Climate dependent contrast in surface mass balance in East
Antarctica over the past 216 ka
F. PARRENIN,1,2
* S. FUJITA,3,4
* A. ABE-OUCHI,5,6
K. KAWAMURA,3,4
V. MASSON-DELMOTTE,7
H. MOTOYAMA,3,4
F. SAITO,6
M. SEVERI,8
B. STENNI,9
R. UEMURA,10
E. W. WOLFF11
1
CNRS, LGGE, F-38041 Grenoble, France
2
University Grenoble Alpes, LGGE, F-38041 Grenoble, France
3
National Institute of Polar Research, Research Organization of Information and Systems, Tokyo, Japan
4
Department of Polar Science, The Graduate University for Advanced Studies (SOKENDAI), Tokyo, Japan
5
Atmosphere and Ocean Research Institute (AORI), University of Tokyo, Chiba, Japan
6
Japan Agency for Marine–Earth Science and Technology, Yokohama, Japan
7
Laboratoire des Sciences du Climat et de l’Environnement, Institut Pierre Simon Laplace, UMR CEA-CNRS-UVSQ-UPS
8212, Gif-sur-Yvette, France
8
Department of Chemistry, University of Florence, Florence, Italy
9
Department of Environmental Sciences, Informatics and Statistics, Ca’ Foscari University Venice, 30123 Venice, Italy
10
Department Chemistry, Biology and Marine Science, Faculty of Science, University of the Ryukyus, Okinawa, Japan
11
Department of Earth Sciences, University of Cambridge, Cambridge, UK
Correspondence: Frédéric Parrenin <parrenin@ujf-grenoble.fr> and Shuji Fujita <sfujita@nipr.ac.jp>
ABSTRACT. Documenting past changes in the East Antarctic surface mass balance is important to
improve ice core chronologies and to constrain the ice-sheet contribution to global mean sea-level
change. Here we reconstruct past changes in the ratio of surface mass balance (SMB ratio) between
the EPICA Dome C (EDC) and Dome Fuji (DF) East Antarctica ice core sites, based on a precise volcanic
synchronization of the two ice cores and on corrections for the vertical thinning of layers. During the
past 216 000 a, this SMB ratio, denoted SMBEDC/SMBDF, varied between 0.7 and 1.1, being small
during cold periods and large during warm periods. Our results therefore reveal larger amplitudes of
changes in SMB at EDC compared with DF, consistent with previous results showing larger amplitudes
of changes in water stable isotopes and estimated surface temperature at EDC compared with DF. Within
the last glacial inception (Marine Isotope Stages, MIS-5c and MIS-5d), the SMB ratio deviates by up to 0.2
from what is expected based on differences in water stable isotope records. Moreover, the SMB ratio is
constant throughout the late parts of the current and last interglacial periods, despite contrasting isotopic
trends.
KEYWORDS: ice cores, surface mass balance, vertical thinning, volcanic synchronization
1. INTRODUCTION
In the context of global warming and sea-level rise, changes
in the mass balance of the ice sheets must be carefully monitored,
understood and anticipated, as they could become
the main contributor of sea-level rise in the coming centuries
(Church and others, 2013). The mass balance of an ice sheet
is approximately the sum of the surface mass balance (SMB)
and the grounding line mass balance, both terms being uncertain
in the context of global warming (Bengtsson and
others, 2011; Rignot and others, 2014).
Present-day field monitoring and atmospheric reanalyses
depict a year-round occurrence of rare snowfall events in
central East Antarctica (e.g. Ekaykin, 2003; Fujita and Abe,
2006). Moisture back-trajectories identify ice core moisture
sources located on average at 45°S today, but with a larger
contribution of lower latitudes for the highest elevation, i.e.
most inland sites (Masson-Delmotte and others, 2008;
Suzuki and others, 2008; Sodemann and Stohl, 2009;
Scarchilli and others, 2011). Different ice cores have different
longitudinal origins of water vapor, depending on their location
(Fig. 1). For instance, Dome C receives moisture predominantly
from the Indian Ocean sector, while it is mostly
advected from the Atlantic Ocean towards Dome F (e.g.
Reijmer and others, 2002; Suzuki and others, 2008;
Sodemann and Stohl, 2009). The SMB in Antarctica is also
a function of the surface elevation of the ice sheet (e.g.
Takahashi and others, 1994; Krinner and Genthon, 1999)
and is affected by the redeposition of snow by wind
(Gallée and others, 2012). For example, it is known that
SMB differs between the windward and leeward sides of
ice divides for strong-wind events (Fujita and others, 2011).
In addition, local variations in the SMB are governed by
the local surface topography, which is influenced by the
bedrock topography (Fujita and others, 2011). Above the
Antarctic plateau, accumulation is not only driven by maritime
intrusions and snowfall events, but also by clear sky
condensation, possibly related to boundary layer dynamics.
Finally, it may also be affected by exchanges of water
vapor between surface snow and the surrounding air
(Hoshina and others, 2014). Such processes have been* These authors contributed equally to this work.
Journal of Glaciology (2016), 62(236) 1037–1048 doi: 10.1017/jog.2016.85
© The Author(s) 2016. This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.
org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
recently highlighted by diurnal variations in vapor isotopic
composition both in Greenland (Steen-Larsen and others,
2014) and Antarctica (Ritter and others, 2016). Considering
the exchanges of water vapor between surface snow and
the surrounding air as well as snow transportation by wind,
not only high-precipitation events associated with strong
winds, but also daily exposure to prevailing wind, E-N-E at
Dome Fuji (DF) and S at EPICA Dome C (EDC), may have significant
effects on SMB, water stable isotope ratios and snow
properties at these dome sites.
Ice sheets also form a rich paleo-climatic archive. For
dating (e.g. Kawamura and others, 2007; Parrenin and
others, 2007) or interpreting the ice core records (e.g. transferring
concentrations of species in ice into atmospheric
fluxes; Wolff and others, 2006), an evaluation of the past
SMB is needed. Indeed, the annual layer thickness along
an ice core is the product of the initial SMB and the vertical
thinning function accounting for the ice flow. Usually, past
SMB is presumed to be exponentially related to the deuterium
(δDice) or oxygen-18 (δ18
Oice) isotopic content of the
ice. The underlying assumption is that condensation is proportional
to the saturation vapor pressure of water in air,
which itself is related to condensation temperature and therefore
to precipitation isotopic composition (Jouzel and others,
1987; Parrenin and others, 2004, 2007; Masson-Delmotte
and others, 2008). This approach has been applied to
obtain past SMB estimates for deep ice cores after correcting
the water stable isotope records for variations in the isotopic
composition of the ocean, and/or for artifacts due to changes
in moisture sources, using the second-order deuterium
excess information (Stenni and others, 2001; Parrenin and
others, 2007; Uemura and others, 2012).
Alternatively, constrains from ice core chronologies can
be used to infer changes in layer thickness, which, after correction
for thinning, provide information on past changes in
SMB. For example, an alternative approach to estimating
past SMB of polar ice sheets is to investigate ice-equivalent
thickness between accurately dated reference horizons in
the ice core stratigraphy, such as volcanic eruption
markers, and to correct for vertical thinning. This approach
is widely used to estimate the SMB of relatively shallow
cores or snow pits (e.g. Frezzotti and others, 2013 and references
therein). However, thus far, this approach has not been
used for the enormous number of volcanic markers found in
very deep ice cores because most of the volcanic markers in
such ice cores have not been dated independently.
