Proceedings of the National Academy of Sciences, Volume 123, Issue 33, August 2026. SignificanceIce shelves provide a key source of stability to the Antarctic Ice Sheet by buttressing the flow of grounded ice. As ice shelves weaken and collapse due to rising temperatures, the glaciers buttressed by them speed up, increasing Antarctica’s ...

Greenland and Antarctic ice sheets are losing mass rapidly, contributing significantly to sea level rise and altering global climate patterns. Freshwater input from these ice sheets influences tropical climate through ocean-atmosphere interactions, but their distinct impacts remain unclear. Here, we use coupled climate model simulations with imposed meltwater fluxes from Greenland, Antarctica, and both combined under global warming scenarios to investigate tropical responses. We find that Greenland meltwater induces a pronounced southward shift of the Intertropical Convergence Zone (ITCZ) and strengthens the Walker circulation, while Antarctic meltwater causes weaker, opposite-signed effects. Combined meltwater forcing partially offsets these responses, with Greenland effects dominating. These asymmetric impacts arise from differential perturbations to ocean heat transport, atmospheric energy transport, and Atlantic–Pacific pressure-wind adjustments, with cloud-radiative changes acting as secondary modulators. Our results suggest that, under symmetric meltwater perturbations, Northern Hemisphere ice sheet melt may disproportionately influence tropical rainfall and circulation patterns, underscoring the need to represent freshwater forcing pathways in climate projections accurately.

Cribroelphidium clavatum is a shallow water benthic Foraminifera transported into bathyal environments of Deep Sea Drilling Program Site 548 offshore Ireland over the last 420,000 years. Their abundance and relationship to periods with Neogloboquadryina pachyderma and Ice Rafted Debris suggests a relationship to British Irish Ice Sheet variability. C. clavatum are common (in the absence of other shallow Foraminifera) during glacial Terminations 4 (~335 ky), 3 (~243 ky), 3a (~221 ky), 2 (~132 ky), and 1 (Heinrich Event (HE) 1, ~15 ky) and MIS 7b (~205 ky). Tests are rare during interglacials and HE 6 (~60 ky) and 3 (~34 ky). Contourite activity dominated interglacials while pelagic facies prevailed during glacials, suggesting Cribroelphidium clavatum deposition was not related to high energy currents, instead, its distribution was likely related to melting during glacial terminations. One way to explain Cribroelphidium clavatum abundance during terminations is warming induced British Irish Ice Sheet ice collapse caused ice shelf melting triggering downslope transport via gravity or debris flows. This does not fully explain the absence of other shallow Foraminifera at this site. However, there may have been common C. clavatum in nearby shallower regions and their test robustness may have favoured preservation. Another way to deposit Cribroelphidium clavatum is to transport tests by iceberg calving from ice shelves close to land, freezing tests into the ice shelf. Warming melted icebergs, releasing tests to the seafloor in the absence of gravity or debris flows. This explains the absence of C. clavatum during HE 6 and 3.

Abstract Late Pleistocene climate in the European Alps was marked by large, orbitally driven fluctuations in temperature, precipitation, and ice extent. Reconstructing mountain glaciation prior to the Last Glacial Maximum remains challenging, however, because glacial deposits are often eroded or reworked. Here we present a 90-kyr speleothem record from Klaus Cramer Cave (western Austria) spanning Marine Isotope Stages (MIS) 12–10. Stable-isotope and trace-element data identify intervals of subglacial speleothem growth associated with warm-based ice cover. Subglacial phases are characterized by low δ 18 O, high δ 13 C, and highly elevated sulfur concentrations, indicating enhanced sulfide oxidation from the infiltration of meltwater through glacially comminuted detritus. Sulfate δ 34 S and δ 18 O values signal that the aerobic oxidation of pyrite was the dominant sulfur source. In the absence of geomorphological evidence, we constrain the timing of warm-based ice retreat to 423.3 ± 1.6 ka and re-advance to 370.8 +2.1 / −1.8 ka, while MIS 11 remained entirely ice-free.

arXiv:2608.19469v1 Announce Type: new Abstract: Paleoclimate archives extend climate knowledge beyond the instrumental era, registering different seasons, variables, time averages, and memory lengths. A longstanding problem is to integrate these heterogeneous sources of information within a unified methodology. Here we present a new data-assimilation framework, Last Millennium Reanalysis 4D-Var (LMR4D-Var), which reconstructs climate trajectories from these heterogeneous datasets while balancing errors in the model, observations, and initial conditions. We compare results using LMR4D-Var to assimilate proxies from PAGES2k, Temp12k, and borehole temperature profiles without treating them as instantaneous equivalents. Instrumental verification shows that LMR4D-Var achieves the highest skill compared with previous reconstructions. Borehole assimilation preserves skill against withheld annually resolved records, increases agreement between reconstructed 300--2000-m ocean heat content and independent estimates, and yields a cooler reconstructed Little Ice Age ocean. Results for Temp12k demonstrate assimilation of decadal-to-millennial records and the potential for Holocene and deeper-time applications with suitable emulators.

arXiv:2512.12142v2 Announce Type: replace-cross Abstract: The Greenland ice sheet is melting at an accelerated rate due to processes that are not fully understood and hard to measure. The distribution of surface meltwater can help understand these processes and is observable through remote sensing, but current maps of meltwater face a trade-off: They are either high-resolution in time or space, but not both. We develop a deep learning model that creates gridded surface meltwater maps at daily 100m resolution by fusing data streams from remote sensing observations and physics-based models. In particular, we spatiotemporally downscale regional climate model (RCM) outputs using synthetic aperture radar (SAR), passive microwave (PMW), and a digital elevation model (DEM) over the Helheim Glacier in Eastern Greenland from 2017-2023. Using SAR-derived meltwater as "ground truth", we show that a deep learning-based method that fuses all data streams is over 10 percentage points more accurate over our study area than existing non deep learning-based approaches that only rely on a regional climate model (83% vs. 95% Acc.) or passive microwave observations (72% vs. 95% Acc.). Alternatively, creating a gridded product through a running window calculation with SAR data underestimates extreme melt events, but also achieves notable accuracy (90%) and does not rely on deep learning. We evaluate standard deep learning methods (UNet and DeepLabv3+), and publish our spatiotemporally aligned dataset as a benchmark, MeltwaterBench, for intercomparisons with more complex data-driven downscaling methods. The code and data are available at github.com/blutjens/hrmelt.

Indicators of biosiliceous production offer unique insights into the silica cycle at dynamic marine-continental interfaces, where seasonal discharges interact with a nutrient-depleted tropical ocean. This study reconstructs variations in siliceous paleoproductivity in the southwestern Gulf of Mexico (SW GoM) over the last 33.0 cal ka BP, investigating the complex interplay between oceanic circulation and continental nutrient forcing. Using a multiproxy approach, we integrated siliceous microfossils (diatoms, radiolarians, silicoflagellates, and sponge spicules) with bulk geochemical indicators (C org , CaCO 3 , and Si opal %wt) and total planktonic foraminifera. Our results reveal that while the SW GoM remained predominantly resource-limited, biogenic production exhibited significant variability across established Quaternary periods, with CONISS analysis delineating three distinct paleoceanographic stages over the Pleistocene-Holocene transition: (1) Last Glacial (33.0–19.0 cal ka BP), characterised by minimal biogenic accumulation across all microfossil groups due to enhanced stratification and reduced fluvial input relative to the current interglacial. (2) Deglaciation or Termination I (19.0–11.7 cal ka BP) was marked by a rise in diatom and sponge spicule abundance alongside peak planktonic foraminifera, indicating a carbonate-dominated system during marine transgression. (3) The interglacial (the last 11.7 cal ka BP) recorded the maximum biosiliceous accumulation, reflected in maximum C org and Si opal values, driven by sea-level stability and increased river discharge relative to the glacial. Our findings demonstrate that the siliceous fraction, although a minor component of total biogenic production, tracks the coupling between the silicon and carbon cycles. This study underscores the sensitivity of siliceous microfossils to land-sea interactions, demonstrating their crucial role in understanding the long-term evolution of tropical ecosystems.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

Understanding past ice sheet behaviours is key to predicting future ice sheets in a warming climate, but meshing geological observations with ice sheet models is challenging. This manuscript models the sedimentary system of the Finnish Lake District Ice Lobe during its retreat about 12 000 years ago, when extensive water flow developed under the ice, and connects models with observations. The approach opens a path to improve knowledge of previously glaciated regions and systems under ice today.

The Earth’s energy imbalance (EEI) that develops at the top of the atmosphere is accommodated by gains or losses of energy in Earth’s heat reservoirs, leading to temperature and sea level change. The global ocean has stored about 90% of the EEI from anthropogenic forcing, but the attribution of past changes of EEI remains largely unknown, thus obscuring our understanding of past climate change. Here, we reconstruct changes in mean ocean temperature over the past 150,000 years that, with a reconstruction of ice sheet–volume changes, allow us to isolate the contributions of the dominant ocean and ice sheet heat reservoirs to the global energy inventory as well as to derive their associated contributions to EEI. We attribute orbital-scale EEI variability to joint precessional and CO 2 forcing that caused changes in rates of ice sheet energy storage. In contrast, millennial-scale EEI variability can be attributed to radiative responses to decreases in the Atlantic meridional overturning circulation and increasing CO 2 during Heinrich stadials that caused changes in rates of ocean energy storage.

This study uses a high-resolution global climate model to simulate future climate, with a particular focus on the Arctic and North Atlantic. The model projects a nearly ice-free Arctic by 2040. The paper also introduces a framework to diagnose deep-water formation (DWF) in the Labrador, Irminger, and Greenland Seas and to quantify their regional contributions to the Atlantic Meridional Overturning Circulation (AMOC), revealing how different ocean regions contribute to the weakening of AMOC in a warming climate.

We present an ensemble of Coupled Model Intercomparison Project Phase 6 (CMIP6)-downscaled regional climate model simulations for Southeast Asia (1960–2100). We (1) evaluate model performance against multiple observational datasets and (2) apply a standardized benchmarking framework to identify high-performing models for further kilometer-scale dynamical downscaling over Southeast Asian megacities. This effort supports the development and use of reliable regional climate projections.

ABSTRACT Subfossil insect communities from the Quartz Creek site provide a detailed record of both Late Pleistocene steppe–tundra and Holocene boreal forest ecosystems in central Yukon Territory, Canada. Fifteen Pleistocene assemblages (marine isotope stages [MIS] 4 to 2) are dominated by steppe–tundra and dry‐tundra taxa, particularly the weevils Connatichela artemisiae and Lepidophorus lineaticollis , indicating a cold and dry environment with widespread Artemisia –grassland habitats. Periodic increases in mesic tundra, riparian and forest‐associated species reflect localised habitat mosaics within an otherwise arid landscape. Separated by a ~20 000‐year stratigraphic hiatus, the Holocene assemblages show a marked increase in taxa associated with shrubs, wetlands, and coniferous forests, including bark beetles such as Scolytus piceae , Polygraphus rufipennis, and Dendroctonus rufipennis , as well as the carpenter ant Camponotus herculeanus . These taxa indicate the presence of boreal forest habitats near the site by at least 6045 cal yr BP. Community composition and biodiversity indices show a clear difference between the Pleistocene and Holocene assemblages, reflecting a shift from relatively species‐poor steppe–tundra communities to more diverse assemblages associated with heterogeneous boreal forest environments. The continued absence of steppe–tundra specialists such as Connatichela artemisiae in our Holocene samples highlights the major restructuring of insect communities associated with the regional collapse of steppe–tundra ecosystems and the subsequent development of boreal forest environments in eastern Beringia.

The Greenland Ice Sheet is an important contributor to global sea level rise. In northern Greenland, floating glacier tongues are primarily melted by ocean currents. These processes are difficult to observe, so we developed a realistic numerical model to study ocean-driven melting at Greenland's largest floating ice tongue, the 79° North Glacier. Our simulation reveals the details of the oceanic currents bringing warm water toward the ice base, melting and shaping the glacier tongue from below.

