CPClimate of the PastCPClim. Past1814-9332Copernicus PublicationsGöttingen, Germany10.5194/cp-13-217-2017An improved north–south synchronization of ice core records around the 41 kyr 10Be peakRaisbeckGrant M.grant.raisbeck@csnsm.in2p3.frCauquoinAlexandrehttps://orcid.org/0000-0002-4620-4696JouzelJeanLandaisAmaellePetitJean-RobertLipenkovVladimir Y.BeerJuerghttps://orcid.org/0000-0001-8765-2041SynalHans-ArnoOerterHansJohnsenSigfus J.SteffensenJorgen P.SvenssonAndershttps://orcid.org/0000-0002-4364-6085YiouFrançoiseCSNSM, CNRS, Université Paris-Saclay, Bats 104–108, 91405 Campus,
Orsay, FranceLSCE/IPSL, CEA-CNRS-UVSQ), CEA Saclay Orme des Merisiers, 91191
Gif-sur-Yvette, FranceLGGE, CNRS, BP 96, 38402, St Martin d'Hères CEDEX, FranceArctic and Antarctic Research Institute, 38 Bering St., St. Petersburg
199397, RussiaEawag, Überlandstrasse 133, Postfach 611, 8600 Dübendorf,
SwitzerlandLaboratory of Ion Beam Physics, ETH Zurich, 8093 Zurich, SwitzerlandAlfred Wegener Institute for Polar and Marine Research, 27570
Bremerhaven, GermanyCentre for Ice and Climate, Niels Bohr Institute, University of
Copenhagen, Copenhagen, DenmarkdeceasedGrant M. Raisbeck (grant.raisbeck@csnsm.in2p3.fr)10March201713321722910July201628July201617December20165February2017This work is licensed under a Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/This article is available from https://cp.copernicus.org/articles/13/217/2017/cp-13-217-2017.htmlThe full text article is available as a PDF file from https://cp.copernicus.org/articles/13/217/2017/cp-13-217-2017.pdf
Using new high-resolution 10Be measurements in the
NGRIP, EDML and Vostok ice cores, together with previously published data
from EDC, we present an improved synchronization between Greenland and
Antarctic ice cores during the Laschamp geomagnetic excursion
∼ 41 kyr ago. We estimate the precision of this
synchronization to be ±20 years, an order of magnitude better than
previous work. We discuss the implications of this new synchronization for
making improved estimates of the depth difference between ice and enclosed
gas of the same age (Δdepth), difference between age of ice and
enclosed gas at the same depth (Δage) in the EDC and EDML ice cores,
spectral properties of the 10Be profiles and phasing between
Dansgaard–Oeschger-10 (in NGRIP) and AIM-10 (in EDML and EDC).
Introduction
In a previous study, Raisbeck et al. (2007) synchronized the Greenland
GRIP and Antarctic EPICA Dome C (EDC) ice cores using the structured peak of
cosmogenic 10Be resulting from a combination of the low geomagnetic
intensity and variable solar activity during the Laschamp geomagnetic
excursion ∼ 41 kyr ago (Lascu et al., 2016, and references therein).
The estimated precision of this synchronization was 200 years, due mainly to
uncertainties associated with the GRIP 10Be record (sample time
resolution of 25–50 years, corrections for loss of 10Be on
0.4 µm filters for some samples, and several missing samples). We
improve this situation significantly here by using a much higher resolution
(5–10 years) 10Be profile in the NGRIP ice core. In addition we report
and synchronize high-resolution 10Be profiles from the peak region of
two other Antarctic cores, EPICA Dronning Maud Land (EDML) and Vostok. Maps
showing the locations and a brief description of drilling operations for the
ice cores studied here can be found in Jouzel (2013). With these
improvements, we estimate the uncertainty on the synchronization to be
±20 years at our new tie points, and ±35 years over the whole
Laschamp event. This precision is supported by the study of Svensson et
al. (2013), who, on the basis of our earlier synchronization, identified
three volcanic signals which they believe are common to the Greenland and
Antarctic cores.
Methods
A continuous series of 55 cm long bag samples from the NGRIP core were
available from 2102 to 2140 m. Each of these were cut into five 11 cm
samples weighing ∼ 100 g and representing 5–10 years. The samples
were treated and measured on the Gif-sur-Yvette Tandetron AMS facility as
described in Raisbeck et al. (2007). AMS measurements were made relative to
NIST 10Be /9Be standard SRM 4320, assuming a nominal value of
2.68×10-11. While the value of 2.79×10-11 has now
been adopted by many groups for this standard, for consistency we continue to
use the value that was used for all the previous measurements at the Gif-sur-Yvette Tandetron. Since this factor is constant, it has no effect on our
synchronization procedures. An average of 400 10Be ions were counted for
each sample, and the estimated precision (not including the uncertainty of
the standard) was ∼ 7 %.
Vostok was one of the sites (the other being the original Dome C site) where
the 10Be peak discussed here was initially found (Raisbeck et al.,
1987). Although the estimated date was younger than presently accepted, and
the origin uncertain, it was already pointed out there that such a peak was a
potentially interesting stratigraphic marker. Those initial measurements were
made on discontinuous samples from the Vostok 3G core. We subsequently made
measurements on a nearly continuous sequence of 1 m samples from the 4G
core, which showed a double-humped peak (Raisbeck et al., 1992). It was in an
effort to look for even finer structure that the measurements shown here were
made. These samples were ∼ 10 cm in length and weighing > 500 g,
taken from the 5G core between 580 and 620 m. As can be seen in Fig. 1d,
the sampling depth is not centered on the peak. This is because the sampling
was based on the initial identification in 3G and, as we discovered, there is
apparently a ∼ 10 m offset between 3G and 5G. Since these measurements
were carried out > 15 years ago, also at the Gif-sur-Yvette AMS facility, the sample
preparation and measurement procedures are those described in Raisbeck et
al. (1987). For almost all samples at least 1000 10Be ions were counted,
leading to an estimated measurement uncertainty of ∼ 6 %.
10Be concentrations, 10Be flux and stable isotope
profiles (δD or δ18O) and their running averages as a
function of depth and age in four ice cores: (a) EPICA Dome C, EDC3 timescale
(Parrenin et al., 2007); (b) NGRIP, GICC05 timescale (Andersen et al.,
2006); (c) EDML, EDML1 timescale (Ruth et al., 2007); and (d) Vostok, FGT1 timescale
(Parrenin et al., 2004). The 10Be flux has been calculated using
accumulation rates used in the above timescales and smoothed with a 50-year
binomial filter. AIM stands for Antarctic isotope maximum.
The EDML 10Be measurements were performed at ETH Zurich with a sample
preparation that is based on the common procedure already applied for the
GRIP ice core (Yiou et al., 1997; Muscheler et al., 2004). Samples of 25 cm
length corresponding to ∼ 15-year resolution and a typical weight of
110 g were melted and processed without any further filtering. The
10Be /9Be ratios were measured using the compact AMS facility
Tandy at ETH Zurich (Müller et al., 2010) resulting in uncertainties
< 3 %. Results were normalized to the internal standard S2007N with a
reference value of
10Be /9Be = (28.1 ± 0.8) × 10-12
(Christl et al., 2013). The EDC 10Be measurements have been described
in detail previously (Raisbeck et al., 2007).
