Antarctic sea-ice cover reached historically low levels in 2023, consistent with the simulated decline in sea-ice extent in response to anthropogenic warming. As Antarctic sea ice is closely linked to multiple components of the Earth system, its demise could precipitate widespread, cascading changes across the cryosphere, atmosphere, and ocean. However, the nature and strength of these interconnections remain poorly understood and are often inadequately represented in models. In this review, we combine modern observations, models, and paleoclimate archives spanning the last glacial cycle to examine how reductions in sea ice may affect other components of the Earth system. We review how Antarctic sea ice interacts with ocean and atmosphere circulation, ice sheets and ice shelves, marine productivity, and the carbon cycle. We find evidence from theory and models that Antarctic sea ice exerts important influences on the Earth system. Paleo-proxy reconstructions provide examples in which changes in sea ice co-occur with changes in the carbon cycle, marine productivity, and ocean circulation. However, isolating the specific impact of sea ice remains challenging in such a highly interconnected system.
This article investigates whether Transformer-based language models could replace the labour-intensive, manual classification of convective weather phenomena found in written texts from the pre-instrumental measurement era. The training set is based on a corpus of 6999 written observations from 494 Central European sources, spanning the period from 1000 to 1817. This corpus has been linguistically normalised, ranging from Middle High German to Contemporary German.For the training set, the text sources containing observations of thunderstorm and hail events are first classified manually using a formalised procedure. Quality assurance is initially carried out at the level of the existing textual information. To this end, evidence classes are introduced. These are based on linguistic evidence regarding the associated phenomena of thunderstorm and hail events. In the next step, the plausibility of the classified thunderstorm and hail events is assessed by checking their consistency with the DWD normal periods (1961–1990, 1991–2020).The seasonal signal is preserved across four language stages and nine source types, from the early 15th to the early 19th century. The classified thunderstorm and hail events exhibit a plausible physical signal and show strong correlations with current observations. The trained models “ThunderstormBERT” and “HailBERT” achieve macro-F1 scores of 0.83 and 0.93 respectively. Misclassifications primarily impact the middle thunderstorm class (moderate thunderstorm), where the classification scheme is least clear-cut.
Hot European summers are often preceded by dry springs such as in 2022, but also in 1473 or 1540, two well-known summers of catastrophic heat. Spring precipitation deficits deplete soil moisture levels, which can amplify extreme summer temperature anomalies in Europe through land-atmosphere feedback mechanisms and altered atmospheric circulation. However, the link is not particularly strong, and hence long time series might help to better elucidate the mechanisms. Starting from documentary data and an atlas of temperature, precipitation, and atmospheric circulation over Europe for climate extremes in the last 600 years, we explore this relationship in more detail. We analyse the extreme heat summer of 1473, which followed dry spring conditions in Southeastern Europe related to a positive East Atlantic pattern. We then use the ModE-RA paleo-reanalysis and combine it with an analysis of 11 760 years of atmospheric model simulations as well as other reanalyses and reconstructions. Using a moving climatology approach with LOESS regressions to calculate anomalies, we identify significant negative correlations between winter-spring precipitation and summer temperatures in all data sets in the latitude band 36–48° N (strongest over southeastern Europe), corresponding to known moisture-limited regions. Moreover, apart from precipitation anomalies, hot summers are preceded by increased blocking over north-central Europe and a positive East Atlantic pattern. Conversely, dry winter-springs are followed by more frequent blocking over northern Europe and, in model simulations, an increase in frequency and intensity of summer heatwaves. A linear regression approach for temperature in the northern Mediterranean region shows that precipitation in April and May has a strong, direct influence that does not vanish when taking the detailed atmospheric circulation in winter, spring, and summer into account, and hence cannot be explained by the effect of circulation on both, spring precipitation and summer temperature.