ABSTRACT The Cadí Massif (42°17′09″ N, 1°38′11″ E; 2649 m a.s.l.) in the southeastern Pyrenees preserves clear evidence of Late Pleistocene glaciation, although previous research has focused almost exclusively on its northern slope. By contrast, the southern slope has remained largely unexplored. In this study, we present the first glacial geochronology for this sector by combining 36 Cl terrestrial cosmogenic nuclide (TCN) surface‐exposure dating with numerical paleoclimate modeling. We analyzed 18 boulder samples collected from moraines in two major cirques to constrain the timing of glacier fluctuations. Although some exposure ages were interpreted as reflecting post‐depositional disturbance or possible nuclide inheritance, the resulting chronology indicates that the upper plateaus and southern cirques remained glaciated during the late stages of deglaciation following the Last Glacial Cycle, while subsequent moraine systems document later glacial advances. These include small snowdrift glaciers, up to ca . 4 km in length, that persisted during the Younger Dryas stadial at ca . 12–13 ka, with their mass balance sustained largely by wind‐driven snow redistribution and avalanche‐fed accumulation. Paleoclimate simulations indicate that temperatures were ca . 3°C lower during that phase, while precipitation was comparable to present‐day conditions. This integrated approach demonstrates the sensitivity of southern‐slope glaciers to abrupt climate shifts and provides a more comprehensive view of the spatial variability of glacial responses across the Cadí Massif and the broader Eastern Pyrenees.

arXiv:2608.21998v1 Announce Type: cross Abstract: Regional climate risk assessment is critical for applications such as infrastructure design, disaster forecasting, and insurance resource allocation. However, estimating regional (i.e., high-spatial-resolution) risk with global climate models (GCMs) remains computationally prohibitive, which has driven the development of downscaling methods for coarse GCM outputs. Downscaling is vital for rare events, since quantifying their extreme properties requires high spatial resolution and very long GCM simulations. These methods non-intrusively increase GCM resolution while correcting statistical biases from unresolved fine-scale processes, thereby improving the accuracy of extreme event statistics with long return periods. A key challenge is preserving dynamical consistency, as freely evolving GCM trajectories are not expected to track the observational dataset used for training the correction operator. This is critical for causal extreme event analyses, where storyline-based risk assessment, i.e., extreme event catalogs, is necessary for effective planning. We address this challenge by introducing Dynamically and Statistically Consistent downscaling (DySCo), a non-intrusive framework yielding high-resolution climate projections consistent with coarse GCM dynamics. DySCo relies on a data-driven reformulation of nudging to create dynamically paired training trajectories without intrusive GCM modifications. Using these paired trajectories, we train a dynamically and statistically consistent, two-stage operator. We evaluate the method by downscaling the Community Earth System Model v2 Large Ensemble (LENS2) in time and space towards historical reanalysis. Results show DySCo achieves superior dynamical consistency with the coarse GCM trajectories, essentially applying a minimal, causal correction to the GCM, preserving top statistical performance comparable to state-of-the-art unsupervised models.

Introduction Maritime Antarctic meltwater lakes are highly sensitive to atmospheric variability, yet their short-term hydrothermal responses remain poorly understood because most studies rely on seasonal or daily observations that cannot resolve sub-daily atmosphere–lake interactions. Methods This study investigated Yanou Lake, a shallow maritime Antarctic meltwater lake, using continuous 5‐min meteorological and hydrological observations collected over 66 days during the 2025–2026 austral summer. Wind speed, air temperature, atmospheric pressure, water temperature, and lake depth were analyzed using an integrated framework combining event‐based atmospheric forcing detection, lagged correlation, wavelet coherence, change‐point detection, and statistically validated threshold analysis. Results A total of 916 five‐min event observations (4.84% of the record) were grouped into 12 independent compound atmospheric forcing events. Cross‐correlation analysis identified weak but measurable delayed atmosphere–lake associations, while autocorrelation indicated persistence in water temperature (75.6 h) and lake depth (165.9 h). Wavelet coherence showed that atmosphere–lake coupling was intermittent and scale dependent, strengthening only during discrete atmospheric events. Event‐level analyses indicated that average lake responses were small and varied among individual events. Change‐point analysis identified repeated thermal and hydrological reorganizations, whereas statistically supported thresholds of 6.22 m s -1 (wind speed) and 2.41 °C (air temperature) were associated with changes in lake behaviour. Discussion These findings demonstrate that atmosphere–lake interactions in Yanou Lake are governed by delayed, episodic, and nonlinear processes rather than persistent atmospheric control. The proposed high‐frequency analytical framework provides a transferable approach for investigating short‐term hydroclimatic dynamics in Antarctic meltwater lakes and improves understanding of cryosphere–hydrology interactions under ongoing polar climate change.

Transfer of ocean heat toward Antarctica’s ice shelves is a key driver of changes in glacial mass balance. Previous work has indicated that transfer of warm, mid-depth waters across the continental slope is elevated close to Antarctica’s largest ice shelves, where dense waters form over the continental shelf. Yet, ice shelves in such regions maintain relatively low area-averaged melt rates, suggesting that ocean processes must shield them from the inflowing heat. In this study, a high-resolution simulation is used to isolate the mechanisms of heat transfer toward the Filchner-Ronne Ice Shelf in the southern Weddell Sea. This study shows that ocean eddies divert almost all heat arriving on the continental shelf up to the ocean surface and away from the ice shelf by stirring heat anomalies along tilted density surfaces. These findings highlight the importance of understanding continental shelf eddies in the context of basal melt variability, for example, in driving future heat delivery to Antarctica’s largest ice shelves.

Abstract Understanding how global climate forcing shapes regional hydroclimate on decadal to millennial timescales is critical for predicting the response of continental ice sheets to future warming. Because instrumental observations capture only a short interval, longer-term perspectives require proxy records. Here we present a high-resolution sedimentary record covering the entire Holocene from southwestern Greenland. We show that after the demise of local glaciers in the Early Holocene, the record allows evaluation of millennial to decadal variations in local precipitation. Comparison with North Atlantic Oscillation (NAO) reconstructions and climate model simulations demonstrate that positive NAO phases correspond to wetter but colder regional conditions. These findings indicate that the observed centennial-scale patterns of ice mass balance during the latest Holocene were driven by precipitation rather than temperature. They further suggest that under future climate scenarios, a shift toward more positive and stable NAO phases could promote localized growth of Greenland Ice Sheet margins.

Nature, Published online: 19 August 2026; doi:10.1038/d41586-026-02376-w Since 2020, snowfall has largely counterbalanced the melting of the Antarctic ice sheet. The atmospheric effect that causes this has been identified — and it’s temporary.

To overcome time-step restrictions, we implement the Free-Surface Stabilisation Algorithm (FSSA) at the ice-ocean interface in Stokes ice-sheet simulations. In 2D experiments, a time step of 10 years is generally numerically stable and accurate, whereas a time step of 50 years is stable, but cannot fully capture grounding-line dynamics. Implementation at the ice-ocean interface increases the applicability of Stokes models and motivates future coupling with adaptive time-stepping schemes.

Sea level is a key indicator of ocean change, but it is hard to measure around Antarctica, where sea ice hides the ocean from satellites. Yet satellites flying over the ice still receive clear echoes from openings in the ice, where water is exposed. We reprocessed these echoes from four satellite missions and combined them with open-water data, building the longest sea level record for the whole Southern Ocean (2003 to 2021). It offers a new way to track how this rapidly changing region evolves.

Proglacial rivers and lakes act as critical meltwater links between the Greenland Ice Sheet (GrIS) and ocean; however, their spatial distribution and hydrographic characteristics remain unquantified at a large scale. In this study, we extract Greenland-wide proglacial water bodies and propose a new method for classifying them into three types: continuous river networks, isolated lakes, and pathway lakes. First, we combine Sentinel-2 imagery from the summer of 2020 with Copernicus digital elevation model to map proglacial water bodies guided by knowledge of water-landscape interactions. Second, we distinguish continuous river networks from isolated lakes, determine pathway lakes by identifying locations with abrupt width changes and excluding wide braided channels. Third, we analyze the hydrographic characteristics of the proglacial river networks and lakes (e.g. number, area, stream order, width, braiding index, and shape). Fourth, we estimate ice meltwater runoff drained into the ocean via the proglacial river networks using regional climate models. The results show that: (1) our method achieves an average Kappa coefficient of 0.91 ± 0.05 and maps a total area of 14,045.7 km2 of proglacial water bodies across Greenland in the summer of 2020, including 6,630 river networks (2,705.5 km2, accounting for 19.3% of the area), and 88,206 lakes (11,340.2 km2, 80.7%); (2) high-order (order ≥ 4) proglacial river networks account for less than 5% in number but over 60% of the total length, area, and river-routed ice meltwater runoff; (3) meltwater-fed rivers widen faster with the stream order and exhibit higher braiding indices compared to non-meltwater-fed rivers, while pathway lakes are generally larger than isolated lakes, both implying that meltwater exerts strong controls on proglacial hydrography; and (4) proglacial river networks route nearly half (~47%) of the ice meltwater runoff to the ocean in the 2020 summer, approximately to marine-terminating glaciers, while extreme imbalance among major basins is also revealed. Viewing collectively, this study presents a detailed high-resolution hydrography dataset of Greenland proglacial rivers and lakes, attempts to advance from open-water mapping to water type classification, and provides a potential comprehensive dataset for future investigations of Greenland proglacial hydrology and its interactions with the ice sheet and ocean.

Surface albedo is crucial for the polar energy balance. Blowing snow (BLSN) frequently perturbs coastal Antarctic albedo, yet its specific radiative effects and observability by medium-resolution satellites remain poorly understood. This study combines multi-source ground observations at Zhongshan Station with Sentinel-2 and MODIS imagery to evaluate the multi-scale albedo response to BLSN. The results show that when snowfall and BLSN occur simultaneously, they generally lead to spatially uniform surface brightening. Conversely, when BLSN occurs after snowfall has ended, strong winds drive snow erosion and redistribution, producing highly dispersed, interleaved patches of local positive and negative albedo changes. Scale-bridging analysis reveals that spatially uniform brightening is reliably captured by medium-resolution satellites. However, these highly dispersed local signals extensively cancel each other out during spatial averaging. In the evaluated events, this cancellation drives the net regional signal toward zero, masking a substantial local surface shortwave radiative forcing that, at solar noon, can reach tens to over a hundred W m−2. This study suggests that the observability of dynamic surface perturbations depends on their sub-pixel spatial structure. Consequently, a near-zero signal at medium resolutions does not guarantee an unchanged surface, highlighting a hidden cancellation effect that obscures critical local energy redistributions.

We created the first empirically calibrated high-resolution map of mountain permafrost distribution in mainland Chile. Using field evidence from rock glaciers and boreholes with climate and terrain data, we found that more than 8000 km², mainly in the dry Andes of northern and central Chile, offer suitable conditions. Permafrost affects water resources, slope stability, and infrastructure, and our model provides a baseline for planning and future research under climate change.

We examine how uncertainty in the initial position of sea ice features (leads, ridges), affects daily-to-weekly winter sea-ice forecasts. Using ensemble simulations with a sea ice–ocean model, we compare two formulations of sea ice mechanics. We show that pack-ice dynamics are highly sensitive to this choice: one formulation strongly amplifies small initial errors, while the other damps them. Our results highlight the need for ensemble forecasts to capture uncertainty and risks in the Arctic.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.

Paleosol climofunction-based methods for paleoclimate reconstruction are an important component of deep-time archives, but current proxies are subject to major gaps in terms of modern dataset quality and the suite of predicted climate parameters. Here we compile an expanded modern soil dataset and add soils from unique climate zones, pair these with updated high-resolution climate datasets, and apply novel partial-least squares and principle components methods to develop an improved predictive soil geochemistry paleoclimate model. This model provides more accurate prediction of common proxy response variables (temperature, precipitation) with lower root mean square predictive error than previous principle components-based methods, and introduces new proxy response variables associated with seasonality (range of temperature, growing season precipitation). We apply the model to key Cretaceous sites from the Western Interior Basin (Two Medicine and Kaiparowits formations) and show that the model predicts comparable mean annual temperature (~13℃) and precipitation (~600 mmyr-1) to other proxies, while also uniquely predicting mean annual range of temperature (~26℃) and growing season precipitation (~450 mmyr-1) that match regional climate model projections. Our model is also broadly available for new paleoclimate studies using open-source coding.

Nature, Published online: 19 August 2026; doi:10.1038/s41586-026-10912-x Recent Antarctic ice mass gain was linked to a recurrent atmospheric teleconnection, but this recurs about once per decade in observations and historical simulations so does not yet reflect a global-warming-driven moistening of Antarctica.

This study examines the impacts of wildfires on permafrost thaw in the Arctic spanning up to ~90 years post-fire and explores the potential of using radar satellite observations to enhance our understanding of environmental change. By assessing diverse high-latitude permafrost landscapes, we show that ground deformation anomalies after fires display similar patterns across regions, while radar backscatter varies depending on predominant ground temperatures.

Preprint under review for ESSD (discussion: open, 0 comments) We present the first 90 m resolution dataset of active layer moisture covering the entire permafrost region of the Qinghai–Tibet Plateau. It was produced from 342 field observations (2009–2024) using four machine learning models combined through bias-aware fusion, and is accompanied by an area-of-applicability mask and per-pixel uncertainty estimates. The dataset is freely available at https://doi.org/10.12072/ncdc.permafrost.db7312.2026.

Reconstructing past temperatures from sparse coral records is challenging. We developed two adaptive methods to improve how model simulations are combined with proxy data. The first adjusts trust in each record based on model consistency; the second adapts how far each record influences the reconstruction. Our approach improved accuracy and better captured El Niño–Southern Oscillation patterns, especially in data‑poor regions. These strategies enable more reliable climate reconstructions

Grounded icebergs anchor Antarctic sea ice and support marine ecosystems, yet their continent-wide distribution was previously unknown. Using satellite radar imagery and an automated artificial intelligence tool, we mapped nearly 40 000 stationary icebergs. We discovered that tiny, frequently overlooked icebergs actually dominate both the total number and area. This public dataset offers a crucial new baseline for modelling coastal ice stability and understanding broader environmental changes.