Here, we propose a similar approach to estimate the past
ratio of SMB between DF and EDC, two remote dome sites in
East Antarctica (Fig. 1). Deep ice cores drilled at the two sites
were first synchronized by identifying 1401 volcanic tie
points over a period covering the past 216 ka (Fujita and
others, 2015). Instead of dating these 1401 volcanic horizons
independently, we derive changes in the relative thickness of
the ice core sections covering the same time periods in the
two cores. After correction of the mechanical thinning
effects due to glacial flow, we are able to derive the SMB
ratio between the two places. The thinning function is relatively
straightforward to evaluate for these drilling sites
located on domes, where horizontal advection is expected
to be negligible. This new approach provides information
on the relative SMB pattern of the East Antarctic ice sheet
over glacial-interglacial cycles from the last 216 ka. We
discuss how spatially inhomogeneous changes of the SMB
may have occurred between the two remote dome sites.
2. METHODS
2.1. The two studied ice core drilling sites
DF and Dome C are two remote dome summits in East
Antarctica located ∼2000 km apart (Table 1; Fig. 1). DF is
located on the polar plateau facing the Atlantic and Indian
Fig. 1. Map of the Antarctic continent with elevation contours every 500 m. The two ice coring sites used in this study, Dome C and DF, are
marked with stars. Other ice coring sites mentioned in this paper are marked with filled circles. The prevailing wind directions at surface for
both the Dome C and DF sites are indicated with arrows.
1038 Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
Ocean sectors and is surrounded by the Dronning Maud
Mountains, a coastal escarpment in particular along longitudes
ranging from ∼20°W to ∼35°E. At DF, prevailing
wind direction is from the east (Table 1). Presently, DF
shows a spatial gradient of SMB decreasing southwards
(Fujita and others, 2011). High-precipitation events associated
with strong winds from the NE direction have a
major influence on the SMB (Fujita and Abe, 2006), but no
seasonality of precipitation has been noted (Fujita and Abe,
2006; Kameda and others, 2008).
Dome C is located at one of the dome summits in East
Antarctica in the Indian Ocean sector and the moisture predominantly
comes from the Indian Ocean. Its elevation is
lower than that of DF by ∼570 m. This part of the East
Antarctic ice sheet has a gentler slope from the coast to
polar plateau compared with the escarpment surrounding
DF (Fig. 1). Reflecting the shape of the East Antarctic ice
sheet, prevailing surface winds are from the south, that is,
from continental inland; surface winds rarely occur from
the west, north or east directions (Aristidi and others, 2005).
It should first be noted that at present, DF is higher by
600 m, slightly colder, and has 30‰ more depleted surface
snow δD than EDC (Table 1; Fig. 1). The EDC-DF isotopic
differences cannot be explained by differences in surface
temperature or relationships with accumulation, based on
Rayleigh distillation relationships. The DF-EDC deuterium
difference would require 6°C difference in condensation
temperature to be accounted for, based on the spatial
isotope/temperature relationship (Jouzel and Merlivat,
1984). In fact, possible explanations for such an isotopic difference
are linked with the origin of precipitation (MassonDelmotte
and others, 2011), with differences in condensation
temperature, with precipitation seasonality and/or intermittency
(Masson-Delmotte and others, 2011), with surface
snow/vapor interactions (Hoshina and others, 2014; SteenLarsen
and others, 2014) or with elevation differences
between the two sites. The more depleted modern value of
δDice at DF may for instance be explained by a higher proportion
of precipitation occurring during winter at this site compared
with EDC (but we have no observational evidence
to confirm this hypothesis at the moment) or by a more
remote moisture source, as expected from the wider winter
expansion of sea ice in the Atlantic compared with the
Indian Ocean sector (Gersonde and others, 2005).
At each of the two sites, two long ice cores have been
drilled. At DF, the first core (DF1) was drilled in 1992–98
to a depth down to 2503 m (Watanabe and others, 2003a).
The second 3035 m long core (DF2), reaching almost to
the ice-sheet bed, was drilled in 2004–07, at a site ∼43 m
from the DF1 borehole (Motoyama, 2007). At Dome C, the
first core (EDC96) was started in the 1996/97 season to a
depth of 790 m. The second 3270 m long core (EDC99)
reaching nearly the ice-sheet bed, was started during the
1999/2000 season at a site 10 m away from the EDC96
core (EPICA community members, 2004). Ice core signals
from these four cores have been used here for volcanic
synchronization.
Here we use as a reference chronology the DFO-2006 timescale
established for the DF ice core (Kawamura and others,
2007). This age scale is only used to plot quantities versus
age, but has no impact on the reconstruction of the SMBEDC/
SMBDF ratio. Comparisons between chronologies for DF and
EDC cores are discussed by Fujita and others (2015).
2.2. EDC-DF volcanic synchronization and synchrobased
SMBEDC/SMBDF ratio
The EDC-DF volcanic matching consists of 1401 depth
tie points (Fujita and others, 2015), down to a depth of
2184 m at DF and 2170 m at EDC, which roughly corresponds
to an age of 216 ka in the DFO-2006 chronology
(Kawamura and others, 2007). On average, that makes one
tie point every 154 a, although their distribution is irregular
(Fig. 2). For the periods of MIS 3, 5a and 5b–5e, a large
number of tie points were found, typically one every 50–
100 a. Identifying unequivocal tie points was difficult in
some cold periods such as MIS 2, 4, 5b and 6. For volcanic
signals, ECM, DEP and sulfate data were used for EDC, and
ECM and ACECM data were used for DF. More details on
the synchronization are given by Fujita and others (2015).
These tie points were then placed on an age scale
(DFO2006) and re-interpolated every ka. The ratio of layer
thickness at EDC and DF of these 1 ka-long intervals ΔzEDC/
ΔzDF can be inferred (Figs 2 and 3). We refer to it as the
layer thickness ratio.
Vertical thinning τ due to ice flow has been estimated both
for the EDC and DF ice cores based on a one-dimensional
(1-D) ice flow model (Parrenin and others, 2007) with a
Table 1. Information on the two drilling sites
Variable DF EDC Data sources
Location 77.32°S, 39.70°E 75°06′S, 123°21′°E (Watanabe and others, 2003a; EPICA community members,
2004)
Elevation on WGS84 (m) 3800 3233 (Watanabe and others, 2003a; EPICA community members,
2004)
Ice thickness (m) 3028 (±15) 3273 (±5) (Fujita and others, 1999; Parrenin and others, 2007)
Distance from the coast (km) ∼930 ∼950 This study
Total length of ice core (m) 3035 3260 (Motoyama and others, 2007; Parrenin and others, 2007)
Annual mean air temperature (°C) −54.8 −51.2a
(Fujita and Abe, 2006; King and others, 2006)
10 m snow temperature (°C) −57.7 −54.8 (Motoyama and others, 2005, Personal communication
L. Arnaud)
Annual accumulation rate (kg m−2
) 27.3 ± 0.4b
26 ± 3 (Frezzotti and others, 2004; Kameda and others, 2008)
Annual mean wind speed (m s−1
) 5.9 at 10 m 3.6 at 3.3 m (Takahashi and others, 2004; Aristidi and others, 2005)
Annual mean atmospheric pressure (hPa) 598.4 644.9 (Parish and Bromwich, 1987; Watanabe and others, 2003a)
Prevailing wind direction at the surface
(from)
NE∼E S (Kameda and others, 1997; Aristidi and others, 2005)
a
Over the period 1984–2013.
b
Snow stake farm.
1039Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
prescribed analytical velocity profile (Lliboutry, 1979) and a
prescribed ice thickness evolution based on a 3-D model of
Antarctic evolution (Ritz and others, 2001). The thinning
ratio between EDC and DF τEDC/τDF is 1 for the present-day
and increases up to ∼1.5 at 200 ka (high thinning ratios
imply that EDC thins less than DF), which is mainly due to
the fact that the ice thickness is larger at EDC than DF
(Table 1). After correcting the layer thickness ratio ΔzEDC/
ΔzDF by the vertical thinning effects, we obtain a synchrobased
SMBEDC/SMBDF ratio (Fig. 2).