Quantitative reconstruction of ocean surface density during the Last Glacial Maximum (LGM) offers valuable insights into the ability of climate models to simulate past climate conditions, when global temperatures were about 4.5 to 6°C colder than today. We assess the performance of the LGM climate simulations, as part of the 3rd and 4th phase of the Paleoclimate Modeling Intercomparisons Project, using a recent ocean surface density reconstruction based on the δ18O of foraminiferal calcite (δ18Oc). We consider the differences between the LGM and the preindustrial climates and each period separately, at both global and regional scales. Because surface density reflects the combined effects of temperature and salinity, we also examined sea surface temperature (SST) to better identify the processes underlying model–data differences.On a global scale, surface density reconstructions generally exhibit greater spatial variability than simulated surface density anomalies (LGM − PI), although part of this difference is reduced when reconstruction uncertainties are taken into account. Model simulations tend to underestimate the magnitude of reconstructed density anomalies and substantial differences between cumulative distribution persist. Part of the mismatch may arise from the uneven spatial distribution of reconstructions, which are mostly located near coastal areas.Density anomaly (LGM − PI) differences between data and models are largely controlled by sea surface salinity (SSS), with SST contributing to a lesser extent. This influence of SSS is directly linked to the reduction in tropical precipitation during the LGM: model simulations that best match the large-scale density anomalies also simulate the strongest reductions in reconstructed low-latitude precipitation during the LGM, highlighting the key role of hydrological cycle changes in shaping surface density.All simulations capture key features of the reconstructed surface densities when the LGM and PI periods are analysed separately at the global scale, with Taylor diagrams and complementary performance metrics indicating moderate to good overall agreement despite differences among simulations.Regional analyses show that most simulations reproduce reconstructed Indian Ocean surface density reasonably well, although model simulations performance is systematically lower in the North Indian Ocean than in the South Indian Ocean. Focusing on the Indo-Pacific Warm Pool, proxy reconstructions indicate a weakened West-East tropical Indian surface density gradient during the LGM, but only 7 out of 14 model simulations (50 %) reproduce this feature.These results highlight the need to improve and better constrain regional hydrological cycle changes in models, as improving their representation is crucial to reduce uncertainties in both paleoclimate simulations and future climate projections.
The Late Ordovician marks the first major continental ice sheet event in the Phanerozoic Eon, despite high CO2
The Southern Hemisphere Westerly Winds (SHWW) play an important role in global climate and Antarctic ice sheet dynamics; however, high-resolution proxy reconstructions are sparse. Conflicting evidence suggests SHWW latitudinal change, weakening and/or strengthening, throughout the Holocene. Dust microparticle characteristics preserved in Antarctic ice cores (i.e., changes in concentration and size) provide a useful physical proxy reflecting wind dynamics. We present South Pole Ice Core (SPC14) dust records spanning the Holocene. Dust concentrations (1.1–5.1 µm) decrease through the Holocene, while the coarse particle percentage (CPP) increase due to a relatively larger decrease in finer (1.1–3.2 µm) dust, highlighting an opposing trend. The South Pole dust trends, consistent with CMIP6-PMIP4 model trends, southern mid-latitude insolation, and mid- to high latitude Southern Hemisphere proxy records, reflect SHWW southward migration and strengthening of wind speed south of 51° S during the mid-Holocene (∼ 7000–3000 years ago). Our findings suggest orbitally-driven warming throughout the Holocene led to strengthening and southward migration of the SHWW beginning ∼ 7000 years BP.