Beryllium 10 fluxes after synchronization to NGRIP (GICC05) using
the Match protocol (Lisiecki and Lisiecki, 2002) for EDC and EDML, or using
five tie points (A–E) and AnalySeries (Paillard et al., 1996) for Vostok. For
each site the detailed fluxes (histograms calculated using 11 cm
accumulation rates for NGRIP, 55 cm for EDC, 50 cm for EDML and 10 cm for
Vostok) and results smoothed with a 50-year binomial filter (curve) are
shown.
The climate records are provided by high-resolution isotopic profiles (either
δD or δ18O) measured along these ice cores. δD is
used as primary climate indicator for Vostok and Dome C, while δ18O
is used for NGRIP and EDML. In addition to existing water isotopic records,
new high-resolution measurements have been performed for this study on the
Vostok 5G ice core (10 cm). These isotopic data are shown in Fig. 1a to
d, with an additional smooth curve
(see legend).
The first two columns report the depth and GICC05 ages
(Svensson et al., 2008) of our 10Be tie points (BeA–BeE) and the
volcanic markers (L1–L3) of Svensson et al. (2013) at the NGRIP site (see
text). The following columns provide the depths of corresponding depths for
EDML, EDC and Vostok ice cores either observed or predicted.
a EDC 96 depth. b From Svensson et al. (2013)
using EDML-EDC correlation of Severi et al. (2007). c From 10Be synchronization (Fig. 3b, c).
From the 10Be concentrations, 10Be fluxes of the four cores were
calculated using accumulation rate estimates from water isotopes, ice
thinning and dating constraints by Parrenin et al. (2007) for EDC, Ruth et
al. (2007) for EDML, and Parrenin et al. (2004) for Vostok, and they were deduced from
layer counting (Svennson et al., 2008) combined with the ss06bm thinning
model of Johnsen et al. (2001) for NGRIP. Alternative accumulation rate
determinations were obtained more recently through dating exercises such as
for the AICC2012 timescale. These determinations do not show significant
differences with previous determinations over our period of interest (Bazin
et al., 2013; Veres et al., 2013; Andersen et al., 2006; Lemieux-Dudon et
al., 2015). Only for the NGRIP ice core was it recently suggested that the
mean accumulation rate of the MIS 3 could be overestimated by 20–30 %
(Guillevic et al., 2013; Kindler et al., 2014; Lemieux-Dudon et al., 2015).
Still, even when using this modified accumulation rate, the shape of the
variations of 10Be flux on the Laschamp event recorded at NGRIP is not
significantly affected. This is because, for our purpose, it is the relative
accumulation rates which are important.
Results
The results for the 10Be concentrations (see Supplement) and
fluxes of the four cores are shown in Fig. 1 as a function of their depth along
with their stable isotope profiles. As previously noted (Yiou et al., 1997;
Raisbeck et al., 2007) all cores show a highly structured peak of 10Be
centered on AIM-10 (EDC, EDML, Vostok) or DO-10 (NGRIP). It is this structure
that allows us to make multiple synchronizations over the whole width of the
peak, as shown in Fig. 2. This synchronization was done in two ways.
Initially we selected five tie points (four peaks and one valley, Table 1 and
Fig. 2) which appeared to us to be clearly common to all the profiles, and
synchronized using the AnalySeries program of Paillard et al. (1996).
Subsequently, in an effort to avoid the subjectivity of choosing tie points,
we did the synchronization using the MATCH protocol of Lisiecki and
Lisiecki (2002). The difference between these two procedures never differed
by more than 42 years for NGRIP-EDC and 27 years for NGRIP-EDML. The maximum
difference for the two procedures was for the period between tie points C and
D, and was very small (< 10 years) near the tie points. This confirms our
visual impression that the chosen tie points are indeed the most robust
common features of the profiles. For Vostok, we were unable to get a
satisfactory synchronization for the MATCH protocol without imposing severe
restrictions. This is probably due to several missing sample sections for
that record.
While we believe there is considerable evidence that flux is a better measure
of 10Be (and other trace species) deposition in polar cores, at the
suggestion of a reviewer we have repeated the Match synchronizations of EDC
and EDML with NGRIP using concentrations. We found that the four peaks and
valley corresponding to the tie points shown in Fig. 2 did not vary by more
than 20 years compared to those found using fluxes. In summary we believe our
10Be synchronization is quite robust to the choice of raw or smoothed
accumulations, or even the use of concentration.
In order to estimate the precision of our synchronization, we have used two
procedures. Initially we adopted a procedure which involved looking at the
correlation of EDC with EDML based on their 10Be synchronization with
NGRIP, and compared it with the direct correlation of EDC and EDML by Severi
et al. (2007) using common volcanic peaks. The results are shown in Fig. 3a.
As can be seen, the agreement is excellent. For the 6 volcanic peaks of Severi
et al. (2007) that fall within our 10Be tie points, the average
difference between the observed and predicted depths corresponds to
4.7 ± 6.2 years (2σ standard deviation). One of the reviewers
(C. Buizert) suggested that this “would provide a true test of the
uncertainty in the 10Be synchronization”. However, we feel that, while
it does strongly support our synchronization procedure, it may underestimate
the uncertainty between NGRIP and the Antarctic cores, as discussed below.
Comparison of the different tie points between (a) EDC and
EDML, (b) NGRIP and Dome C and (c) NGRIP and EDML, as derived from
10Be from this work (red squares and line), from Loulergue et al. (2007) (green full circles) from Raisbeck et al. (2007) (green empty
circles) with additional volcanic EDML/EDC tie points (open blue circles from Severi et al., 2007) and proposed bipolar volcanic points (full black squares from Svensson et al., 2013). We have indicated the uncertainty
for the tie points used in Loulergue et al. (2007). For the new
10Be and volcanic tie points, the uncertainties are
smaller than their symbols.
As a second estimate of the precision of our synchronization, we can look at
the volcanic correlations of Svensson et al. (2013). Starting from an earlier
10Be synchronization, and layer counting results in both NGRIP and EDML,
those authors identified three volcanic spikes they believed were common to
NGRIP-EDML, two of which are linked to EDC by Severi et al. (2007). In Table 1
and Fig. 3b and c, we show the position of those events as found by our improved
10Be synchronization. For EDML the differences are 66 cm
(∼ 44 years) for L1, 51 cm (∼ 28 years) for L2 and 44 cm
(∼ 24 years) for L3. In EDC the differences compared to those given by
Svensson et al. (2013) are 27 cm (∼ 23 years) for L2 and 36 cm
(∼ 31 years) for L3. As mentioned above, we believe our tie points have
the most robust synchronized ages. It is therefore reasonable to expect the
difference in predicted ages to increase with distance from those tie points.
