Search

The Last Interglacial (LIG) period, which had higher summer solar insolation than today, has been suggested as the last time that Arctic summers were ice free. However, the latest suite of Coupled Modelling Intercomparison Project 6 Paleoclimate (CMIP6-PMIP4) simulations of the LIG produce a wide range of Arctic summer minimum sea ice area (SIA) results, with a 30 % to 96 % reduction from the pre-industrial (PI) period. Sea ice proxies are also currently neither abundant nor consistent enough to determine the most realistic state. Here we estimate LIG minimum SIA indirectly through the use of 21 proxy records for LIG summer surface air temperature (SSAT) and 11 CMIP6-PMIP4 models for the LIG. We use two approaches. First, we use two tests to determine how skilful models are at simulating reconstructed ΔSSAT from proxy records (where Δ refers to LIG-PI). This identifies a positive correlation between model skill and the magnitude of ΔSIA: the most reliable models simulate a larger sea ice reduction. Averaging the two most skilful models yields an average SIA of 1.3×106 km2 for the LIG. This equates to a 4.5×106 km2 or 79 % SIA reduction from the PI to the LIG. Second, across the 11 models, the averaged ΔSSAT at the 21 proxy locations and the pan-Arctic average ΔSSAT are inversely correlated with ΔSIA (r=-0.86 and −0.79, respectively). In other words, the models show that a larger Arctic warming is associated with a greater sea ice reduction. Using the proxy-record-averaged ΔSSAT of 4.5±1.7 K and the relationship between ΔSSAT and ΔSIA suggests an estimated sea ice reduction of 4.2±1.4×106 km2 or about 74 % less sea ice than the PI period. The mean proxy-location ΔSSAT is well correlated with the Arctic-wide ΔSSAT north of 60∘ N (r=0.97), and this relationship is used to show that the mean proxy record ΔSSAT is equivalent to an Arctic-wide warming of 3.7±1.5 K at the LIG compared to the PI period. Applying this Arctic-wide ΔSSAT and its modelled relationship to ΔSIA, results in a similar estimate of LIG sea ice reduction of 4.1±1.2×106 km2. These LIG climatological minimum SIA of 1.3 to 1.5×106 km2 are close to the definition of a summer ice-free Arctic, which is a maximum sea ice extent of less than 1×106 km2. The results of this study thus suggest that the Arctic likely experienced a mixture of ice-free and near-ice-free summers during the LIG.

Southern Ocean bathymetry constrains the path of the Antarctic Circumpolar Current (ACC), but the bathymetric influence on the coupled ocean–atmosphere system is poorly understood. Here, we investigate this impact by respectively flattening large topographic barriers around the Kerguelen Plateau, Campbell Plateau, Mid-Atlantic Ridge, and Drake Passage in four simulations in a coupled climate model. The barriers impact both the wind and buoyancy forcing of the ACC transport, which increases by between 4% and 14% when barriers are removed individually and by 56% when all barriers are removed simultaneously. The removal of Kerguelen Plateau bathymetry increases convection south of the plateau and the removal of Drake Passage bathymetry reduces convection upstream in the Ross Sea. When the barriers are removed, zonal flattening of the currents leads to sea surface temperature (SST) anomalies that strongly correlate to precipitation anomalies, with correlation coefficients ranging between r = 0.92 and r = 0.97 in the four experiments. The SST anomalies correlate to the surface winds too in some locations. However, they also generate circumpolar waves of sea level pressure (SLP) anomalies, which induce remote wind speed changes that are unconnected to the underlying SST field. The meridional variability in the wind stress curl contours over the Mid-Atlantic Ridge, the Kerguelen Plateau, and the Campbell Plateau disappears when these barriers are removed, confirming the impact of bathymetry on surface winds. However, bathymetry-induced wind changes are too small to affect the overall wave-3 asymmetry in the Southern Hemisphere westerlies. Removal of Southern Hemisphere orography is also inconsequential to the wave-3 pattern. Significance Statement Several studies in the past have pointed to the controlling effect of bathymetry on currents in the Southern Ocean circulation, but a clear idea of the importance of the major topographic barriers in the Southern Ocean is lacking. By removing these barriers systematically in a coupled climate model, we can evaluate their impact on several important components of the climate system, such as the Antarctic Circumpolar Current (ACC) pathways and strength, Antarctic Bottom Water formation, sea surface temperature, overlying winds, air–sea fluxes, and even precipitation. This helps in our understanding of what controls the pathways of the ACC and how much it matters for climate.

The lig127k-H11 simulation of the Paleoclimate Modelling Intercomparison Project (PMIP4) is run using the HadGEM3-GC3.1 model. We focus on the coupled system response to the applied meltwater forcing. We show here that the coupling between the atmosphere and the ocean is altered in the hosing experiment compared to a Last Interglacial simulation with no meltwater forcing applied. Two aspects in particular of the atmosphere–ocean coupling are found to be affected: Northern Hemisphere (NH) gyre heat transport and Antarctic sea ice area. We apply 0.2 Sv of meltwater forcing across the North Atlantic during a 250-year-long simulation. We find that the strength of the Atlantic Meridional Overturning Circulation (AMOC) is reduced by 60 % after 150 years of meltwater forcing, with an associated decrease of 0.2 to 0.4 PW in meridional ocean heat transport at all latitudes. The changes in ocean heat transport affect surface temperatures. The largest increase in the meridional surface temperature gradient occurs between 40–50∘ N. This increase is associated with a strengthening of 20 % in 850 hPa winds. The jet stream intensification in the Northern Hemisphere in return alters the temperature structure of the ocean by increasing the gyre circulation at the mid-latitudes and the associated heat transport by +0.1–0.2 PW, and it decreases the gyre circulation at high latitudes with a decrease of ocean heat transport of −0.2 PW. The changes in meridional surface temperature and pressure gradients cause the Intertropical Convergence Zone (ITCZ) to move southward, leading to stronger westerlies and a more positive Southern Annual Mode (SAM) in the Southern Hemisphere (SH). The positive SAM influences sea ice formation, leading to an increase in Antarctic sea ice.

The Hadley Centre Global Environment Model version 3 (HadGEM3) is the first coupled climate model to simulate an ice-free Arctic during the Last Interglacial (LIG), 127 000 years ago. This simulation appears to yield accurate Arctic surface temperatures during the summer season. Here, we investigate the causes and impacts of this extreme simulated ice loss. We find that the summer ice melt was predominantly driven by thermodynamic processes: atmospheric and ocean circulation changes did not significantly contribute to the ice loss. We demonstrate these thermodynamic processes were significantly impacted by melt ponds, which formed on average 8 d earlier during the LIG than during the pre-industrial control (PI) simulation. This relatively small difference significantly changed the LIG surface energy balance and impacted the albedo feedback. Compared to the PI simulation: in mid-June, of the absorbed flux at the surface over ice-covered cells (sea-ice concentration > 0.15), ponds accounted for 45 %–50 %, open water 35 %–45 %, and bare ice and snow 5 %–10 %. We show that the simulated ice loss led to large Arctic sea surface salinity and temperature changes. The sea surface temperature and salinity signals we identify here provide a means to verify, in marine observations, if and when an ice-free Arctic occurred during the LIG. Strong LIG correlations between spring melt pond and summer ice area indicate that, as Arctic ice continues to thin in future, the spring melt pond area will likely become an increasingly reliable predictor of the September sea-ice area. Finally, we note that models with explicitly modelled melt ponds seem to simulate particularly low LIG sea-ice area. These results show that models with explicit (as opposed to parameterised) melt ponds can simulate very different sea-ice behaviour under forcings other than the present day. This is of concern for future projections of sea-ice loss.

