Wefer, G. (Eds.), Use of Proxies in
sity records. This detail demonstrates the global coherence and Paleoceanography: Examples from the South Atlantic. SpringereVerlag, Berlin,
dynamism of field variability, with the field often collapsing to low pp. 635e654.
Boudreau, B.P., 1994. Is burial velocity a master parameter for bioturbation? Geo-
values; these paleointensity minima are often accompanied by
chim. Cosmochim. Acta 58, 1243e1249.
geomagnetic excursions (e.g. Laj and Channell, 2007; Roberts, Boudreau, B.P., 1998. Mean mixed depth of sediments: the wherefore and the why.
2008), which supports the view that excursions are much more Limnol. Oceanogr. 43, 524e526.
frequent features of geomagnetic variability than was once thought Brachfeld, S., Domack, E., Kissel, C., Laj, C., Leventer, A., Ishman, S., Gilbert, R.,
Camerlenghi, A., Eglington, L.B., 2003. Holocene history of the Larsen-A Ice Shelf
(Fig. 1). Our knowledge of geomagnetic field behaviour would be constrained by geomagnetic paleointensity dating. Geology 31, 749e752.
much poorer without detailed relative paleointensity records. Cande, S.C., Kent, D.V., 1995. Revised calibration of the geomagnetic polarity
Ongoing work will inevitably lead to development of global timescale for the late Cretaceous and Cenozoic. J. Geophys. Res. 100, 6093e
6095.
paleointensity stacks further back in time, with greater global Canfield, D.E., Berner, R.A., 1987. Dissolution and pyritization of magnetite in anoxic
coverage to enable assessment of dipolar versus non-dipolar marine sediments. Geochim. Cosmochim. Acta 51, 645e659.
signals, with greater resolution to understand the effects of Carter-Stiglitz, B., Valet, J.-P., LeGoff, M., 2006. Constraints on the acquisition of
remanent magnetization in fine-grained sediments imposed by redeposition
smoothing on the paleomagnetic record and to reconstruct better experiments. Earth Planet. Sci. Lett. 245, 427e437.
the frequency spectrum of ancient geomagnetic variations. This Channell, J.E.T., Guyodo, Y., 2004. The Matuyama Chronozone at ODP Site 982
task will require special cases with a combination of desirable (Rockall Bank): evidence for decimeter-scale magnetization lock-in depths. AGU
Geophys. Monogr. Ser. 145, 205e219.
factors, such as high deposition rates without significant reductive
Channell, J.E.T., Kleiven, H.F., 2000. Geomagnetic palaeointensities and astrochro-
diagenesis, moderate concentrations of ideal magnetite particles, nological ages for the MatuyamaeBrunhes boundary and the boundaries of the
and superb chronologies, much like the outstanding records ob- Jaramillo Subchron: palaeomagnetic and oxygen isotope records from ODP Site
983. Phil. Trans. R. Soc. Lond. A 358, 1027e1047.
tained from the North Atlantic Ocean (Channell and Kleiven, 2000;
Channell, J.E.T., Hodell, D.A., McManus, J., Lehman, B., 1998. Orbital modulation of
Channell et al., 2000, 2009; Laj et al., 2000). the Earth’s magnetic field intensity. Nature 394, 464e468.
Despite the successes of sedimentary paleointensity studies, we Channell, J.E.T., Stoner, J.S., Hodell, D.A., Charles, C.D., 2000. Geomagnetic paleo-
remain remarkably ignorant of the physical processes by which intensity for the last 100 kyr from the sub-Antarctic South Atlantic: a tool for
inter-hemispheric correlation. Earth Planet. Sci. Lett. 175, 145e160.
sediments record paleomagnetic signals. We urgently need new Channell, J.E.T., Curtis, J.H., Flower, B.P., 2004. The MatuyamaeBrunhes boundary
knowledge to enable us to understand how sediments record interval (500e900 ka) in the North Atlantic drift sediments. Geophys. J. Int. 158,
information about the intensity of the geomagnetic field. We need 489e505.
Channell, J.E.T., Xuan, C., Hodell, D.A., 2009. Stacking paleointensity and oxygen