Determining the climatic conditions under which the Greenland Ice Sheet (GrIS) was smaller than present is important to quantify GrIS sensitivity to climate change. We use biomarkers in sediment collected beneath the GrIS at Camp Century, northwestern Greenland, to reconstruct summer temperature and atmospheric circulation 416 000 ± 38 000 years ago, at some point during Marine Isotope Stage 11 (MIS 11). We find that northwestern Greenland summer climate during MIS 11 was similar to the middle Holocene but different than the 20th century. Air temperature was 4.7 ± 3.2 °C warmer, and atmospheric water vapor isotope values were 22 ± 18 ‰ 2H-enriched, indicating a greater contribution of locally evapotranspired moisture. These conditions are similar to or slightly warmer than during peak Holocene warmth on Greenland, when the Camp Century site remained ice-covered, and cooler than the Last Interglacial (LIG). Biomarkers from lower in the section likely represent an earlier ice-free interval, potentially during the Pliocene or early Pleistocene, and record climatic conditions similar to MIS 11. These data add to the sparse available climate data for the early and middle Pleistocene on Greenland, and suggest interglacial periods had similar temperature throughout the Pleistocene. We interpret the reduced GrIS extent in northwestern Greenland during MIS11 compared to the Holocene, despite similar temperature, to indicate ice-sheet response to prolonged warmth, as reconstructed in southern Greenland. Therefore, efforts to reduce both the magnitude and duration of summer warmth in the coming centuries will be important to curbing ice-sheet retreat.
Sea Surface Temperature reconstructions derived from alkenone biomarker (SST-alk) reveal a cooling trend in the North Atlantic during the late Holocene (the last 5000 years), contrary to the warming simulated by transient climate models driven by 20 ppm increase in greenhouse gas concentrations. In this study, we present new SST-alk time series from the inter-gyre region of the North Atlantic. Our results, together with existing data, indicate that late Holocene cooling is marked by an increasing zonal SST gradient across the mid-latitude North Atlantic, with more pronounced cooling in the west than in the east. Complementary planktonic foraminifera assemblage data, interpreted alongside other circulation proxies, suggest that a late Holocene reorganization of inter-gyre ocean circulation as a likely driver of this zonal asymmetry. Finally, we compare these proxy-based reconstructions with transient climate model simulations (TraCE-21k), and find that models do not reproduce the observed zonally asymmetric cooling and the inferred circulation changes. This misrepresentation of spatial and temporal variability likely explains the data-model discrepancy in the mid-latitude North Atlantic.
A recent study identified an east-west dipole pattern in Asian Summer Monsoon precipitation and in its oxygen-isotope composition variability in response to orbital forcing (Wen et al., 2024). The Korean Peninsula (KP) is situated near the nodal line of this dipole and should therefore exhibit muted precessional variability in both. So far, this conjecture has not been tested using paleoclimatic data. Here, we present speleothem oxygen isotope (δ18O) records from the KP, which support the notion of suppressed orbital-scale hydroclimate variability. Conducting a transient model simulation with the isotope-enabled Community Earth System Model (iCESM1.2) covering the past 130 000 years, along with tagging experiments for low and high insolation conditions, we show that, on precessional timescales, isotopic contributions from oceanic and continental moisture sources compensate each other over the KP, resulting in only a weak regional signal in precipitation δ18O. Based on iCESM1.2 simulations, we further demonstrate that deuterium excess (d-excess) variability over the KP would still capture the moisture source region's zonal seesaw response to precessional forcing. We propose that, given the region's strong sensitivity to changes in moisture sources, paleo-water d-excess reconstructions from East Asia speleothems could provide valuable additional constraints on the drivers of regional monsoon systems. This study offers new insights into the spatiotemporal variability of Pan-Asian hydroclimates and its links to changes in moisture source.