Changes in the Greenland ice sheet (GIS) affect global sea level. Greenland stable water isotope (δ18O) records from ice cores offer information on past changes in the surface of the GIS. Here, we use the isotope-enabled Hadley Centre Coupled Model version 3 (HadCM3) climate model to simulate a set of last interglacial (LIG) idealised GIS surface elevation change scenarios focusing on GIS ice core sites. We investigate how δ18O depends on the magnitude and sign of GIS elevation change and evaluate how the response is altered by sea ice changes. We find that modifying GIS elevation induces changes in Northern Hemisphere atmospheric circulation, sea ice and precipitation patterns. These climate feedbacks lead to ice-core-averaged isotopic lapse rates of 0.49 ‰ (100 m)−1 for the lowered GIS states and 0.29 ‰ (100 m)−1 for the enlarged GIS states. This is lower than the spatially derived Greenland lapse rates of 0.62–0.72 ‰ (100 m)−1. These results thus suggest non-linearities in the isotope–elevation relationship and have consequences for the interpretation of past elevation and climate changes across Greenland. In particular, our results suggest that winter sea ice changes may significantly influence isotope–elevation gradients: winter sea ice effect can decrease (increase) modelled core-averaged isotopic lapse rate values by about −19 % (and +28 %) for the lowered (enlarged) GIS states, respectively. The largest influence of sea ice on δ18O changes is found in coastal regions like the Camp Century site.

The Last Interglacial (LIG), a warmer period 130,000–116,000 years before present, is a potential analogue for future climate change. Stronger LIG summertime insolation at high northern latitudes drove Arctic land summer temperatures 4–5 °C higher than in the pre-industrial era. Climate model simulations have previously failed to capture these elevated temperatures, possibly because they were unable to correctly capture LIG sea-ice changes. Here, we show that the latest version of the fully coupled UK Hadley Center climate model (HadGEM3) simulates a more accurate Arctic LIG climate, including elevated temperatures. Improved model physics, including a sophisticated sea-ice melt-pond scheme, result in a complete simulated loss of Arctic sea ice in summer during the LIG, which has yet to be simulated in past generations of models. This ice-free Arctic yields a compelling solution to the long-standing puzzle of what drove LIG Arctic warmth and supports a fast retreat of future Arctic summer sea ice. Arctic climate in the Last Interglacial (LIG)—a warm period 130,000–116,000 years ago—is poorly simulated by modern climate models. A model with improved sea-ice melt-pond physics reproduces LIG Arctic temperatures, suggests an ice-free Arctic during this period and predicts the same by 2035.

When the same weather or climate simulation is run on different high-performance computing (HPC) platforms, model outputs may not be identical for a given initial condition. While the role of HPC platforms in delivering better climate projections is to some extent discussed in the literature, attention is mainly focused on scalability and performance rather than on the impact of machine-dependent processes on the numerical solution. Here we investigate the behaviour of the Preindustrial (PI) simulation prepared by the UK Met Office for the forthcoming CMIP6 (Coupled Model Intercomparison Project Phase 6) under different computing environments. Discrepancies between the means of key climate variables were analysed at different timescales, from decadal to centennial. We found that for the two simulations to be statistically indistinguishable, a 200-year averaging period must be used for the analysis of the results. Thus, constant-forcing climate simulations using the HadGEM3-GC3.1 model are reproducible on different HPC platforms provided that a sufficiently long duration of simulation is used. In regions where El Niño–Southern Oscillation (ENSO) teleconnection patterns were detected, we found large sea surface temperature and sea ice concentration differences on centennial timescales. This indicates that a 100-year constant-forcing climate simulation may not be long enough to adequately capture the internal variability of the HadGEM3-GC3.1 model, despite this being the minimum simulation length recommended by CMIP6 protocols for many MIP (Model Intercomparison Project) experiments. On the basis of our findings, we recommend a minimum simulation length of 200 years whenever possible.

Antarctic sea ice plays a critical role in the Earth system, influencing energy, heat and freshwater fluxes, air–sea gas exchange, ice shelf dynamics, ocean circulation, nutrient cycling, marine productivity and global carbon cycling. However, accurate simulation of recent sea-ice changes remains challenging and, therefore, projecting future sea-ice changes and their influence on the global climate system is uncertain. Reconstructing past changes in sea-ice cover can provide additional insights into climate feedbacks within the Earth system at different timescales. This paper is the first of two review papers from the Cycles of Sea Ice Dynamics in the Earth system (C-SIDE) working group. In this first paper, we review marine- and ice core-based sea-ice proxies and reconstructions of sea-ice changes throughout the last glacial–interglacial cycle. Antarctic sea-ice reconstructions rely mainly on diatom fossil assemblages and highly branched isoprenoid (HBI) alkenes in marine sediments, supported by chemical proxies in Antarctic ice cores. Most reconstructions for the Last Glacial Maximum (LGM) suggest that winter sea ice expanded all around Antarctica and covered almost twice its modern surface extent. In contrast, LGM summer sea ice expanded mainly in the regions off the Weddell and Ross seas. The difference between winter and summer sea ice during the LGM led to a larger seasonal cycle than today. More recent efforts have focused on reconstructing Antarctic sea ice during warm periods, such as the Holocene and the Last Interglacial (LIG), which may serve as an analogue for the future. Notwithstanding regional heterogeneities, existing reconstructions suggest that sea-ice cover increased from the warm mid-Holocene to the colder Late Holocene with pervasive decadal- to millennial-scale variability throughout the Holocene. Studies, supported by proxy modelling experiments, suggest that sea-ice cover was halved during the warmer LIG when global average temperatures were ∼2 ∘C above the pre-industrial (PI). There are limited marine (14) and ice core (4) sea-ice proxy records covering the complete 130 000 year (130 ka) last glacial cycle. The glacial–interglacial pattern of sea-ice advance and retreat appears relatively similar in each basin of the Southern Ocean. Rapid retreat of sea ice occurred during Terminations II and I while the expansion of sea ice during the last glaciation appears more gradual especially in ice core data sets. Marine records suggest that the first prominent expansion occurred during Marine Isotope Stage (MIS) 4 and that sea ice reached maximum extent during MIS 2. We, however, note that additional sea-ice records and transient model simulations are required to better identify the underlying drivers and feedbacks of Antarctic sea-ice changes over the last 130 ka. This understanding is critical to improve future predictions.

Marine Isotope Stage (MIS) 5e (130–116 ka) represents an important ‘process analogue’ for understanding the climatic feedbacks and responses likely active under future anthropogenic warming. Reconstructing the Southern Ocean (SO) palaeoenvironment during MIS 5e and comparing it to the present day provides insights into the different responses of the SO sectors to a warmer climate. This study presents new records from seven marine sediment cores for MIS 5e together with their surface sediment records; all cores are located south of 55 oS. We investigate changes in diatom species assemblage and the accompanying variations in sea surface temperatures, winter sea-ice extent (WSIE) and glacial meltwater flux. All records show warmer conditions and a reduced WSIE during MIS 5e relative to the surface sediments. While the Pacific and Indian Sector records present very stable conditions throughout MIS 5e, the Atlantic Sector records display much more changeable conditions, particularly with respect to the WSIE. These variable conditions are attributed to higher iceberg and glacial meltwater flux in the Weddell Sea. This evidence for increased iceberg and glacial meltwater flux in the Weddell Sea during MIS 5e may have significant implications for understanding the stability of the West Antarctic Ice Sheet, both during MIS 5e and under future warming.

Environmental conditions during Marine Isotope Stage (MIS) 5e (130–116 ka) represent an important “process analogue” for understanding the climatic responses to present and future anthropogenic warming. The response of Antarctic sea ice to global warming is particularly uncertain due to the short length of the observational record. Reconstructing Antarctic winter sea-ice extent during MIS 5e therefore provides insights into the temporal and spatial patterns of sea-ice change under a warmer-than-present climate. This study presents new MIS 5e records from nine marine sediment cores located south of the Antarctic Polar Front between 55 and 70∘ S. Winter sea-ice extent and sea-surface temperatures are reconstructed using marine diatom assemblages and a modern analogue technique transfer function, and changes in these environmental variables between the three Southern Ocean sectors are investigated. The Atlantic and East Indian sector records show much more variable MIS 5e winter sea-ice extent and sea-surface temperatures than the Pacific sector records. High variability in the Atlantic sector winter sea-ice extent is attributed to high glacial meltwater flux in the Weddell Sea, indicated by increased abundances of the diatom species Eucampia antarctica and Fragilariopsis cylindrus. The high variability in the East Indian sector winter sea-ice extent is conversely believed to result from large latitudinal migrations of the flow bands of the Antarctic Circumpolar Current, inferred from latitudinal shifts in the sea-surface temperature isotherms. Overall, these findings suggest that Pacific sector winter sea ice displays a low sensitivity to warmer climates. The different variability and sensitivity of Antarctic winter sea-ice extent in the three Southern Ocean sectors during MIS 5e may have significant implications for the Southern Hemisphere climatic system under future warming.