However, L1 is an exception, having the largest difference (66 cm) despite
being the nearest (∼ 1 m) to a tie point (BeC). If that tie point is
correct, it is thus reasonable to question whether L1 is indeed a bipolar
event. In fact L1 was not listed as one of those used by Severi et al. (2007)
to correlate EDC with EDML.
For the other two volcanic peaks, it can be noted that the age differences
are systematic, having an average value of 27 ± 7 (2σ) years
for L2 and L3. Thus, the relative predicted ages between EDML and EDC for L2
and L3 differ by less than 10 years, consistent with their identification as
the same event by Severi et al. (2007). However, 27 ± 7 years is
significantly larger than the 4.7 ± 6.2-year precision found above. We
must therefore conclude that either L2 and L3 of Svensson et al. (2013) are
also not bipolar, or that 4.7 ± 6.2 years underestimates our real
uncertainty between NGRIP and EDC/EDML. We tend to favor the second
explanation. The reason is that the 10Be peaks involved are irregular
and have durations of the order of 100 years. Given this, and the 10Be
sampling resolution, it seems unlikely to us that they can be synchronized to
a precision of less than 10 years. Our synchronization procedure, either by
eye (AnalySeries) or Match protocol, will maintain a tight correlation
between EDC and EDML when synchronizing with any feature in NGRIP. However,
if for some reason (higher and more variable ice accumulation rate, higher
sampling resolution) the form of the 10Be flux peaks at NGRIP differs
slightly from those in EDC/EDML, then the correlation between NGRIP and the
Antarctic cores may be offset by several decades. For this reason, we prefer
to be conservative and adopt as our uncertainty the value implied assuming
that the volcanic peaks L2 and L3 of Svensson et al. (2013) are indeed
bipolar.
Fourier (left) and wavelet (right) analyses of our 10Be flux
records at (a) NGRIP, (b) EDC and (c) EDML from 40.480 to 42.320 kyr b2k on
the GICC05 age scale. The solid black lines on the wavelet panels indicate
the regions which are significant at the 95 % level.
In summary, if L2 and L3 are indeed bipolar events, and L1 is not, we
estimate from the above that the uncertainty in our synchronization is ≤ 35 years over the whole Laschamp event, and ≤ 20 years at our tie
points. This is an order of magnitude better than in Raisbeck et al. (2007).
It can also be compared with the 2σ uncertainty in the methane
interpolar synchronization (±73 years) estimated by Buizert et al. (2015)
at this age. This shows that, at least in regions where there is significant
structure in 10Be profiles, it is possible to link Greenland and
Antarctic climate records with decadal precision. As pointed out by Svensson
et al. (2013), using such correlation as a framework, it may be possible to
find other common volcanic signals in ice cores from the two hemispheres.
Spectral properties of the 10Be profiles
As indicated in the Introduction, the centennial variations in the 10Be
profiles are believed to be caused by variations in solar activity. For
example, during the Holocene, Steinhilber et al. (2012) found periodicities
of 88 years (Gleissberg cycle) and 210 years (de Vries
cycle) in a composite of tree ring 14C and ice core 10Be records.
For the Laschamp period, Wagner et al. (2001) reported a ∼ 205-year
periodicity of 10Be in the GRIP core. It is thus interesting to do a
spectral analysis on the higher resolution and better quality NGRIP 10Be
record reported here. This is shown in Fig. 4, where a very significant
(> 99 %) 200-year signal is found in the Fourier spectrum (REDFIT
program of Schulz and Mudelsee, 2002). As can be seen in the wavelet spectra,
however (MATLAB package of Grinsted et al., 2004), this periodicity is really
only significant for a short interval near the end of the time period
studied. While there is also a short interval where an 88-year periodicity
appears, this is not significant in the Fourier spectrum.
Since we now have the EDC and EDML 10Be profiles synchronized on the
NGRIP timescale, we might expect these to show similar periodicities.
Somewhat to our surprise, however, this is only partially the case. While
both show a ∼ 200-year signal in the same time region of the wavelet
analyses, these are barely significant (∼ 90 %) in the Fourier
spectra. This is because that peak appears to be distorted by longer signals
of ∼ 290 and 350 years in EDC and EDML, respectively. This is consistent
with the observation of Cauquoin et al. (2014), who found that the 210-year
periodicity in the EDML Holocene data used by Steinhilber et al. (2012), the
only Antarctic core in their composite, was only significant over short and
sporadic time periods. It is also consistent with the fact that Cauquoin et
al. (2014) did not find any evidence for this periodicity in 10Be from
EDC during the interval 336–325 ka BP. It thus appears that there may be
meteorological effects in the Antarctic (smaller ice accumulation rate?)
which are relatively more important than in Greenland, and which tend to
diminish or interfere with the 10Be production signal recorded in the
ice, compared to Greenland.
Δage and Δdepth derived for the five 10Be tie
points derived in this work compared with their values as used in scenario 1
and 4 (Loulergue et al., 2007) and for AICC2012 (Bazin et al., 2013).
Implications for Δdepth and Δage estimates
Using preliminary values of our previous 10Be synchronization, together
with methane correlations, Loulergue et al. (2007) estimated the difference
in depth (Δdepth) and age (Δage) between the ice and gas
records in EDC and EDML. The discrepancy they found at EDC compared with the
classical firn densification model (Goujon et al., 2003) was in fact the
first hard evidence that this model was inadequate for glacial ice from low
accumulation sites in Antarctica. This had important implications for the
EDC4 gas age construction. Loulergue et al. (2007) hence proposed that the
Δage obtained by the Goujon et al. (2003) model using temperature and
accumulation rate deduced from water isotopes (scenario 1 in Fig. 5 and
Loulergue et al., 2007) was not appropriate. Instead, Δage for the
EDC gas age was calculated with the Goujon et al. (2003) model but with an
artificially higher accumulation rate than the one used in the construction
of the EDC3 ice timescale (scenario 4 in Fig. 5 and Loulergue et al., 2007),
so that the calculated Δage at 41 kyr was in agreement with the
Δage determination using 10Be and CH4 synchronization.
In the construction of the AICC2012 chronology (Bazin et al., 2013; Veres et
al., 2013), a different approach for the Δage calculation has been
chosen. The construction of AICC2012 relies on the use of a Bayesian tool,
DATICE (Lemieux-Dudon et al., 2010), to optimize the chronology of five ice
cores using absolute dating constraints and stratigraphic links in the gas
and ice phases. The AICC2012 Δage calculation for EDC is hence
constrained by the CH4 and 10Be data as in Loulergue et al. (2007).
It also benefits from the recent finding of Parrenin et al. (2012) showing
that the firn lock-in depth at Dome C during the last deglaciation was best
determined from δ15N measurements in air trapped in ice cores
rather than by outputs of firn modeling. As a consequence, the background
scenario used for the lock-in depth and hence the Δage in the
construction of the AICC2012 chronology for EDC was based on δ15N
measurements performed at Dome C (Dreyfus et al., 2010; Landais et al.,
2013),
leading to a significantly smaller Δage than the firn model output
for EDC during the glacial period. It should be emphasized that no firn model
has been used for the background scenario of the EDC lock-in depth in the
construction of AICC2012, which makes the comparison of Δage and
Δdepth between the AICC2012 and EDC3 gas timescales not
straightforward.