2.3. Surface temperature and mass balance at EDC
and DF
A simple Rayleigh model can link the variations of the isotopic
composition of vapor in an air mass with the variations of
its temperature (Dansgaard, 1964). Estimates of surface temperature
can therefore be evaluated from the isotopic content
of the ice δ18
Oice and δDice, first corrected for the temporal
variations in isotopic content of the ocean δ18
OSW and
δDSW (Jouzel and others, 2003):
δ18
Ocorr ¼ δ18
Oice À δ18
OSW
1 þ δ18
Oice
1 þ δ18
OSW
; ð1Þ
δDcorr ¼ δDice À δDSW
1 þ δDice
1 þ δDSW
: ð2Þ
δ18
OSW is taken as reconstructed by Bintanja and others
(2005) based on the exhaustive LR04 oceanic stack
(Lisiecki and Raymo, 2005). δDSW is calculated on the assumption
that δDSW = 8δ18
OSW. From this, we derive a first
reconstruction of accumulation, called ocean-corrected
(Fig. 4):
aoc ¼ A0
exp β δDcorr À δD0
corr
À ÁÀ Á
ð3Þ
with β = 0.015, δD0
corr = −390.9‰ and A0
= 3.1 cm (ice) a−1
for EDC and β = 0.013, δD0
corr = −403.1‰ and A0
= 3.8 cm
(ice) a−1
for DF. The values of these parameters were chosen
for a best fit (Fig. 4) with the published accumulation reconstructions
that are compatible with the age scales of the two
cores (Parrenin and others, 2007). The ratio of these reconstructed
SMB histories is hereafter named ocean-corrected
SMBEDC/SMBDF ratio.
SMB estimates can be refined by correcting the water
stable isotope records for variations in the temperature of
the ocean where the moisture forms and/or for artifacts due
to changes in moisture sources using the second-order deuterium
excess record. Site and source temperature estimates
over time ΔTsite and ΔTsource can be deduced using a set of
two linear equations (Stenni and others, 2001; Uemura and
Fig. 2. Ratio of layer thickness (green) or surface mass balance (red) after correcting for the EDC–DF thinning ratio (blue). The density of tie
points (violet) is indicated in the lower panel (No. of tie points ka−1
). The DFO-2006 (Kawamura and others, 2007) and AICC2012 (Bazin and
others, 2013; Veres and others, 2013) age scales are indicated.
Fig. 3. Scheme illustrating the derivation of the ΔzEDC/ΔzDF ratio
from the volcanic links in between the EDC and DF ice cores.
1040 Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
others, 2012):
δDcorr À δD0
corr ¼ γsiteΔTsite À γsourceΔTsource; ð4Þ
dcorr À d0
corr ¼ ÀβsiteΔTsite þ βsourceΔTsource; ð5Þ
where dcorr, the corrected deuterium excess, is dcorr = δDcorr–
8δ18
Ocorr. The coefficients of Eqs. (4, 5), namely γsite, γsource,
βsite and βsource, have been previously inferred at both sites
using a simple Rayleigh-based model, constrained with
present-day surface data on the trajectories of air masses
and their values can be found in Uemura and others
(2012). Site temperature is calculated from ΔTsite using the
10 m snow temperature (Table 1) as the average temperature
for the 0–1 ka b1950 interval. From this, we derive a second
accumulation reconstruction, called source-corrected
(Fig. 4), which is not based directly on δDice but on the
source-corrected site temperature reconstruction:
asc ¼ A0
exp β γsite À βsite
γsource
βsource
 
ΔTsite
 
; ð6Þ
which translates into:
asc ¼ A0
exp β δDcorr À δD0
corr
À Á
þ
γsource
βsource
dcorr À d0
corr
À Á
  
;
ð7Þ
with β = 0.016, δD0
corr = −390.9‰, d0
corr = 10.1‰ and A0
=
3.2 cm (ice) a−1
for EDC and β = 0.015, δD0
corr = −403.1‰,
d0
corr = 13.4‰ and A0
= 3.7 cm (ice) a−1
for DF. The values
of these parameters were chosen for a best fit (Fig. 4)
with the published accumulation reconstructions that are
compatible with the age scales of the two cores (Parrenin
and others, 2007). The ratio of these reconstructed SMB histories
is hereafter named source-corrected SMBEDC/SMBDF
ratio.
The formulas that we used here for temperature reconstructions
are the same as in Uemura and others (2012), but slightly
different from those used by Parrenin and others (2007). At the
time of this second study, EDC deuterium excess was not
available, and therefore the accumulation was only inferred
from the deuterium data after ocean correction. Here, we
use the EDC deuterium excess data (Stenni and others,
2010) to provide a coherent approach for the two sites.
2.4. Simplified ice thickness models at EDC and DF
Ice thickness variations on the East Antarctic plateau can be
mainly explained by variations in surface accumulation
(Parrenin and others, 2007). To test the implications of
various accumulation reconstructions on the ice thickness
history at the two sites, we used a 1-D conceptual model
(Parrenin and others, 2007) fitted onto 3-D simulations
(Ritz and others, 2001). This model is a simple linear perturbation
model, where the vertical velocity of ice at surface with
respect to bedrock is written:
a À
∂H
∂t
¼ k þ kHH þ kSS; ð8Þ
and where the bedrock follows a simple relaxation law to an
equilibrium:
∂B
∂t
¼
ðB0 À H=kBÞ À B
τB
; ð9Þ
where a is the accumulation rate, B is the bedrock elevation,
H is the ice thickness, S = B + H is the surface elevation of
the ice sheet, and k, kH, kS, kB, B0 and τB are parameters,
whose values for EDC and DF are given in Table 2 of
Parrenin and others (2007). B0 corresponds to a bedrock elevation
without isostatic effect. There is a systematic bias of
the model toward lower elevation. Therefore, we only consider
the elevation changes with respect to the present.
Fig. 4. Comparison of various accumulation reconstructions for the EDC and DF ice cores: ocean-corrected (light blue), source-corrected
(green) and published (pink, Parrenin and others, 2007).
1041Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
3. RESULTS
3.1. Synchro-based SMBEDC/SMBDF ratio and
comparison with the δD records
The ratio of layer thickness ΔzEDC/ΔzDF inferred from the volcanic
synchronization (Fig. 2, green curve) exhibits an increasing
trend towards the past. We correct the layer
thickness ratio for the vertical thinning ratio as deduced
from ice flow modeling at both sites (Fig. 2, blue curve)
and obtain the SMBEDC/SMBDF ratio (Fig. 2, orange curve).
Our synchro-based SMBEDC/SMBDF ratio depicts large variations
of up to 0.4. These variations resemble the variations
of the δDice profiles. The correlation coefficient is 0.74
between the average deuterium profile at EDC and DF and
the SMBEDC/SMBDF ratio for 1 ka re-sampled time series.
Figure 5 suggests a correlation for minor troughs and peaks,
as indicated by the thin black vertical dashed lines, although
the correlation is not systematic. For example, during the
early optimum of the last interglacial period occurring at
∼128 ka on the DFO-2006 age scale, our method reconstructs
SMBEDC/SMBDF ratio to be 1.1. However, there are
also periods when no relationship between SMBEDC/SMBDF
and δDice is observed. This is the case during the glacial inception
cooling (MIS 5d and 5c). Indeed, the lowest values
of the synchro-based SMBEDC/SMBDF ratio (0.7) are reached
for MIS 5d–5c, while the lowest values of the δDice profiles
are reached at the Last Glacial Maximum (LGM).