Water resources are crucial for the survival and development of human societies in arid Central Asia. However, our understanding of regional hydroclimate variability and moisture dynamics remains limited. Here, we present a precisely dated, high-resolution speleothem δ 18 O record from caves in northern Uzbekistan that reconstructs changes in precipitation δ 18 O over the last 47,000 years. During the glacial interval, relatively low speleothem δ 18 O values align with precipitation isotope records from high northern latitudes. The close correspondence between the Uzbekistan record and Greenland ice-core δ 18 O during Heinrich Stadial 1 indicates a strong temperature control on regional precipitation δ 18 O, although Dansgaard-Oeschger-scale variability is weakly expressed. During the Holocene, speleothem δ 18 O reached minimum values near 8 ka BP and then increased toward the present, broadly consistent with records from southern Uzbekistan and the Asian monsoon region, but with less coherent centennial- to millennial-scale variability. Supported by model simulations, we propose that during the glacial and mid-Holocene, when winter temperatures were low, the regional precipitation was dominated by northerly and northwesterly moisture transport linked to enhanced high-latitude temperature gradients and southward-shifted westerlies. In contrast, late Holocene hydroclimate increasingly reflected southerly to southwesterly cyclonic moisture transport from the eastern Mediterranean and the Near East, driven by northward-shifted westerlies. These results show that arid Central Asian hydroclimate and precipitation δ 18 O have been shaped by the position and strength of the westerlies through major reorganizations in moisture-source trajectories, with direct implications for understanding future water availability as subtropical and temperate circulation zones shift northward.

The Yellowstone geo-ecosystem has been the subject of much research, but the ecological history of the Yellowstone Plateau volcanic field and its iconic geyser basins is less known. In this investigation, paleoenvironmental analyses of sediment cores from lakes in Lower Geyser Basin were compared with regional records and paleoclimate model simulations to reconstruct the vegetation, wildfire, limnology, hydrothermal dynamics, and climate drivers since deglaciation, 15,000 to 14,000 years ago. Pollen data from Lower Geyser Basin lakes reveal the strong influence of infertile rhyolitic soils on vegetation history: an initial late-glacial steppe was replaced by lodgepole pine forest from 12.8 to 11.0 ka, with little change in forest composition or cover thereafter despite changing climate. This stability contrasts with the more dynamic vegetation response on nonrhyolite substrates in the Yellowstone region where nutrient and moisture availability is greater. Highest wildfire activity and low lake nutrient levels in Lower Geyser Basin occurred from 12 to 4 ka, when summers were substantially warmer and drier and fire-inducing vapor pressure deficits were 29 to 56% higher. The hydrothermal history, inferred from sedimentary arsenic and cesium abundances, was spatially and temporally variable but lake-forming hydrothermal events align with periods of abundant moisture. Thus, long-term changes in wildfire, limnology, and, to some extent, hydrothermal activity were governed by insolation-driven climate variations, whereas the vegetation response was muted and constrained by geologic processes. These findings suggest that warmer, drier conditions in the future could result in less hydrothermal activity yet little change in forest cover across the Yellowstone Plateau volcanic field despite more wildfires.

Science Advances, Volume 12, Issue 34, August 2026.

arXiv:2608.19008v1 Announce Type: new Abstract: Scenario planning for rare, high-impact events often requires massive ensembles to stochastically sample relevant trajectories. Although autoregressive weather emulators can efficiently generate such ensembles, isolating trajectories of interest requires sifting through petabytes of data, a challenge that grows exponentially with lead time and rarity. In contrast, a non-autoregressive foundation model like Climate in a Bottle video (cBottle-video) can directly sample trajectories terminating in extremes, avoiding large-ensemble search. We use cBottle-video to generate 1000-member ensembles with start- and/or end-conditioning across three extreme events---the 2021 Pacific Northwest (PNW) heatwave, Superstorm Sandy, and Hurricane Ian. Antecedent 500 hPa geopotential height ($z_{500}$) spread at the free end of end-conditioned ensembles reaches 84--89\% of the final-state spread of start-conditioned ensembles, revealing substantial diversity consistent with each extreme event. For the 2021 PNW heatwave, end-conditioned ensemble members begin uniformly warmer than reanalysis and stay warm, replacing the observed rapid intensification with persistent antecedent heat. For Superstorm Sandy, the leading modes of $z_{500}$ at the antecedent end of the end-conditioned ensemble explain 44\% of the variance in track latitude, and roughly 10\% of ensemble members begin as stronger hurricanes than Sandy. For Hurricane Ian, variation in the first landfall location among end-conditioned trajectories underscores the importance of accounting for intermediate hazard exposure in risk planning.

arXiv:2507.10916v2 Announce Type: replace Abstract: Snow cover on Antarctic sea ice is highly heterogeneous, yet climate models have historically simplified it to a uniform layer in winter and bare ice in summer, introducing uncertainty in simulated surface energy fluxes. Here, using shipborne, high-resolution, close-range imagery of the sea ice surface, visual observations, and concurrent atmospheric data from five expeditions across the Antarctic marginal ice zone (2019-2024), we show that fine-scale (sub-grid) snow cover heterogeneity persists across all seasons, is largely independent of ice concentration, and is governed by ice morphology and regional accumulation history. When the heterogeneity is incorporated into flux calculations, it introduces systematic biases in surface energy fluxes relative to uniform-surface assumptions. These findings, based on observations from two Southern Ocean sectors, suggest that the inclusion of sub-grid snow distribution should be considered for simulating Antarctic sea ice evolution and its broader polar climate feedback.

This study evaluates an updated version of the ALARO (ALADIN–AROME) regional climate model over Belgium at multiple resolutions, using long-term climate simulations. Coupling the model to the externalised surface scheme SURFEX (Surface Externalisée) and running at higher resolutions improved simulated temperature, precipitation, and precipitation extremes, supporting the value of high-resolution modelling for representing local climate extremes and informing adaptation.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

Sapropel S1 corresponds to the latest deoxygenation event that occurred in the Mediterranean Sea between ∼10.2 and 6.8 cal kyr BP. However, terrestrial deposition coeval with Mediterranean Sapropel S1 remains poorly understood. Through the integration of sedimentological, paleontological, geochemical (TOC, TN, δ 13 C, δ 15 N) and chronological data within a regional onshore-offshore framework, this study investigates the Upper Pleistocene-Holocene fill of the Biferno paleovalley system (Southern Italy) and characterizes the S1-equivalent (S1 eq ) succession. The results document a continuous S1 eq record comprising three stratigraphic intervals (S1a eq , S1break eq , and S1b eq ) that can be correlated with the corresponding offshore S1 phases. Facies evolution from poorly-drained floodplain to estuarine deposits reflects Holocene relative sea-level rise and paleovalley drowning. Bulk geochemical proxies display strong stratigraphic control, with δ 13 C and δ 15 N trends closely tracking facies shifts and Organic Matter (OM) sources. Onshore-offshore correlation and comparison with other Adriatic paleovalley systems reveal a marked decoupling in organic-carbon accumulation during the S1 break, consistent with the progressive landward migration of river mouths during the Holocene transgression. This evolution promoted enhanced terrigenous OM storage within paleovalleys while reducing its export toward deeper basin sectors. These findings indicate that paleovalley systems constitute key transient repositories of sediment and organic carbon during rapid transgression and provide a valuable terrestrial archive for reconstructing the evolution of the Adriatic source-to-sink system during Mediterranean Sapropel S1 deposition. Their stratigraphic record offers a critical link between continental and marine systems, providing new constraints on the environmental processes accompanying S1 deposition in the Adriatic and Mediterranean regions.

arXiv:2603.07227v3 Announce Type: replace Abstract: Estimating changes in extremes quantiles from environmental processes non-stationary in time, from small samples is challenging since it is difficult to characterise tail non-stationarity adequately. Using annual maxima and minima of near-surface temperature (tas) from CMIP6 output for some Earth desert regions, we use generalised extreme value (GEV) regression to model changes in extreme quantiles in time. We consider candidate models with different parametric forms for the variation of GEV parameters with time, estimating parameters using Bayesian inference. We select optimal candidate models using model selection criteria, including the Akaike, Bayesian, divergence and widely-applicable information criteria. In an extreme value setting, the performance of different criteria is unreliable. We therefore undertake a simulation study using ground truth models generating data similar to our extrema, to assess performance of the criteria to minimise error in prediction of change DeltaQ in the 100-year return value of tas over (2015,2125). The Bayesian information criterion (BIC) provides best performance, out-performing the divergence and widely-applicable Information criteria. We compare BIC model selection with a stacked Bayesian model average. Using BIC-selected GEV regression, we estimate joint posterior distributions of DeltaQ, coupled over three climate scenarios, for different combinations of desert region, global climate model and climate ensemble. We find significant increases in DeltaQ for regional annual maxima under stronger forcing scenarios for all desert regions. Similar but weaker trends are observed for regional annual minima. There is evidence that extreme minima are warming more rapidly than extreme maxima in Antarctica, whereas most hot desert regions exhibit the opposite effect. An ancillary file of supplementary material is provided.

arXiv:2605.29976v3 Announce Type: replace Abstract: We evaluate the climate simulation capabilities of ArchesWeather and ArchesWeatherGen, two machine learning models originally trained for weather forecasting and evaluated up to a 10-day lead time. ArchesWeather is a deterministic model, while ArchesWeatherGen is a probabilistic flow-matching model leveraging ArchesWeather's forecasts, enabling ensemble-based uncertainty quantification. In this work, we adapt these models to act as forced atmospheric models by using additional conditioning on the monthly mean sea surface temperature (SST) and sea ice cover (SIC) as boundary conditions. In particular, we follow the AI Model Intercomparison Project (AIMIP) Phase 1 protocol, which, analogous to the Atmospheric Model Intercomparison Project (AMIP), proposes a standardized experimental setup to evaluate the climate skill of ML-based forced atmospheric models. We present a comprehensive evaluation of both models under these conditions, including comparison against numerical climate models, ablation studies that examine key design choices in the extension, and an analysis of forced versus unforced configurations. Despite being originally developed for weather forecasting, we demonstrate that forced configurations of ArchesWeather and ArchesWeatherGen produce stable long-term climate simulations, have a stable annual cycle, and capture the drift of many climate variables. The models faithfully reproduce ERA5's climatology, large-scale circulations and interannual variability, and they capture the tails of the distributions.

Nature, Published online: 19 August 2026; doi:10.1038/s41586-026-10938-1 A 2,000-year Peruvian ice core record of atmospheric CH4 concentrations contains the first historical global CH4 record from the low latitudes, and implementation of a four-box model indicates higher equatorial CH4 strength than previous estimates from polar-only records.

Nature Communications, Published online: 19 August 2026; doi:10.1038/s41467-026-76772-1 Antarctic ice shelves face increasing collapse risk as environmental stresses intensify. Long-term weakening, ocean-induced thinning, and loss of stabilizing sea ice can drive disintegration yet are mostly missing from models which only consider surface melt-driven fracturing, limiting accuracy of projections.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.

Journal of Advances in Modeling Earth Systems, Volume 18, Issue 8, August 2026.

Understanding which parts of climate models cause uncertainty requires many large computer experiments. We developed a new workflow that greatly improves the speed and efficiency of these studies. It can analyse millions of model variations up to 25 times faster without losing accuracy, allowing scientists to explore uncertainty in more detail and make climate predictions more reliable.

Journal of Geophysical Research: Atmospheres, Volume 131, Issue 16, 28 August 2026.

arXiv:2608.22454v1 Announce Type: new Abstract: Current particle flow filters rely on Gaussian prior assumptions that fail to capture the non-Gaussian attractor structure of chaotic systems. This study proposes a Score-based Particle Flow Filter (Score-PFF) that replaces the parametric prior gradient with a neural network-learned score function via denoising score matching. This enables flexible characterization of multimodal, skewed, and complex prior distributions in chaotic dynamics. Pure prior adjustment experiments demonstrate correct gradient directions toward the attractor (26.9%-29.0% error reduction over Gaussian priors). Under linear observations, Score-PFF significantly outperforms both Gaussian PFF and EAKF (Cohen's d = 1.03 and 1.40), preserving non-Gaussian structure that EAKF progressively Gaussianizes. Under nonlinear observation operators, Score-PFF maintains robust performance with up to 60% RMSE reduction in strongly non-Gaussian regimes. On the 1000-dimensional Lorenz-96 system, Score-PFF achieves 49.5% RMSE reduction over PFF while reducing per-assimilation cost by replacing SVD-based covariance inversion with neural network inference. Score-PFF establishes a computationally tractable, non-Gaussian data assimilation framework suitable for high-dimensional geophysical systems.

Abstract This study investigates the characteristics of east Pacific (EPAC) easterly waves (EWs) in 10 ensemble members of the Community Earth System Model version 2 Large Ensemble (CESM2-LENS) under the current climate and their projected changes by the end of the 21st century under the Shared Socioeconomic Pathway 3–7.0 (SSP370) warming scenario. Under the current climate, the CESM2 ensemble mean produces realistic mean precipitation, horizontal winds, and EW-filtered vorticity variability over the EPAC, although the EW-filtered precipitation amplitude is underestimated relative to observations. CESM2 generally reproduces the distribution of EW tracks identified using 700-hPa curvature vorticity in the Caribbean and EPAC, along with the horizontal structure, amplitude, and propagation of EW disturbances. However, CESM2 EWs exhibit a more top-heavy vertical profile and a more pronounced eastward-tilted vertical velocity structure relative to ERA5, suggesting biases in convective vertical structure and diabatic heating processes. Under the future warming scenario, mean precipitation over the EPAC strengthens and shifts southward, while the increase in EW-filtered precipitation variability is modest. EW track density and anomalous vorticity amplitude decrease along the Central American coast but increase within the southwestern ITCZ region. Increased static stability with warming and weaker upper-level vertical velocity per unit precipitation largely explain the weakened midlevel vorticity in EWs. The weakened dynamical signal of the EWs may limit convective activity, thus resulting in only modest changes in EW precipitation amplitude, which should be treated with caution given the CESM2 biases in simulating diabatic heating processes that controls the evolution of EW moisture.