a better physical understanding of how sediments acquire a rema- isotope data for the last 1.5 Myr (PISO-1500). Earth Planet. Sci. Lett. 283, 14e23.
nent magnetization (with theoretical, numerical and experimental Channell, J.E.T., 1999. Geomagnetic paleointensity and directional secular variation
approaches). We need robust rock magnetic methods to enable at Ocean Drilling Program (ODP) Site 984 (Bjorn Drift) since 500 ka: compari-
sons with ODP Site 983 (Gardar Drift). J. Geophys. Res. 104, 22937e22951.
screening to help identify whether sediments under investigation Cui, Y.L., Verosub, K.L., Roberts, A.P., 1994. The effect of maghemitization on large
can be expected to have a linear relationship between magnetizing multi-domain magnetite. Geophys. Res. Lett. 21, 757e760.
field and the recorded magnetization. We need to find ways to deMenocal, P.B., Ruddiman, W.F., Kent, D.V., 1990. Depth of post-depositional
remanence acquisition in deep-sea sediments e a case-study of the
assess and take into account the presence and size distribution of BrunheseMatuyama reversal and oxygen isotopic stage-19.1. Earth Planet. Sci.
flocs in order to understand their effects on paleointensity Lett. 99, 1e13.
normalizations. Much work remains to be done to place relative Egli, R., Chen, A.P., Winklhofer, M., Kodama, K.P., Horng, C.-S., 2010. Detection of
noninteracting single domain particles using first-order reversal curve
paleointensity analyses on a secure theoretical and empirical
diagrams. Geochem. Geophys. Geosyst. 11, Q01Z11. http://dx.doi.org/10.1029/
foundation. 2009GC002916.
Egli, R., 2004. Characterization of individual rock magnetic components by analysis
of remanence curves, 1. Unmixing natural sediments. Stud. Geophys. Geod. 48,
Acknowledgements 391e446.
Elsasser, W., Ney, E.P., Winkler, J.R., 1956. Cosmic-ray intensity and geomagnetism.
We are grateful to the Editors for soliciting this paper. Our work Nature 178, 1226e1227.
Evans, H.F., Channell, J.E.T., Stoner, J.S., Hillaire-Marcel, C., Wright, J.D., Neitzke, L.C.,
has benefitted from the support of the Australian Research Council Mountain, G.S., 2007. Paleointensity-assisted chronostratigraphy of detrital
(through grant DP120103952) and the U.S. National Science Foun- layers on the Eirik Drift (North Atlantic) since marine isotope stage 11. Geo-
dation (through grant EAR 1013192). We thank an anonymous chem. Geophys. Geosyst. 8, Q11007. http://dx.doi.org/10.1029/2007GC11720.
Frank, M., Schwarz, B., Baumann, S., Kubik, P.K., Suter, M., Mangini, A., 1997.
referee for constructive review comments and Prof. Claude Hillaire- A 200 kyr record of cosmogenic radionuclide production rate and geomagnetic
Marcel for efficient editorial handling. field intensity from 10Be in globally stacked deep-sea sediments. Earth Planet.
Sci. Lett. 149, 121e129.
Froelich, P.N., Klinkhammer, G.P., Bender, M.L., Luedtke, N.A., Heath, G.R., Cullen, D.,
References Dauphin, P., Hammond, D., Hartman, B., Maynard, V., 1979. Early oxidation of
organic matter in pelagic sediments of the eastern equatorial Atlantic: suboxic
Abrajevitch, A., Kodama, K., 2009. Biochemical vs. detrital mechanism of remanence diagenesis. Geochim. Cosmochim. Acta 43, 1075e1090.
acquisition in marine carbonates: a lesson from the KeT boundary interval. Gage, J.D., Tyler, P.A., 1991. Deep-sea Biology: a Natural History of Organisms at the
Earth Planet. Sci. Lett. 286, 269e277. Deep-sea Floor. Cambridge Univ. Press, 504 pp.
Galdi, G.P., Vaidya, A., 2001. Translational steady fall of symmetric bodies in Lisé-Provonost, A., St-Onge, G., Brachfeld, S., Barletta, F., Darby, D., 2009. Paleo-
a NaviereStokes liquid, with application to particle sedimentation. J. Math. magnetic constraints on the Holocene stratigraphy of the Arctic Alaskan
Fluid Mech. 3, 183e211. margin. Glob. Planet. Change 68, 85e99.