Using the five tie points from the present synchronization, we have used a
slightly modified version of the technique described by Loulergue et
al. (2007) to calculate Δdepth of EDC and EDML. Instead of
correlating the methane profiles at a single point, we have correlated the
whole profiles using the Match protocol (Fig. 6). Then, a Δage has
been determined for each 10Be peak. The resulting Δages and
Δdepths (Fig. 5) are compared to the Δdepth and Δage
provided by Loulergue et al. (2007) for the construction of the EDC3 gas
timescale. Our results are in reasonably good agreement with the Δage
and Δdepths calculated using scenario 4 in Loulergue et al. (2007).
Our results are also in good agreement with the AICC2012 Δage at EDC,
hence validating the chosen approach for delta age calculation. It is clear
from Fig. 6 that the resolution of the methane data in EDC and EDML,
particularly in the period 41.3–42.5 kyr, is the limiting factor in
calculating the Δage uncertainty.
Synchronization of the methane profile from NGRIP with EDC and
EDML profiles. The methane data are from Baumgartner et al. (2014) for
NGRIP, Loulergue et al. (2007) for EDC and Schilt et al. (2010) for EDML.
Implications for bipolar seesaw and stable isotope interpretation
The last glacial period is characterized by a succession of millennial-scale
climatic events identified both in Greenland and in Antarctica. In Greenland,
the climate shifts are very abrupt with a temperature shift between a cold
Greenland stadial (GS) to a warm Greenland interstadial (GIS) designated as
Dansgaard–Oeschger (DO) events. The DO succession and amplitude seems quite
unpredictable even if it has already been suggested that the sequence of the
abrupt variability of the last glacial period (length and frequency of GS and
GIS) can be affected by the long-term (orbital) climatic variability (Schulz
et al., 2002; Capron et al., 2010). In Antarctica the temperature changes of
the corresponding climatic events are much smoother, and are designated as
Antarctic isotopic maxima (AIM). A clear relationship has been demonstrated
between DO and AIM events: each DO event is associated with an AIM (EPICA
community members, 2006; Jouzel et al., 2007) and the temperature/isotopic
maximum of the AIM occurs approximately synchronously with the Greenland
abrupt temperature increase (Fig. 7). The observed relationship between
Greenland and Antarctica temperature evolutions led to the proposed theory of
the bipolar seesaw (Broecker, 1998; Stocker, 1998). The global
picture of the bipolar seesaw can be explained by a simple modeling of a slow
thermal response of Antarctic temperature to Greenland abrupt warmings and
coolings through a heat reservoir in the Southern Ocean (Stocker and Johnsen,
2003).
Water isotopic records for NGRIP and four Antarctic ice cores over DO
events 5 to 12 on the GICC05 timescale with the transition between GS and
GIS indicated by the vertical orange lines. The synchronization between
NGRIP (red), EDML (blue), EDC (black) and Vostok (brown) is based on the
AICC2012 timescale, while the WDC (green) record has been transferred on the
GICC05 timescale as WD2014 in Buizert et al. (2015).
The knowledge of the exact relative timing between DO events in Greenland
and AIM events in Antarctica is limited by the ice core chronologies.
Usually, Greenland and Antarctic ice cores are synchronized using the global
atmospheric signals in the air trapped in ice core (CH4, δ18O of O2). However, the climatic signals are recorded in the ice
phase through water isotopic composition. Because the entrapped air is
systematically younger than the entrapping ice, by centuries to millennia,
accurately determining this age difference is crucial for the Greenland vs.
Antarctica phasing.
Recent studies have allowed for the age uncertainty
on Greenland vs. Antarctica climatic sequences to be significantly decreased. This has been done through an
increase in the number of stratigraphic links between Greenland and Antarctic
ice cores (e.g., Schüpbach et al., 2011; Svensson et al., 2013) and new
high-resolution Antarctic records at relatively high accumulation sites, and
hence with small ice–air age differences (WAIS Divide Project Members,
2015). These recent improvements have revealed a refined sequence of
Greenland vs. lower latitudes and Antarctic climatic changes during the last
glacial period. First, for GS associated with an occurrence of Heinrich
events, it has been evidenced that a decoupling exists between the Greenland
or high-latitude climate and lower latitudes, with Greenland remaining in a
cold stable phase, while lower or Southern Hemisphere latitudes exhibit
significant climatic variations (Barker et al., 2015; Guillevic et al., 2014;
Rhodes et al., 2015; Landais et al., 2015). Second, a comparison of several
climatic records in Antarctica with the Greenland sequence has highlighted
regional differences in the shape of the AIM events and in the phasing
between the temperature increase in Antarctica and the abrupt temperature
change in Greenland over the long GSs that are associated with a strong
discharge in the North Atlantic, i.e., Heinrich stadials (Buiron et al., 2012;
Landais et al., 2015). Third, an accurate dating exercise on the west
Antarctica WDC ice core over the last 60 kyr suggests that the decrease in
Antarctic temperature occurs ∼ 200 years after the abrupt temperature
increase in Greenland (WAIS Divide Project Members, 2015). This has been
interpreted as a “northern push for bipolar seesaw” (van Ommen, 2015).
Still these studies suffer from the same limitation associated with the
ice–air age difference and rely on determination of the past depth of the
firn, based on firn densification and/or measurements of δ15N of
N2 in the trapped air.
Here we use the direct synchronization of NGRIP with EDC, EDML and Vostok
based on the 10Be records. In Fig. 8 we show the 10Be-synchronized
climate records of the four cores, as represented by the δD and δ18O profiles. Comparing the three Antarctic records, one can observe that,
while they have the same general features, there are also significant
differences in detail. These are probably due to meteorological effects such
as seasonal variation in isotopic ratios and depositional effects. This
shows that one must be careful to not over-interpret details of individual
stable isotope profiles in low-accumulation regions of the Antarctic as a
climate proxy. This is particularly evident in the Vostok 5G profile, which
differs significantly from the previously published profile from the 3G core
(Jouzel et al., 1987).
Phasing between the GS–GIS transition in Greenland
(vertical orange line) and the maximum of the AIM 10 in 4 different
Antarctic ice cores on the GICC05 age scale. The synchronicity between
Vostok, EDC, EDML and NGRIP is based on the 10Be
records while the WDC isotopic record has been drawn on the GICC05 timescale
using the correspondence between the WSD timescale and GICC05 given in
Buizert et al. (2015). A stack (purple curve) is drawn using the 3 δ18O records from EDML, EDC and WDC, and the black
lines correspond to its inflections identified by eye.
Phasing between NGRIP GS–GIS transition and AIM for different
Antarctic ice cores over DO-AIM 10.