Nature, Published online: 18 August 2026; doi:10.1038/d41586-026-02577-3 Heatwaves expose weak links in food systems

Nature Geoscience, Published online: 17 August 2026; doi:10.1038/s41561-026-02057-3 The Labrador Sea plays a crucial role in transporting oxygen into the deep limb of the Atlantic Meridional Overturning Circulation, despite its limited contribution to density transformation, according to an analysis of 2 years of mooring data.

Nature Communications, Published online: 21 August 2026; doi:10.1038/s41467-026-77087-x Heatwaves reshape plant monoterpene emissions by preferentially increasing highly reactive acyclic isomers, altering atmospheric composition and increasing ozone and OH reactivity beyond changes in total monoterpene amounts

We studied how accessible and reliable the computer code behind major climate models has been over time. By reviewing different phases of the Coupled Model Intercomparison Project, we found improvements in transparency and coding practices, but also gaps that limit reproducibility. Our work suggests practical steps to make future climate research more open, traceable, and trustworthy for scientists and society.

by Carolina Estevez, Kelly Feke, Vivek Srikrishnan, Tony E. Wong Sea-level rise has large impacts on coastal areas. Many approaches to quantify the net benefits of regional or global coastal adaptation rely on strong assumptions about economically efficient decision-making and may neglect critical uncertainties about future extreme water levels and socioeconomic development. In particular, the deep and dynamic uncertainties associated with future sea levels complicate efforts to adapt coastal areas and infrastructure to account for sea-level rise. Robust decision-making (RDM) provides a way to identify adaptation strategies that perform well across uncertain sea-level futures. In this work, we quantified the impacts of an RDM approach on regional and global coastal adaptation using an economic regret, defined as the difference in cost of a strategy compared to the “optimal” outcome. We modeled decision-making using a regret-based criterion and computed the economic regret of each adaptation decision candidate, which we then compared to classical decision-making approaches. We found that the majority of coastal segments that changed strategies under regret-based criteria opted for a higher level of adaptation, primarily by expanding their retreat elevation. Although the total adaptation costs remained comparable to those under the cost-minimizing criterion, the use of regret led to a 3–50% reduction in flood damage, highlighting the advantages of prioritizing robust outcomes over purely cost optimization. While our analysis assumed immediate implementation of adaptation and evaluated a limited set of decision criteria, the results demonstrate that incorporating regret-based decision-making into coastal impact models can provide policy-relevant insights for designing more robust adaptation strategies under deep uncertainty.

The Younger Dryas (YD) terminated the last glaciation and initiated the Holocene, which was accompanied by widespread glacial advances. Traditional scaling models, production rates, and outlier screening approaches for 10Be exposure dating often yield inconsistent formation ages for the same moraine, hindering a comprehensive understanding of spatiotemporal patterns and climatic controls of YD glaciation across the Tibetan Plateau. In this study, we reprocessed 10Be exposure ages using the Probabilistic Cosmogenic Age Analysis Tool 2.2 (P-CAAT) to constrain YD moraine chronologies. Three glacial events were dated to 12.5 ka, 11.9 ka, and 11.5 ka. Glacier-climate simulations showcased cold dry conditions in the monsoon domain and cold-wet conditions in the westerly domain during the YD. The modelled temperature reductions matched global climatic records. Glacial advances were primarily forced by cooling in monsoon areas, whereas joint cooling and increased precipitation dominated in westerly regions. This study clarifies the temporal characteristics and driving mechanisms of YD glaciation on the plateau, though additional dating information is required for further verification.

arXiv:2512.13899v2 Announce Type: replace Abstract: Accurately forecasting winter precipitation type and its transitions is critical for high-impact decision making. However, existing methods struggle in thermodynamically ambiguous regimes, and most do not quantify forecast uncertainty from a single model run. We developed an evidential neural network that predicts calibrated probabilities for four winter precipitation types (rain, snow, freezing rain, and ice pellets) along with epistemic uncertainty estimates at the computational cost of a standard neural network. The model was trained on quality-controlled and curated observations from the crowd-sourced mPING dataset paired with vertical thermodynamic profiles from the NOAA Rapid Refresh model analyses. Rigorous physical quality control removed thermodynamically implausible reports. Bulk evaluation against held-out mPING observations from June 2020 through June 2022 shows the ML model outperforms area-based deterministic methods in success ratio for freezing rain and ice pellets while maintaining comparable or better performance for rain and snow. A reduced freezing rain probability of detection reflects genuinely ambiguous thermodynamic environments rather than a uniform model deficiency and is more robustly represented through the full probability distribution than through the dominant predicted class alone. Thermodynamic regime analysis demonstrates that model prediction errors are physically structured and concentrated in interpretable regions of diagnostic space consistent with the known difficulty of freezing rain and ice pellet discrimination. We further demonstrate the model's physical consistency and operational utility through two contrasting mid-western U.S. winter storm case studies and an interactive visualization tool that enables dynamic interrogation of model predictions and uncertainty in real time.

DSDP Site 568, off the west coast of Guatemala, was chosen to fill the palynological data gap in the Eastern Equatorial Pacific (EEP) region and improve tropical-to-subtropical regional paleoclimatic models of the Northern Hemisphere. This study analyses palynological assemblages and geochemical proxies, including Fe/Al, Ti/Al, and TOC, to evaluate precipitation and marine productivity. Statistical analysis of these variables indicates an oxic depositional environment with low marine primary productivity. The findings also show how dinoflagellate cyst assemblages (DCS) respond to global paleoceanographic and climatic events. These DCS are mainly composed of the phototrophic dinoflagellate taxa Polysphaeridium, Operculodinium , and Spiniferites , which are typical of subtropical and tropical settings. By combining DSDP data on diatoms, planktonic foraminifera, and calcareous nannofossils with dinoflagellate cyst data from this study, the depositional age is estimated from the Early Miocene (∼ 19 Ma) to the Early Pleistocene (> 0.62 Ma). A moderate rise in palynological assemblages is associated with the Middle Miocene Climatic Optimum. Lower TOC values and palynological concentrations were noted during the Middle-to-Late Miocene Carbonate Crash Event. Conversely, palynological concentrations increased somewhat during the Late Miocene-Early Pliocene Global Biogenic Bloom. Lastly, the Pliocene-Pleistocene Northern Hemisphere Glaciation is evident in the area through notable rises in TOC values and palynomorph concentrations.

The study proposes a diffusion-based framework for uncertainty quantification in wind speed nowcasting by learning forecast error distributions. By randomly generating errors and adding them to a physics-based wind nowcast, multiple forecast scenarios can be produced. The results improve forecast accuracy and provide reliable estimates of forecast uncertainty.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.

Earth's Future, Volume 14, Issue 8, August 2026.

Wildfire can significantly alter karst surface environments, yet its impact on cave systems and speleothem archives remains poorly understood. This study investigates the relationship between surface fire events and their imprint on inverted ages recorded in soil and flowstone in the Postojna (Slovenia) karst area. We present new radiocarbon ( 14 C) ages from charcoal preserved in a well-developed Luvisol and from a carbon-rich black layer enclosed within cave flowstone, complemented by U/Th dating of flowstone and micromineralogical analyses of soil horizons. Charcoal ages range from modern (110 years BP) to the Late Pleistocene (∼31,000 years BP). Reversed age sequences are observed in the deepest Bt3 soil horizon and in the cave flowstone and are linked to downslope movement triggered by a major palaeofire around 92,000 years BP. This process transported older (∼31,000 years BP) charcoal downslope, hindered soil respiration, and altered meteoric-water infiltration, resulting in anomalously high dead carbon fractions (DCF ≈ 40–90%). δ 13 C signatures support shifts between open Late Pleistocene steppe-tundra and denser Holocene forest vegetation, while highly depleted δ 13 C values in the flowstone black layer reflect soot inputs from high-temperature combustion. Evidence of magnetite, maghemite, and mullite confined to the Bt3 horizon confirms high-temperature wildfire. Integrating soil, cave, and mineralogical data demonstrates that major wildfire events can generate synchronous surface–subsurface signals, affect sediment redistribution on karst slopes, and produce identifiable fire horizons within speleothems. The study highlights the potential of combined soil charcoal and speleothem archives to reconstruct palaeofire activity and landscape evolution in karst environments.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.

Journal of Advances in Modeling Earth Systems, Volume 18, Issue 8, August 2026.

Resolving the origin and flowpaths of meltwaters in Arctic glacierized catchments, beyond the traditional atmospheric versus water-rock framework, is crucial for understanding how weathering processes evolve with glacier retreat and regulate solute export. We investigated the hydrochemistry and isotopic composition of glacial meltwaters and proglacial streams in the Kongsfjorden region (Svalbard), integrating major ions and stable water isotopes with coupled 87 Sr/ 86 Sr and δ 11 B analyses during the early melt season. Major ion systematics reveal a dominant marine aerosol imprint, indicating that early meltwater chemistry is largely controlled by snowpack flushing of sea-spray deposits accumulated in the winter snowpack. Strong Na–Cl correlations and B/Na ratios close to those of seawater support this interpretation. Nevertheless, sea-salt corrections and isotopic tracers demonstrate measurable crustal contributions. Radiogenic 87 Sr/ 86 Sr values (>0.720 in the Midtre Lovénbreen and Kongsvegen basins) cannot be explained by marine input or carbonate dissolution alone and instead require interaction with rubidium (Rb)-rich siliciclastic or metamorphic lithologies. Boron isotopes further constrain the relative contribution of these sources, with estimated marine B fractions ranging from approximately 30%–80% across catchments. Coupled 87 Sr/ 86 Sr–δ 11 B isotope systematics identify mixing between a marine aerosol component and at least two crustal end-members (carbonate and metamorphic/siliciclastic), highlighting glacier-specific differences linked to hydrological routing and local geology. Our results show that the combined use of Sr and B isotopes provides a powerful geochemical framework to disentangle atmospheric marine inputs from crustal weathering processes in Arctic glacial meltwaters. They further indicate that, even under early melt-season conditions dominated by snowpack flushing, water–rock interaction still modulates meltwater chemistry. This study represents the first application of δ 11 B isotopes to Arctic glacial and proglacial waters and one of the few investigations employing Sr isotopes in Arctic continental meltwater systems. The integrated 87 Sr/ 86 Sr–δ 11 B framework provides a robust isotopic baseline for assessing evolving weathering regimes and solute export pathways in rapidly changing polar environments.

Abstract. Living and fossil benthic foraminifera are widely used as bioindicators of ecosystem quality, but their analysis is time-consuming. With the help of artificial intelligence, models for the automatic identification of fossil planktonic foraminifera and, more recently, benthic foraminifera are emerging. Although large datasets of images can now be acquired automatically, efficient processing pipelines are still needed. Here we use convolutional neural networks (CNNs) to automatically identify living benthic foraminifera at the species level. Rose Bengal staining is widely used in ecological and biomonitoring studies to identify living foraminifera, which reflect present environmental conditions. We investigate both low-diversity assemblages from an intertidal mudflat on the French Atlantic coast (Bourgneuf Bay) and high-diversity assemblages from the French Mediterranean coast. Samples were imaged using a modified 3D printer equipped with a camera. The three trained CNN models are efficient in identifying species of the total community (i.e., living and dead foraminifera together) in both low- and high-diversity settings with an accuracy of 96.0 % and 82.2 %, respectively. The results compare well to published indices for ecosystem quality in high-diversity ecosystems. One CNN model can distinguish both species and vital status (i.e., living and dead specimens) for the low-diversity spot with an accuracy of 94.2 %. The results of this first application of machine learning for identifying living benthic foraminifera are very encouraging for improving the efficiency and applicability of these bio-indicators in biomonitoring studies. This approach highlights the potential of automated image-based identification to accelerate the use of benthic foraminifera as bioindicators in large-scale environmental monitoring programs.

Science Advances, Volume 12, Issue 34, August 2026.

This research aims to improve dry-snow slab avalanche forecasts, particularly the propagation of cracks in weak layers, as current models rely on uncertain and difficult-to-measure parameters. We combined snowpack simulations with field measurements to link fracture energy to measurable and modelable snow properties. This new approach better reproduces observed crack lengths and allows weak layers to be tracked throughout the season, thereby improving operational avalanche risk assessment.

Preprint under review for ESSD (discussion: open, 0 comments) Closing the sea‑level budget on annual and longer time scales is a cornerstone of physical oceanography because sea‑level rise is one of the best indicators of climate change, and a closed budget shows we have identified and quantified all major drivers. We examined it from 1993 to 2022, finding an accelerated rise that matched ice melt and warm water until 2015. Afterwards an unexplained gap appears. Better deep‑ocean observations and refined gravity processing are needed to close the budget.