Gee, J.S., Kent, D.V., 2007. Source of oceanic magnetic anomalies and the geomag- Lisiecki, L.E., Raymo, M.E., 2005. A PlioceneePleistocene stack of 57 globally
netic polarity timescale. In: Kono, M. (Ed.), Treatise on Geophysics. Geomag- distributed benthic d18O records. Paleoceanography 20, PA1003. http://
netism, vol. 5. Elsevier, pp. 455e507. dx.doi.org/10.1029/2004PA001071.
Gee, J.S., Cande, S.C., Hildebrand, J.A., Donnelly, K., Parker, R.L., 2000. Geomagnetic Liu, Q.S., Roberts, A.P., Rohling, E.J., Zhu, R.X., Sun, Y.B., 2008. Post-depositional
intensity variations over the past 780 kyr obtained from near-seafloor magnetic remanent magnetization lock-in and the location of the MatuyamaeBrunhes
anomalies. Nature 408, 827e832. geomagnetic reversal boundary in marine and Chinese loess sequences. Earth
Guyodo, Y., Channell, J.E.T., 2002. Effects of variable sedimentation rates and age Planet. Sci. Lett. 275, 102e110.
errors on the resolution of sedimentary paleointensity records. Geochem. Løvlie, R., 1976. The intensity pattern of post-depositional remanence acquired in
Geophys. Geosyst. 3 (8). http://dx.doi.org/10.1029/2001GC000211 some marine sediments deposited during a reversal of the external magnetic
Guyodo, Y., Valet, J.-P., 1996. Relative variations in geomagnetic intensity from field. Earth Planet. Sci. Lett. 30, 209e214.
sedimentary records: the past 200 thousand years. Earth Planet. Sci. Lett. 143, Lowrie, W., Kent, D.V., 2004. Geomagnetic polarity timescales and reversal
23e36. frequency regimes. AGU Geophys. Monogr. Ser. 145, 117e130.
Guyodo, Y., Valet, J.-P., 1999. Global changes in intensity of the Earth’s magnetic field Lu, R., Banerjee, S.K., Marvin, J., 1990. Effects of clay mineralogy and the electrical
during the past 800 kyr. Nature 399, 249e252. conductivity of water on the acquisition of depositional remanent magnetiza-
Guyodo, Y., Valet, J.-P., 2006. A comparison of relative paleointensity records of the tion in sediments. J. Geophys. Res. 95, 4531e4538.
Matuyama Chron for the period 0.75e1.25 Ma. Phys. Earth Planet. Inter. 156, Lund, S.P., Keigwin, L., 1994. Measurement of the degree of smoothing in sediment
205e212. paleomagnetic secular variation records: an example from late Quaternary
Guyodo, Y., Gaillot, P., Channell, J.E.T., 2000. Wavelet analysis of relative geomag- deep-sea sediments of the Bermuda Rise, western North Atlantic Ocean. Earth
netic paleointensity at ODP Site 983. Earth Planet. Sci. Lett. 184, 109e123. Planet. Sci. Lett. 122, 317e330.
Hamano, Y., 1980. An experiment on the post-depositional remanent magnetization Macrí, P., Sagnotti, L., Dinarès-Turell, J., Carburlotto, A., 2005. A composite record of
in artifical and natural sediments. Earth Planet. Sci. Lett. 51, 221e232. Late Pleistocene relative geomagnetic paleointensity from the Wilkes Land
Hartl, P., Tauxe, L., 1996. A precursor to the Matuyama/Brunhes transition-field Basin (Antarctica). Phys. Earth Planet. Inter. 151, 223e242.
instability as recorded in pelagic sediments. Earth Planet. Sci. Lett. 138, Macrí, P., Sagnotti, L., Lucchi, R.G., Rebesco, M., 2006. A stacked record of relative
121e135. geomagnetic paleointensity for the past 270 kyr from the western continental
Heslop, D., Witt, A., Kleiner, T., Fabian, K., 2006. The role of magnetostatic inter- rise of the Antarctic Peninsula. Earth Planet. Sci. Lett. 252, 162e179.
actions in sediment suspensions. Geophys. J. Int. 171, 1029e1035. Malkus, W.V.R., 1968. Precession of the Earth as a cause of geomagnetism. Science
Heslop, D., 2007a. Are hydrodynamic shape effects important when modelling the 160, 259e264.
formation of depositional remanent magnetization? Geophys. J. Int. 165, 775e785. Meynadier, L., Valet, J.-P., 1996. Post-depositional realignment of magnetic grains