Improving the understanding of changes in the mean and variability of climate variables as well as their interrelation is crucial for reliable climate change projections. Comparisons between general circulation models and paleoclimate archives using indirect proxies for temperature or precipitation have been used to test and validate the capability of climate models to represent climate changes. The oxygen isotopic ratio δ18O, a proxy for many different climate variables, is routinely measured in speleothem samples at decadal or higher resolution, and single specimens can cover full glacial–interglacial cycles. The calcium carbonate cave deposits are precisely dateable and provide well preserved (semi-)continuous albeit multivariate climate signals in the lower and mid-latitudes, where the measured δ18O in the mineral does not directly represent temperature or precipitation. Therefore, speleothems represent suitable archives to assess climate model abilities to simulate climate variability beyond the timescales covered by meteorological observations (101–102 years). Here, we present three transient isotope-enabled simulations from the Hadley Center Climate Model version 3 (iHadCM3) covering the last millennium (850–1850 CE) and compare them to a large global dataset of speleothem δ18O records from the Speleothem Isotopes Synthesis and AnaLysis (SISAL) database version 2 (Comas-Bru et al., 2020b). We systematically evaluate offsets in mean and variance of simulated δ18O and test for the main climate drivers recorded in δ18O for individual records or regions. The time-mean spatial offsets between the simulated δ18O and the speleothem data are fairly small. However, using robust filters and spectral analysis, we show that the observed archive-based variability of δ18O is lower than simulated by iHadCM3 on decadal and higher on centennial timescales. Most of this difference can likely be attributed to the records' lower temporal resolution and averaging or smoothing processes affecting the δ18O signal, e.g., through soil water residence times. Using cross-correlation analyses at site level and modeled grid-box level, we find evidence for highly variable but generally low signal-to-noise ratios in the proxy data. This points to a high influence of cave-internal processes and regional climate particularities and could suggest low regional representativity of individual sites. Long-range strong positive correlations dominate the speleothem correlation network but are much weaker in the simulation. One reason for this could lie in a lack of long-term internal climate variability in these model simulations, which could be tested by repeating similar comparisons with other isotope-enabled climate models and paleoclimate databases.

The 8.2 ka event is a well-known cooling event in the Northern Hemisphere, but is poorly understood in Madagascar. Here, we compare paleoclimate data and outputs from paleoclimate simulations to better understand it. Records from Madagascar suggest two distinct sub-events (8.3 ka and 8.2 ka), that seem to correlate with records from northern high latitude. This could indicate causal relationships via changes in the Atlantic Meridional Overturning Circulation (AMOC) with changes in moisture source's δ18O, and changes in the mean position of the Inter-Tropical Convergence Zone (ITCZ), as climate modelling suggests. These two sub-events are also apparent in other terrestrial records, but the climatic signals are different. The prominent 8.2 ka sub-event records a clear antiphase relationship between the northern and southern hemisphere monsoons, whereas such relationship is less evident during the first 8.3 ka sub-event. Data–model comparison have also shown a mismatch between the paleoclimate data and the model outputs, the causes of which are more or less understood and may lie in the proxies, in the model, or in both data and model. Knowing that paleoclimate proxies and climate models produce different sets of variables, further research is needed to improve the data–model comparison approach, so that both paleoclimate data and paleoclimate models will better predict the likely climate status of a region during a specified time in the past with minimal uncertainties.

Ice core records from Greenland provide evidence for multiple abrupt warming events recurring at millennial time scales during the last glacial interval. Although climate transitions strongly resembling these Dansgaard-Oeschger (DO) transitions have been identified in several speleothem records, our understanding of the climate and ecosystem impacts of the Greenland warming events in lower latitudes remains incomplete.Here, we investigate the influence of DO transitions on the global atmospheric circulation pattern. We comprehensively analyse d18O changes during DO transitions in a globally distributed dataset of speleothems (SISALv2; Comas-Bru et al., 2020). Speleothem d18O signals mostly reflect changes in precipitation amount and moisture source. Thereby this proxy allows us to infer spatially resolved changes in global atmospheric dynamics that are characteristically linked to DO transitions. We confirm the previously proposed shift of the Intertropical Convergence Zone towards more northerly positions. In addition, we find evidence for a similar northward shift of the westerly winds of the Northern Hemisphere. Furthermore, we identify a decreasing trend in the transition amplitudes with increasing distances from the North Atlantic region. This confirms previous suggestions of this region being the core and origin of these past abrupt climate changes. References:Comas-Bru et al., 2020, Earth System Science Data 12, 2579–2606

The modeling of paleoclimate, using physically based tools, is increasingly seen as a strong out-of-sample test of the models that are used for the projection of future climate changes. New to the Coupled Model Intercomparison Project (CMIP6) is the Tier 1 Last Interglacial experiment for 127 000 years ago (lig127k), designed to address the climate responses to stronger orbital forcing than the midHolocene experiment, using the same state-of-the-art models as for the future and following a common experimental protocol. Here we present a first analysis of a multi-model ensemble of 17 climate models, all of which have completed the CMIP6 DECK (Diagnostic, Evaluation and Characterization of Klima) experiments. The equilibrium climate sensitivity (ECS) of these models varies from 1.8 to 5.6 ∘C. The seasonal character of the insolation anomalies results in strong summer warming over the Northern Hemisphere continents in the lig127k ensemble as compared to the CMIP6 piControl and much-reduced minimum sea ice in the Arctic. The multi-model results indicate enhanced summer monsoonal precipitation in the Northern Hemisphere and reductions in the Southern Hemisphere. These responses are greater in the lig127k than the CMIP6 midHolocene simulations as expected from the larger insolation anomalies at 127 than 6 ka. New synthesis for surface temperature and precipitation, targeted for 127 ka, have been developed for comparison to the multi-model ensemble. The lig127k model ensemble and data reconstructions are in good agreement for summer temperature anomalies over Canada, Scandinavia, and the North Atlantic and for precipitation over the Northern Hemisphere continents. The model–data comparisons and mismatches point to further study of the sensitivity of the simulations to uncertainties in the boundary conditions and of the uncertainties and sparse coverage in current proxy reconstructions. The CMIP6–Paleoclimate Modeling Intercomparison Project (PMIP4) lig127k simulations, in combination with the proxy record, improve our confidence in future projections of monsoons, surface temperature, and Arctic sea ice, thus providing a key target for model evaluation and optimization.

The modeling of paleoclimate, using physically based tools, is increasingly seen as a strong out-of-sample test of the models that are used for the projection of future climate changes. New to CMIP6 is the Tier 1 lig127k experiment, designed to address the climate responses to stronger orbital forcing than the midHolocene experiment, using the same state-of-the-art models and following a common experimental protocol. We present a multi-model ensemble of 17 climate models, all of which (except for two) have also completed the CMIP6 DECK experiments. The Equilibrium Climate Sensitivity (ECS) of these models varies from 2.1 to 5.6 °C. The seasonal character of the insolation anomalies results in strong warming over the Northern Hemisphere (NH) continents in the lig127k ensemble as compared to the piControl in June–July–August and a much-reduced minimum (August–September) summer sea ice extent in the Arctic. The multi-model results indicate enhanced summer monsoonal precipitation and areal extent in the Northern Hemisphere and reductions in the Southern Hemisphere. These responses are greater in the lig127k than midHolocene simulations as expected from the larger insolation anomalies at 127 ka than 6 ka. New syntheses for surface temperature and precipitation, targeted for 127 ka, have been developed for comparison to the multi-model ensemble. The lig127k model ensemble and data reconstructions are in good agreement for summer temperature anomalies over Canada, Scandinavia, and the North Atlantic and precipitation over the Northern Hemisphere continents. The model-data comparisons and mismatches point to further study of the sensitivity of the simulations to uncertainties in the specified boundary conditions and of the uncertainties and sparse coverage in current proxy reconstructions. The CMIP6-PMIP4 lig127k simulations, in combination with the proxy record, have potential implications for confidence in future projections of monsoons, surface temperature, Arctic sea ice, and the stability of the Greenland ice sheet.