Abstract. On 20–21 June 2024, an unprecedented flood of the Etançons river caused important damage to the village of La Bérarde (Écrins, France). An analysis of the event showed that the flood was caused by the combination of an intense rain-on-snow event at high altitude and the subglacial drainage of a supraglacial lake from Glacier de Bonne Pierre. In this study, we quantify the water volume that could have also been trapped beneath the glacier in local minima of the hydraulic head, i.e., in locations that could host so-called glacier water pockets impounded by hydraulic barriers. In the absence of direct observations of water pockets, we use a numerical, steady-state approach that computes the subglacial hydraulic head from surface and bedrock topography of Glacier de Bonne Pierre. As of June 2024, hydraulic barriers at Glacier de Bonne Pierre could, in theory, have impounded water volumes on the order of 105 m3, with the largest modeled water pocket beneath a surface depression that temporarily hosted a supraglacial lake. These results provide a first-order estimate of the potential subglacial water storage capacity prior to the June 2024 flood. We propagate uncertainties in surface elevation, bedrock elevation, and flotation fraction (the ratio of basal water pressure to ice overburden pressure) through a stochastic framework and show that spatial variability in the flotation fraction dominates the uncertainty in the resultant water-pocket volume. This highlights the strong sensitivity of subglacial water-routing results to poorly constrained basal water pressure conditions. While acknowledging that the actual presence and contribution of such water pockets cannot be confirmed from available observations, our study highlights the glacial flood potential of debris-covered glaciers with pronounced surface topographic depressions, which can promote both supraglacial and subglacial water storage.

The article describes the implementation of the nitrogen cycle in the land surface model ISBA (Interaction-Soil-Biosphere-Atmosphere). The model is evaluated using Free Air CO2 Enrichment experiments. A comparison with a multi-model analysis shows that the nitrogen model version performs better than the carbon-only version, notably a reduced sensitivity to elevated CO2, and smaller C stocks. We also present a detailed analysis of the simulated N dynamics in the soil.

Ocean circulation plays a key role in regulating atmospheric carbon dioxide levels during Pleistocene glacial-interglacial cycles. However, spatial responses of the oceans remain poorly constrained because reliable records from important carbon exchange regions, such as the equatorial Indian Ocean, are scarce. Here we show environmental magnetic records from the western equatorial Indian Ocean spanning the past 200,000 years. The magnetic signals primarily reflect variations in magnetofossils, the magnetic remains of magnetotactic bacteria, with higher abundances during interglacials. Combined with geochemical indicators, these records reveal enhanced productivity associated with increased nutrient availability during warm periods. Productivity variations closely follow glacial-interglacial cycles and reflect changes in deep ocean circulation. Glacial carbon burial resulted from reduced productivity caused by expanded Antarctic Bottom Water and enhanced stratification, which suppressed the upward transport of deep-sourced nutrients and carbon dioxide. In contrast, weakened Antarctic Bottom Water influence during warm interglacials strengthened deep-surface exchange, enhancing productivity and contributing to carbon dioxide release. Deep circulation changes driven by Antarctic Bottom Water regulate equatorial Indian Ocean productivity and carbon cycling across glacial-interglacial cycles over the past 200,000 years, as shown using environmental magnetic records from the western equatorial Indian Ocean.

arXiv:2608.23197v1 Announce Type: new Abstract: In this article, we present extreme value statistics of temperature extremes over continental France with a focus on their implications for electrical power infrastructure. These are obtained by fitting a non-stationary generalised extreme value distribution using a Bayesian setup, which also provides errors or uncertainty values for all estimates. Within this method, a combination of simulated data from a collection of 28 CMIP6 models and measured records from the E-OBS dataset is used. The investigated climate scenarios are SSP2-4.5, SSP3-7.0 and SSP5-8.5, considering both historical and future climates spanning the years from 1850 to 2099. The method provides full spatial resolution on a 0.25 degree grid, allowing to asses extreme temperatures at arbitrary locations. Leveraging this particular aspect, various maps revealing the spatial structure of annual maximum high temperature extremes over continental France for the median and the statistical upper bound are shown together with summary information for the administrative regions in France towards the end of the century. It is statistically possible that the south west region of Occitanie could reach temperatures of up to 57{\deg}C under a high emission scenario by 2080, which is the highest compared to all other regions in continental France. We also perform a comparison to the reference climate adaptation trajectory for France (TRACC - Trajectoire de r\'echauffement de R\'ef\'erence pour l'Adaptation au Changement Climatique), showing that it potentially underestimates the stated maximum temperatures which could be surpassed by up to +8{\deg}C. Furthermore, five reference electrical power infrastructure locations are investigated on how they are potentially affected by temperature extremes with the quantified intensities.

arXiv:2608.24651v1 Announce Type: new Abstract: We introduce AICON, a global machine learning weather prediction (MLWP) model which generates forecasts at 13 km spatial resolution with a 3-hour time step, trained on the high-resolution, non-hydrostatic ICON-DREAM dataset. AICON is in full operational use at Deutscher Wetterdienst since 2nd of March 2026. The model employs a graph neural network (GNN) architecture with an encoder-processor-decoder structure, where node updates are performed using a graph attention mechanism. A key feature of AICON is its use of an icosahedral multi-mesh derived from the native grid of the ICON model, ensuring consistency with the training data. ICON's terrain-following vertical SLEVE coordinate is one of the major distinctions from existing emulators. AICON's training strategy prioritizes small-scale fidelity by avoiding autoregressive multi-step rollout and longer forecast horizons during training, a design choice motivated by the hypothesis that this approach preserves fine-scale features often damped in models optimized for longer-range forecasts. We describe the prognostic and diagnostic variables used for training, the transfer learning protocol employed to accelerate convergence, and the model's performance across a range of evaluation metrics. An extensive evaluation, including routine verification against observation, a tropical cyclone case and spectral analysis reveal the strengths and limitations in the representation of atmospheric variability across scales. Routine verification against observations demonstrates competitive skill relative to the operational ICON model, particularly for near-surface variables in the short to medium forecast range.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.

Preprint under review for ESSD (discussion: open, 0 comments) We present a global hazard-oriented dataset of storm surges and significant wave heights, including bulk and spectral peak wave parameters, spanning 1950–2023 on a high-resolution in coastal areas. Significant wave heights are bias-corrected using satellite altimetry, while storm surges are derived from sea surface heights through frequency-based filtering. The dataset is tailor made for regional downscaling and assessment of compound-events for multi hazard mapping.

by Alexander R. Kaye, Lantao Sun, William S. Hart, James W. Hurrell, Michael J. Tildesley, Robin N. Thompson Climate factors such as temperature and rainfall influence vector-borne disease dynamics. Consequently, climate change is affecting the global distribution of diseases such as dengue. Natural climate variability (NCV), which adds noise to the climate signal, alters the future climate trajectory, but is rarely analysed in climate-health analyses. In this study, extending a published model of climate suitability for Ae. aegypti , we consider a climate-sensitive dengue transmission model. Using 100 climate projections up to 2100 from the Community Earth System Model (a climate model including climate change and NCV), each run under the same shared socioeconomic pathway scenario, we generate 100 equally plausible simulations of the future basic reproduction number of dengue globally and the population at risk. We quantify the difference in transmission suitability between the most-suitable and least-suitable projections (i.e., uncertainty due to NCV) and demonstrate that NCV affects future suitability for transmission, dominating epidemiological parameter uncertainty in many locations. While the global population at risk from dengue (and, as we show, other vector-borne diseases) is expected to increase, NCV affects its precise value. Our findings demonstrate that NCV should be routinely incorporated into climate-sensitive epidemiological projections, and a key output of our research is the provision of a methodological framework for this. Accounting for NCV will enable public health policy decisions to be informed by the range of possible future outcomes.

Proceedings of the National Academy of Sciences, Volume 123, Issue 33, August 2026. SignificanceOn orbital timescales, terrestrial rainfall and Arabian Sea upwelling records of South Asian summer monsoon (SASM) evolve largely out-of-phase, with the former and latter in-phase with Northern Hemisphere (NH) and Southern Hemisphere (SH) ...

arXiv:2607.02213v2 Announce Type: replace Abstract: Midlatitude storm tracks play a central role in Earth's climate by transporting heat from low to high latitudes. Their strength is therefore set by the meridional temperature gradients, traditionally thought to be maintained solely by differential solar heating. This framework predicts a substantial seasonal reduction in storm activity, as meridional insolation gradients vanish during summer. Yet in the Southern Hemisphere, storm activity decreases by only 30% from its seasonal maximum. Here, we show that cloud radiative effects are essential for the seasonal maintenance of storm activity by reinforcing the temperature gradients that sustain it. Satellite observations reveal that sunlight reflected by midlatitude clouds in early summer creates a substantial meridional gradient in surface heating, despite the nearly uniform summer insolation. Idealized aquaplanet simulations then show that shortwave cloud radiative effects reinforce meridional sea-surface temperature gradients, thereby strengthening storm activity primarily during late summer and autumn, while longwave cloud effects partly offset this response. To interpret these results, we develop a simple theoretical model linking storms, clouds, and sea-surface temperature gradients. The model reproduces the simulated seasonal response and identifies two emergent cloud properties that control the feedback strength: the maximum attainable cloud albedo and the sensitivity of cloud cover to storm activity. Together, these findings indicate that cloud radiative feedbacks are key to maintaining the thermal gradients that sustain storm activity. More broadly, they reveal a strong coupling among storms, clouds, and the ocean spanning distinct spatial and temporal scales.

Why rivers have not deeply incised the interior of the Tibetan Plateau over geologic time scales remains an open question. Resolving this issue requires clarifying and quantifying the respective influences of different geological factors on headward river erosion. Focusing on the Eastern Himalayan syntaxis, we reveal that glacial damming occurred frequently along an ∼70 km reach of the Yarlung Tsangpo River, based on 10Be exposure dating and paleoglacial modeling. Glacial-damming episodes limited headward erosion via a cycle in which prolonged damming inhibited erosion during glacial periods and short-lived damming delayed it during interglacial periods. Moreover, these damming processes triggered extensive aggradation upstream and protected bedrock from erosion over long time scales, while they enhanced downstream incision, reshaping the longitudinal profile of the Yarlung Tsangpo River. Repeated glacial-damming processes cumulatively raised the effective local base level by ∼300 m since the middle Pleistocene. Overall, we show that the coupling of glaciation and tectonics is the key factor maintaining the long-term stability of the southeastern Tibetan Plateau margin.

by Katherine E. Dale, Evie S. Brahmstedt, Joshua Culpepper, Kyle A. Emery, Alexander W. Mott, Walter A. Rich, Xinyu Sun, Audrey Thellman, Tianfei Xue, Atefeh Hosseini, Hadis Miraly Two types of extremes are often used as indicators of future change: Extreme climate events and extreme habitats. These extremes can be useful gauges of environmental change, but only if biological, spatial, and temporal scopes are aligned between the extreme and the future change it is hypothesized to mirror. In general, short-lived extreme events provide insights towards acute responses but are not as suitable for assessing long-term ecological or evolutionary change. Extreme habitats can reveal potential long-term adaptations, but community structure and trophic interactions can be challenging to assess due to species often following habitat-dependent evolutionary paths, and these environments tend to undergo gradual rather than rapid ecological change. Community structure and species interactions may further differ for both types of extremes between biological response levels. In cases of a mismatch between the size of an extreme habitat or event and its comparative system, it will be more appropriate to study biological responses at smaller geographic scales than the habitat or event occurs. Here, we explore the opportunities and limitations of a variety of extremes in marine and freshwater systems.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.

The paleoenvironment and centennial-scale climate fluctuations during the late Holocene in Lake Nakaumi were inferred through ostracod and geochemical analyses of a new core NKU23-01 obtained from the central portion of the lake. The paleoenvironment was divided into five intervals based on Q-mode cluster analyses of ostracod assemblages and Q-mode factor analyses of ITRAX results. Lake Nakaumi was an enclosed bay with seaweed beds and rich in organic matter before 3.9 cal kyr BP. Then, it became a relatively open bay affected by strong marine influence during 3.9–3.6 cal kyr BP. The open bay contained rich organic matter and was affected by inflow from the open sea during 3.6–2.6 cal kyr BP. The terrestrial influence increased during 2.6–0.8 cal kyr BP. Thereafter, it became an enclosed brackish lagoon after 0.8 cal kyr BP. Three strong marine influence periods, which may have been associated with the rise in sea levels, occurred at approximately 4–3.5, 1.8–1.6, and 0.3–0.1 cal kyr BP, respectively. Seven low Ti periods are identified in the standardized values of Ti, which occurred at approximately 0.2, 0.8, 1.2, 2.0, 2.7, 3.5, and 3.9 cal kyr BP, which may reflect the weak East Asian Winter Monsoon period of the Late Holocene.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.