We now focus on the present and last interglacial periods. We
first observe that the synchro-based SMBEDC/SMBDF is stable
during the late part of the interglacial periods (125–118 and
8–0 ka b1950), despite different δD trends (a long-term decrease
at DF versus stable levels at EDC). Second, we note very similar
levels in the synchro-based SMBEDC/SMBDF during the present
and during the warmer last interglacial period.
3.2. Comparison of the synchro-based and isotopebased
SMBEDC/SMBDF ratios
The synchro-based and isotope-based (ocean-corrected and
source-corrected) SMBEDC/SMBDF ratios are compared in
Figure 5. The large-scale variations of all SMB ratios display
glacial-interglacial variations. The source correction has
only a minor effect when compared with the ocean-corrected
SMBEDC/SMBDF, with differences generally < 0.1. The differences
are larger between the synchro-based and isotopebased
SMBEDC/SMBDF ratios. In particular during MIS 5d
and 5c, the difference reaches 0.2, the synchro-based
SMBEDC/SMBDF ratio being less than what can be inferred
based on water isotopes. The difference also reaches 0.1 for
periods at 60 and 90 ka b1950. Another noticeable difference
is that the synchro-based SMBEDC/SMBDF ratio displays peaks
at the beginning of the Holocene and Eemian periods while
the isotope-based SMBEDC/SMBDF ratios do not.
3.3. Consequences on elevation variations and
thinning functions
As Parrenin and others (2007) have shown, the thinning function
is influenced by the elevation variations at the drilling
sites, which is itself influenced by accumulation variations.
We can therefore ask ourselves how robust are the thinning
functions that we used to calculate the synchro-based
SMBEDC/SMBDF ratio. For EDC, we calculate an alternative
Fig. 5. (top) DF (Watanabe and others, 2003a) (blue) and EDC (Jouzel and others, 2007) (pink) δD ice variations. Top labels indicate the
Marine Isotope Stages and bottom labels indicate the Antarctic Isotopic Maxima (AIMs) events. (middle top) DF (blue) and EDC (pink)
source-corrected T site reconstructions (this study). (middle bottom) DF (blue) and EDC (pink) source-corrected surface accumulation rate
(this study). (bottom) Ratio of ocean-corrected (light blue), source-corrected (green) and synchro-based (red) surface mass balances. The
DFO-2006 (Kawamura and others, 2007) and AICC2012 age scales (Bazin and others, 2013; Veres and others, 2013) are used. The thin
vertical black dashed lines mark correspondences in millennial scale events. Note that we did not plot the ratio of the published SMBs at
EDC and DF (Parrenin and others, 2007), since they used inconsistent formulas (Section 2).
1042 Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
accumulation reconstruction by multiplying the sourcecorrected
SMBDF by our synchro-based SMBEDC/SMBDF
ratio. We call this the synchro-based SMBEDC. Similarly, we
calculate the synchro-based SMBDF by dividing the sourcecorrected
SMBEDC by our synchro-based SMBEDC/SMBDF
ratio. Then we calculate the elevation variations at EDC and
DF from our conceptual model described in Section 2 and
from the source-corrected or synchro-based SMB scenarios.
The results are displayed on Figure 6. For both EDC and DF,
the two accumulation scenarios lead to the same qualitative
features, with an ice thickness smaller during glacial periods
and higher during interglacial periods. For EDC, the LGM ice
thickness is ∼160 m (resp. ∼150 m) lower than the presentday
thickness for the source-corrected (resp. synchro-based)
SMB scenario. The standard deviation of the difference
between both scenarios is ∼10.5 m. The largest difference is
∼35 m for MIS5d, i.e. ∼1% of the ice thickness. For DF, the
LGM ice thickness is ∼145 m (resp. ∼160 m) lower than
the present-day thickness for the source-corrected (resp.
synchro-based) SMB scenario. The standard deviation of the
difference between both scenarios is ∼14.5 m. The largest difference
is ∼60 m for MIS5d, i.e. ∼2% of the ice thickness.
4. DISCUSSIONS
4.1. Reliability of the thinning corrections and of the
volcanic match
A first argument for the reliability of the thinning evaluation
comes from the fact that the SMBEDC/SMBDF curve, after the
thinning correction, has a negligible decreasing trend
toward the past. Indeed, we do not expect the SMBEDC/
SMBDF ratio to have a trend over several glacial-interglacial
cycle. This suggests that the main trends of the thinning
functions at EDC and DF have been captured by the ice
flow models. The part of the ΔzEDC/ΔzDF curve that varies
at the glacial/interglacial scale, could also be due in part
to the vertical thinning, with glacial layers relatively more
thinned at EDC, and not correctly accounted for in the ice
flow modeling exercises that we used. There is indeed a correlation
between climate and some ice physical properties,
such as fabric (Durand and others, 2009) or impurities
(Watanabe and others, 2003b; Fujita and others, 2015),
which can have an impact on ice flow properties.
However, this hypothesis seems unlikely for several
reasons. First, by mass conservation, an abnormally
thinned layer at some place can only be explained if this
layer is abnormally thickened at a neighboring place, but
no irregularity is observed in the isochronal layers observed
by ice sounding radars at DF (Fujita and others, 1999, 2012)
or EDC (Siegert and others, 1998; Tabacco and others,
1998; Cavitte and others, 2016). Second, if glacial ice is
softer than interglacial ice, the relative difference in cumulated
vertical thinning with interglacial ice should increase
with the age of the ice layers, as is shown by mechanical
simulations (Durand and others, 2007), but no such effect
is observed in the ΔzEDC/ΔzDF curve. Third, mechanical
simulations do not suggest that ice layers with different viscosities
can lead to a very irregular thinning function
(Durand and others, 2007). As outlined in Section 3, the
Fig. 6. Elevation variations at EDC (top) and DF (bottom) based on two different scenarios of SMB variations: source-corrected (blue) and
synchro-based (pink).
1043Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
ice thickness reconstructions at EDC and DF are also partially
uncertain due in particular to uncertainties in the accumulation
scenarios. These relative uncertainties on the ice
thickness scenarios of ∼1% for EDC and ∼2% for DF lead
to approximately the same relative uncertainties on the thinning
functions. This can explain neither the glacial-interglacial
variations of the synchro-based SMBEDC/SMBDF ratio of
up to 0.4, nor the 0.2 difference between the source-corrected
and synchro-based SMBEDC/SMBDF ratios at MIS5d.
Finally, the interplay between the dome movement and
the different strain rates at different locations could lead to
irregularities in the thinning function. Indeed, the strain
regime is different right at a dome than a few kilometers
downstream on a flank due to the Raymond effect
(Raymond, 1983) and the strain regime is also a function
of the ice thickness. But there is no obvious reason why
these effects would lead to a thinning function correlated
with the deuterium profiles.
These elements therefore suggest that the main characteristics
of our synchro-based SMBEDC/SMBDF ratio are not due
to error in the evaluations of the thinning functions at both
sites. However, we can expect some errors in the modeling
of the thinning functions at both sites to affect the details of
our synchro-based SMBEDC/SMBDF ratio. For example, as
outlined previously, the error in the SMB reconstruction
used at both sites to model the thinning functions (Parrenin
and others, 2007) can affect ice-sheet thickness evaluation
by ∼2% at maximum, which translates into a 2% error in
the thinning function. Also, spatial variations in the ice thickness
around both drilling sites can affect the thinning functions
if the ice flow is not purely vertical (Parrenin and
others, 2004). This is, however, difficult to assess since we
do not have robust estimates of the dome movements
during the past.