The shape and extent of the Greenland Ice Sheet (GIS) during the Last Interglacial (LIG) is a matter of controversy, with different studies proposing a wide range of reconstructions. Here, for the first time, we combine stable water isotopic information from ice cores with isotope-enabled climate model outputs to investigate the problem. Exploring the space of possible ice sheet geometries by simulation is prohibitively expensive. We address this problem by using a Gaussian process emulator as a statistical surrogate of the full climate model. The emulator is calibrated using the results of a small number of carefully chosen simulations and then permits fast, probabilistic predictions of the simulator outputs at untried inputs. The inputs are GIS morphologies, parameterized through a dimension-reduction technique adapted to the spherical geometry of the setting. Based on the emulator predictions, the characteristics of morphologies compatible with the available ice core measurements are explored, leading to a reduction in uncertainty on the LIG GIS morphology. Moreover, a scenario-based approach allows to assess the gains in uncertainty reduction which would result from the availability of better dated LIG measurements in Greenland ice cores.

The Last Interglacial (LIG) is a period of great importance as an analog for future climate change. Global sea level was 6-9 m higher than present. Stronger LIG summertime insolation at high northern latitudes drove Arctic land summer temperatures around 4-5 K higher than during the preindustrial era. Climate-model simulations have previously failed to capture these elevated temperatures. This may be because these models failed to correctly capture LIG sea ice changes.Here, we show that the latest version of the UK Hadley Center coupled ocean-atmosphere climate model (HadGEM3) simulates a much improved Arctic LIG climate, including the observed high temperatures. Improved model physics in HadGEM3, including a sophisticated sea ice melt-pond scheme, results in the first-ever simulation of the complete loss of Arctic sea ice in summer during the LIG.Our ice-free Arctic yields a compelling solution to the long-standing puzzle of what drove LIG Arctic warmth. The LIG simulation result is a new independent constraint on the strength of Arctic sea ice decline in climate-model projections, and provides support for a fast retreat of Arctic summer sea ice in the future.

The Last interglacial (LIG) is a period with increased summer insolation at high northern latitudes, which results in strong changes in the terrestrial and marine cryosphere. Understanding the mechanisms for this response via climate modelling and comparing the models’ representation of climate reconstructions is one of the objectives set up by the Paleoclimate Mod-elling Intercomparison Project for its contribution to the sixth phase of the Coupled Model Intercomparison Project. Here we analyse the results from 12 climate models in terms of Arctic sea ice. The mean pre-industrial to LIG reduction in minimum sea ice area (SIA) reaches 59 % (multi-model mean LIG area is 2.21 mill. km2, compared to 5.85 mill. km2 for the PI), and the range of model results for LIG minimum sea ice area (from 0.02 to 5.65 mill. km2) is larger than for PI (from 4.10 to 8.30 mill. km2). On the other hand there is little change for the maximum sea ice area (which is 12 mill. km2 for both the PI and the LIG, with a standard deviation of 1.04 mill. km2 for PI and 1.21 mill. km2 for LIG). To evaluate the model results we synthesize LIG sea ice data from marine cores collected in the Arctic Ocean, Nordic Seas and northern North Atlantic. South of 78° N in the Atlantic and Nordic seas the LIG was seasonally ice-free. North of 78° N there are some discrepancies between sea-ice re-constructions based on dinocysts/foraminifers/ostracods and IP25: some sites have both seasonal and perennial interpretations based on the same core, but different indicators. Because of the conflicting interpretations it is not possible for any one model to match every data point in our data synthesis, or say whether the Arctic was seasonally ice-free. Drivers for the inter-model differences are: different phasing of the up and down short-wave anomalies over the Arctic ocean, associated with differences in model albedo; possible cloud property differences, in terms of optical depth; LIG ocean circulation changes which occur for some, but not all, LIG simulations. Finally we note that inter-comparisons between the LIG simulations, and simulations with moderate CO2 increase (during the transition to high CO2 levels), may yield insight into likely 21C Arctic sea ice changes using these LIG simulations.

Palaeoclimate model simulations are an important tool to improve our understanding of the mechanisms of climate change. These simulations also provide tests of the ability of models to simulate climates very different to today. Here we present the results from two simulations using the latest version of the UK’s physical climate model, HadGEM3-GC3.1; the mid-Holocene (~ 6 ka) and Last Interglacial (~ 127 ka) simulations, both conducted under the auspices of CMIP6/PMIP4. These periods are of particular interest to PMIP4 because they represent the two most recent warm periods in Earth history, where atmospheric concentration of greenhouse gases and continental configuration is similar to the pre-industrial period but where there were significant changes to the Earth’s orbital configuration, resulting in a very different seasonal cycle of radiative forcing. Results for these simulations are assessed against proxy data, previous versions of the UK model, and models from the previous CMIP5 exercise. When the current version is compared to the previous generation of the UK model, the most recent version suggests limited improvement. In common with these previous model versions, the simulations reproduce global land and ocean temperatures (both surface and at 1.5 m) and a West African monsoon that is consistent with the latitudinal and seasonal distribution of insolation. The Last Interglacial simulation appears to accurately capture Northern Hemisphere temperature changes, but without the addition of Last Interglacial meltwater forcing cannot capture the magnitude of Southern Hemisphere changes. Model-data comparisons indicate that some geographical regions, and some seasons, produce better matches to the palaeodata (relative to pre-industrial) than others. Model-model comparisons, relative to previous generations same model and other models, indicate similarity between generations in terms of both the intensity and northward enhancement of the mid-Holocene West African monsoon, both of which are underestimated. On the Saharan greening which occurred the mid-Holocene African Humid Period, simulation results are likewise consistent with other models. The most recent version of the UK model appears to still be unable to reproduce the amount of rainfall necessary to support grassland across the Sahara.

Millennial scale climate variations called Dansgaard-Oeschger cycles occurred frequently during the last glacial, with their central impact on climate in the North Atlantic region. These events are, for example, well captured by the stable oxygen isotope composition in continental ice from Greenland, but also in records from other regions. Recently, it has been shown that a water isotope enabled general circulation model is able to reproduce those millennial-scale oxygen isotope changes from Greenland (Sime et al., 2019). On a global scale, this isotope-enabled model has not been tested in its performance, as stable oxygen isotope records covering those millennial scale variability were so far missing or not systematically compiled.In the continental realm, speleothems provide an excellent archive to store the oxygen isotope composition in precipitation during those rapid events. Here, we use a newly established speleothem data base (SISAL, Atsawawaranunt et al., 2018) from which we extracted 126 speleothems, growing in some interval during the last glacial period. We established an automated method for identification of the rapid onsets of interstadials. While the applied method seems to be not sensitive enough to capture all warming events due to the diverse characteristics of speleothem data (temporal resolution, growth stops and dating uncertainties) and low signal-to-noise-ratio, we are confident that our method is not detecting variations in stable oxygen isotopes that do not reflect stadial-interstadial transitions. Finally, all found transitions were stacked for individual speleothem records in order to provide a mean stadial-interstadial transition for various continental locations. This data set could be useful for future comparison of isotope enabled model simulations and corresponding observations, and to test their ability in modelling millennial scale variability. ReferencesAtsawawaranunt, et al. (2018). The SISAL database: A global resource to document oxygen and carbon isotope records from speleothems. Earth System Science Data 10, 1687–1713Sime, L. C., Hopcroft, P. O., Rhodes, R. H. (2019). Impact of abrupt sea ice loss on Greenland water isotopes during the last glacial period. PNAS 116, 4099-4104.