Abstract In cold and arid regions, climate change has altered precipitation patterns and accelerated glacierized retreat, which imposes severe risks on the stability of regional hydrological systems. Combined with the strong interactions between surface water (SW) and groundwater (GW) regulated by distinctive geological conditions, clarifying the baseline hydrological regime and predicting future hydrological changes is critical to maintaining local ecosystem stability. The headwater catchment of the Bortala River in Northwest China represents a typical glacierized watershed dominated by spring-fed discharge. Under the continuous influence of climate change, quantitative investigations into surface water-groundwater interactions in this catchment are still inadequate. In this study, a coupled SWAT-MODFLOW model embedded with a glacier module was developed to clarify the surface water-groundwater interactions relationship. Based on the Budyko framework, under CMIP6 climate scenarios, to establish an optimized decomposition and reconstruction simulation system, furthermore, to explore the changing characteristics of temperature, precipitation, and runoff. Observed hydrometeorological parameters were utilized to calibrate and validate the established model. A set of statistical indices was employed to assess the modeling performance, and the coupled model attained favorable simulation accuracy. The main results indicated that although surface water served as the dominant recharge source for groundwater from 1971 to 2020, representing a cumulative reduction of 26.7% from the initial value of the historical period, and an increased average annual decline rate of 0.45%. Future scenario projections revealed an increasing trend in total watershed runoff under both SSP2-4.5 and SSP5-8.5, with growth rates of 13.6% and 14.2%, respectively. A runoff inflection point is projected to occur around the 2070s (± 10 years), which will appear approximately 10 years earlier under the high-emission SSP5-8.5 scenario. In addition, intensified precipitation extremes may trigger severe hydrological risks, which are expected to concentrate around the 2040s and 2070s.

The growing population, food security, and climate change pose future threats to the sustainability of water resources around the globe. Therefore, projecting future water consumption is crucial for decision-making in water governance, policy reforms, and management planning. In this study, we estimated Colombia's current (2005–2016) and future (2031–2090) water demand under climate change Representative Concentration Pathways (RCP2.6 and RCP8.5) using CORDEX data, and evaluated alternative, spatially-explicit water-saving strategies based on a temporal water footprint assessment approach. The results show that the average harvested area is 3.65 million hectares. In this area, coffee takes the largest share, followed by corn, plantain, and rice. Sugarcane showed the highest average yield (19.9 t ha⁻¹), followed by potato (16.0 t ha⁻¹) and yucca (10.5 t ha⁻¹), highlighting substantial differences in crop productivity. Currently, the total water consumption is 27 km³ /year, consisting of 86% green water (precipitation) and 14% blue water (irrigation water from surface and/or groundwater sources). Oil palm, plantain, and coffee are highly water-intensive crops that account for 80% of total blue water use. Future projections under climate change scenarios indicate that green water consumption could rise slightly, while blue water consumption could increase by up to + 26% in RCP2.6 and up to + 124% in RCP8.5. This increase will be gradual in the mid-term (2031–2060) and will become extreme in the late century (2061–2090). However, the future rise in blue water demand could be reduced by up to −78% by replacing water-intensive crops with crops using less blue water per unit area. A further reduction of up to −91% is possible through the combined implementation of improved irrigation technologies and optimized cropping patterns, rather than traditional surface irrigation practices.

Relatively warm and wet conditions prevailed in the southwestern Mediterranean during the Early–Middle Holocene transition, as indicated by multiple paleoclimate archives. However, the impact of the global 8.2 ka cooling event, which may have disrupted such climatic stability, remains poorly documented in the southern Iberian Peninsula. Here, we combine the stable isotope composition of gypsum (δ 17 O, δ 18 O and δ 2 H) and coeval biogenic carbonate (δ 18 O) in sediments from Medina Lake to reconstruct the hydroclimate evolution in southwestern Iberia from 9.5 to 7.0 cal ka BP, including the 8.2 ka event. By applying an isotope mass-balance model, we found that the lake functioned as a (semi-)permanent water body through this period, except for a short-lived interval at 8.3–8.1 cal ka BP. At that time, the lake shifted to a more ephemeral state with a reduced hydroperiod. This transition was associated with decreased precipitation and lower atmospheric relative humidity. Paleotemperature estimates obtained from tandem measurements of δ 18 O in gypsum hydration water and ostracod shells reveal that warm-season lake water temperatures at Medina Lake during the 8.1–8.3 cal ka BP interval (25.2 ± 3.1 ° C) were comparable to current instrumental summer lake water temperatures (27.2 ± 4.6 ° C) in short-hydroperiod lakes of southern Iberia. This suggests limited or no significant summer cooling in response to the 8.2 ka event compared to present-day conditions. Overall, our results indicate that the 8.2 ka event in southern Iberia was primarily characterized by enhanced warm-season aridity rather than pronounced cooling, in agreement with other paleoclimate records from the western Mediterranean.

arXiv:2608.18712v1 Announce Type: new Abstract: Detection and attribution studies typically rely on free running climate models with modified anthropogenic forcings, yet they fail to reproduce key observed decadal trends. Despite reduced biases, higher predictive skill, large ensembles, and improved trends through frequent reinitialization, seasonal forecast models have not been used for this purpose. We introduce a new set of counterfactual seasonal hindcasts using the ECMWF coupled seasonal forecasting system based on alternative forcing scenarios by modifying atmospheric forcing and ocean and sea ice initial conditions. An observation based estimate of the forced ocean temperature signal is derived to amplify or remove this signal from the ocean initial conditions. Retrospective forecasts include a control configuration together with enhanced and reduced forcing experiments in which either CO2 increases together with doubled anthropogenic ocean warming or CO2 is fixed and the anthropogenic ocean-warming signal is removed. Additional experiments isolate aerosol forcing. The counterfactual hindcasts substantially alter long term temperature trends while largely preserving seasonal prediction skill, interannual variability, and model drift. Enhanced forcing strengthens several observed climate trends underestimated by the control, including top of atmosphere radiative fluxes and aspects of tropical Pacific atmospheric circulation. However, the coupled atmosphere ocean system fails to sustain the strengthened tropical Pacific temperature gradient through Bjerknes feedbacks, suggesting a fundamental model limitation. Aerosol forcing has little impact, likely because indirect aerosol effects are omitted. These results establish counterfactual seasonal hindcasts as a powerful tool for dynamic attribution and diagnosing model deficiencies in responses to anthropogenic climate forcing.

Preprint under review for ESSD (discussion: open, 0 comments) This study provides a new winter Arctic sea-ice dataset of penetration-aware emission properties from satellite microwave observations for 2012–2025. It offers physically interpretable information on how snow-covered sea ice emits microwave radiation, helping to describe the Arctic sea-ice environment as a multivariate geophysical system. The dataset can serve as reusable geophysical information for characterizing sea-ice variability and supporting remote sensing and modelling applications.

Climate-driven glacier retreat is reorganizing sediment cascades across Alpine catchments. However, what remains poorly quantified is how their structural connectivity—the spatial configuration of landforms and pathways along which sediment can be transferred—evolves over deglaciation timescales. We reconstruct structural connectivity evolution in three glaciated catchments of the European Alps (i.e. Grastal, Kaunertal, Martelltal) over ~150 years since the Little Ice Age, using graph-based analysis of sediment cascade networks in which landforms are nodes and sediment pathways directed edges. Graphs were derived from multitemporal geomorphological maps combining historical maps, terrestrial photographs, Structure-from-Motion multi-view stereo (SfM-MVS) photogrammetry, and Airborne Laser Scanning (ALS)–derived DEMs, extending reconstruction to the 1870s–1890s. Despite similar deglaciation forcing, the three catchments followed divergent trajectories, with sources, links, and barriers evolving along site-specific rather than common paths. Changes were non-linear and episodic: glacier recession produced threshold-like reorganizations at Kaunertal and Grastal near the turn of the millennium, once retreating glaciers crossed topologically significant boundaries. Barrier positioning, not abundance, controlled system-scale filtering: a single strategically located lake in Grastal intercepted more sediment pathways than 13 lakes in Martelltal. Event-based routing showed connectivity increased in Kaunertal but declined in the other two via different mechanisms, indicating that network restructuring can outweigh source expansion in setting potential sediment delivery.

Operational marine ecosystem forecasts traditionally rely on combining phytoplankton chlorophyll observations with model forecasts. However, using our newly developed ensemble data assimilation system, we demonstrate that assimilating phytoplankton carbon data leads to more balanced phytoplankton estimates and adjusts estimates of global ocean carbon.

arXiv:2608.17986v1 Announce Type: new Abstract: Regional climate zones are expected to shift under global warming. Whether AI climate models have learned to generalize climate-zone distributions under warming in a physically meaningful way affects their suitability for climate projection. We address this question by applying a K\"oppen-Geiger climate-zone decomposition to AIMIP Phase 1 models under prescribed +4K SST forcing and comparing their responses to physics-based AMIP models. Using this diagnostic, we compare baseline classification skill, per-zone responses in temperature, precipitation, and near-surface specific humidity, and the spatial structure of departures from physics-based models. All AI models considered reproduce the 1979-2014 ERA5 climatology within the physics-based models' range, but only the hybrid physics-AI model NeuralGCM-HRD reorganizes zones in agreement with established thermodynamic and hydrological scaling relations. The remaining emulators have distinct failure modes traceable to their architectural treatment of land cells. A physically consistent climate-zone response is therefore necessary for AI models intended for climate projection.

Abstract Mediterranean hurricanes, or “Medicanes,” are rare, tropical-like cyclones producing intense rainfall and posing significant hazards to coastal populations. This study estimates medicane precipitation hazards using a statistical-deterministic approach that generates synthetic tracks from reanalysis data and global climate models. Coupling these tracks with the Tropical Cyclone Rainfall (TCR) algorithm, the precipitation field for each synthetic event is calculated, enabling spatially and temporally resolved hazard assessments. Validation against ERA5, satellite products, and surface observations confirms the model’s performance, particularly in coastal regions, despite a localized rainfall underestimation in the Central Mediterranean Sea. Sensitivity analyses (via track sub-sampling and TCR-ERA5 coupling) reveal that this discrepancy is not driven by sample size or the rainfall algorithm, but likely by synthetic track-generation or input data limitations. Under the RCP8.5 scenario, projected future changes indicate a marked increase in rainfall extremes: 100-year return period rainfall is expected to increase by up to 140 mm along the Adriatic coast, 250-year events by up to 160 mm, and 500-year events may exceed 800 mm total rainfall in some areas. In particular, regions such as southern Italy, northern Algeria, Sardinia, and Corsica are expected to see an increase of 100–160 mm in extreme Medicane-induced rainfall. Conversely, no substantial change is projected over southern Greece. This methodology offers a robust framework for quantifying hydrological impacts of low-frequency, high-impact storms in the Mediterranean. Our findings underscore the potential for more destructive rainfall events under climate change, highlighting the urgent need for enhanced monitoring, improved local observational networks, and targeted adaptation strategies.

Nature Climate Change, Published online: 19 August 2026; doi:10.1038/s41558-026-02717-7 Adaptation to climate risks requires integrating knowledge across IPCC working groups. This study presents a climate impact taxonomy that connects climatic impact-drivers from Working Group I to representative key risks from Working Group II and provides more direct guidance for risk assessment and adaptation strategies.

Proceedings of the National Academy of Sciences, Volume 123, Issue 33, August 2026. SignificanceIrreversible cellular processes play critical roles in diverse biological phenomena, including normal cell differentiation and pathological progression of mutated cells. These processes are driven by a select set of feedback loops within the ...

Climate change is progressively eroding the operational viability of ski tourism, while measures intended to sustain winter seasons can generate new vulnerabilities. This study examines how technical interventions, infrastructure reinforcement, and activity-expansion strategies shape adaptation and maladaptation risk in Japan's ski sector. We develop an exploratory Snow Reliability Line (SRL) screening framework that links elevation-based snow-reliability exposure, publicly documented resort actions, and maladaptation coding. The exposure framework is calibrated with Japan Meteorological Agency (JMA) 1991–2020 station climatology and tested through multi-parameter sensitivity analysis. It combines Alpine SRL research with a Japan-oriented regional adjustment and a document-based assessment of adaptation and maladaptation pathways. The maladaptation evaluation relies exclusively on conservative, document-based policy analysis of publicly available resort materials. The exploratory classification reveals a spatial and operational divide. High-elevation and northern resorts retain relatively stable winter conditions, whereas low- to mid-elevation sites are more frequently classified as snowmaking-dependent or critically marginal. Maladaptation risks are concentrated in the A1-B1-Dead and A1-B1-C1 pathways, where activity expansion, extended operations, accessibility incentives, and fixed winter infrastructure are associated with higher aggregate energy demand, opportunity costs, and path dependence. The findings show why ski-destination adaptation should be evaluated together with mitigation, anticipatory governance, investment flexibility, and year-round carbon accounting. Among the 14 resorts in the two highest-risk pathways, six show no documented transition planning, indicating potential continuity-driven maladaptive exposure distinct from action-driven lock-in.

