Invited reviewA review of the bipolar see–saw from synchronized and high resolution ice core water stable isotope records from Greenland and East Antarctica
Introduction
This introductory section summarizes the history of the identification of the bipolar seesaw pattern from Greenland and Antarctic ice cores (Section 1.1), and the ongoing debate on its causes and mechanisms, combining information from other natural archives, conceptual models, and a hierarchy of climate models (Section 1.2). From these open questions, it motivates the need for improved chronological constraints and high resolution, synchronized climate records documenting the spatial structure of changes in Greenland and Antarctica. The last Section 1.3 finally explains the structure of this manuscript.
Abrupt events punctuating climate variability of the Last Glacial period have been identified worldwide in highly resolved terrestrial, marine and ice core records (Voelker, 2002, Clement and Peterson, 2008). Since the 1960s, successive deep Greenland ice core records have provided continuous and extremely highly detailed records of climate variability, now encompassing the whole Last Glacial period, from 116 000 to 11 700 years ago recorded in GRIP (Dansgaard et al., 1993), GISP2 (Grootes et al., 1993), NorthGRIP (NorthGRIP Comm. Members, 2004) and NEEM ice cores (NEEM Comm. Members, 2013). During this time interval, 25 rapid events, called “Dansgaard-Oeschger events” (hereafter DO events), have been identified in numerous measurements performed along these Greenland ice cores (NorthGRIP Community Members, 2004). Greenland abrupt temperature variations are qualitatively recorded at high resolution in water stable isotopes, while their magnitude is estimated using the thermal fractionation of gases inside the firn with an uncertainty of ∼2 °C (Severinghaus and Brook, 1999, Kindler et al., 2014). Lasting a few centuries to a few millennia, DO events are characterized by the succession of a cold phase (Greenland Stadial, GS), an abrupt warming of 5–16 °C in a few years or decades, followed by a warm phase (Greenland Interstadial, GI) marked by a gradual cooling before a relatively abrupt cooling into the next GS, taking place within a few centuries. The widespread extent of DO events is reflected in parallel changes in the atmospheric composition (CH4 concentration, as well as inflexions in the atmospheric δ18O of O2, hereafter δ18Oatm) (Chappellaz et al., 1993, Chappellaz et al., 2013, Landais et al., 2007). The strong abrupt temperature and CH4 increases occur in phase, within 10 years (Severinghaus et al., 1998, Rosen et al., 2014). This abrupt variability in the atmospheric composition, being recorded in the air trapped in Greenland and Antarctic ice cores, has provided a critical tool for the transfer of the accurate Greenland age scales based on annual layer counting towards Antarctic ice core chronologies (Blunier et al., 1998, Schüpbach et al., 2011).
Since the 1970s, East Antarctic ice cores have also depicted millennial climate variability during the Last Glacial period, albeit with limitations in temporal resolution emerging from lower accumulation rates, and less accurate chronologies when annual layer counting is not possible. In Antarctic water stable isotope records, this millennial variability is marked by Antarctic Isotopic Maxima (AIM), initially identified in the central East Antarctic plateau as symmetric gradual isotopic enrichment (warming) and depletion (cooling) trends. Using a first synchronization of the Greenland GRIP and GISP2 ice cores with the Antarctic Vostok ice core through the alignment of δ18Oatm, Bender et al. (1994a) evidenced a recurrent relationship between Greenland and Antarctic water stable isotope millennial events for the nine longest GS. This Greenland and Antarctica pattern was also shown in parallel by Jouzel et al. (1994). A refined synchronization of Greenland (GRIP, GISP2) and Antarctic (Byrd) ice core records was built by Blunier et al. (1998) and Blunier and Brook (2001) based on the alignment of CH4 records over the last 90,000 years. 7 main Antarctic warm events were identified (called A events) as Antarctic counterparts of major Greenland DO events. During each of these 7 events, Antarctic temperatures increased gradually during GS, and the end of Antarctic warming coincided with the onset of rapid warming in Greenland.
Using higher resolution data as well as an improved synchronization, it has been further evidenced that each DO event has an Antarctic Isotopic Maximum counterpart (EPICA Community Members, 2006, Jouzel et al., 2007), except for the first DO event of the Last Glacial period identified in the NorthGRIP ice core, DO25 (Capron et al., 2012). The same bipolar characteristic was also identified at the sub-millennial scale, during GS precursors of DO, or rebound events at the end of GIS, lasting only a few centuries (Capron et al., 2010a), albeit with the restrictions associated with the accuracy of the chronology, a few hundred years at best.