The peak of the Last Interglacial, Marine Isotope Stage (MIS) 5e (130–116 ka), provides a valuable ‘process analogue’ for validating the climatic feedbacks and forcings likely active under future anthropogenic warming. Reconstructing exact timings of MIS 5e peak warming and minimum winter sea-ice extent (WSIE) throughout the Southern Ocean (SO) will help to identify the interactions and feedbacks within the ice-ocean system. Here we present a new MIS 5e marine sediment record from the SW Atlantic sector together with 28 published core records (chronologies standardised to the LR04 δ18O benthic stack; Lisiecki and Raymo, 2005) to investigate the timing and sequence of minimum WSIE and peak warming across the SO. Sea-surface temperatures (SSTs) peaked earliest in the Indian (20oE–150oE) and Atlantic (70oW–20oE) sectors, at 128.7 ± 0.8 ka and 127.4 ± 1.1 ka respectively, followed by the Pacific sector (150oE–70oW) at 124.9 ± 3.6 ka. The interval of minimum WSIE for all three sectors occurred within the period from 129–125 ka, consistent with the ∼128 ka sea salt flux minimum in Antarctic ice cores. Minimum WSIE appears to have coincided with peak July insolation at 55 oS, suggesting it could be linked with the mildest winters. The reduced WSIE during MIS 5e would have likely reduced the production of deep- and bottom water masses, inhibiting storage of CO2 in the abyssal ocean and lowering nutrient availability in SO surface waters. Examining a wide spatial range of proxy records for MIS 5e is a critical step forward in understanding climatic interactions and processes that will be active under warmer global temperatures.

Improving our knowledge of the temporal and spatial variability of the Antarctic Ice Sheet (AIS) surface mass balance (SMB) is crucial to reduce the uncertainties of past, present, and future Antarctic contributions to sea level rise. An examination of the surface air temperature–SMB relationship in model simulations demonstrates a strong link between the two. Reconstructions based on ice cores display a weaker relationship, indicating a model–data discrepancy that may be due to model biases or to the non-climatic noise present in the records. We find that, on the regional scale, the modeled relationship between surface air temperature and SMB is often stronger than between temperature and δ18O. This suggests that SMB data can be used to reconstruct past surface air temperature. Using this finding, we assimilate isotope-enabled SMB and δ18O model output with ice core observations to generate a new surface air temperature reconstruction. Although an independent evaluation of the skill is difficult because of the short observational time series, this new reconstruction outperforms the previous reconstructions for the continental-mean temperature that were based on δ18O alone. The improvement is most significant for the East Antarctic region, where the uncertainties are particularly large. Finally, using the same data assimilation method as for the surface air temperature reconstruction, we provide a spatial SMB reconstruction for the AIS over the last 2 centuries, showing large variability in SMB trends at a regional scale, with an increase (0.82 Gt yr−2) in West Antarctica over 1957–2000 and a decrease in East Antarctica during the same period (−0.13 Gt yr−2). As expected, this is consistent with the recent reconstruction used as a constraint in the data assimilation.

El Niño–Southern Oscillation (ENSO) is the strongest mode of interannual climate variability in the current climate, influencing ecosystems, agriculture, and weather systems across the globe, but future projections of ENSO frequency and amplitude remain highly uncertain. A comparison of changes in ENSO in a range of past and future climate simulations can provide insights into the sensitivity of ENSO to changes in the mean state, including changes in the seasonality of incoming solar radiation, global average temperatures, and spatial patterns of sea surface temperatures. As a comprehensive set of coupled model simulations is now available for both palaeoclimate time slices (the Last Glacial Maximum, mid-Holocene, and last interglacial) and idealised future warming scenarios (1 % per year CO2 increase, abrupt four-time CO2 increase), this allows a detailed evaluation of ENSO changes in this wide range of climates. Such a comparison can assist in constraining uncertainty in future projections, providing insights into model agreement and the sensitivity of ENSO to a range of factors. The majority of models simulate a consistent weakening of ENSO activity in the last interglacial and mid-Holocene experiments, and there is an ensemble mean reduction of variability in the western equatorial Pacific in the Last Glacial Maximum experiments. Changes in global temperature produce a weaker precipitation response to ENSO in the cold Last Glacial Maximum experiments and an enhanced precipitation response to ENSO in the warm increased CO2 experiments. No consistent relationship between changes in ENSO amplitude and annual cycle was identified across experiments.

Improving our knowledge of the temporal and spatial variability of the Antarctic Ice Sheet (AIS) Surface Mass Balance (SMB) is crucial to reduce the uncertainties of past, present and future Antarctic contributions to sea level rise. Here, we show that Global Climate Models (GCMs) can reproduce the present-day (1979–2005) AIS SMB and the temporal variations over the last two centuries. An examination of the surface temperature–SMB relationship in model simulations demonstrates a strong link between the two. Reconstructions based on ice cores display a weaker relationship, indicating a model-data discrepancy that may be due to model biases or to the non-climatic noise present in the records. We find that, on the regional scale, the modelled temperature-SMB relationship is stronger than the relationship between δ18O-temperature. This suggests that SMB data can be used to reconstruct past surface temperatures. Using this finding, we assimilate isotope-enabled model SMB and δ18O output with ice-core observations, to generate a new surface temperature reconstruction. Although an independent evaluation of the skill is difficult because of the short observational time series, this new reconstruction outperforms the previous reconstructions for the continental-mean temperature that were based on δ18O alone with a linear correlation coefficient with the observed surface temperatures (1958–2010 CE) of 0.73. The improvement is largest for the East Antarctic region, where the uncertainties are particularly large. Finally, we provide a spatial SMB reconstruction of the AIS over the last two centuries showing 1) large variability in SMB trends at regional scale; and 2) a large SMB increase (0.82 Gt year−2) in West Antarctica over 1957–2000 while at the same time, East Antarctica has experienced a large SMB decrease (−3.3 Gt year−2), which is consistent with a recent reconstruction.

When the same weather or climate simulation is run on different High Performance Computing (HPC) platforms, model outputs may not be identical for a given initial condition. While the role of HPC platforms in delivering better climate projections is often discussed in literature, attention is mainly focused on scalability and performance rather than on the impact of machine-dependent processes on the numerical solution. At the same time, machine dependence is an overlooked source of uncertainty when it comes to discussing the model spread observed within the Coupled Model Intercomparison Projects (CMIP). Here we investigate the impact of machine dependence on model results and quantify, for a selected case study, the magnitude of the uncertainty. We consider the Preindustrial (PI) simulation prepared by the UK Met Office for the forthcoming CMIP6. We compare key climate variables between PI control simulations run on the UK Met Office supercomputer and the ARCHER HPC platform. Discrepancies strongly depend on the timescale. Decadal means show substantial differences of up to 0.2 °C for global mean air temperature, 1 W/m2 for TOA outgoing longwave flux and 1.2 million km2 for Southern Hemisphere sea ice area. However, on multi-centennial timescales the differences are not significant and the long-term statistics of the two runs are similar. Differences between the two simulations can be linked to variations in the strongest modes of climate variability. In the Southern Hemisphere, this results in large SST anomalies where ENSO teleconnection patterns are expected that can reach 0.6 °C (and SNR > 1) even on centennial timescales.

Significance The Dansgaard–Oeschger events contained in Greenland ice cores constitute the archetypal record of abrupt climate change. An accurate understanding of these events hinges on interpretation of Greenland records of oxygen and nitrogen isotopes. We present here the important results from a suite of modeled Dansgaard–Oeschger events. These simulations show that the change in oxygen isotope per degree of warming becomes smaller during larger events. Abrupt reductions in sea ice also emerge as a strong control on ice core oxygen isotopes because of the influence on both the moisture source and the regional temperature increase. This work confirms the significance of sea ice for past abrupt warming events., Greenland ice cores provide excellent evidence of past abrupt climate changes. However, there is no universally accepted theory of how and why these Dansgaard–Oeschger (DO) events occur. Several mechanisms have been proposed to explain DO events, including sea ice, ice shelf buildup, ice sheets, atmospheric circulation, and meltwater changes. DO event temperature reconstructions depend on the stable water isotope (δ18O) and nitrogen isotope measurements from Greenland ice cores: interpretation of these measurements holds the key to understanding the nature of DO events. Here, we demonstrate the primary importance of sea ice as a control on Greenland ice core δ18O: 95% of the variability in δ18O in southern Greenland is explained by DO event sea ice changes. Our suite of DO events, simulated using a general circulation model, accurately captures the amplitude of δ18O enrichment during the abrupt DO event onsets. Simulated geographical variability is broadly consistent with available ice core evidence. We find an hitherto unknown sensitivity of the δ18O paleothermometer to the magnitude of DO event temperature increase: the change in δ18O per Kelvin temperature increase reduces with DO event amplitude. We show that this effect is controlled by precipitation seasonality.