Age difference of the inflection point with respect to the mid-slope of abrupt warming in NGRIPMATLAB algorithm (WAIS Divide Project Members, 2015)BREAKFIT software (Mudelsee, 2010)EDML (this study)-213 ± 40 years-207 ± 50 yearsEDC (this study)+130 ± 30 years+150 ± 50 yearsVostok (this study)Non-significantNon-significantWDC, WAIS Divide Project Members (2015)-156 ± 50 years-129 ± 50 years
We confirm the Greenland vs. Antarctica relationship observed by Raisbeck et
al. (2007) with the increase in δD or δ18O in Antarctica
leading the abrupt δ18O increase in NGRIP by several centuries,
which is a classical feature of the bipolar seesaw (Blunier and Brook, 2001),
operating here for a small DO (AIM) event. The high resolution of our water
isotopes and 10Be records also enables us to look at the fine-scale
Greenland vs. Antarctica temporal relationship around the abrupt warming
recorded at NorthGRIP. Although it is impossible to identify by eye any clear
inflection point corresponding to AIM 10 in Vostok, a significant peak of
δ18O is observed at WDC and EDML about 200 years before the abrupt
δ18O increase at NorthGRIP, while the most significant peak in EDC
is observed about 150 years after the abrupt Greenland δ18O
increase. The signal is not unambiguous at EDML since two other prominent
peaks (but peaking at lower δ18O values) are also identified in the
following centuries. The EDC, WDC and EDML δ18O profiles also
display a δ18O decrease occurring about 200 years after the
NorthGRIP δ18O increase, as was extensively discussed in WAIS
Divide Project Members (2015).
For a proper comparison with the results displayed on WDC in WAIS Divide
Project Members (2015), we have used the same statistical approaches,
i.e., the BREAKFIT software by Mudelsee (2010) and a similar automated
routine (referred to as MATLAB routine in Table 2) using a second-order
polynomial (rather than a linear for BREAKFIT) fit to the data. The water
isotope profiles for EDML and EDC show several wiggles associated with a
variability of ∼ 100–200 years. Despite this variability, it is still
possible to identify a maximum for AIM 10 using the aforementioned
statistical tools, keeping in mind limitations linked to the chosen range of
detection and short-term variability in δ18O signal. The two
statistical tools identify inflection points on the EDML and EDC
δ18O records. Using both methods, the maximum is reached at EDML,
∼ 210 years before the abrupt Greenland warming, while the maximum
detected at Dome C appears to be ∼ 140 years later. To complete this
study, using the same statistical tools we have determined the shift between
the abrupt Greenland warming and the maximum in the water isotopic record at
WDC, and found that the WDC δ18O maximum is ∼ 142 years
earlier. Such analysis leads to the conclusion that all ice cores do not
display the same isotopic signal over an AIM, a result in line with previous
studies highlighting a strong regional variability at the AIM scale in
Antarctica (e.g., Buiron et al., 2012; Landais et al., 2015).
As suggested by a referee (C. Buizert), a stack can be drawn using the 3
δ18O records from EDML, EDC and WDC (Fig. 8). Because we could not
get a 10Be synchronization with the Match protocol, and unlike EDC and EDML
have no independent evidence from volcanic spikes to support our estimated
precision, we prefer not to include Vostok in the stack, although tests
including it show no significant influence on our conclusions. We emphasize
that constructing such a stack is appropriate only if one assumes “a
priori” that the “true” climatic part of the δ18O records for
the three core locations are synchronous at the studied timescale, which remains
to be shown. This stack clearly shows the δ18O decrease observed
200 years after the abrupt δ18O increase in Greenland, but it also
depicts an inflection point ∼ 200 years (185–219 years according to the
MATLAB routine) before the Greenland abrupt δ18O increase, so that
on the stack, δ18O displays a ∼ 400-year-wide plateau at the
time of the abrupt Greenland δ18O increase. Such a plateau is not
unexpected and follows from our observation that in some ice cores (WDC,
EDML) the maximum of the AIM was before the abrupt δ18O increase in
Greenland, while in another (EDC) the maximum of the AIM occurs after the
abrupt Greenland δ18O increase.
This result seems at first sight to nuance the conclusions presented in WAIS
Divide Project Members (2015) using the same approach for the determination
of the breakpoint. In both the EDML and WDC ice cores, we find that the
maximum in Antarctica for AIM 10 is reached more than 140 years before the
Greenland warming. The stack of Antarctic δ18O records over AIM 10
also suggests that the maximum δ18O level is reached
∼ 200 years before the Greenland warming. This suggests one must be
cautious in drawing a conclusion about a general mechanism for AIM vs. DO
dynamics based on a stack of several cores for a given DO-AIM event.
Similarly, one must be cautious about drawing conclusions based on a stack of
different DO-AIM events in the same core, as proposed in WAIS Divide Project
Members (2015). Indeed, the lead/lag of Greenland vs. Antarctica may be
different from one event to another. This is evident from Fig. 7, where the
maximum of AIM 7, 8 and 10 clearly leads the abrupt warming in Greenland both
at WDC and EDML. Both stacking several AIM together on one core or stacking
one AIM using several cores can be misleading when discussing short-term
variability that may be different from one core to another, or one event to
another.
Finally, even if the location of the maximum of the AIM, hence the statement
of “northern push for bipolar seesaw”, can be disputed, we confirm that the
main δ18O decrease in Antarctica occurs ∼ 200 years after the
Greenland temperature increase. This particular result is robust in different
Antarctic sectors and for different DO/AIM events and should be retained to
refine our understanding of bipolar seesaw.
Conclusion
We have shown here that, in periods where there are significant production
variations of 10Be, it is possible to synchronize Greenland and
Antarctic ice cores with decadal precision. This in turn means that the
climate and other environmental parameters registered in these ice cores can
be synchronized at this same precision, thus allowing different models and
mechanisms to be more finely tested.
The data in the Supplement are also available in the PANGAEA
database: https://doi.pangaea.de/10.1594/PANGAEA.872454 (NGRIP, EDML,
Vostok) and https://doi.pangaea.de/10.1594/PANGAEA.872445
(EDC).
The Supplement related to this article is available online at doi:10.5194/cp-13-217-2017-supplement.
Acknowledgements
This project was initiated during discussions between the first author and
Sigfus Johnsen at an EPICA meeting at San Feliu, Spain, in 2003. Sigfus was
always a strong supporter of 10Be measurements in ice cores, and
enthusiastically responded positively to the idea of using high-resolution
NGRIP samples to overcome limitations encountered with GRIP samples. We
therefore dedicate this paper to Sigfus' memory. We thank C. Buizert and an
anonymous reviewer for helpful comments. We would like to
acknowledge the many people involved in logistics, drill development and
implementation, ice core processing and analysis; without them this work
would not have been possible. This includes NorthGRIP, directed and organized
by the Department of Geophysics at the Niels Bohr Institute for Astronomy,
Physics and Geophysics, University of Copenhagen, and supported by funding
agencies and institutions from participating countries, and EPICA, the
European Project for Ice Coring in Antarctica, 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. We thank all Russian and French participants of the
Vostok drilling, field work and ice sampling. The research leading to these
results has received funding from the European Research Council under the
European Union's Seventh Framework Programme (FP7/20072013)/ERC grant
agreement no. 30604. Edited by: K.