While there is growing evidence for the recurrence of abrupt climate change with similar characteristics during earlier glacial periods from high resolution Antarctic, terrestrial and deep sea records (e.g. McManus et al., 1999, Loulergue et al., 2007, Martrat et al., 2007, Barker et al., 2011, Lambert et al., 2012), we will focus here on the Last Glacial period for which the bipolar structure of events can be accurately characterized from high resolution and well dated records at both poles.
In parallel to ice core records highlighting millennial scale variability during the Last Glacial period, deep-sea sediments from the North Atlantic have revealed the recurrence of iceberg rafted debris in marine cores during GS, associated with iceberg discharges from glacial ice sheets, changes in sea ice extent, surface temperature and salinity, and reorganizations of the thermohaline circulation (e.g. Heinrich, 1988, Bond et al., 1992, Broecker et al., 1992, Grousset et al., 1993, McManus et al., 1998, Labeyrie et al., 1999, Elliot et al., 2002). Six major iceberg discharge episodes were identified as Heinrich events, corresponding to collapses of the Laurentide and/or European ice sheets (see review by Hemming, 2004). A Heinrich stadial was therefore defined as a Greenland cold phase during which a Heinrich event occurred (Barker et al., 2009, Sanchez-Goni and Harrison, 2010). This feature led to the hypothesis that cold phases during Heinrich events (and, implicitly, all GS) were caused by changes in large scale Atlantic ocean circulation, driven by massive inflows of freshwater linked with glacial ice sheet collapses (e.g. Broecker, 1991; Paillard and Labeyrie, 1994; Ganopolski and Rahmstorf, 2001).
During the last decade, glacial abrupt events have been investigated using coupled ocean-atmosphere models of varying complexity (e.g. Stouffer et al., 2006, Kageyama et al., 2013). Several aspects of the observed patterns can be captured through the response of the Earth system to imposed freshwater perturbations in the North Atlantic (Ganopolski and Rahmstorf, 2001, Liu et al., 2009, Kageyama et al., 2010, Roche et al., 2010), mimicking Heinrich events. Depending on the background state of the climate (glacial or interglacial, orbital context …), of the simulated Atlantic Meridional Oceanic Circulation (AMOC), and the magnitude of the freshwater forcing, these models can produce a complete shutdown of the AMOC (Heinrich-like state) or a reduction of the strength of the AMOC (GS-like state) (e.g. Menviel et al., 2014). In all models, the injection of freshwater robustly produces a significant cooling of the North Atlantic region. The amplitude of the associated temperature change is probably affected by the simulated change in sea-ice extent and feedbacks between sea-ice and temperature that vary in the different models (Kageyama et al., 2013). These hosing experiments also produce an inter-hemispheric seesaw temperature pattern and impacts on the position of the intertropical convergence zone, hereafter ITCZ (e.g. Dahl et al., 2005, Broccoli et al., 2006, Krebs and Timmermann, 2007, Swingedouw et al., 2009) through changes in meridional heat transport. In response to freshwater forcing, climate models simulate a decrease of the NADW (North Atlantic Deep Water) export and a possible increase of the AABW (Antarctic Bottom Water) export in the Southern Ocean (Rind et al., 2001). The alternation between NADW and AABW formation is supported by paleoceanographic deep circulation tracers (e.g. review by Adkins, 2013), as well as by changes in 14C of CO2 measurements (Broecker, 1998; Anderson et al., 2009). The different models confirm the robustness of the bipolar seesaw signature of the climate response to AMOC weakening with the South Atlantic systematically warming in response to a freshwater discharge applied in the North Atlantic. There are still regional differences in the simulated Southern Ocean response (Clement and Peterson, 2008, Kageyama et al., 2010, Timmermann et al., 2010). Some models simulate a quasi-uniform warming (e.g. Otto-Bliesner and Brady, 2010) while others show contrasted patterns with a West Pacific cooling associated with the Southern Indian Ocean sector warming.