Creating more accurate reconstructions of past Antarctic ice sheet shapes allows us to better predict how they will vary in the changing climate and contribute to future sea level changes. In this research, we use expert elicitation to create a subjective prior distribution of the Antarctic ice sheets at the Last Glacial Maximum (LGM), 21Ka. A design of shapes from this distribution will be run through the global climate model HadCM3, providing us with output that we can compare with proxy data to find a better estimate of the ice sheet shape at the LGM.

Last Interglacial (LIG), stable water isotope values (δ18O) measured in Greenland deep ice cores are at least 2.5‰ higher compared to the present day. Previous isotopic climate simulations of the LIG do not capture the observed Greenland δ18O increases. Here, we use the isotope-enabled HadCM3 (UK Met Office coupled atmosphere-ocean general circulation model) to investigate whether a retreat of Northern Hemisphere sea ice was responsible for this model-data disagreement. Our results highlight the potential significance of sea ice changes on the LIG Greenland isotopic maximum. Sea ice loss in combination with increased sea surface temperatures, over the Arctic, affect δ18O: water vapour enriched in heavy isotopes and a shorter distillation path may both increase δ18O values over Greenland. We show, for the first time, that simulations of the response to Arctic sea ice reduction are capable of producing the likely magnitude of LIG δ18O increases at NEEM, NGRIP, GIPS2 and Camp Century ice core sites. However, we may underestimate δ18O changes at the Renland, DYE3 and GRIP ice core locations. Accounting for possible ice sheet changes is likely to be required to produce a better fit to the LIG ice core δ18O values.

The Southern Hemisphere westerly winds (SHW) play an important role in regulating the capacity of the Southern Ocean carbon sink. They modulate upwelling of carbon-rich deep water and, with sea ice, determine the ocean surface area available for air–sea gas exchange. Some models indicate that the current strengthening and poleward shift of these winds will weaken the carbon sink. If correct, centennial- to millennial-scale reconstructions of the SHW intensity should be linked with past changes in atmospheric CO2, temperature and sea ice. Here we present a 12,300-year reconstruction of wind strength based on three independent proxies that track inputs of sea-salt aerosols and minerogenic particles accumulating in lake sediments on sub-Antarctic Macquarie Island. Between about 12.1 thousand years ago (ka) and 11.2 ka, and since about 7 ka, the wind intensities were above their long-term mean and corresponded with increasing atmospheric CO2. Conversely, from about 11.2 to 7.2 ka, the wind intensities were below their long-term mean and corresponded with decreasing atmospheric CO2. These observations are consistent with model inferences of enhanced SHW contributing to the long-term outgassing of CO2 from the Southern Ocean. The strength of the Southern Hemisphere westerly winds varied throughout the Holocene, according to a reconstruction from lake sediments, with periods of stronger winds coincident with higher atmospheric CO2 levels.

We compare multi-ice core data with δ18O model output for the early last interglacial Antarctic sea ice minimum. The spatial pattern of δ18O across Antarctica is sensitive to the spatial pattern of sea ice retreat. Local sea ice retreat increases the proportion of winter precipitation, depleting δ18O at ice core sites. However, retreat also enriches δ18O because of the reduced source-to-site distance for atmospheric vapor. The joint overall effect is for δ18O to increase as sea ice is reduced. Our data-model comparison indicates a winter sea ice retreat of 67, 59, and 43% relative to preindustrial in the Atlantic, Indian, and Pacific sectors of the Southern Ocean. A compilation of Southern Ocean sea ice proxy data provides weak support for this reconstruction. However, most published marine core sites are located too far north of the 128,000 years B.P. sea ice edge, preventing independent corroboration for this sea ice reconstruction. Plain Language Summary The Antarctic isotope and temperature maximum, which occurred approximately 128,000 years before present (B.P.) during the warmer than present last interglacial period, is associated with a major retreat of Antarctic sea ice. Understanding the details of this major sea ice retreat is crucial in order to understand the sensitivity of the Southern Hemisphere sea ice system and to evaluate the performance of climate model simulations in response to future warming. This work uses a multi-ice and ocean core data-model evaluation to assess the magnitude and spatial pattern of this sea ice retreat. Our results suggest that sea ice retreat was greatest in the Atlantic and Indian sectors of the Southern Ocean and less in the Pacific sector. These results may have had serious implications for the stability of marine terminating glaciers around the Antarctic Ice Sheet and their contribution to the last interglacial sea level rise. These results also support a hypothesized slowdown in northward ocean heat transport during the early last interglacial.

Warmer than present Antarctic and Southern Ocean temperatures during the last interglacial, approximately 128,000 years ago, have been attributed to changes in north-south ocean heat transport, causing opposing hemispheric temperature anomalies. We investigate the magnitude of Antarctic warming and Antarctic ice core isotopic enrichment in response to Northern Hemisphere meltwater input during the early last interglacial. A 1,600 year HadCM3 simulation driven by 0.25 Sv of meltwater input reproduces 50-60 % of the peak Southern Ocean summer sea surface temperature anomaly, sea ice retreat and ice core isotope enrichment. We also find a robust increase in the proportion of cold season precipitation during the last interglacial, leading to lower isotopic values at the Antarctic ice core sites. These results suggest that a HadCM3 simulation including 0.25 Sv for 3,000-4,000 years would reconcile the last interglacial observations, providing a potential solution for the last interglacial 'missing heat' problem.

The evaluation of ice sheet models is one of the pressing problems in the study of ice sheet dynamics. Here we examine the question of how much isochronous information is contained within the publicly available Center for Remote Sensing of Ice Sheets (CReSIS) Greenland airborne radio echo soundings data set. We identify regions containing isochronous reflectors using automatic radio echo sounding processing (ARESP) algorithms. We find that isochronous reflectors are present within 36% of the CReSIS radio echo sounding englacial data by location and 41% by total number of data. Between 1000 and 3000 m in depth, isochronous reflectors are present along more than 50% of the data set flight path. Lower volumes of cold glacial period ice also correspond with more isochronous reflectors. We find good agreement between ARESP and continuity index results, providing confidence in these findings. Ice structure data sets, based on data identified here, will be of use in evaluating ice sheet simulations and the assessment of past rates of snow accumulation.

Changes in the strength and position of Southern Hemisphere westerly winds during the Last Glacial cycle have been invoked to explain both millennial and glacial–interglacial climate fluctuations. However, neither paleo models nor paleodata agree on the magnitude, or even the sign, of the change in wind strength and latitude during the most studied glacial period, the Last Glacial Maximum (LGM), compared to the recent past. This paper synthesizes paleo-environmental data that have been used to infer changes in LGM winds. Data compilations are provided for changes in terrestrial moisture, dust deposition, sea surface temperatures and ocean fronts, and ocean productivity, and existing data on Southern Hemisphere ocean circulation changes during the LGM are summarized. We find that any hypothesis of LGM wind and climate change needs to provide a plausible explanation for increased moisture on the west coast of continents, cooler temperatures and higher productivity in the Subantarctic Zone, and reductions in Agulhas leakage around southern Africa. Our comparison suggests that an overall strengthening, an equatorward displacement, or no change at all in winds could all be interpreted as consistent with observations. If a single cause related to the southern westerlies is sought for all the evidence presented, then an equatorward displacement or strengthening of the winds would be consistent with the largest proportion of the observations. However, other processes, such as weakening or poleward shifts in winds, a weakened hydrological cycle, extended sea-ice cover, and changed buoyancy fluxes, cannot be ruled out as potential explanations of observed changes in moisture, surface temperature, and productivity. We contend that resolving the position and strength of westerly winds during the LGM remains elusive based on data reconstructions alone. However, we believe that these data reconstructions of environmental conditions can be used in conjunction with model simulations to identify which processes best represent westerly wind conditions during the LGM.