The robustness of the volcanic match used to deduce the
synchro-based SMBEDC/SMBDF ratio can also be discussed,
especially during the 105–113 ka b1950 time interval (MIS
5c and 5d), when the synchro-based and isotope-based
SMBEDC/SMBDF ratios deviate by as much as 0.2. Given the
number of tie points in this interval (95), it seems very unlikely
that the volcanic match is entirely wrong. Moreover, if the
low values of the synchro-based SMBEDC/SMBDF ratio during
this time period were due to an incorrect volcanic match, it
would be compensated by too high values before 105 ka
b1950 or after 113 ka b1950. But no such high values are
observed, suggesting that an incorrect volcanic match is
not the cause of the synchro-based and isotope-based
SMBEDC/SMBDF ratios difference during this time period.
Moreover, we tried to guide our volcanic match during this
time interval, so that the synchro-based SMBEDC/SMBDF
ratio would be in agreement with the isotope-based
SMBEDC/SMBDF ratios, but no satisfying volcanic match
could be achieved.
Sulfate (Watanabe and others, 2003b; Wolff and others,
2006; Iizuka and others, 2012) and beryllium-10
(Cauquoin and others, 2015) have been proposed to have a
nearly temporally constant flux above Antarctica and could
be used for an independent check of our synchro-based
SMBEDC/SMBDF ratio. Among those two, only sulfate is available
for our studied time period and for both EDC (Wolff and
others, 2006) and DF (Watanabe and others, 2003b; Iizuka
and others, 2012). Unfortunately, it appears that the sulfate
flux is not constant enough at DF (Iizuka and others, 2012;
Figure 2d) for sulfate dilution to be a useful SMB proxy.
4.2. Relative change of SMB and relative change of
local temperature
Glacial climatic conditions coincide with a reduced synchrobased
SMBEDC/SMBDF ratio. At the LGM, we note that the
20% lower accumulation at EDC than at DF roughly corresponds
to 1/5 of the full magnitude of the Holocene-LGM
relative accumulation variations (factor of 2; Fig. 5). How
would this difference in SMB translate into temperature differences,
assuming a constant accumulation/temperature relationship?
It would correspond to a 2°C temperature
anomaly, scaled to a 10°C Holocene-LGM contrast of
surface temperature (Parrenin and others, 2013). This is consistent
with a recent estimate of 2.5°C for the difference in
LGM-present precipitation-weighted temperature change
between the two sites, with larger amplitude estimated at
EDC than at DF (Uemura and others, 2012). We therefore
conclude that our inferred synchro-based SMB ratio change
may be a consequence of a difference of precipitationweighted
temperature change between both sites.
4.3. Relative change of SMB and relative change of
δD, implications for ice-sheet modeling
The differences between the variations of SMBEDC/SMBDF
ratio and the δD variations as well as the differences
between the synchro-based and isotope-based SMB ratios,
both described in Section 3, challenge the assumption of
close relationships between water stable isotope and accumulation
anomalies, especially during MIS5c and MIS5d.
This feature has already been suggested during the early
Holocene (Parrenin and others, 2007), during the last deglaciation
(WAIS Divide Project Members, 2013) and from
climate simulations of the last interglacial (Sime and others,
2009). Our study also indicates that this decoupling is site dependent.
During the late part of the Holocene and the
Eemian, our synchro-based SMBEDC/SMBDF ratio is more
constant than what is inferred from δD variations.
Clearly, the isotope-based SMB at either EDC or DF or
both are associated with relative uncertainties of at least
10%. Indeed, larger uncertainty ranges as possible because
our synchronization method cannot detect correlated
errors. This was also suggested by Fujita and others (2015)
based on the same volcanic synchronization. They hypothesize
that a cause of the systematic DFO2006/AICC2012 age
differences in MIS 5 are associated with differences in the
dating approaches, either the age-markers-based dating or
the glaciological dating. They further hypothesize that
major sources of the discrepancies were systematic errors
in SMB estimation.
Inaccurate estimates of SMB based on water stable isotope
records can cause important errors for chemical flux reconstructions
and ice core chronologies but also for firn
(Goujon and others, 2003) and ice-sheet (Ritz and others,
2001) modeling. For example, if the SMB evaluation at one
of the two sites is wrong by as much as 20%, this would
lead to a 20% error in fluxes reconstructions, in events duration
and in Δage evaluation (i.e. 500–1000 a error in the
ice age/gas age difference). Concerning ice-sheet modeling,
we showed (Fig. 6) that our ice thickness reconstructions
may be wrong by as much as 35 m for EDC and 60 m
for DF (the maximum error occurring for MIS5d). Taking
into account a vertical gradient of temperature of 1°C
(100 m)−1
, such ice sheet thinning would lead to an
1044 Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
underestimation of the magnitude of temperature decrease
‘at fixed elevation’ by ∼0.5°C.
4.4. Atmospheric process, dome movement or
elevation change artifact?
In the following, we discuss three different hypotheses to
explain the changes in the SMBEDC/SMBDF ratio: regional differences
in climate (at constant geometry of the ice sheet),
changes of dome position affecting snow redeposition by
wind, or differences in elevation changes.
Different atmospheric processes may explain the variations
in the SMBEDC/SMBDF ratio: (1) effects related to moisture
sources and distillation along transport paths; (2)
different glacial sea ice expansions in the Atlantic and
Indian Ocean sectors (Gersonde and others, 2005), enhancing
accumulation and temperature changes at EDC compared
with those at DF; (3) effects associated with
precipitation intermittency and/or seasonality (e.g. Suzuki
and others, 2013), precipitation amounts being expected to
be reduced during cold periods; (4) less frequent blocking
events at EDC (Massom and others, 2004) than at DF
(Hirasawa and others, 2000) during glacial periods, these
warm events being responsible of a large proportion of the
total annual accumulation today (Hirasawa and others,
2000); and (5) differences in sublimation, which is an important
process since surface snow/vapor exchanges can alter the
snowfall signal in-between snowfall events (Hoshina and
others, 2014; Steen-Larsen and others, 2014).
We now explore the dome movement hypothesis. Today,
we observe a spatial gradient of accumulation at Dome C
(Urbini and others, 2008) and DF (Fujita and others, 2011),
due to orographic precipitation or to snow redeposition by
winds linked with surface curvature. Under glacial/interglacial
climatic changes (e.g. migration of the grounding line), it
seems natural that the ice divides locations migrate. A movement
of the domes (Saito, 2002, provides information on the
movement of DF during the past) could therefore create an
apparent change of accumulation in the ice core records. A
movement of the domes can also modify the trajectories and
therefore the upstream origin of ice particles in the ice cores.
Given that accumulation varies spatially, in particular due to
surface topographic variations related to bedrock reliefs
(Fujita and others, 2011), this second process can also create
an apparent change of accumulation in the ice cores. In this
case, we would not expect any constant relationship
between water stable isotopes and accumulation rate, except
if the dome movement is itself correlated to processes affecting
the isotopic composition of water vapor and precipitation (e.g.
via sea-level changes and grounding line migration).