Heslop, D., 2007b. A wavelet investigation of possible orbital influences on past and asymmetrical saw-tooth patterns of magnetization intensity. Earth Planet.
geomagnetic field intensity. Geochem. Geophys. Geosyst. 8, Q03003. http:// Sci. Lett. 140, 123e132.
dx.doi.org/10.1029/2006GC001498. Mitra, R., Tauxe, L., 2009. Full vector model for magnetization in sediments. Earth
Irving, E., Major, A., 1964. Post-depositional detrital remanent magnetization in Planet. Sci. Lett. 286, 535e545.
a synthetic sediment. Sedimentology 3, 135e143. Muxworthy, A.R., Williams, W., Virdee, D., 2003. Effect of magnetostatic interactions
Johnson, E.A., Murphy, T., Torreson, O.W., 1948. Pre-history of the Earth’s magnetic on the hysteresis properties of single-domain and pseudo-single domain grains.
field. Terr. Magn. Atmos. Electr. 53, 349e372. J. Geophys. Res. 108, 2517. http://dx.doi.org/10.1029/2003JB002588.
Karlin, R., Levi, S., 1983. Diagenesis of magnetic minerals in recent haemipelagic Nagata, T., 1961. Rock Magnetism. Maruzen.
sediments. Nature 303, 327e330. Néel, L., 1955. Some theoretical aspects of rock-magnetism. Adv. Phys. 4, 191e243.
Karlin, R., Levi, S., 1985. Geochemical and sedimentological control of the magnetic Opdyke, N.D., Henry, K.W., 1969. A test of the dipole hypothesis. Earth Planet. Sci.
properties of hemipelagic sediments. J. Geophys. Res. 90, 10373e10392. Lett. 6, 139e151.
Katari, K., Bloxham, J., 2001. Effects of sediment aggregate size on DRM intensity: Otofuji, Y., Sasajima, S., 1981. A magnetization process of sediments: laboratory
a new theory. Earth Planet. Sci. Lett. 186, 113e122. experiments on post-depositional remanent magnetization. Geophys. J. R.
Katari, K., Tauxe, L., 2000. Effects of pH and salinity on the intensity of magneti- Astron. Soc. 66, 241e259.
zation in redeposited sediments. Earth Planet. Sci. Lett. 181, 489e496. Pike, C.R., Roberts, A.P., Verosub, K.L., 1999. Characterizing interactions in fine
Katari, K., Tauxe, L., King, J., 2000. A reassessment of post-depositional remanent magnetic particle systems using first order reversal curves. J. Appl. Phys. 85,
magnetism: preliminary experiments with natural sediments. Earth Planet. Sci. 6660e6667.
Lett. 183, 147e160. Pike, C.R., Roberts, A.P., Dekkers, M.J., Verosub, K.L., 2001. An investigation of multi-
Kent, D.V., Schneider, D.A., 1995. Correlation of paleointensity variation records in domain hysteresis mechanisms using FORC diagrams. Phys. Earth Planet. Inter.
the Brunhes/Matuyama polarity transition interval. Earth Planet. Sci. Lett. 129, 126, 13e28.
135e144. Poulton, S.W., Krom, M.D., Raiswell, R., 2004. A revised scheme for the reactivity of
Kent, D.V., 1973. Post-depositional remanent magnetisation in deep sea sediment. iron (oxyhydr)oxide minerals towards dissolved sulfide. Geochim. Cosmochim.
Nature 246, 32e34. Acta 68, 3703e3715.
Kiefer, T., Sarnthein, M., Erlenkeuser, H., Grootes, P.M., Roberts, A.P., 2001. North Raisbeck, G.M., Yiou, F., Cattani, O., Jouzel, J., 2006. 10Be evidence for the
Pacific response to millennial-scale changes in ocean circulation over the last MatuyamaeBrunhes geomagnetic reversal in the EPICA Dome C ice core. Nature
60 kyr. Paleoceanography 16, 179e189. 444, 82e84.
King, J.W., Banerjee, S.K., Marvin, J., 1983. A new rock-magnetic approach to Ramsey, C.B., 2008. Deposition models for chronological records. Quat. Sci. Rev. 27,
selecting sediments for geomagnetic paleointensity studies: application to 42e60.
paleointensity for the last 4000 years. J. Geophys. Res. 88, 5911e5921. Roberts, A.P., Turner, G.M., 1993. Diagenetic formation of ferrimagnetic iron
King, R.F., 1955. The remanent magnetism of artificially deposited sediments. Mon. sulphide minerals in rapidly deposited marine sediments, South Island, New
Not. R. Astron. Soc. Geophys. Suppl. 7, 115e134. Zealand. Earth Planet. Sci. Lett. 115, 257e273.