[1] Oxygen isotope palaeoclimate records, preserved in moss tissue cellulose, are complicated by environmental influences on the relationships between source water inputs and evaporative conditions. We carried out stable isotope analyses of precipitation collected from the maritime Antarctic and cellulose extracted from co-located Chorisodontium aciphyllum dominated moss peat bank deposits accumulated since 1870 A.D. Analyses of stable oxygen and hydrogen isotope composition of summer precipitation on Signy Island (60.7°S, 45.6°W) established a local meteoric water line (LMWL) similar to both the global MWL and other LMWLs, and almost identical to the HadAM3 isotope-enabled global circulation model output. The oxygen isotopic composition of cellulose (δ18OC) revealed little temporal variation between four moss peat banks on Signy Island since 1870. However, δ18OC followed two patterns with Sites A and D consistently 3‰ enriched relative to δ18OC values from Sites B and C. The growing moss surfaces at Sites A and D are likely to have been hydrated by isotopically heavier summer precipitation, whilst at Sites B and C, the moss banks are regularly saturated by the isotopically depleted snow melt streams. Laboratory experiments revealed that evaporative enrichment of C. aciphyllum moss leaf water by 5‰ occurred rapidly following saturation (ecologically equivalent to post-rainfall or snow melt periods). In addition to the recognized source water-cellulose fractionation extent of 27 ± 3‰, such a shift would account for the 32‰ difference measured between δ18O of Signy Island precipitation and cellulose.

Latitudinal shifts in the Southern Ocean westerly wind jet could drive changes in the glacial to interglacial ocean CO2 inventory. However, whilst CMIP5 model results feature consistent future-warming jet shifts, there is considerable disagreement in deglacial-warming jet shifts. We find here that the dependence of pre-industrial (PI) to Last Glacial Maximum (LGM) jet shifts on PI jet position, or state dependency, explains less of the shifts in jet simulated by the models for the LGM compared with future-warming scenarios. State dependence is also weaker for intensity changes, compared to latitudinal shifts in the jet. Winter sea ice was considerably more extensive during the LGM, along with state dependence, changes in surface heat fluxes due this sea ice change probably had a large impact on the jet. Models which both simulate realistically large expansions in sea ice and feature PI jets which are south of 50 °S, show an increase in wind speed around 55 °S, and can show a poleward shift in the jet between the PI and the LGM. However models with the PI jet positioned equatorwards of around 47 °S do not show this response: the sea ice edge is too far from the jet for it to respond. In models with accurately positioned PI jets; a +1° difference in the latitude of the sea ice edge tends to be associated with a −0.85° shift in the 850 hPa jet. However, it seems that around 5° of expansion of LGM sea ice is necessary to hold the jet in its PI position. Since observational data supports an expansion of more than 5°, this result suggests that a slight poleward shift and intensification, was the most likely jet change between the PI and the LGM. Without the effect of sea ice, models simulated polewards shifted westerlies in warming climates and equatorward shifted westerlies in colder climates. However, the feedback of sea ice counters and reverses the equatorward trend in cooler climates so that the LGM winds were more likely to have also been shifted slightly poleward.

Reconstructions of salinity are used to diagnose changes in the hydrological cycle and ocean circulation. A widely used method of determining past salinity uses oxygen isotope (δOw) residuals after the extraction of the global ice volume and temperature components. This method relies on a constant relationship between δOw and salinity throughout time. Here we use the isotope-enabled fully coupled General Circulation Model (GCM) HadCM3 to test the application of spatially and time-independent relationships in the reconstruction of past ocean salinity. Simulations of the Late Holocene (LH), Last Glacial Maximum (LGM), and Last Interglacial (LIG) climates are performed and benchmarked against existing compilations of stable oxygen isotopes in carbonates (δOc), which primarily reflect δOw and temperature. We find that HadCM3 produces an accurate representation of the surface ocean δOc distribution for the LH and LGM. Our simulations show considerable variability in spatial and temporal δOw-salinity relationships. Spatial gradients are generally shallower but within ~50% of the actual simulated LH to LGM and LH to LIG temporal gradients and temporal gradients calculated from multi-decadal variability are generally shallower than both spatial and actual simulated gradients. The largest sources of uncertainty in salinity reconstructions are found to be caused by changes in regional freshwater budgets, ocean circulation, and sea ice regimes. These can cause errors in salinity estimates exceeding 4 psu. Our results suggest that paleosalinity reconstructions in the South Atlantic, Indian and Tropical Pacific Oceans should be most robust, since these regions exhibit relatively constant δOw-salinity relationships across spatial and temporal scales. Largest uncertainties will affect North Atlantic and high latitude paleosalinity reconstructions. Finally, the results show that it is difficult to generate reliable salinity estimates for regions of dynamic oceanography, such as the North Atlantic, without additional constraints.

Rapid warming over the past 50 years on the Antarctic Peninsula is associated with the collapse of a number of ice shelves and accelerating glacier mass loss1, 2, 3, 4, 5, 6, 7. In contrast, warming has been comparatively modest over West Antarctica and significant changes have not been observed over most of East Antarctica8, 9, suggesting that the ice-core palaeoclimate records available from these areas may not be representative of the climate history of the Antarctic Peninsula. Here we show that the Antarctic Peninsula experienced an early-Holocene warm period followed by stable temperatures, from about 9,200 to 2,500 years ago, that were similar to modern-day levels. Our temperature estimates are based on an ice-core record of deuterium variations from James Ross Island, off the northeastern tip of the Antarctic Peninsula. We find that the late-Holocene development of ice shelves near James Ross Island was coincident with pronounced cooling from 2,500 to 600 years ago. This cooling was part of a millennial-scale climate excursion with opposing anomalies on the eastern and western sides of the Antarctic Peninsula. Although warming of the northeastern Antarctic Peninsula began around 600 years ago, the high rate of warming over the past century is unusual (but not unprecedented) in the context of natural climate variability over the past two millennia. The connection shown here between past temperature and ice-shelf stability suggests that warming for several centuries rendered ice shelves on the northeastern Antarctic Peninsula vulnerable to collapse. Continued warming to temperatures that now exceed the stable conditions of most of the Holocene epoch is likely to cause ice-shelf instability to encroach farther southward along the Antarctic Peninsula.

The sensitivity of interglacial Greenland temperature to orbital and CO2 forcing is investigated using the NorthGRIP ice core data and coupled ocean-atmosphere IPSL-CM4 model simulations. These simulations were conducted in response to different interglacial orbital configurations, and to increased CO2 concentrations. These different forcings cause very distinct simulated seasonal and latitudinal temperature and water cycle changes, limiting the analogies between the last interglacial and future climate. However, the IPSL-CM4 model shows similar magnitudes of Arctic summer warming and climate feedbacks in response to 2 × CO2 and orbital forcing of the last interglacial period (126 000 years ago). The IPSL-CM4 model produces a remarkably linear relationship between TOA incoming summer solar radiation and simulated changes in summer and annual mean central Greenland temperature. This contrasts with the stable isotope record from the Greenland ice cores, showing a multi-millennial lagged response to summer insolation. During the early part of interglacials, the observed lags may be explained by ice sheet-ocean feedbacks linked with changes in ice sheet elevation and the impact of meltwater on ocean circulation, as investigated with sensitivity studies. A quantitative comparison between ice core data and climate simulations requires stability of the stable isotope – temperature relationship to be explored. Atmospheric simulations including water stable isotopes have been conducted with the LMDZiso model under different boundary conditions. This set of simulations allows calculation of a temporal Greenland isotope-temperature slope (0.3–0.4‰ per °C) during warmer-than-present Arctic climates, in response to increased CO2, increased ocean temperature and orbital forcing. This temporal slope appears half as large as the modern spatial gradient and is consistent with other ice core estimates. It may, however, be model-dependent, as indicated by preliminary comparison with other models. This suggests that further simulations and detailed inter-model comparisons are also likely to be of benefit. Comparisons with Greenland ice core stable isotope data reveals that IPSL-CM4/LMDZiso simulations strongly underestimate the amplitude of the ice core signal during the last interglacial, which could reach +8–10 °C at fixed-elevation. While the model-data mismatch may result from missing positive feedbacks (e.g. vegetation), it could also be explained by a reduced elevation of the central Greenland ice sheet surface by 300–400 m.