Goto-Azuma Reviewed by: C. Buizert and one anonymous referee
ReferencesAndersen, K. K., Svensson, A., Rasmussen, S. O., Steffensen, J. P., Johnsen,
S. J., Bigler, M., Rothlisberger, R., Ruth, U., Siggaard-Andersen, M.-L.,
Dahl-Jensen, D., Vinther, B. M., and Clausen, H. B.: The Greenland Ice Core
Chronology 2005, 15–42 ka. Part 1: constructing the time scale, Quaternary
Sci. Rev., 25, 3246–3257, 10.1016/j.quascirev.2006.08.002, 2006.Baumgartner, M., Kindler, P., Eicher, O., Floch, G., Schilt, A., Schwander,
J., Spahni, R., Capron, E., Chappellaz, J., Leuenberger, M., Fischer, H., and
Stocker, T. F.: NGRIP CH4 concentration from 120 to 10 kyr before present
and its relation to a δ15N temperature reconstruction from the same
ice core, Clim. Past, 10, 903–920, 10.5194/cp-10-903-2014, 2014.Barker, S., Chen, J., Gong, X., Jonkers, L., Knorr, G., and Thornalley, D.:
Icebergs not the trigger for North Atlantic cold events, Nature, 520,
333–336, 10.1038/nature14330, 2015.Bazin, L., Landais, A., Lemieux-Dudon, B., Toyé Mahamadou Kele, H.,
Veres, D., Parrenin, F., Martinerie, P., Ritz, C., Capron, E., Lipenkov, V.,
Loutre, M.-F., Raynaud, D., Vinther, B., Svensson, A., Rasmussen, S. O.,
Severi, M., Blunier, T., Leuenberger, M., Fischer, H., Masson-Delmotte, V.,
Chappellaz, J., and Wolff, E.: An optimized multi-proxy, multi-site Antarctic
ice and gas orbital chronology (AICC2012): 120–800 ka, Clim. Past, 9,
1715–1731, 10.5194/cp-9-1715-2013, 2013.Blunier, T. and Brook, E. J.: Timing and millennial-scale climate change in
Antarctica and Greenland during the last glacial period, Science, 291,
109–112, 10.1126/science.291.5501.109, 2001.Broecker, W. S.: Paleocean circulation during the last deglaciation: A
bipolar seesaw?, Paleoceanography, 13, 119–121, 10.1029/97PA03707, 1998.Buiron, D., Stenni, B., Chappellaz, J., Landais, A., Baumgartner, M.,
Bonazza, M., Capron, E., Frezzotti, M., Kageyama, M., Lemieux-Dudon, B.,
Masson-Delmotte, V., Parrenin, F., Schilt, A., Selmo, E., Severi, M.,
Swingedouw, D., and Udisti, R.: Regional imprints of millennial variability
during the MIS 3 period around Antarctica, Quaternary Sci. Rev., 48, 99–112,
10.1016/j.quascirev.2012.05.023, 2012.Buizert, C., Cuffey, K. M., Severinghaus, J. P., Baggenstos, D., Fudge, T.
J., Steig, E. J., Markle, B. R., Winstrup, M., Rhodes, R. H., Brook, E. J.,
Sowers, T. A., Clow, G. D., Cheng, H., Edwards, R. L., Sigl, M., McConnell,
J. R., and Taylor, K. C.: The WAIS Divide deep ice core WD2014 chronology –
Part 1: Methane synchronization (68–31 ka BP) and the gas age–ice age
difference, Clim. Past, 11, 153–173, 10.5194/cp-11-153-2015, 2015.Capron, E., Landais, A., Chappellaz, J., Schilt, A., Buiron, D., Dahl-Jensen,
D., Johnsen, S. J., Jouzel, J., Lemieux-Dudon, B., Loulergue, L.,
Leuenberger, M., Masson-Delmotte, V., Meyer, H., Oerter, H., and Stenni, B.:
Millennial and sub-millennial scale climatic variations recorded in polar ice
cores over the last glacial period, Clim. Past, 6, 345–365,
10.5194/cp-6-345-2010, 2010.Cauquoin, A., Raisbeck, G. M., Jouzel, J., Bard, E., and ASTER Team: No
evidence for planetary influence on solar activity 330 ka ago, Astron.
Astrophys., 561, A132, 10.1051/0004-6361/201322879, 2014.Christl, M., Vockenhuber, C., Kubik, P. W., Wacker, L., Lachner, J., Alfimov,
V., and Synal, H. A.: The ETH Zurich AMS facilities: Performance parameters
and reference materials, Nucl. Instrum. Meth. B, 294, 29–38,
10.1016/j.nimb.2012.03.004, 2013.Dreyfus, G., Jouzel, J., Bender, M. L., Landais, A., Masson-Delmotte, V., and
Leuenberger, M.: Firn processes and δ15N: potential for a gas-phase
climate proxy, Quaternary Sci. Rev., 29, 28–42,
10.1016/j.quascirev.2009.10.012, 2010.EPICA community members: One-to-one coupling of glacial climate variability
in Greenland and Antarctica, Nature, 444, 195–198, 10.1038/nature05301,
2006.Goujon, C., Barnola, J. M., and Ritz, C.: 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, 101–1018, 10.1029/2002JD003319, 2003.Grinsted, A., Moore, J. C., and Jevrejeva, S.: Application of the cross
wavelet transform and wavelet coherence to geophysical time series, Nonlin.
Processes Geophys., 11, 561–566, 10.5194/npg-11-561-2004, 2004.Guillevic, M., Bazin, L., Landais, A., Kindler, P., Orsi, A.,
Masson-Delmotte, V., Blunier, T., Buchardt, S. L., Capron, E., Leuenberger,
M., Martinerie, P., Prié, F., and Vinther, B. M.: Spatial gradients of
temperature, accumulation and δ18O-ice in Greenland over a series
of Dansgaard–Oeschger events, Clim. Past, 9, 1029–1051,
10.5194/cp-9-1029-2013, 2013.Guillevic, M., Bazin, L., Landais, A., Stowasser, C., Masson-Delmotte, V.,
Blunier, T., Eynaud, F., Falourd, S., Michel, E., Minster, B., Popp, T.,
Prié, F., and Vinther, B. M.: Evidence for a three-phase sequence during
Heinrich Stadial 4 using a multiproxy approach based on Greenland ice core
records, Clim. Past, 10, 2115–2133, 10.5194/cp-10-2115-2014, 2014.Johnsen, S., Dahl-Jensen, D., Gundestrup, N., Steffenssen, J. P., Clausen, H.