Conceptual models, paleoceanographic data and climate models of varying complexity all converge to show that DO events are associated with changes in AMOC. However, a number of physical mechanisms allowing quasi–periodic transitions between different modes of operation of the AMOC have been proposed. Among them one must distinguish between those where abrupt millennial climate shifts result from changes in external forcing (e.g freshwater cycle, solar cycle) from those where either internal instabilities of the large-scale ocean circulation or nonlinear sea ice – ocean – ice sheet interactions play a fundamental role. Recent modeling studies suggest that the relatively weak Atlantic northward heat transport that prevails under cold background conditions is the key to the existence of such instabilities (Arzel et al., 2010, Colin de Verdière and te Raa, 2010, Arzel et al., 2012). In those studies, ice-sheet ocean interactions, atmospheric noise or time-varying external forcing are not essential to the emergence of millennial climate shifts. Glaciological studies have stressed that calving due to internal Laurentide ice sheet instabilities can deliver massive meltwater fluxes albeit with large uncertainties on the exact timing, magnitude and rate of delivery (MacAyeal, 1993, Marshall and Clarke, 1997). Such calving events and associated meltwater – induced climate variability can be reproduced in climate models of reduced complexity (Ganopolski, 2003, Ganopolski et al., 2010). Whether the iceberg discharge is a cause or a consequence of changes in AMOC is however an open question. Indeed, a reduced AMOC can also trigger subsurface warming and instabilities of ice streams (Shaffer et al., 2004, Alvarez-Solas and Ramstein, 2011, Alvarez-Solas et al., 2013, Marcott et al., 2011). Changes in atmospheric circulation in relationship with changes in sea ice extent and/or changes in ice sheet topography may also cause abrupt glacial climate shifts (Wunsch, 2006, Li et al., 2010, Zhang et al., 2014).
The initial trigger of instabilities may not lie within the North Atlantic, which could just act as an amplifier (Cane and Clement, 1999). Several authors have explored the possible role of Antarctic freshwater fluxes on AMOC instabilities. For instance, Weaver et al. (2003) suggested an Antarctic origin of meltwater pulse 1A. This 14–18 m global mean sea level rise occurred during the abrupt Bølling-Allerød warming (Deschamps et al., 2012). While there is evidence for West Antarctic ice retreat coeval with MWP1A (Kilfeather et al., 2011, Smith et al., 2011), the magnitude of the Antarctic contribution remains disputed (Clark et al., 2009, Bentley et al., 2010, Mackintosh et al., 2011, Golledge et al., 2014), as glaciological studies indicate possible large contributions from North America (Carlson and Winsor, 2012, Gregoire et al., 2012). Idealized Southern Ocean hosing simulations do not produce large Greenland warming (Seidov et al., 2005, Stouffer et al., 2007, Swingedouw et al., 2008) and are thus suggesting that Antarctica cannot be the driver of DO events. Intrinsic instabilities of the Southern Ocean stratification have also been found in climate models of intermediate complexity (Meissner et al., 2006). These instabilities generate abrupt multi-millennial oscillations whose mechanism is essentially captured by the Welander (1982) two-box model. Corresponding changes in surface air temperature reach a few degrees in the Southern Ocean with little impact in the Northern Hemisphere.
Finally discriminating the respective role of changes in AMOC with respect to changes in sea ice extent and atmospheric circulation and identifying the trigger for the millennial variability calls for very high resolution paleoclimate records, an accurate identification of the north-south timing of changes, and the characterization of regional patterns of changes.
In this manuscript, we focus on the last 60 000 years where our bipolar chronological framework is most accurate, thanks to layer counting for Greenland ice cores and numerous age markers, using the latest available common chronology for Greenland and Antarctic ice core, AICC2012 (Bazin et al., 2013, Veres et al., 2013). The accuracy and limitations of the chronology are specifically addressed in Section 2. The temporal relationships between Antarctic and Greenland temperature over the Last Glacial cycle will be discussed in Section 3 using the AICC2012 chronology. This will include new highly resolved measurements of water stable isotopes from two East Antarctic ice cores (EDC and TALDICE). The comprehensive picture of the see–saw sequence, including regional variability among East Antarctic sites is discussed in Section 4. Finally, Section 5 addresses perspectives to progress in the understanding of mechanisms driving Greenland-Antarctic abrupt climate variability through the use of multiple tracers of climate at lower latitudes, as well as insights expected from earlier glacial periods.