Central Asia is a global climate-change hotspot where future water availability remains uncertain, as hydrological forcing data are ambiguous and observations provide limited constraints on model parameters. We reassessed future runoff changes in the glacierized Ala-Archa basin by reducing model equifinality with a multi-objective calibration strategy that explicitly incorporates winter low-flow runoff, total runoff, snow cover fraction, and glacier mass balance. The calibrated model was used to project annual and seasonal river discharge under four Shared Socioeconomic Pathways. Results indicate a transition from a nival–glacial to a more pluvial-driven runoff regime. Rising temperatures advance the melt season, increase the proportion of liquid precipitation, and shift peak runoff from July to June. Although glacier and snowmelt contributions decline substantially as cryospheric storage shrinks, total annual runoff decreases only modestly by the late 21st century. This muted annual response reflects compensating effects from increased rainfall and a sustained rise in summer baseflow. Improved low-flow calibration suggests that increased groundwater recharge can partly offset meltwater losses, leading to a less severe summer discharge reduction than previously reported. Neglecting low-flow constraints may therefore overestimate future water scarcity in glacier-fed catchments.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.

Understanding of the local effects of sea-ice concentration variations on the Arctic atmosphere is a prerequisite for assessing the role of Arctic sea-ice decline in the climate system, including its influence on mid-latitudes. In our study, using data from atmospheric reanalysis, we present how the relationships of sea-ice concentration, temperature, and specific humidity and their direction change depending on region and season over the Arctic.

by Ian Hellebosch, Niels Souverijns, Sara Top, Dirk Lauwaet, Steven Caluwaerts, Koen De Ridder Climate change and ongoing urbanization are increasing human exposure to heat stress in cities, where strong spatial contrasts in shading and imperviousness generate large variations at the scale of streets and public spaces. Urban adaptation planning therefore requires high-resolution information on outdoor heat stress. Microscale climate models provide such information, but their reliability must be demonstrated at meter-scale spatial resolution. This study presents a validation approach for such models through a detailed evaluation of the coupled UrbClim-HiREx modeling framework. UrbClim is a fast urban boundary layer model operating at 100 m resolution, while HiREx is a radiation-based model that simulates wet-bulb globe temperature (WBGT) at 1 m resolution. Model performance is assessed against a dense observational dataset collected during 7 consecutive heat wave days in June 2023 in Ghent, Belgium, spanning multiple urban microenvironments within a 300 m × 300 m area. UrbClim reproduces air temperature and humidity with high accuracy and provides realistic pedestrian-level wind speeds after in-canopy extrapolation, yielding a reliable forcing for HiREx. HiREx reproduces the diurnal evolution of WBGT with root mean squared errors of 0.95 °C at an open site and 0.51 °C beneath tree canopy. It resolves transient shading effects and meter-scale spatial variability, with hourly daytime root mean squared errors across the microenvironments remaining below 1.59 °C. The model reproduces the observed reductions in peak daytime heat stress under tree shade of approximately 4 °C WBGT. These findings provide the most detailed validation of the UrbClim-HiREx modeling framework to date and establish a transferable approach for evaluating microscale heat stress models. The results confirm the framework’s suitability for identifying urban heat stress hotspots and assessing adaptation strategies at meter-scale resolution.

Science Advances, Volume 12, Issue 34, August 2026.

Science Advances, Volume 12, Issue 34, August 2026.

The Paleocene−Eocene Thermal Maximum (PETM) was a brief yet extreme warming event that profoundly affected marine ecosystems, whereas its impact on contemporaneous lacustrine systems remains less well constrained. This event highlights the fundamentally different responses of marine and terrestrial ecosystems to rapid perturbations in the carbon cycle. Here, we investigate a Paleogene lacustrine succession from the Bohai Bay Basin (BBB) in northeastern China, integrating high-resolution biogeochemical, cyclostratigraphic, and statistical analyses to reconstruct the coupled evolution of carbon-cycle perturbations, environmental change, and ecological turnover across the PETM. During the core-PETM interval, the BBB lacustrine system experienced rapid ecosystem expansion under warm, humid conditions, characterized by elevated primary productivity and enhanced organic carbon burial. This phase was marked by increased abundance of algae and aquatic macrophytes, accompanied by the development of hypoxic to anoxic conditions. In the subsequent recovery stage, sustained warming and increasing salinity likely restructured the ecological dynamics. The abundance of aquatic macrophytes and algae communities declined, whereas cyanobacteria progressively increased among primary producers. This ecological shift may have promoted localized increases in harmful algal taxa, intensified oxygen consumption, and consequently reduced biodiversity and destabilized the ecosystem. Following the core-PETM, the BBB lacustrine depositional environment gradually recovered. However, this post-crisis state did not simply return to pre-PETM conditions; cyanobacteria continued to increase, while algal diversity remained relatively low. These patterns suggest that the ecosystem established a new ecological equilibrium characterized by elevated primary productivity and enhanced organic matter preservation, but reduced ecosystem complexity. Our findings document the ecological succession of a continental lacustrine system under hyperwarming conditions, highlighting its nonlinear, phased response to climatic forcing. These results reveal complex feedbacks among climate change, hydrological processes, and ecosystem dynamics during the PETM and provide important insights into the potential responses of modern lacustrine systems to anthropogenic global warming, particularly in the context of increasing eutrophication and harmful algal blooms.

Abstract Climate change detection and attribution plays a central role in establishing the influence of human activities on global warming. The widely used optimal fingerprinting framework can be formulated as a linear errors-in-variables regression in which the covariance of internal climate variability is estimated from climate model simulations. Existing regularized optimal fingerprinting approaches often underestimate the variance of the scaling factors, leading to confidence intervals that are too narrow and fail to achieve nominal coverage. In this paper, we first propose consistent variance estimators for the regression coefficients within the general class of linear shrinkage weight matrices, thereby addressing the undercoverage problem in conventional methods. Building on this result, we further derive a linearly optimal weight matrix that directly minimizes the asymptotic variances of the estimated scaling factors. Numerical studies confirm improved empirical coverage and shorter interval lengths. When applied to annual mean near-surface air temperature (1951–2020), the proposed method produces narrower and more reliable intervals and provides new insights in detection and attribution across different regional scales.

Journal of Geophysical Research: Atmospheres, Volume 131, Issue 16, 28 August 2026.

Journal of Geophysical Research: Atmospheres, Volume 131, Issue 16, 28 August 2026.

Journal of Geophysical Research: Atmospheres, Volume 131, Issue 16, 28 August 2026.

Journal of Advances in Modeling Earth Systems, Volume 18, Issue 8, August 2026.

Global streamflow reanalyses such as GloFAS-ERA5 are available everywhere yet lose fidelity in small, snow- and glacier-fed headwaters that feed Central Asia’s transboundary rivers. We present an explainable, calibrated deep ensemble framework that corrects the GloFAS-ERA5 log-residual at gauges of the Syr Darya and Amu Darya systems using the CA-discharge archive. An entity-aware long short-term memory (LSTM) backbone drives a regime-gated mixture of experts trained under a closed-form mixture continuous ranked probability score (CRPS) and augmented with snow physics constraints and a regime-conditional (Mondrian) conformal layer; skill was assessed under temporal holdout, leave-one-basin-out and prediction in ungauged region protocols, with grouped Shapley value attribution. Correction rendered all 74 gauges skillful, raising the median modified Kling–Gupta efficiency (KGE′) from 0.386 (raw) to 0.825 (flagship); on temporal point skill the framework is statistically tied with gradient boosting (paired Wilcoxon p = 0.49). Under-dispersed raw intervals (90% coverage 0.68) were recalibrated to near-nominal coverage (~0.90), and high-flow exceedance decision skill was moderate (Q90 Brier skill score 0.34, ROC-AUC 0.92), while low-flow (Q10) exceedance showed no skill over climatology. Transfer to ungauged, more glacierized catchments was a measured limit that degraded with glacier fraction and basin area. The framework’s value is calibration, an inspectable (supervised) regime structure and regional physical insight, not point skill superiority.

Journal of Geophysical Research: Atmospheres, Volume 131, Issue 16, 28 August 2026.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.

arXiv:2608.24428v1 Announce Type: new Abstract: Using monthly Ni\~no-3.4 anomalies through July 2026, we investigate how much predictive information is contained in delayed observations of the index. Ridge regression identifies informative delays, while multilayer perceptron and sparse identification of nonlinear dynamics (SINDy) models test whether nonlinear complexity provides additional direct forecast skill; gated recurrent unit (GRU) and long short-term memory (LSTM) networks provide a complementary test in which the temporal representation is learned internally. Delayed observations substantially improve forecasts over persistence and climatology at leads of up to six months, but increasing model complexity provides no systematic improvement. Historical recursive experiments favor a simple explicit SINDy recurrence and select shallow recurrent architectures, with no appreciable gain from learning the temporal representation internally. These results support a compact predictive representation of Ni\~no-3.4 evolution in which the representation of past information is more consequential than model complexity. As a prospective application, the selected models are used to forecast the developing 2026 event beyond the last available observation and to compare its predicted evolution with completed historical El Ni\~no events.

Climate change is becoming an increasingly important source of economic and health vulnerability in developing countries. This study examines the dynamic relationship between climate change and poverty across 22 countries in the Middle East and North Africa (MENA) region from 2000 to 2023. Poverty is captured through two indicators: household consumption expenditure, the monetary dimension, and infant mortality, the non-monetary dimension. Methodologically, the analysis relies on a panel autoregressive distributed lag (panel ARDL) model, estimated using the Pooled Mean Group (PMG) and Mean Group (MG) approaches. The results reveal a long-run relationship among climatic variables, macroeconomic factors, and poverty-related indicators. In the long run, precipitation is associated with a decline in household consumption expenditure, while temperature is associated with higher infant mortality, indicating a deterioration in both monetary and health-related well-being under changing climatic conditions. In the short run, rising temperatures are also associated with lower household consumption expenditure, revealing the immediate vulnerability of living standards to climate shocks. In addition, GDP per capita is associated with higher household consumption and lower infant mortality, while education is associated with lower health-related poverty. Inflation appears to exacerbate poverty, whereas the positive association between health expenditure and infant mortality suggests reverse causality or inefficiencies in the allocation of health resources. These findings highlight the need to articulate climate adaptation strategy, macroeconomic stability, education investment, and improved efficiency of health spending in order to achieve sustainable reduction in poverty in the MENA region.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

ABSTRACT High‐altitude alpine lakes (HAALs) of the Kashmir Himalaya are critical mountain freshwater systems that regulate headwater hydrology, sustain fragile ecosystems, and support nature‐based livelihoods. Despite their environmental and socio‐economic importance, comprehensive landscape‐based assessments integrating geomorphological processes, hydrological functioning, and sustainability remain limited. This study develops a comparative framework for evaluating seven alpine lake systems across the Greater Himalayan Range (GHR) and Pir Panjal Range (PPR) by integrating geomorphological characterization, quantitative morphometry, multi‐temporal analysis, hydrological connectivity, and a sustainability assessment. The results reveal pronounced spatial variability in basin geometry, shoreline complexity, and hydrological organization, reflecting contrasting influences of Quaternary glaciation, tectonic inheritance, lithological heterogeneity, and post‐glacial landscape adjustment. The investigated lakes encompass diverse geomorphic settings, including valley‐confined troughs, moraine‐dammed basins, compact cirque systems, and cascade‐linked lacustrine complexes. Within the PPR, the marked morphometric contrast between the elongated, shoreline‐complex Kousarnag Lake and the compact, near‐circular Bramsar Lake provides quantitative evidence of substantial intra‐regional geomorphic heterogeneity. Multi‐temporal analysis indicates that the surface area of Gangabal and Nundkol lakes increased by ∼11.6% and ∼12.1%, respectively, between 1980 and 2025, reflecting differential responses to hydro‐climatic variability and catchment‐scale geomorphic processes. Hydrologically, these alpine lakes function as critical headwater components of the Jhelum and Chenab river systems, regulating downstream water availability, sediment transfer, and ecosystem services. Their increasing sensitivity to glacier retreat, hydrological variability, geomorphic instability, and anthropogenic pressures highlights the need for adaptive, basin‐specific conservation and management strategies. The principal contribution of this study is the development of a transferable landscape‐based framework that explicitly integrates geomorphic controls, lake morphometry, hydrological connectivity, environmental sensitivity, stakeholder perspectives, and sustainability‐oriented management within a unified assessment framework. The framework provides a scientific basis for climate‐sensitive monitoring, conservation prioritization, sustainable tourism planning, and evidence‐based policy while contributing to Sustainable Development Goals (SDGs) 6, 8, 11, 13, and 15, and offers a transferable approach for the sustainable management of glacier‐fed alpine lake systems in rapidly changing mountain environments.

Glacial lakes are sensitive indicators of climate-driven cryospheric change, and their accurate mapping provides fundamental spatial information for water-resource assessment and glacial lake outburst flood (GLOF) hazard assessment. In southeastern Tibet, automatic extraction remains difficult because glacial lakes are small and easily confused with snow, mountain shadows, dark bedrock, riverine wetlands, and non-glacial water bodies. In this study, we integrate optical bands and water indices from Sentinel-2, topographic information derived from a digital elevation model, and radar backscatter from Sentinel-1 into a nine-channel multimodal dataset, and develop an improved YOLO11-seg model that combines spatial-to-depth downsampling, multi-scale attention, long-range context modeling, and content-aware upsampling to enhance small-lake detection, background suppression, and boundary delineation. Compared with U-Net, DeepLabV3+, YOLOv8-seg, YOLO11-seg, YOLO12-seg, and YOLO26-seg, the proposed model achieved the highest F1-Score of 0.9205 and an mAP50(M) of 0.9331, while its F1-Score and IoU on the independent test set reached 0.9209 and 0.8533, respectively. Using remote sensing imagery acquired in 2024, the model extracted 4766 glacial lakes in southeastern Tibet, covering 428.13 km2; 80.84% of these lakes were smaller than 0.10 km2. The results demonstrate an effective and reproducible framework for automatic glacial lake mapping in complex alpine environments.