Kopp, R.E., Weiss, B.P., Maloof, A.C., Vali, H., Nash, C.Z., Kirschvink, J.L., 2006. Chains, Roberts, A.P., Winklhofer, M., 2004. Why are geomagnetic excursions not always
clumps, and strings: magnetofossil taphonomy with ferromagnetic resonance recorded in sediments? Constraints from post-depositional remanent magne-
spectroscopy. Earth Planet. Sci. Lett. 247, 10e25. tization lock-in modelling. Earth Planet. Sci. Lett. 227, 345e359.
Korhonen, K., Donadini, F., Riisager, P., Pesonen, L.J., 2008. GEOMAGIA50: an Roberts, A.P., Lehman, B., Weeks, R.J., Verosub, K.L., Laj, C., 1997. Relative paleo-
archeointensity database with PHP and MySQL. Geochem. Geophys. Geosyst. 9, intensity of the geomagnetic field over the last 200,000 years from ODP sites
Q04029. http://dx.doi.org/10.1029/2007GC001893. 883 and 884, North Pacific Ocean. Earth Planet. Sci. Lett. 152, 11e23.
Laj, C., Channell, J.E.T., 2007. Geomagnetic excursions. In: Kono, M. (Ed.), Treatise on Roberts, A.P., Pike, C.R., Verosub, K.L., 2000. First-order reversal curve diagrams:
Geophysics. Geomagnetism, vol. 5. Elsevier, pp. 373e416. a new tool for characterizing the magnetic properties of natural samples.
Laj, C., Kissel, C., Mazaud, A., Channell, J.E.T., Beer, J., 2000. North Atlantic palae- J. Geophys. Res. 105, 28461e28475.
ointensity stack since 75 ka (NAPIS-75) and the duration of the Laschamp event. Roberts, A.P., Winklhofer, M., Liang, W.-T., Horng, C.-S., 2003. Testing the hypothesis
Phil. Trans. R. Soc. Lond. A 358, 1009e1025. of orbital (eccentricity) influence on Earth’s magnetic field. Earth Planet. Sci.
Laj, C., Kissel, C., Beer, J., 2004. High resolution global paleointensity stack since Lett. 216, 187e192.
75 kyr (GLOPIS-75) calibrated to absolute values. AGU Geophys. Monogr. Ser. Roberts, A.P., Florindo, F., Villa, G., Chang, L., Jovane, L., Bohaty, S.M., Larrasoaña, J.C.,
145, 255e265. Heslop, D., Fitz Gerald, J.D., 2011. Magnetotactic bacterial abundance in pelagic
Lal, D., 1988. Theoretically Expected Variations in the Terrestrial Cosmic-ray marine environments is limited by organic carbon flux and availability of dis-
Production Rates of Isotopes. Corso XCV, Soc. Ital. Fis., Bologna, pp. 215e233. solved iron. Earth Planet. Sci. Lett. 310, 441e452.
Lambe, T.W., Whitman, R.V., 1969. Soil Mechanics. John Wiley. Roberts, A.P., Chang, L., Heslop, D., Florindo, F., Larrasoaña, J.C., 2012. Searching for
Larrasoaña, J.C., Roberts, A.P., Chang, L., Schellenberg, S.A., Fitz Gerald, J.D., single domain magnetite in the ‘pseudo-single-domain’ sedimentary haystack:
Norris, R.D., Zachos, J.C., 2012. Magnetotactic bacterial response to Antarctic implications of biogenic magnetite preservation for sediment magnetism and
dust supply during the PalaeoceneeEocene thermal maximum. Earth Planet. relative paleointensity determinations. J. Geophys. Res. 117, B08104. http://
Sci. Lett. 333e334, 122e133. dx.doi.org/10.1029/2012JB009412.
Levi, S., Banerjee, S.K., 1976. On the possibility of obtaining relative paleointensities Roberts, A.P., 2008. Geomagnetic excursions: knowns and unknowns. Geophys. Res.
from lake sediments. Earth Planet. Sci. Lett. 29, 219e226. Lett. 35, L17307. http://dx.doi.org/10.1029/2008GL034719.
Sagnotti, L., Macrí, P., Camerlenghi, A., Rebesco, M., 2001. Environmental magnetism Tauxe, L., 2006. Depositional remanent magnetization: toward an improved theo-
of Antarctic late Pleistocene sediments and interhemispheric correlation of retical and experimental foundation. Earth Planet. Sci. Lett. 244, 515e529.
climatic events. Earth Planet. Sci. Lett. 192, 65e80. Thellier, E., Thellier, O., 1959. Sur l’intensité du champ magnétique terrestre dans la
Sagnotti, L., Budillon, F., Dinarès-Turell, J., Iorio, M., Macrí, P., 2005. Evidence for passé historique et géologique. Ann. Géophys. 15, 285e376.