Concerning the elevation hypothesis, it is not impossible
that EDC and DF experienced different changes in elevation,
since the ice flow at those two sites should not react in the
same way to sea-level changes (Saito and Abe-Ouchi,
2010). On one hand, DF is relatively insensitive to sealevel
changes since it is well protected by the Dronning
Maud Mountains. On the other hand, EDC is very sensitive
to sea-level changes since grounding lines in Wilkes Lands
can advance and retreat over large distances (Mengel and
Levermann, 2014). Therefore, it is expected that elevation
change should be different at EDC and DF. Quantitatively,
if inter-site temperature differences are driven only by ice
thickness change, then using a vertical temperature gradient
of 1°C (100 m)−1
(Krinner and Genthon, 1999), 2°C colder
glacial conditions at EDC will translate into a 200 m relative
elevation difference at the LGM between EDC and DF. This is
quite large compared with the current estimates of the central
East Antarctic ice sheet LGM topography, as ice-sheet simulations
suggest an overall lowering of surface elevation by
∼120 m (Ritz and others, 2001), driven by the lower glacial
accumulation. We however stress that these ice-sheet simulations
were driven by a homogeneous scenario of accumulation
changes (a hypothesis challenged by our findings), and
that they have intrinsic limitations in the representation of dynamical
effects associated with grounding line migration
(Pattyn and others, 2012). Moreover, we remark that there
are abrupt variations of the SMB ratio, for example ∼0.2
during just 1 ka at 113 ka b1950, which cannot be due to elevation
variations since the later are a slow integrator of
surface accumulation variations and dynamical effects.
5. CONCLUSIONS
Our study suggests that the vertical thinning functions evaluated
by ice flow models at EDC and DF are valid for the depth
range covered here (the past 216 ka). We produce a new
paleoclimatic record, the SMBEDC/SMBDF ratio, which
varies at the glacial/interglacial scale. Regional differences
in climate are identified, with EDC characterized by an
enhanced (+20%) amplitude of glacial cooling and drying,
compared with DF. The data show that interglacial changes
in SMBEDC/SMBDF ratio do not always scale with those of
water stable isotopes, challenging classical hypotheses
used for ice core chronologies and for ice-sheet modeling.
The SMB ratio reduces strongly in MIS 5d and 5c, 20%
lower than what would be deduced from the isotopes.
Moreover, the SMB ratio is almost constant during the late
parts of the current and last interglacial periods, in contradiction
with contrasting isotopic trends at EDC and DF. Our
reconstructions of SMB at EDC and DF might be improved
in the future by using new aerosols records with nearly constant
fluxes above Antarctica, for example, beryllium-10
(Cauquoin and others, 2015). Changes in the SMBEDC/
SMBDF ratio may be due to regional climate differences at
the two sites, to an artifact of dome movement influencing
snow redistribution by wind, or to a different change of elevation
at the two sites. Further studies are necessary to discriminate
between these three hypotheses. In particular,
new simulations of Antarctic ice sheet evolution using a
new generation of ice-sheet models with a realistic representation
of grounding line migrations (e.g. Pollard and
DeConto, 2012) will allow exploration of the movements
and elevation variations of the domes. There is also a need
for more accurate atmospheric models able to reproduce
the measured present or past spatial pattern of accumulation.
Expanding this approach towards other sites will also give
more regional information on the past SMB pattern. The inferred
regional differences in SMB variations should also be
taken into account in glacial-interglacial Antarctic ice sheet
modeling.
AUTHOR CONTRIBUTION
The writing of this paper was led by the two first authors:
F. Parrenin and S. Fujita. They contributed equally and
share the responsibilities for this paper. They carried out
the synchronization work, led discussions and oversaw
the writing of this paper. E. Wolff and M. Severi provided
1045Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
the EDC electrical profile data and EDC sulfate data,
respectively. S. Fujita and H. Motoyama provided the
entire electrical profile data of the DF core. All authors
joined in the scientific discussions.
ACKNOWLEDGEMENTS
We thank Gerhard Krinner, Catherine Ritz, Kazue Suzuki,
Jean Jouzel, Michiel Van Den Broeke and the GLACE team
at LGGE for helpful discussions. We wish to also thank all
participants to the field seasons at Dome C. The main logistic
support was provided by IPEV and PNRA at Dome C. This
work is a contribution to the European Project for Ice
Coring in Antarctica (EPICA), a joint European Science
Foundation/European Commission scientific program,
funded by the EU and by national contributions from
Belgium, Denmark, France, Germany, Italy, The
Netherlands, Norway, Sweden, Switzerland and the UK.
This is EPICA publication nb 305. This study was supported
by a grant from the CNRS/INSU/LEFE program (‘IceChrono’
proposal). We also thank all the Dome Fuji Deep Ice Core
Project members who contributed to obtaining the ice core
samples, either through logistics, drilling or core processing.
The main logistics support was provided by the Japanese
Antarctic Research Expedition (JARE), managed by the
Ministry of Education, Culture, Sports, Science and
Technology (MEXT). This study was supported in part by a
Grant-in-Aid for Scientific Research (A) (20241007) and a
Grant-in-Aid for Young Scientists (S) (21671001) from the
Japan Society for the Promotion of Science (JSPS). The manuscript
was prepared with the support of National Institute of
Polar Research (NIPR) publication subsidy.
REFERENCES
Aristidi E and 11 others (2005) An analysis of temperatures and wind
speeds above Dome C, Antarctica. Astron. Astrophys. 430(2),
739–746 (doi: 10.1051/0004-6361:20041876)
Bazin L and 21 others (2013) An optimized multi-proxy, multi-site
Antarctic ice and gas orbital chronology (AICC2012): 120–800
ka. Clim. Past 9(4), 1715–1731 (doi: 10.5194/cp-9-1715-2013)
Bengtsson L, Koumoutsaris S and Hodges K (2011) Large-scale
surface mass balance of ice sheets from a comprehensive atmospheric
model. Surv. Geophys. 32(4), 459–474 (doi: 10.1007/
s10712-011-9120-8)
Bintanja R, de Wal RSW and Oerlemans J (2005) Modelled atmospheric
temperatures and global sea levels over the past million
years. Nature 437, 125–128 (doi: 10.1038/nature03975)
Cauquoin A and 9 others (2015) Comparing past accumulation rate
reconstructions in East Antarctic ice cores using 10Be, water isotopes
and CMIP5-PMIP3 models. Clim. Past 11(3), 355–367 (doi:
10.5194/cp-11-355-2015)
Cavitte MG and 7 others (2016) Deep radiostratigraphy of the East
Antarctic plateau: connecting the Dome C and Vostok ice core
sites. J. Glaciol. 62(232), 323–334 (doi: 10.1017/jog.2016.11)
Church JA and 13 others (2013) Sea level change. In Stocker TF,
Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J,
Nauels A, Xia Y, Bex V and Midgley PM eds. Climate change
2013: the physical science basis. contribution of working group
I to the fifth assessment report of the intergovernmental panel
on climate change. Cambridge, United Kingdom and
New York, NY, USA. http://ipcc.ch/report/ar5/wg1/
Dansgaard W (1964) Stable isotopes in precipitation. Tellus 16, 436–
468 (doi: 10.1111/j.2153-3490.1964.tb00181.x)
Durand G and 8 others (2007) Change in ice rheology during
climate variations – implications for ice flow modelling and
dating of the EPICA Dome C core. Clim. Past 3, 155–167 (doi:
10.5194/cp-3-155-2007)
Durand G and 7 others (2009) Evolution of the texture along
the EPICA Dome C ice core. Low Temp. Sci. 68(Supplement),
91–105
Ekaykin A (2003) Régime météorologique de l’Antarctique central et
son rôle dans la composition isotopique de la neige. Faculté de
géographie de Saint Pétersbourg. http://tel.archives-ouvertes.fr/
tel-00701466
EPICA community members (2004) 8 glacial cycles from an
Antarctic ice core. Nature 429, 623–628 (doi: 10.1038/
nature02599)
Frezzotti M and 13 others (2004) New estimations of precipitation
and surface sublimation in East Antarctica from snow accumulation
measurements. Clim. Dyn. 23(7–8), 803–813 (doi: 10.1007/
s00382-004-0462-5)
Frezzotti M, Scarchilli C, Becagli S, Proposito M and Urbini S (2013)
A synthesis of the Antarctic surface mass balance during the last
800 yr. Cryosphere 7(1), 303–319 (doi: 10.5194/tc-7-303-2013)
Fujita K and Abe O (2006) Stable isotopes in daily precipitation at
Dome Fuji, East Antarctica. Geophys. Res. Lett. 33(18), L18503
(doi: 10.1029/2006GL026936)
Fujita S and 6 others (1999) Nature of radio echo layering in the
Antarctic Ice Sheet detected by a two-frequency experiment. J.