Marine Isotope Stage (MIS) 11 has long been considered a unique Quaternary interglacial stage due to its orbital similarities with the Holocene, persistent high atmospheric CO22, ice volume, and millennial-scale variability in driving vegetation and summer monsoon changes, depending on the changing boundary conditions throughout MIS 11. Site U1446 pollen data, supported by model simulations, indicate that the maximum tropical forest expansion and strongest monsoon activity during the interglacial period (MIS 11c) followed boreal summer insolation, revealing the dominant role of insolation under a warm climate state with high CO224
Iron (Fe) availability modulates phytoplankton blooms in High-Nutrient Low-Chlorophyll (HNLC) regions, i.e., ocean areas characterized by an abundance of major nutrients but low marine productivity. Fe can be delivered to the oceans through atmospheric dust deposition, making ice cores unique archives for reconstructing past changes in aeolian Fe deposition. However, while it is known that during dustier periods atmospheric Fe depositions increased, uncertainties remain regarding the fraction of Fe actually available to phytoplankton. Here, we present evidence from the EGRIP ice core (Greenland) during the Pleistocene-Holocene transition (10.3–13.0 ka), which provides insights into atmospheric aerosol deposition over the Fe-limited North Pacific Ocean. Results show that, in contrast to the 17-fold enhancement in total Fe determined by Inductively Coupled Plasma Mass Spectrometry, the iron fraction more likely available for phytoplankton (i.e., labile iron) increased only modestly (+29 %) during the Younger Dryas compared to the Early Holocene, likely due to prevailing alkaline aerosol conditions reducing its solubility. This finding supports the hypothesis that factors other than atmospheric Fe deposition (e.g., stronger water stratification, sea-ice extent, volcanic eruptions, iron remobilization from sediments), played a more relevant role in regulating marine net primary productivity in the HNLC North Pacific Ocean over the last glacial transition.
Dust deposits in ice cores provide a valuable archive of past atmospheric circulation, offering insights into climate dynamics during key climate transitions such as the last glacial termination. Here, we present a novel high-resolution reconstruction of dust provenance in the EPICA Dome C (EDC) ice core from 33.7 to 2.8 kyr BP, based on Rare Earth Element (REE) patterns. Using a statistical unmixing algorithm on 279 samples, we quantify, for the first time, contributions from key Southern Hemisphere dust sources. During the late Marine Isotope Stage 3 (MIS3), the Last Glacial Maximum (LGM) and Heinrich Stadial 1 (HS1), dust at EDC was dominated by Patagonian sources, ∼ 65 %–75 % of total deposition, with secondary inputs from Australia (∼ 9 %–11 %), Southern Africa (∼ 5 %–9 %), New Zealand (∼ 6 %–14 %), and Puna-Altiplano (∼ 2 %–4 %). After ∼ 14.5 kyr BP, and especially during the Holocene, the dust assemblage shifted toward greater contributions from low-latitude sources including Australia (21 %–28 %), Southern Africa (∼ 8 %–9 %), and the Puna-Altiplano (∼ 8 %–9 %) – while the Patagonian contribution decreased to an average of ∼ 43 %–53 %, New Zealand (∼ 4 %–10 %) – and source compositions became more variable. This transition in dust provenance contribution is supported by our modeled Sr-Nd isotope composition of EDC dust, which also aligns with changes observed in other East Antarctic ice cores. Comparison with the EDML ice core provenance record reveals overall agreement in major sources and timing of the shifts but also highlights regional variations in the secondary contributions, with EDC showing more consistent inputs from Australia and EDML from Southern Africa. The coherence of the provenance shift recorded at EDC and EDML ice cores points to a common large-scale control, in line with the influence of eustatic sea-level rise and postglacial reorganization of fluvial sediment routing and availability in southern South America.
The ocean mixed layer plays an essential role in the climate system, regulating energy fluxes at the ocean-atmosphere interface. Its representation in climate models is thus critical. Here, we evaluate the mixed layer depth (MLD) in 15 models from the Paleoclimate Modelling Intercomparison Project 4 (PMIP4) against dinocyst-based MLD reconstructions from the subpolar North Atlantic for the mid-Holocene (MH, 6 ka BP). We find a large spread in MLD responses to MH forcings across the models in the present-day deep-water formation areas, underscoring the importance of model uncertainty. Most models fail to capture the direction of MLD change, and the ensemble mean does not necessarily outperform individual models. While the ensemble mean aligns broadly with proxy data in the Nordic Seas, pronounced proxy-model discrepancy in the Labrador Sea suggests that meltwater forcing is a missing parameter, and that deep-water formation in the Labrador Sea may be particularly vulnerable under a future scenario of global warming and ice-sheet melting.