Section snippets
Methods for Greenland and Antarctic age scale synchronization and AICC2012
For a discussion of bipolar seesaw, we concentrate here on the relative uncertainty between Antarctic and Greenland chronologies.
A critical limitation for the description of the sequence of Greenland versus Antarctic climate change is linked to the difficulty of synchronizing different ice cores at high temporal precision. Through time, a collection of absolute and relative ice core dating constraints has been accumulated. For instance, the identification of the Laschamp geomagnetic excursion
Stack Greenland and Antarctic records on AICC2012
In order to extract the common East Antarctic signal, we have combined the water isotopic records for the 4 Antarctic ice cores synchronized on AICC2012 to obtain an East Antarctic isotopic stack. From the available resolution of individual records, the East Antarctic stack has been produced with a 100 year resolution. For Greenland, we have used the available synchronization of the Greenland ice cores (GRIP, GISP2 and NorthGRIP) performed during the construction of the GICC05 timescale (
Discussion
The structure of Antarctic δ18O records depicts a variability which does not follow a simple bipolar seesaw scheme, both in the sharp events depicted in high resolution data from EDC and TALDICE, and in the two phase structure observed during major Antarctic warmings (AIM 8 and AIM 12 warming phases). These patterns can therefore not be explained by a simple seesaw mechanism implying a slow response of Antarctic temperature to abrupt North Atlantic climate shifts, modulated by the thermal
Conclusions and perspectives
Despite uncertainties associated with chronologies of East Antarctic ice cores, the AICC2012 approach provides an accurate framework to investigate the bipolar patterns of glacial climate variability at the millennial time scale, and at the multi-centennial time scale when sufficient stratigraphic links are available, such as for the period close to the Laschamp event.
Common features of Antarctic climate variability are evidenced in a δ18O stack record. Precise synchronization of the different
Acknowledgments
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 United Kingdom. The main logistic support was provided by IPEV and PNRA (at Dome C) and AWI (at Dronning Maud Land). The Talos Dome Ice core Project (TALDICE), a joint European
References (140)
- et al.
Patagonian origin of glacial dust deposited in East Antarctica (Vostok and Dome C) during glacial stages 2, 4 and 6
Earth Planet. Sci. Lett.
(1997) - et al.
Regional imprints of millennial variability during the MIS 3 period around Antarctica
Quat. Sci. Rev.
(2012) - et al.
Synchronising EDML and NorthGRIP ice cores using δ18O of atmospheric oxygen (δ18Oatm ) and CH4 measurements over MIS5 (80-123 kyr)
QSR
(2010) - et al.
Terrigenous plant wax inputs to the Arabian sea: implications for the reconstruction of winds associated with the Indian Monsoon
Geochim. Cosmochim. Acta
(2005) - et al.
Changes in North Atlantic deep-water formation associated with the Dansgaard-Oeschger temperature oscillations (10–60 ka)
Quat. Sci. Rev.
(2002) - et al.
On the age of the Laschamp geomagnetic excursion
Earth Planet. Sci. Lett.
(2004) Origin and consequences of cyclic ice rafting in the Northeast Atlantic Ocean during the past 130000 years
Quat. Res.
(1988)- et al.
On the impact of impurities on the densification of polar firn
Earth Planet. Sci. Lett.
(2012) - et al.
Isotope calibrated Greenland temperature record over Marine Isotope Stage 3 and its relation to CH4
Earth Planet. Sci. Lett.
(2006) - et al.
A reconstruction of sea surface warming in the northern North Atlantic during MIS 3 ice-rafting events
Quat. Sci. Rev.
(2010)
Modelling glacial climatic millennial-scale variability related to changes in the Atlantic meridional overturning circulation: a review
Quat. Sci. Rev.
Firn-air d15N in modern polar sites and glacial-interglacial ice: a model-data mismatch during glacial periods in Antarctica?
Quat. Sci. Rev.
Triple isotopic composition of oxygen in surface snow and water vapor at NEEM (Greenland)
Geochim. Cosmochim. Acta
Consistent dating for Antarctic and Greenland ice cores
Quat. Sci. Rev.
Radiometrically determined sedimentary fluxes in the sub-polar North Atlantic during the last 140,000 years
Earth Planet. Sci. Lett.
The sensitivity of the climate response to the magnitude and location of freshwater forcing: last glacial maximum experiments
Quat. Sci. Rev.