a variable paleomagnetic lock-in depth in the Holocene sequence from the Salerno Thellier, E., 1938. Sur l’aimantation des terres cuites et ses applications géo-
Gulf (Italy): implications for “high-resolution” paleomagnetic dating. Geochem. physiques. Ann. Inst. Phys. Globe Univ. Paris 16, 157e302.
Geophys. Geosyst. 6, Q11013. http://dx.doi.org/10.1029/2005GC001043. Thomson, J., Brown, L., Nixon, S., Cook, G.T., McKenzie, A.B., 2000. Bioturbation and
Shcherbakov, V., Shcherbakova, V., 1983. On the theory of depositional remanent Holocene sediment accumulation fluxes in the north-east Atlantic Ocean
magnetization in sedimentary rocks. Geophys. Surv. 5, 369e380. (Benthic Boundary Layer experiment sites). Mar. Geol. 169, 21e39.
Shipboard Scientific Party, 2005. North Atlantic Climate: Ice SheeteOcean Atmo- Trauth, M.H., Sarnthein, M., Arnold, M., 1997. Bioturbational mixing depth and
sphere Interactions on Millennial Timescales During the Late Neogenee carbon flux at the seafloor. Paleoceanography 12, 517e526.
Quaternary Using a Paleointensity-assisted Chronology for the North Atlantic. Tucker, P., 1980. Stirred remanent magnetization: a laboratory analogue of post-
IODP Prel. Rept. 303. http://iodp.tamu.edu/publications/PR/303PR/303PR.PDF. depositional realignment. J. Geophys. 48, 153e157.
Skinner, L.C., Shackleton, N.J., 2005. An Atlantic lead over Pacific deep-water change Valet, J.-P., Meynadier, L., Guyodo, Y., 2005. Geomagnetic dipole strength and
across Termination I: implications for the application of the marine isotope reversal rate over the past two million years. Nature 435, 802e805.
stage stratigraphy. Paleoceanography 24, 571e580. Valet, J.-P., 2003. Time variations in geomagnetic intensity. Rev. Geophys. 41, 1004.
Spassov, S., Valet, J.-P., 2012. Detrital magnetizations from redeposition experiments http://dx.doi.org/10.1029/2001RG000104.
of different natural sediments. Earth Planet. Sci. Lett. 351e352, 147e157. Van Vreumingen, M., 1993. The influence of salinity and flocculation upon the
Spassov, S., Heller, F., Evans, M.E., Yue, L.P., von Dobeneck, T., 2003. A lock-in model acquisition of remanent magnetization in some artificial sediments. Geophys. J.
for the complex MatuyamaeBrunhes boundary record of the loess/palaeosol Int. 114, 607e614.
sequence at Lingtai (Central Chinese Loess Plateau). Geophys. J. Int. 155, 350e Venz, K.A., Hodell, D.A., Stanton, C., Warnke, D.A., 1999. A 1.0 Myr record of glacial
366. North Atlantic intermediate water variability from ODP Site 982 in the north-
Stoner, J.S., Channell, J.E.T., Hillaire-Marcel, C., Kissel, C., 2000. Geomagnetic pale- east Atlantic. Paleoceanography 14, 42e52.
ointensity and environmental record from Labrador Sea core MD95-2024: Verosub, K.L., Ensley, R.A., Ulrick, J.S., 1979. The role of water content in the
global marine sediment and ice core chronostratigraphy for the last 110 kyr. magnetization of sediments. Geophys. Res. Lett. 6, 226e228.
Earth Planet. Sci. Lett. 183, 161e177. Verosub, K.L., 1977. Depositional and post-depositional processes in the magneti-
Stoner, J.S., Laj, C., Channell, J.E.T., Kissel, C., 2002. South Atlantic and North Atlantic zation of sediments. Rev. Geophys. Space Phys. 15, 129e143.
geomagnetic paleointensity stacks (0e80 ka): implications for inter- Weiss, B.P., Kim, S.S., Kirschvink, J.L., Kopp, R.E., Sankaran, M., Kobayashi, A.,
hemispheric correlation. Quat. Sci. Rev. 21, 1141e1151. Komeili, A., 2004. Ferromagnetic resonance and low-temperature magnetic test
Stott, L., Poulsen, C., Lund, S., Thunell, R., 2002. Super ENSO and global climate for biogenic magnetite. Earth Planet. Sci. Lett. 224, 73e89.
oscillations at millennial time scales. Science 297, 222e226. Willmott, V., Domack, E.W., Canals, M., Brachfeld, S., 2006. A high-resolution rela-