The major climate change recorded during the Early-Middle Pleistocene Transition (1.4–0.4 Ma) is characterized by an increase in the length of climatic cycles from 41 to 100 ka. During this period, the Mediterranean climate underwent aridification associated with a drop in temperatures. This period also corresponds to the arrival of the first hominins in Western Europe. This study aims to establish the climatic framework at the end of the Early Pleistocene between MIS 37-31 (∼ 1.25 to 1.06 Ma). To this end, a multiproxy approach was applied to assemblages of planktonic foraminifera, pollen and clay mineralogy, constituting a multi-method approach to climate reconstruction. Comparisons with other sequences from the Central and Western Mediterranean show major differences between the Iberian Peninsula and southern Italy. Indeed, the MIS 31 appears cooler and dryer in the south of the Iberian Peninsula than in central Mediterranean. Clay mineralogy analysis highlights a sudden change in oceanic and atmospheric circulation in the Alboran Basin around 1140 ka. Climate reconstructions follow climatic cycles, with temperatures that appear to be underestimated compared to temperature reconstructions based on fauna from southern Spain. Precipitation reconstructions, on the other hand, are more consistent, suggesting that the climate and vegetation of this region are more likely to be influenced by variations in precipitation. These new data enhance our understanding of the climate of the South-western Mediterranean at the beginning of the Early-Middle Pleistocene Transition.
The Eocene–Oligocene Transition (EOT dated at ∼ 34 Ma) represents one of the most significant climatic shifts of the Cenozoic, marking the transition from the last warmhouse state to a coolhouse state. This global cooling had major consequences for terrestrial ecosystems and was synchronous with the dispersal of numerous Asian mammalian clades towards western Europe. However, the terrestrial expression of the EOT exhibits strong regional heterogeneity questioning its role in establishing dispersal corridors associated with floral and faunal turnovers.Here, we describe, date, and document the paleoenvironments of a continental sedimentary section from Balkanatolia, a biogeographic province corresponding to the present-day Balkans (NE Mediterranean region) and Anatolia (Türkiye). This region most likely functioned as a critical stepping stone for the dispersal of Asian mammals toward western Europe. Our sedimentary record represents a fluvio-lacustrine system exposed over a ∼ 200 m section located in Büyükteflek (Çiçekdağı/Kırşehir area, Türkiye), dated by magnetostratigraphy to the Priabonian and the lower Rupelian, including the Oi-1 glaciation (∼ 33.65 Ma). Clumped isotope analyses on pedogenic carbonates across our record show evidence for a Late Eocene Warming starting during the middle Priabonian (ca. 37 Ma), followed by a marked cooling event at the Eocene–Oligocene Glacial Maximum (EOGM). Stable isotopic data and sedimentary facies further indicate that this complete interval is associated with a long-term aridification trend, starting during the Late Eocene warming and culminating at the EOT. Our results provide the first quantitative record of late Eocene warming on land, and our temperature estimates for the earliest Oligocene cooling are consistent with other Eurasian clumped-isotope records. These temperature shifts and associated aridification steps may have acted as contributing drivers of the late Eocene decline of Balkanatolian endemic taxa and likely facilitated the westward expansion of Asia-derived mammals ultimately resulting in the colonization of western Europe.