Geophys. Res. Solid Earth 104(B6), 13013–13024 (doi:
10.1029/1999JB900034)
Fujita S and 9 others and others (2011) Spatial and temporal variability
of snow accumulation rate on the East Antarctic ice divide
between Dome Fuji and EPICA DML. Cryosphere 5, 1057–
1081 (doi: 10.5194/tc-5-1057-2011)
Fujita S and 7 others (2012) Radar diagnosis of the subglacial conditions
in Dronning Maud Land, East Antarctica. Cryosphere 6(5),
1203–1219 (doi: 10.5194/tc-6-1203-2012)
Fujita S, Parrenin F, Severi M, Motoyama H and Wolff EW (2015)
Volcanic synchronization of Dome Fuji and Dome C Antarctic
deep ice cores over the past 216 kyr. Clim. Past 11(10), 1395–
1416 (doi: 10.5194/cp-11-1395-2015)
Gallée H and 5 others (2012) Transport of snow by the wind: a comparison
between observations in Adélie Land, Antarctica, and
simulations made with the regional climate model MAR.
Bound. Layer Meteorol. 146(1), 133–147 (doi: 10.1007/
s10546-012-9764-z)
Gersonde R, Crosta X, Abelmann A and Armand L (2005) Sea-surface
temperature and sea ice distribution of the Southern Ocean at the
EPILOG Last Glacial Maximum—a circum-Antarctic view based
on siliceous microfossil records. Quat. Sci. Rev. 24(7–9), 869–
896 (doi: 10.1016/j.quascirev.2004.07.015)
Goujon C, Barnola J-M and Ritz C (2003) Modeling the densification
of polar firn including heat diffusion: application to close-off
characteristics and gas isotopic fractionation for Antarctica and
Greenland sites. J. Geophys. Res. 108 (D24), ACL10/1–10 (doi:
10.1029/2002JD003319)
Hirasawa N, Nakamura H and Yamanouchi T (2000) Abrupt changes
in meteorological conditions observed at an inland Antarctic
Station in association with wintertime blocking. Geophys. Res.
Lett. 27(13), 1911–1914 (doi: 10.1029/1999GL011039)
Hoshina Y and 8 others (2014) Effect of accumulation rate on water
stable isotopes of near-surface snow in inland Antarctica. J.
Geophys. Res. Atmos. 119(1), 2013JD020771 (doi: 10.1002/
2013JD020771)
Iizuka Y and 6 others (2012) Sulphate-climate coupling over the past
300 000 a in inland Antarctica. Nature 490(7418), 81–84 (doi:
10.1038/nature11359)
Jouzel J and Merlivat L (1984) Deuterium and oxygen 18 in precipitation:
modeling of the isotopic effects during snow formation. J.
Geophys. Res. Atmos. 89(D7), 11749–11757 (doi: 10.1029/
JD089iD07p11749)
Jouzel J and 6 others (1987) Vostok ice core: a continuous isotope
temperature record over the last climatic cycle (160 000 a).
Nature 329(6138), 403–408 (doi: 10.1038/329403a0)
1046 Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
Jouzel J and 6 others (2003) Magnitude of isotope/temperature scaling
for interpretation of central Antarctic ice cores. J. Geophys. Res.
Atmos. 108(D12), 4361 (doi: 10.1029/2002JD002677)
Jouzel J and 31 others (2007) Orbital and millennial Antarctic
climate variability over the past 800 000 a. Science 317(5839),
793–796 (doi: 10.1126/science.1141038)
Kameda T and 9 others (1997) Meteorological observations along a
traverse route from coast to Dome Fuji Station, Antarctica,
recorded by Automatic Weather Stations in 1995. Proc. of the
NIPR Symp. on Polar Meteorology and Glaciology. 35–50
Kameda T, Motoyama H, Fujita S and Takahashi S (2008) Temporal
and spatial variability of surface mass balance at Dome Fuji, East
Antarctica, by the stake method from 1995 to 2006. J. Glaciol. 54
(184), 107–116 (doi: 10.3189/002214308784409062)
Kawamura K and 11 others (2007) Northern hemisphere forcing of
climatic cycles over the past 360 000a implied by absolute
dating of Antarctic ice cores. Nature 448, 912–917 (doi:
10.1038/nature06015)
King JC, Argentini SA and Anderson PS (2006) Contrasts between the
summertime surface energy balance and boundary layer structure
at Dome C and Halley stations, Antarctica. J. Geophys.
Res. Atmos. 111(D2), D02105 (doi: 10.1029/2005JD006130)
Krinner G and Genthon C (1999) Altitude dependence of the ice
sheet surface climate. Geophys. Res. Lett. 26(15), 2227–2230
(doi: 10.1029/1999GL900536)
Lisiecki LE and Raymo ME (2005) A plio-pleistocene stack of 57 globally
distributed benthic δ18
O records. Paleoceanography 20(1),
PA1003 (doi: 10.1029/2004PA001071)
Lliboutry L (1979) A critical review of analytical approximate solutions
for steady state velocities and temperature in cold ice
sheets. Z. Gletscherkd. Glacialgeol. 15(2), 135–148
Massom RA and 6 others (2004) Precipitation over the Interior East
Antarctic ice sheet related to midlatitude blocking-high activity.
J. Clim. 17(10), 1914–1928 (doi: 10.1175/1520-0442(2004)
017<1914:POTIEA>2.0.CO;2)
Masson-Delmotte V and 35 others (2008) A review of Antarctic
surface snow isotopic composition: observations, atmospheric
circulation, and Isotopic modeling. J. Clim. 21(13), 3359–3387
(doi: 10.1175/2007JCLI2139.1)
Masson-Delmotte V and 18 others (2011) A comparison of the
present and last interglacial periods in six Antarctic ice cores.
Clim. Past 7(2), 397–423 (doi: 10.5194/cp-7-397-2011)
Mengel M and Levermann A (2014) Ice plug prevents irreversible
discharge from East Antarctica. Nat. Clim. Change 4(6), 451–
455 (doi: 10.1038/nclimate2226)
Motoyama H (2007) The Second Deep Ice Coring Project at Dome
Fuji, Antarctica. Sci. Drill. 5, 41–43 (doi: 10.2204/iodp.
sd.5.05.2007)
Motoyama H, Hirasawa N, Satow K and Watanabe O (2005)
Seasonal variations in oxygen isotope ratios of daily collected
precipitation and wind drift samples and in the final snow
cover at Dome Fuji Station, Antarctica. J. Geophys. Res.