Suganuma, Y., Yokoyama, Y., Yamazaki, T., Kawamura, K., Horng, C.-S., Matsuzaki, H., tive paleointensity record from the GerlacheeBoyd paleo-ice stream region,
2010. 10Be evidence for delayed acquisition of remanent magnetization in northern Antarctic Peninsula. Quat. Res. 66, 1e11.
marine sediments: implication for a new age for the MatuyamaeBrunhes Xuan, C., Channell, J.E.T., 2008. Origin of orbital periods in the sedimentary relative
boundary. Earth Planet. Sci. Lett. 296, 443e450. paleointensity records. Phys. Earth Planet. Inter. 169, 140e151.
Suganuma, Y., Okuno, J., Heslop, D., Roberts, A.P., Yamazaki, T., Yokoyama, Y., 2011. Xuan, C., Channell, J.E.T., 2010. Origin of apparent magnetic excursions in deep-sea
Post-depositional remanent magnetization lock-in for marine sediments sediments from Mendeleev-Alpha Ridge, Arctic Ocean. Geochem. Geophys.
deduced from 10Be and paleomagnetic records through the Matuyamae Geosyst. 11, Q02003. http://dx.doi.org/10.1029/2009GC002879.
Brunhes boundary. Earth Planet. Sci. Lett. 311, 39e52. Yamazaki, T., Oda, H., 2002. Orbital influence on Earth’s magnetic field: 100,000-
Sugiura, N., 1979. ARM, TRM and magnetic interactions: concentration dependence. year periodicity in inclination. Science 295, 2435e2438.
Earth Planet. Sci. Lett. 42, 451e455. Yamazaki, T., Oda, H., 2005. A geomagnetic paleointensity stack between 0.8 and
Tarduno, J.A., Tian, W., Wilkison, S., 1998. Biogeochemical remanent magnetization 3.0 Ma from equatorial Pacific sediment cores. Geochem. Geophys. Geosyst. 6,
in pelagic sediments of the western equatorial Pacific Ocean. Geophys. Res. Lett. Q11H20. http://dx.doi.org/10.1029/2005GC001001.
25, 3987e3990. Yamazaki, T., Solheid, P., 2011. Maghemite-to-magnetite reduction across the Fe-
Tauxe, L., Kent, D.V., 2004. A simplified statistical model for the geomagnetic field redox boundary in a sediment core from the Ontong-Java Plateau: influence
and the detection of shallow bias in paleomagnetic inclinations: was the on relative palaeointensity estimation and environmental magnetic application.
ancient magnetic field dipolar? AGU Geophys. Monogr. Ser. 145, 101e116. Geophys. J. Int. 185, 1243e1254.
Tauxe, L., Shackleton, N.J., 1994. Relative palaeointensity records from the Ontong- Yamazaki, T., 1999. Relative paleointensity of the geomagnetic field during the
Java Plateau. Geophys. J. Int. 117, 769e782. Brunhes Chron recorded in North Pacific deep-sea sediment cores: orbital
Tauxe, L., Yamazaki, T., 2007. Paleointensities. In: Kono, M. (Ed.), Treatise on influence? Earth Planet. Sci. Lett. 169, 23e35.
Geophysics. Geomagnetism, vol. 5. Elsevier, pp. 509e563. Yamazaki, T., 2008. Magnetostatic interactions in deep-sea sediments inferred from
Tauxe, L., Herbert, T., Shackleton, N.J., Kok, Y.S., 1996. Astronomical calibration of the first-order reversal curve diagrams: implications for relative paleointensity
MatuyamaeBrunhes boundary: consequences for magnetic remanence acqui- normalization. Geochem. Geophys. Geosyst. 9, Q02005. http://dx.doi.org/
sition in marine carbonates and the Asian loess sequences. Earth Planet. Sci. 10.1029/2007GC001797.
Lett. 140, 133e146. Ziegler, L.B., Constable, C.G., Johnson, C.L., Tauxe, L., 2011. PADM2M: a penalized
Tauxe, L., Steindorf, J.L., Harris, A., 2006. Depositional remanent magnetization: maximum likelihood model of the 0e2 Ma palaeomagnetic axial dipole
toward an improved theoretical and experimental foundation. Earth Planet. Sci. moment. Geophys. J. Int. 184, 1069e1089.
Lett. 244, 515e529. Zimmerman, S.H., Hemming, S.R., Kent, D.V., Searle, S.Y., 2006. Revised chronology
Tauxe, L., 1993. Sedimentary records of relative paleointensity of the geomagnetic for the late Pleistocene Mono Lake sediments based on paleointensity corre-
field: theory and practice. Rev. Geophys. 31, 319e354. lation to the global reference curve. Earth Planet. Sci. Lett. 252, 94e106.