We present a novel automated processing method for obtaining layer dip from radio-echo sounding (RES) data. The method is robust, easily applicable and can be used to process large (several terabytes) ground and airborne RES datasets using modest computing resources. We give test results from the application of the method to two Antarctic datasets: the Fletcher Promontory ground-based radar dataset and the Wilkes Subglacial Basin airborne radar dataset. The automated RES processing (ARESP) method comprises the basic steps: (1) RES noise reduction; (2) radar layer identification; (3) isolation of individual ‘layer objects’; (4) measurement of orientation and other object properties; (5) elimination of noise in the orientation data; and (6) collation of the valid dip information. The apparent dip datasets produced by the method will aid glaciologists seeking to understand ice-flow dynamics in Greenland and Antarctica: ARESP could enable a shift from selective regional case studies to ice-sheet-scale studies.

Reconstructions of temperature variations from Antarctic ice cores rely on the assumption that the relationship between hydrogen and oxygen isotopic ratios and temperature are stable in space and time. Sime et al. analyse three 340,000-year-old ice cores from East Antarctica and use an isotope-enabled general circulation model to show that instead, the relationship is nonlinear. During warm periods, the ratios are less sensitive to temperature, so previous estimates of interglacial temperatures are likely to be about 3 °C too low. This is consistent with peak Antarctic interglacial temperatures at least 6 °C higher than today. This work suggests that there are serious deficiencies in our understanding of climates that are warmer than today's. Reconstructions of temperature variations from Antarctic ice cores rely on the assumption that the relationship between hydrogen and oxygen isotope ratios and temperature are stable in space and time. Three East Antarctic ice core records are now analysed alongside input from general circulation models to reveal that during warmer interglacial periods the isotope ratios are less sensitive to temperature than during colder interglacials; consequently, previous estimates of interglacial temperatures are probably too cold. Stable isotope ratios of oxygen and hydrogen in the Antarctic ice core record have revolutionized our understanding of Pleistocene climate variations and have allowed reconstructions of Antarctic temperature over the past 800,000 years (800 kyr; refs 1, 2). The relationship between the D/H ratio of mean annual precipitation and mean annual surface air temperature is said to be uniform ±10% over East Antarctica3 and constant with time ±20% (refs 3–5). In the absence of strong independent temperature proxy evidence allowing us to calibrate individual ice cores, prior general circulation model (GCM) studies have supported the assumption of constant uniform conversion for climates cooler than that of the present day3,5. Here we analyse the three available 340 kyr East Antarctic ice core records alongside input from GCM modelling. We show that for warmer interglacial periods the relationship between temperature and the isotopic signature varies among ice core sites, and that therefore the conversions must be nonlinear for at least some sites. Model results indicate that the isotopic composition of East Antarctic ice is less sensitive to temperature changes during warmer climates. We conclude that previous temperature estimates from interglacial climates are likely to be too low. The available evidence is consistent with a peak Antarctic interglacial temperature that was at least 6 K higher than that of the present day —approximately double the widely quoted 3 ± 1.5 K (refs 5, 6).

Several studies have suggested that sea-level rise during the last interglacial implies retreat of the West Antarctic Ice Sheet (WAIS). The prevalent hypothesis is that the retreat coincided with the peak Antarctic temperature and stable water isotope values from 128,000 years ago (128 ka); very early in the last interglacial. Here, by analysing climate model simulations of last interglacial WAIS loss featuring water isotopes, we show instead that the isotopic response to WAIS loss is in opposition to the isotopic evidence at 128 ka. Instead, a reduction in winter sea ice area of 65±7% fully explains the 128 ka ice core evidence. Our finding of a marked retreat of the sea ice at 128 ka demonstrates the sensitivity of Antarctic sea ice extent to climate warming., The peak in Antarctic ice core isotope values, 128,000 years before present, was concurrent with a significantly warmer-than-present Antarctic climate. Here, the authors show that this isotope maximum was associated with a major retreat of sea ice and not a collapse of the West Antarctic Ice Sheet.

International audience; Measurements of Last Interglacial stable water isotopes in ice cores show that central Greenland d18O increased by at least 3‰ compared to present day. Attempting to quantify the Greenland interglacial temperature change from these ice core measurements rests on our ability to interpret the stable water isotope content of Greenland snow. Current orbitally driven interglacial simulations do not show d18O or temperature rises of the correct magnitude, leading to difficulty in using only these experiments to inform our understanding of higher interglacial d18O. Here, analysis of greenhouse gas warmed simulations from two isotope-enabled general circulation models, in conjunction with a set of Last Interglacial sea surface observations, indicates a possible explanation for the interglacial d18O rise. A reduction in the winter time sea ice concentration around the northern half of Greenland, together with an increase in sea surface temperatures over the same region, is found to be sufficient to drive a >3‰ interglacial enrichment in central Greenland snow. Warm climate d18O and dD in precipitation falling on Greenland are shown to be strongly influenced by local sea surface condition changes: local sea surface warming and a shrunken sea ice extent increase the proportion of water vapour from local (isotopically enriched) sources, compared to that from distal (isotopically depleted) sources. Precipitation intermittency changes, under warmer conditions, leads to geographical variability in the d18O against temperature gradients across Greenland. Little sea surface warming around the northern areas of Greenland leads to low d18O against temperature gradients (0.1-0.3‰ per °C), whilst large sea surface warmings in these regions leads to higher gradients (0.3-0.7‰ per °C). These gradients imply a wide possible range of present day to interglacial temperature increases (4 to >10 °C). Thus, we find that uncertainty about local interglacial sea surface conditions, rather than precipitation intermittency changes, may lead to the largest uncertainties in interpreting temperature from Greenland ice cores. We find that interglacial sea surface change observational records are currently insufficient to enable discrimination between these different d18O against temperature gradients. In conclusion, further information on interglacial sea surface temperatures and sea ice changes around northern Greenland should indicate whether +5 °C during the Last Interglacial is sufficient to drive the observed ice core d18O increase, or whether a larger temperature increases or ice sheet changes are also required to explain the ice core observations.

The Southern Hemisphere (SH) westerly winds are thought to be critical to global ocean circulation, productivity, and carbon storage. For example, an equatorward shift in the winds, though its affect on the Southern Ocean circulation, has been suggested as the leading cause for the reduction in atmospheric CO2 during the Last Glacial period. Despite the importance of the winds, it is currently not clear, from observations or model results, how they behave during the Last Glacial. Here, an atmospheric modelling study is performed to help determine likely changes in the SH westerly winds during the Last Glacial Maximum (LGM). Using LGM boundary conditions, the maximum in SH westerlies is strengthened by ∼+1 m s−1 and moved southward by ∼2° at the 850 hPa pressure level. Boundary layer stabilisation effects over equatorward extended LGM sea-ice can lead to a small apparent equatorward shift in the wind band at the surface. Further sensitivity analysis with individual boundary condition changes indicate that changes in sea surface temperatures are the strongest factor behind the wind change. The HadAM3 atmospheric simulations, along with published PMIP2 coupled climate model simulations, are then assessed against the newly synthesised database of moisture observations for the LGM. Although the moisture data is the most commonly cited evidence in support of a large equatorward shift in the SH winds during the LGM, none of the models that produce realistic LGM precipitation changes show such a large equatorward shift. In fact, the model which best simulates the moisture proxy data is the HadAM3 LGM simulation which shows a small poleward wind shift. While we cannot prove here that a large equatorward shift would not be able to reproduce the moisture data as well, we show that the moisture proxies do not provide an observational evidence base for it.