The cause of the fundamental reorganisation of the climate system during the Middle Pleistocene Transition (MPT; ∼ 1.25–0.65 million years ago (Ma)), when glacial cycles intensified and lengthened from 41 thousand year (kyr) to quasi-100 kyr periodicity, remains one of the enduring unsolved questions in palaeoclimate science. Increasing attention has focused on the role of the marine carbon cycle, with enhanced carbon storage in the deep sea linked to changes in deep-water temperature and salinity that increase abyssal ocean density stratification. Here we report new high-resolution reconstructions of deep-water temperature and the isotopic composition of seawater (δ18Oseawater), a proxy for ice volume and salinity, from North Atlantic Site U1385, using paired benthic foraminiferal
The North African desert margin is considered one of the most sensitive areas to future climate changes, yet the periodicities, coupling mechanisms and external forcing of Holocene environmental variability remain poorly understood. To investigate millennial- to centennial-scale periodicities in Holocene climatic variability and geomorphological processes, we use a Holocene sediment record from Lake Sidi Ali in the semiarid to sub-humid Middle Atlas with a robust 210Pb 
Pollen records are among the highest-resolution spatial and temporal proxies for reconstructing past vegetation dynamics, environmental changes and climate variability. Over the past decade, a large variety of methods based on different ecological or mathematical concepts has been used to reconstruct paleoclimatic conditions from pollen assemblages. However, the accuracy of these climate reconstructions strongly depends on the choice of the modern calibration dataset, the taxonomic resolution, and/or the modelling assumptions. The lack of a univocal response still limits the application of pollen-based climate reconstructions to assess key climate changes over multiple time periods especially during the Last Glacial Maximum (LGM, ∼ 23–19 ka BP). Here, we present a multi-method approach, including the Modern Analogue Technique (MAT), the Weighted Averaging Partial Least Squares regression (WA-PLS) and the probability density function-based Climate REconstruction SofTware (CREST), to reconstruct European climates during the LGM. The quality and performance of our climate reconstructions show strong heterogeneity when based on large calibration datasets encompassing wide climatic and vegetation gradients, making local sampling for climate reconstructions difficult. Instead of sampling the global calibration dataset, we test the effect of the latest biomization and megabiomization methods (local calibrations based on megabiomization approaches) on climate reconstructions by introducing a new biome-based approach. Unlike previous studies, we use the weighted mean of climate variables from all megabiome scores rather than only considering the dominant (i.e., highest score) megabiome. This significantly reduces some of the statistical noise of climate reconstructions, drastically minimizing threshold and non-linear effects associated with megabiome classification changes. With these methodological advancements and our multi-method comparison, we evaluate the uncertainties (RMSEP) of the paleoclimate reconstructions for the LGM in Europe. Across climate reconstruction methods (MAT, WA-PLS and CREST methods), European LGM annual temperatures from the biomization method were on average 6.4±2.0 °C (mean SD) colder than today, consistent with megabiomization results (6.8±2.2 °C colder). Winter temperature (mean temperature of the coldest month, MTCO) results exhibit substantial spatial variability across Europe. Local calibration techniques significantly reduce uncertainties in LGM MTCO reconstructions, but they remain highly sensitive to the choice of calibration datasets.
It is believed that the atmospheric circulation on Snowball Earth produced a net ablation zone exposing bare sea ice. Under sufficiently low temperatures, salt begins to precipitate out of sea ice, forming a lag deposit of crystals with high albedo as the ice sublimates. This could have resulted in a salt-albedo feedback that has not previously been included in modeling studies of Snowball Earth. We implement a salt-albedo feedback in a simple climate model and show that, once initiated, this mechanism could have intensified global cooling in the initial phase of Snowball Earth. Our results suggest that salt precipitation may have played a role in shaping the early climate of Snowball Earth.
Marine anoxia has been implicated as a key environmental driver of the end-Permian mass extinction (EPME) and the subsequent prolonged recovery. However, the spatial and temporal extent of oxygen limitation during the EPME interval remains contentious. Here, we present iron speciation, pyrite framboid and molybdenum–uranium (Mo–U) covariation data from two palaeogeographically distinct settings: the Tethyan Chibi section (South China) and the Panthalassian Ursula Creek section (Western Canada) to evaluate redox dynamics across the Permian-Triassic transition. Our data suggest that bottom waters were predominantly dysoxic during the late Changhsingian at both sites. Later, the prevalence of small pyrite framboids, elevated Mo and U enrichment factors (MoEFEF), and high MoEF