Roberts Et Al. (2013) - Paleointensidade

Uploaded by

Copyright:

Available Formats

Roberts Et Al. (2013) - Paleointensidade

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

Roberts Et Al. (2013) - Paleointensidade

Uploaded by

Copyright:

Available Formats

Quaternary Science Reviews 61 (2013) 1e16

Contents lists available at SciVerse ScienceDirect

Quaternary Science Reviews

Magnetic paleointensity stratigraphy and high-resolution Quaternary

1. Introduction global signal that can potentially be used to provide a high-

measurements made by Gauss in 1835. Determination of geomag-

Brunhes precursor (797 ka)

Cobb Mt. (1190-1215 ka)

Jaramillo (987-1068 ka)

Olduvai (1778-1945 ka)

Iceland Basin (188 ka)

Pre-Olduvai (1977 ka)

Kamikatsura (883 ka)

Gilsa (1567-1575 ka)

Santa Rosa (932 ka)

Punaruu (1115 ka)

Stage 17 (670 ka)

Bjorn (1255 ka)

is based on the assumption that the magnetization of a rock will be

related linearly to the geomagnetic ﬁeld strength. It is, therefore,

CR0 (260 ka)

West Eiffel 4 (555 ka)

Intra-Jaramillo (1048 ka)

where aanc is a constant of proportionality. For certain igneous

rocks, archaeological artefacts or other materials that have cooled

Banc ¼ ðTRManc =TRMlab ÞBlab :

4.5 normalization of a sedimentary NRM that is analogous to that for

2.0 4. Geomagnetic relative paleointensity in high-resolution

Quaternary (e.g. Valet et al., 2005; Yamazaki and Oda, 2005;

independent methods that can constrain such problems are

Sedimentation rates in most marine environments are not high

ODP983 (60°N) 140 |Lat.| > 50°

70 70 that only small decreases in water content are necessary to lock-in

0 20 40 60 80 100 0 20 40 60 80 100 120 0 200 400 600 800

magnetic polarity records, such offsets are usually unimportant.

a Paleointensity (µT) 10 cm Gaussian lock-in 20 cm Gaussian lock-in

b Paleointensity (µT) 10 cm cubic lock-in 20 cm cubic lock-in

You might also like