Atmos. 110(D11), D11106 (doi: 10.1029/2004JD004953)
Motoyama H and others (2007) A New 3035.22 m deep ice core at
dome Fuji, Antarctica and reconstruction of global climate and
environmental change over past 720 kyr. AGU Fall Meeting
Abstracts. 76
Parish TR and Bromwich DH (1987) The surface windfield over the
Antarctic ice sheets. Nature 328(6125), 51–54 (doi: 10.1038/
328051a0)
Parrenin F, Rémy F, Ritz C, Siegert M and Jouzel J (2004) New modelling
of the Vostok ice flow line and implication for the glaciological
chronology of the Vostok ice core. J. Geophys. Res. 109,
D20102 (doi: 10.1029/2004JD004561)
Parrenin F and 15 others (2007) 1-D-ice flow modelling at EPICA
Dome C and Dome Fuji, East Antarctica. Clim. Past 3(2), 243–
259 (doi: 10.5194/cp-3-243-2007)
Parrenin F and 9 others (2013) Synchronous change of Atmospheric
CO2 and Antarctic temperature during the last deglacial
warming. Science 339(6123), 1060–1063 (doi: 10.1126/
science.1226368)
Pattyn F and 18 others (2012) Results of the Marine Ice Sheet Model
Intercomparison Project, MISMIP. Cryosphere 6(3), 573–588
(doi: 10.5194/tc-6-573-2012)
Pollard D and DeConto RM (2012) Description of a hybrid ice sheetshelf
model, and application to Antarctica. Geosci. Model Dev. 5
(5), 1273–1295 (doi: 10.5194/gmd-5-1273-2012)
Raymond CF (1983) Deformation in the vicinity of ice divides. J.
Glaciol. 29(103), 357–373
Reijmer CH, van den Broeke MR and Scheele MP (2002) Air parcel
trajectories and snowfall related to five deep drilling locations in
Antarctica based on the ERA-15 Dataset. J. Clim. 15(14), 1957–
1968 (doi: 10.1175/1520-0442(2002)015<1957:APTASR>2.0.
CO;2)
Rignot E, Mouginot J, Morlighem M, Seroussi H and Scheuchl B
(2014) Widespread, rapid grounding line retreat of Pine Island,
Thwaites, Smith, and Kohler glaciers, West Antarctica, from
1992 to 2011. Geophys. Res. Lett. 41(10), 3502–3509 (doi:
10.1002/2014GL060140)
Ritter F and 9 others (2016) Isotopic exchange on the diurnal scale
between near-surface snow and lower atmospheric water vapor
at Kohnen station, East Antarctica. Cryosph. Discuss., 1–35
(doi: 10.5194/tc-2016-4)
Ritz C, Rommelaere V and Dumas C (2001) Modeling the evolution
of Antarctic ice sheet over the last 420 000 a: implications for
altitude changes in the Vostok region. J. Geophys. Res. Atmos.
106(D23), 31943–31964 (doi: 10.1029/2001JD900232)
Saito F (2002) Development of a three dimensional ice sheet model for
numerical studies of Antarctic and Greenland ice sheet. University
of Tokyo, Department of Earth and Planetary Science, Tokyo, Japan
Saito F and Abe-Ouchi A (2010) Modelled response of the volume
and thickness of the Antarctic ice sheet to the advance of the
grounded area. Ann. Glaciol. 51(55), 41–48 (doi: 10.3189/
172756410791392808)
Scarchilli C, Frezzotti M and Ruti PM (2011) Snow precipitation at
four ice core sites in East Antarctica: provenance, seasonality
and blocking factors. Clim. Dyn. 37(9–10), 2107–2125 (doi:
10.1007/s00382-010-0946-4)
Siegert MJ, Hodgkins R and Dowdeswell JA (1998) A chronology for
the Dome C deep ice-core site through radio-echo layer
Correlation with the Vostok Ice Core, Antarctica. Geophys.
Res. Lett. 25(7), 1019–1022 (doi: 10.1029/98GL00718)
Sime LC, Wolff EW, Oliver KIC and Tindall JC (2009) Evidence for
warmer interglacials in East Antarctic ice cores. Nature 462
(7271), 342–345 (doi: 10.1038/nature08564)
Sodemann H and Stohl A (2009) Asymmetries in the moisture origin
of Antarctic precipitation. Geophys. Res. Lett. 36(22), L22803
(doi: 10.1029/2009GL040242)
Steen-Larsen HC and 18 others (2014) What controls the isotopic
composition of Greenland surface snow? Clim. Past 10(1),
377–392 (doi: 10.5194/cp-10-377-2014)
Stenni B and 7 others (2001) An oceanic cold reversal during the last
deglaciation. Science 293(5537), 2074–2077 (doi: 10.1126/
science.1059702)
Stenni B and 14 others (2010) The deuterium excess records of
EPICA Dome C and Dronning Maud Land ice cores (East
Antarctica). Quat. Sci. Rev. 29(1–2), 146–159 (doi: 10.1016/j.
quascirev.2009.10.009)
Suzuki K, Yamanouchi T and Motoyama H (2008) Moisture transport
to Syowa and Dome Fuji stations in Antarctica. J. Geophys. Res.
Atmos. 113(D24), D24114 (doi: 10.1029/2008JD009794)
Suzuki K, Yamanouchi T, Kawamura K and Motoyama H (2013) The
spatial and seasonal distributions of air-transport origins to the
Antarctic based on 5-day backward trajectory analysis. Polar
Sci. 7(3–4), 205–213 (doi: 10.1016/j.polar.2013.08.001)
Tabacco IE, Passerini A, Corbelli F and Gorman M (1998)
Determination of the surface and bed topography at Dome C,
East Antarctica. J. Glaciol. 44, 185–191
1047Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms
Takahashi S, Ageta Y, Fujii Y and Vvataabe O (1994) Surface mass
balance in east Dronning Maud Land, Antarctica, observed by
Japanese Antarctic research expeditions. Ann. Glaciol. 20(1),
242–248
Takahashi S, Kameda T, Enomoto H, Motoyama H and Watanabe O
(2004) Automatic Weather Station (AWS) data collected by the
33rd to 42nd Japanese Antarctic Research Expeditions during
1993-2001. JARE Data Rep. Meteorol. 36, 1–416
Uemura R and 5 others (2012) Ranges of moisture-source temperature
estimated from Antarctic ice cores stable isotope records
over glacial–interglacial cycles. Clim. Past 8(3), 1109–1125
(doi: 10.5194/cp-8-1109-2012)
Urbini S and 6 others (2008) Historical behaviour of Dome C and
Talos Dome (East Antarctica) as investigated by snow accumulation
and ice velocity measurements. Glob. Planet. Change
60(3–4), 576–588 (doi: 10.1016/j.gloplacha.2007.08.002)
Veres D and 15 others (2013) The Antarctic ice core chronology
(AICC2012): an optimized multi-parameter and multi-site
dating approach for the last 120 thousand years. Clim. Past 9
(4), 1733–1748 (doi: 10.5194/cp-9-1733-2013)
WAIS Divide Project Members (2013) Onset of deglacial warming in
West Antarctica driven by local orbital forcing. Nature 500
(7463), 440–444 (doi: 10.1038/nature12376)
Watanabe O and 5 others (2003a) Homogeneous climate variability
across East Antarctica over the past three glacial cycles. Nature
422(6931), 509–512 (doi: 10.1038/nature01525)
Watanabe O and 11 others (2003b) General tendencies of stable isotopes
and major chemical constituents of the Dome Fuji deep ice
core. Mem. Natl. Inst. Polar Res. Spec. 57, 1–24
Wolff EW and 27 others (2006) Southern Ocean sea-ice extent,
productivity and iron flux over the past eight glacial cycles.
Nature 440(7083), 491–496 (doi: 10.1038/nature04614)
MS received 3 March 2016 and accepted in revised form 9 June 2016; first published online 1 August 2016
1048 Parrenin and others: Climate dependent contrast in surface mass balance in East Antarctica over the past 216 ka
. https://doi.org/10.1017/jog.2016.85
Downloaded from https:/www.cambridge.org/core. University of South Bohemia, on 15 Feb 2017 at 13:52:26, subject to the Cambridge Core terms of use, available at https:/www.cambridge.org/core/terms