The role of deep ocean circulation in setting glacial climates
Paleoceanography
Abrupt change in atmospheric CO2 during the last ice age
Geophys. Res. Lett.
On the triggering mechanism of Heinrich events
Proc. Natl. Acad. Sci. U. S. A.
Iceberg discharges of the last glacial period driven by oceanic circulation changes
Proc. Natl. Acad. Sci. U. S. A.
Wind-driven upwelling in the Southern Ocean and the deglacial rise in atmospheric CO2
Science
The role of oceanic heat transport and wind-stress forcing in abrupt millennial-scale climate transitions
J. Clim.
Abrupt millennial variability and interdecadal-interstadial oscillations in a global coupled model: sensitivity to the background climate state
Clim. Dyn.
Interhemispheric Atlantic seesaw response during the last deglaciation
Nature
800,000 years of abrupt climate variability
Science
An optimized multi-proxy, multi-site Antarctic ice and gas orbital chronology (AICC2012): 120–800 ka
Clim. Past
Optimisation of glaciological parameters for ice core chronology by implementing counted layers between identified depth levels
Clim. Past Discuss.
Climate connections between Greenland and Antarctica during the last 100,000 years
Nature
The Dole effect and its variations during the last 130,000 years as measured in the Vostok ice core
Global Biogeochem. Cycles
Deglacial history of the West Antarctic Ice Sheet in the Weddell Sea embayment: constraints on past ice volume change
Geology
Timing of millennial-scale climate change in Antarctica and Greenland during the last glacial period
Science
Asynchrony of Antarctic and Greenland climate change during the last glacial period
Nature
Hydrogen isotopes preclude marine hydrate CH4 emissions at the onset of Dansgaard-Oeschger events
Science
Evidence for massive discharges of icebergs into the North Atlantic Ocean during the last glacial period
Nature
The response of the ITCZ to Northern Hemisphere cooling
Geophys. Res. Lett.
The great ocean conveyor
Oceanography
Paleocean circulation during the last deglaciation: a bipolar seesaw?
Paleoceanography
Origin of the northern Atlantic's Heinrich events
Clim. Dyn.
TALDICE-1 age scale of the Talos Dome deep ice core, East Antarctica
Clim. Past
Greenland temperature response to climate forcing during the Last Deglaciation
Science
A Role for the Tropical Pacific Coupled Ocean-atmosphere System on Milankovitch and Millennial Timescales. Part I: a Modeling Study of Tropical Pacific Variability, Mechanism of Global Climate Change at Millennial Time Scales
Millennial-scale climatic variability over the last glacial period: Greenland-Antarctic sequences of events over Marine Isotopic Stage (MIS) 5 compared to MIS 3
Clim. Past
A global picture of the first abrupt climatic event occurring during the last glacial inception
GRL
Northern Hemisphere ice-sheet responses to past climate warming
Nat. Geosci.
Synchronous changes in atmospheric CH4 and Greenland climate between 40 and 8 kyr BP
Nature
High-resolution glacial and deglacial record of atmospheric methane by continuous flow and laser spectrometer analysis along the NEEM ice core
Clim. Past
The Last Glacial Maximum
Science
Mechanisms of abrupt climate change of the last glacial period
Rev. Geophys.
Weak oceanic heat transport as a cause of the instability of glacial climates
Clim. Dyn.
Changes in the atmospheric CH4 gradient between Greenland and Antarctica during the Last Glacial and the transition to the Holocene
Geophys. Res. Lett.
Cited by (64)
Subsurface warming in the tropical Atlantic for the last 3 deglaciations: Insights from organic molecular proxies
2023, Quaternary Science ReviewsMillennial-scale changes in the Asian monsoon during the MIS12 period as recorded by a Chinese stalagmite
2023, Palaeogeography, Palaeoclimatology, PalaeoecologyIn situ measurement of water vapor isotope ratios in air with a laser-based spectrometer
2022, Spectrochimica Acta - Part A: Molecular and Biomolecular SpectroscopyThree-phase structure of the East Asia summer monsoon during Heinrich Stadial 4 recorded in Xianyun Cave, southeastern China
2021, Quaternary Science ReviewsA gradual transition into Greenland interstadial 14 in southeastern China based on a sub-decadally-resolved stalagmite record
2021, Quaternary Science Reviews