B., Miller, H., Masson-Delmotte, V., Sveinbjörnsdottir, A. E., and White,
J.: Oxygen isotope and palaeotemperature records from six Greenland ice-core
stations: Camp Century, Dye-3, GRIP, GISP2, Renland and NorthGRIP, J.
Quaternary Sci., 16, 299–307, 10.1002/jqs.622, 2001.Jouzel, J.: A brief history of ice core science over the last 50 yr, Clim.
Past, 9, 2525–2547, 10.5194/cp-9-2525-2013, 2013.
Jouzel, J., Lorius, C., Petit, J. R., Genthon, C., Barkov, N. I., Kotlyakov,
V. M., and Petrov, V. N.: Vostok ice core: a continuous isotope temperature
record over the last climatic cycle (160 000 years), Nature, 329, 403–408,
1987.Jouzel, J., Masson-Delmotte, V., Cattani, O., Dreyfus, G., Falourd, S.,
Hoffmann, G., Minster, B., Nouet, J., Barnola, J.-M., Chappellaz, J.,
Fischer, H., Gallet, J. C., Johnsen, S. J., Leuenberger, M., Loulergue, L.,
Luethi, D., Oerter, H., Parrenin, F., Raisbeck, G., Raynaud, D., Schilt, A.,
Schwander, J., Selmo, E., Souchez, R., Spahni, R., Stauffer, B., Steffensen,
J. P., Stenni, B., Stocker, T., Tison, J.-L., Werner, M., and Wolff, E.:
Orbital and millennial Antarctic climate variability over the past 800,000
years, Science, 317, 793–797, 10.1126/science.1141038, 2007.Kindler, P., Guillevic, M., Baumgartner, M., Schwander, J., Landais, A., and
Leuenberger, M.: Temperature reconstruction from 10 to 120 kyr b2k from the
NGRIP ice core, Clim. Past, 10, 887–902, 10.5194/cp-10-887-2014, 2014.Landais, A., Dreyfus, G., Capron, E., Jouzel, J., Masson-Delmotte, V., Roche,
D. M., Prie, F., Caillon, N., Chappellaz, J., Leuenberger, M., Lourantou, A.,
Parrenin, F., Raynaud, D., and Teste, G.: Two-phase change in CO2,
Antarctic temperature and global climate during Termination II, Nat. Geosci.,
6, 1062–1065, 10.1038/ngeo1985, 2013.Landais, A., Masson-Delmotte, V., Stenni, B., Selmo, E., Roche, D. M.,
Jouzel, J., Lambert, F., Guillevic, M., Bazin, L., Arzel, O., Vinther, B.,
Gkinis, V., and Popp, T.: A review of the bipolar see–saw from synchronized
and high resolution ice core water stable isotope records from Greenland and
East Antarctica, Quaternary Sci. Rev., 114, 18–32,
10.1016/j.quascirev.2015.01.031, 2015.Lascu, I., Feinberg, J. M., Dorale, J. A., Cheng, H., and Edwards, R. L.: Age
of the Laschamp excursion determined by U-Th dating of a speleotherm
geomagnetic record from North America, Geology, 44, 139–142,
10.1130/G37490.1, 2016.Lemieux-Dudon, B., Blayo, E., Petit, J. R., Waelbroeck, C., Svensson, A.,
Ritz, C., Barnola, J. M., Narcisi, B. M., and Parrenin, F.: Consistent dating
for Antarctic and Greenland ice cores, Quaternary Sci. Rev., 29, 8–20,
10.1016/j.quascirev.2009.11.010, 2010.Lemieux-Dudon, B., Bazin, L., Landais, A., Toyé Mahamadou Kele, H.,
Guillevic, M., Kindler, P., Parrenin, F., and Martinerie, P.: Implementation
of counted layers for coherent ice core chronology, Clim. Past, 11, 959–978,
10.5194/cp-11-959-2015, 2015.Lisiecki, L. E. and Lisiecki, P. A.: Application of dynamic programming to
the correlation of paleoclimate records, Paleoceanography, 17, 1049,
10.1029/2001PA000733, 2002.Loulergue, L., Parrenin, F., Blunier, T., Barnola, J.-M., Spahni, R., Schilt,
A., Raisbeck, G., and Chappellaz, J.: New constraints on the gas age-ice age
difference along the EPICA ice cores, 0–50 kyr, Clim. Past, 3, 527–540,
10.5194/cp-3-527-2007, 2007.Müller, A. M., Christl, M., Lachner, J., Suter, M., and Synal, H.-A.:
Competitive 10Be measurements below 1 MeV with the upgraded ETH-TANDY
AMS facility, Nucl. Instrum. Meth. B, 268, 2801–2807,
10.1016/j.nimb.2010.05.104, 2010.
Mudelsee, M.: Climate time series analysis: classical statistical and
bootstrap methods, Springer, Dordrecht, 2010.Muscheler, R., Beer, J., Wagner, G., Laj, C., Kissel, C., Raisbeck, G. M.,
Yiou, F., and Kubik, P. W.: Changes in the carbon cycle during the last
deglaciation as indicated by the comparison of 10Be and 14C
records, Earth Planet. Sc. Lett., 219, 325–340,
10.1016/S0012-821X(03)00722-2, 2004.Paillard, D., Labeyrie, L., and Yiou, P.: Macintosh program performs
time-series analysis, EOS T. Am. Geophys. Un., 77, p. 379,
10.1029/96EO00259, 1996.Parrenin, F., Remy, F., Ritz, C., Siegert, M. J., and Jouzel, J.: New
modelling of the Vostok ice flow line and implication for the glaciological
chronology of the Vostok ice core, J. Geophys. Res., 109, D20102,
10.1029/2004JD004561, 2004.Parrenin, F., Barnola, J.-M., Beer, J., Blunier, T., Castellano, E.,
Chappellaz, J., Dreyfus, G., Fischer, H., Fujita, S., Jouzel, J., Kawamura,
K., Lemieux-Dudon, B., Loulergue, L., Masson-Delmotte, V., Narcisi, B.,
Petit, J.-R., Raisbeck, G., Raynaud, D., Ruth, U., Schwander, J., Severi, M.,
Spahni, R., Steffensen, J. P., Svensson, A., Udisti, R., Waelbroeck, C., and
Wolff, E.: The EDC3 chronology for the EPICA Dome C ice core, Clim. Past, 3,
485–497, 10.5194/cp-3-485-2007, 2007.Parrenin, F., Barker, S., Blunier, T., Chappellaz, J., Jouzel, J., Landais,
A., Masson-Delmotte, V., Schwander, J., and Veres, D.: On the gas-ice depth
difference (Δdepth) along the EPICA Dome C ice core, Clim. Past, 8,
1239–1255, 10.5194/cp-8-1239-2012, 2012.Raisbeck, G. M., Yiou, F., Bourlès, D., Lorius, C., Jouzel, J., and
Barkov, N. I.: Evidence for two intervals of enhanced Be-10 deposition in
Antarctic ice during the last glacial period, Nature, 326, 273–277,
10.1038/326273a0, 1987.Raisbeck, G. M., Yiou, F., Jouzel, J., Petit, J. R., Bard, E., and Barkov, N.
I.: 10Be deposition at Vostok, Antarctica, during the last 50 000 years
and its relationship to possible cosmogenic production variations during this
period, in: The timing of the last deglaciation : absolute and radiocarbon
chronologies, edited by: Bard, E. and Broecker, W. S., NATO publication,
127–139, 1992.Raisbeck, G. M., Yiou, F., Jouzel, J., and Stocker, T. F.: Direct north-south
synchronization of abrupt climate change record in ice cores using Beryllium
10, Clim. Past, 3, 541–547, 10.5194/cp-3-541-2007, 2007.Rhodes, R. H., Brook, E. J., Chiang, J. C. H., Blunier, T., Maselli, O. J.,
McConnell, J. R., Romanini, D., and Severinghaus, J. P.: Enhanced tropical
methane production in response to iceberg discharge in the North Atlantic,
Science, 348, 1016–1019, 10.1126/science.1262005, 2015.Ruth, U., Barnola, J.-M., Beer, J., Bigler, M., Blunier, T., Castellano, E.,
Fischer, H., Fundel, F., Huybrechts, P., Kaufmann, P., Kipfstuhl, S.,
Lambrecht, A., Morganti, A., Oerter, H., Parrenin, F., Rybak, O., Severi, M.,
Udisti, R., Wilhelms, F., and Wolff, E.: “EDML1”: a chronology for the
EPICA deep ice core from Dronning Maud Land, Antarctica, over the last
150 000 years, Clim. Past, 3, 475–484, 10.5194/cp-3-475-2007, 2007.Schilt, A., Baumgartner, M., Schwander, J., Buiron, D., Capron, É.,
Chappellaz, J., Loulergue, L., Schüpbach, S., Spahni, R., Fischer, H.,
and Stocker, T. F.: Atmospheric nitrous oxide during the last 140,000 years,
Earth Planet. Sc. Lett., 300, 33–43, 10.1016/j.epsl.2010.09.027, 2010.Schulz, M. and Mudelsee, M.: REDFIT: estimating red-noise spectra directly
from unevenly spaced paleoclimatic time series, Comput. Geosci., 28,
421–426, 10.1016/S0098-3004(01)00044-9, 2002.Schulz, M., Paul, A., and Timmermann, A.: Relaxation oscillators in concert:
A framework for climate change at millennial timescales during the late
Pleistocene, Geophys. Res. Lett., 29, 2193, 10.1029/2002GL016144, 2002.Schüpbach, S., Federer, U., Bigler, M., Fischer, H., and Stocker, T. F.:
A refined TALDICE-1a age scale from 55 to 112 ka before present for the
Talos Dome ice core based on high-resolution methane measurements, Clim.
Past, 7, 1001–1009, 10.5194/cp-7-1001-2011, 2011.Severi, M., Becagli, S., Castellano, E., Morganti, A., Traversi, R., Udisti,
R., Ruth, U., Fischer, H., Huybrechts, P., Wolff, E., Parrenin, F., Kaufmann,
P., Lambert, F., and Steffensen, J. P.: Synchronisation of the EDML and EDC
ice cores for the last 52 kyr by volcanic signature matching, Clim. Past, 3,
367–374, 10.5194/cp-3-367-2007, 2007.Steinhilber, F., Abreu, J. A., Beer, J., Brunner, I., Christl, M., Fischer,
H., Heikkila, U., Kubik, P. W., Mann, M., McCracken, K. G., Miller, H.,
Miyahara, H., Oerter, H., and Wilhelms, F.: 9,400 years of cosmic radiation
and solar activity from ice cores and tree rings, P. Natl. Acad. Sci. USA,
109, 5967–5971, 10.1073/pnas.1118965109, 2012.Stocker, T. F.: Climate change: the seesaw effect, Science, 282, 61–62,
10.1126/science.282.5386.61, 1998.Stocker, T. F. and Johnsen, S. J.: A minimum thermodynamic model for the
bipolar seesaw, Paleoceanography, 18, 1087, 10.1029/2003PA000920, 2003.Svensson, A., Andersen, K. K., Bigler, M., Clausen, H. B., Dahl-Jensen, D.,
Davies, S. M., Johnsen, S. J., Muscheler, R., Parrenin, F., Rasmussen, S. O.,
Röthlisberger, R., Seierstad, I., Steffensen, J. P., and Vinther, B. M.:
A 60 000 year Greenland stratigraphic ice core chronology, Clim. Past, 4,
47–57, 10.5194/cp-4-47-2008, 2008.Svensson, A., Bigler, M., Blunier, T., Clausen, H. B., Dahl-Jensen, D.,
Fischer, H., Fujita, S., Goto-Azuma, K., Johnsen, S. J., Kawamura, K.,
Kipfstuhl, S., Kohno, M., Parrenin, F., Popp, T., Rasmussen, S. O.,
Schwander, J., Seierstad, I., Severi, M., Steffensen, J. P., Udisti, R.,
Uemura, R., Vallelonga, P., Vinther, B. M., Wegner, A., Wilhelms, F., and
Winstrup, M.: Direct linking of Greenland and Antarctic ice cores at the Toba
eruption (74 ka BP), Clim. Past, 9, 749–766, 10.5194/cp-9-749-2013,
2013.Van Ommen, T.: Palaeoclimate: Northern push for the bipolar see-saw, Nature,
520, 630–631, 10.1038/520630a, 2015.Veres, D., Bazin, L., Landais, A., Toyé Mahamadou Kele, H.,
Lemieux-Dudon, B., Parrenin, F., Martinerie, P., Blayo, E., Blunier, T.,
Capron, E., Chappellaz, J., Rasmussen, S. O., Severi, M., Svensson, A.,
Vinther, B., and Wolff, E. W.: The Antarctic ice core chronology (AICC2012):
an optimized multi-parameter and multi-site dating approach for the last 120
thousand years, Clim. Past, 9, 1733–1748, 10.5194/cp-9-1733-2013, 2013.Wagner, G., Beer, J., Masarik, J., Muscheler, R., Kubik, P. W., Mende, W.,
Laj, C., Raisbeck, G. M., and Yiou, F.: Presence of the solar de Vries cycle
(∼ 205 years) during the last ice age, Geophys. Res. Lett., 28,
303–306, 10.1029/2000GL006116, 2001.WAIS Divide Project Members: Precise interpolar phasing of abrupt climate
change during the last ice age, Nature, 520, 661–665,
10.1038/nature14401, 2015.
Yiou, F., Raisbeck, G. M., Baumgartner, S., Beer, J., Hammer, C., Johnsen,
S., Jouzel, J., Kubik, P. W., Lestringuez, J., Stievenard, M., Suter, M., and
Yiou, P.: Beryllium 10 in the Greenland Ice Core Project ice core at Summit,
Greenland, J. Geophys. Res., 102, 26783–26794, 10.1029/97JC01265, 1997.