Flood frequency and magnitude are expected to increase under future climate scenarios, yet long-term records of flood variability remain sparse beyond the instrumental period. Lake sediments provide valuable archives of past floods, but existing detection methods often struggle to identify both minerogenic and organic flood layers at high resolution. Here, we present a new non-destructive approach based on computed tomography (CT) scanning combined with thresholding techniques to isolate flood deposits in sediment cores. Using a 1.48 m section from Lake Vangsvatnet (Western Norway), a flood-prone catchment, we demonstrate that CT-derived density contrasts enable precise identification of event layers at sub-centimeter resolution. Volume-per-cent measurements of thresholded CT data reveal multiple flood layers, including organic-rich horizons that are not captured by conventional proxies such as magnetic susceptibility (MS) or X-ray fluorescence (XRF). This method significantly improves detection compared to traditional approaches and offers a rapid, reproducible workflow for distinguishing flood deposits from background sedimentation. Our findings highlight CT scanning as a powerful tool for paleoflood reconstruction and suggest its potential for refining long-term flood chronologies in regions sensitive to hydrological extremes.

Abstract Wave breaking is the dominant physical process responsible for energy dissipation of ocean surface waves and is critical for exchanges of momentum, energy and mass across the air-sea interface. Over the past two decades, extensive research efforts have been devoted to investigating the climatology and trends of wind speed and wave height in global oceans, revealing their upward trends over the past several decades. Nonetheless, research on global-scale wave breaking statistics remains relatively limited. It is not immediately clear whether wave steepness and wave breaking have intensified because the trends of wavelength are unknown. In this paper, we first present a simplified theoretical analysis based on the empirical wind-sea fetch laws and Toba’s 3/2 power law. A climatological relation between wind-sea wave steepness and wind speed is obtained, which suggests that as a consequence of accelerating winds on a climatological timescale, wave height grows faster than wavelength does and thus results in amplified steepness and intensified wave breaking. We then conduct a thorough analysis of global climatology and trends of the dominant wave breaking probability b T , an important proxy of wave breaking estimated empirically from a recently produced global wave hindcast over the period from 1992 to 2019. It is confirmed that in the Southern Ocean (global oceans), b T has increased by approximately 4% (2%) over the past 28 years, and these relative trends of b T are comparable to those for wind speed and wave height.

Nature Climate Change, Published online: 19 August 2026; doi:10.1038/s41558-026-02727-5 Extreme heat events are increasing, and urban adaptation is urgently needed for humans and non-humans to function safely under heat. Here we present a thermal justice framework and highlight how its use facilitates equitable adaptation and minimizes heat-risk displacement.

Nature Communications, Published online: 20 August 2026; doi:10.1038/s41467-026-76837-1 Here authors show patient-derived neural models of childhood dementia exhibit key disease features. A drug screening platform integrating machine learning, imaging, transcriptomics and electrophysiology, identifies repurposed compounds that mitigate adverse effects.

Nature Reviews Earth & Environment, Published online: 19 August 2026; doi:10.1038/s43017-026-00815-w Suspended particulate matter (SPM) influences the transport and fate of carbon and environmental contaminants in natural aquatic systems. This Review explores SPM dynamics along the land–ocean continuum and discusses how human activities and climate change might affect its role in the carbon cycle and sediment transport.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

We reconstruct interbasinal temperature and salinity gradients using stacked Mg/Ca and benthic δ¹⁸O records for the past 1.5 Myr. Across the Middle Pleistocene Transition, the deep Atlantic cooled and the Pacific became more saline, increasing deep ocean density stratification. The glacial ocean became a more effective carbon trap, which lowered atmospheric pCO2, and led to the growth of larger ice sheets. Results support a physical role for abyssal ocean stratification in explaining the MPT.

Nature, Published online: 18 August 2026; doi:10.1038/d41586-026-02600-7 Heatwave-driven sleepless nights are robbing people of a crucial aspect of health and recovery. Plus, dolphin calves learn hunting techniques from their mothers and how slavery is linked to health disparities in the United States today.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

arXiv:2608.19899v1 Announce Type: new Abstract: Machine learning (ML) is widely used for the development of ocean colour algorithms, but most studies focus on model parameter training and hyperparameter tuning. The optimisation of the data that feeds the models - i.e., Feature Engineering (FE) - is not fully explored. We assess the impact of FE in ocean colour machine learning models and we propose an optimisation framework that includes seven sequenced levels of data transformation: i. band choice, ii. log scaling, iii. spectral shape normalisation, iv. index extraction, v. principal component analysis, vi. feature scaling, and vii. zero-to-one scaling. We demonstrate the application for Multi-layer perceptron, Support Vector Machines, and eXtreme Gradient Boosting Trees on Sentinel-3 OLCI observations in the Norwegian coastal waters. The models are trained to estimate Chlorophyll-a concentration [Chl-a] and Secchi disk depth (Zsd). Results show that accuracy is highly variable among FE found in six studies using Sentinel-3 OLCI and the ones that we optimise. The R range from 0.01 to 0.55 for [Chl-a] and from 0.15 to 0.68 for Zsd, where the optimised FE shows the top results. The ML models with optimised FE could also improve by two times the R and reduce up to 63% of the mean absolute error when compared to CHL_OC4ME and CHL_NN standard algorithms. Nevertheless, no common optimised FE is found for all target variables and ML models, suggesting that FE optimisation is necessary for each application. Therefore, our proposed framework can be key for improving the accuracy of water quality monitoring in coastal waters.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

Abstract Near-surface extreme winds profoundly affect human society, yet process-based understanding of their changes under climate forcings remains limited. This study systematically investigates the responses of high (HWE) and low (LWE) wind extremes (10-meter) to climate forcings using a hierarchy of climate model experiments from multiple general circulation models that participated in the Cloud Feedback Model Intercomparison Project. We analyze idealized atmosphere-only aquaplanet (Aqua) simulations and more realistic land-atmosphere (AMIP) simulations to identify robust responses to climate forcings and trace the sources of structural uncertainty. In Aqua simulations, tropical LWE changes exhibit large inter-model spread, which can be partly traced to model-dependent responses of low-pressure systems. In contrast, extratropical HWE intensify robustly with surface warming, linked to the strengthening of high-latitude extratropical cyclones. The AMIP simulations confirm the robust intensification of extratropical HWE. The more realistic boundary conditions in AMIP simulations act as a constraint, reducing inter-model spread in tropical zonal means compared to Aqua simulations. A comparison of uniform and patterned 4-K warming experiments suggests that the global magnitude of warming, rather than the specific warming pattern, dominates the large-scale responses of wind extremes. However, regional projections of extreme wind changes, especially over land, remain highly uncertain due to divergences in model formulations. Case studies reveal that major disagreements in HWE changes can stem from fundamental differences in representing the type and seasonality of extreme-producing weather systems. Our results underscore that reducing uncertainty in regional wind projections requires constraining the physical representation of weather systems in climate models.

Nature Geoscience, Published online: 21 August 2026; doi:10.1038/s41561-026-02077-z Reply to: Uncertain internal variability in Arctic warming

Nature Geoscience, Published online: 21 August 2026; doi:10.1038/s41561-026-02078-y Uncertain internal variability in Arctic warming

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

arXiv:2608.20947v1 Announce Type: new Abstract: The Indian Summer Monsoon shows significant spatial variation. While prior work primarily focused on forecasting rainfall amounts, little attention has been given to how consistently a location's seasonal rainfall trajectory repeats from year to year. We introduce a graph-based machine learning framework to classify locations across India by this inter-annual consistency. Using 2001 to 2022 GSMaP ISRO data (excluding 2012), we constructed graphs for 29,026 grid points where nodes represent individual years and edges denote cosine similarity. A Graph Convolutional Network classified locations as either consistent or erratic with 96.8% accuracy. Applied to the Indian landmass, the model successfully identified the Western Ghats, Northeast India, and parts of central India as consistent regions. This classification was rigorously validated through statistical testing and temporal stability analysis, showing 93.6% agreement across two independent timeframes. Crucially, the results reveal a previously unreported coupling: regions with higher rainfall volumes are also the most temporally repeatable year-to-year, demonstrating an emergent, spatially coherent structure.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

Journal of Geophysical Research: Solid Earth, Volume 131, Issue 8, August 2026.

When sea ice moves along the ocean surface, it can generate waves below the ocean surface. Because these processes occur at spatial and temporal scales smaller than those captured by climate models, they need to be approximated. Here, we (1) find the relevant value ranges and spatial distributions of the parameters that describe this problem to improve this approximation and (2) examine the wave generation and turbulence using numerical experiments for a selected set of parameter values.

We studied 59 years of daily river flow, snow depth, and air temperature for a sub-Arctic river in Finland to see how the timing of seasonal change has shifted. Spring events now arrive earlier and together, while autumn events stay irregular and lag behind. The cold season is shrinking from both ends, and the atmosphere is increasingly running ahead of the snow and water response. The amounts of water and snow, however, have not changed, only their timing.

by Sven Grüner, Oliver Mußhoff This paper presents the findings of a global survey involving 1,510 researchers regarding climate change, focusing on behavioral patterns perceived as relevant to adaptation and mitigation efforts. Similar behavioral tendencies can be observed across both areas: inequity aversion, guilt aversion, and climate anxiety are seen as primary drivers (positive effects), while status quo bias, information avoidance, and mass numbing are regarded as considerable barriers (negative effects). Additionally, the paper finds that researchers’ diverse backgrounds partly explain differing perceptions, indicating the potential for intersectional collaboration in academia to address climate change.

We present a new open-source software called AQUA designed to handle and evaluate the vast amounts of data produced by high-resolution climate simulations. As climate models become more detailed, their output can become too large and complex for traditional analysis methods. We introduce a new framework for data access, processing, and comparison that is fast, reliable, and easy to automate. Our results show that it can efficiently manage very large datasets and support real-time quality checks.

Earth's Future, Volume 14, Issue 8, August 2026.

Methane is a powerful driver of climate change, yet emissions from wetlands and farming remain uncertain. We improved a global vegetation model that links water, plants and soils to methane production and release. The model now captures waterlogged areas, methane creation and escape, and flood-tolerant plants. It reproduces global wetland patterns and emissions more realistically, helping assess how climate and land-use change may affect future methane release and improve climate projections.

Accurate measurements of water stored in mountain snow are essential for managing water resources and understanding climate change. We evaluated a satellite-based method using airborne laser surveys and ground observations across several snow-covered regions. The method performed well under dry snow conditions and identified the factors that affect its accuracy, supporting improved monitoring of mountain snow from future satellite missions.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.

How reliable are probabilistic seasonal snow forecasts in winter? Although they are routinely issued by operational prediction centres, their reliability has never been evaluated. We close this gap by analyzing 30 years of seasonal snow re-forecasts and evaluating them against two snow datasets. Our results provide comprehensive assessment of seasonal snow forecast reliability and offer new insights into their performance in different parts of the Northern Hemisphere.

GCAM-Europe is an open-access modelling tool designed to assess the impacts of climate and environmental policies across European countries. Built on a global framework, it adds detailed regional and sectoral representation for Europe, including energy, land use, water, and emissions. This allows for analysis of policy effects within and between countries, as well as global spillover effects, to support transparent, evidence-based decision-making.

Computer models for hurricane prediction are increasingly detailed but require large computing resources. We developed a flexible approach that follows hurricanes, applying high-resolution simulations only where needed. This method reduces computing costs by up to an order of magnitude compared to conventional limited-area domains, while achieving resolutions down to 300 m. The approach enables more efficient hurricane research and improved understanding of tropical dynamics.

Science Advances, Volume 12, Issue 34, August 2026.

We evaluated how different configurations of a large-scale river basin model affect simulations of streamflow, evaporation, soil moisture, and groundwater in the Po River Basin in Italy. We tested alternative soil depths and the inclusion or removal of subsurface flow pathways, and compared results with observations and with an established long-term water balance relationship. Setups that best matched river flow often underestimated evaporation and overestimated deep groundwater recharge.

arXiv:2307.11040v2 Announce Type: replace Abstract: Between 1945 and 1962, the United States conducted 101 atmospheric nuclear weapon tests in New Mexico and Nevada, dispersing radioactive fallout across the contiguous United States. We reconstruct the first 10 days of fission-product deposition from the 94 non-zero-yield tests using U.S. government test data, benchmarked particle-size distributions, and the HYSPLIT atmospheric transport model driven by high-resolution ERA5 reanalysis. Comparisons with county-level iodine-131 deposition estimates from the National Cancer Institute show good agreement for tower and surface tests and motivate refinement of the particle-size distribution parameters associated with airbursts. We aggregate per-test reconstructions into a cumulative deposition map that includes long-range fallout from the 1945 Trinity test. Trinity fallout alone reached 46 states within 10 days, consistent with contamination observed in Indiana and Iowa soon after the test. An ensemble analysis bounds meteorological uncertainty for Trinity, showing robust large-scale transport patterns despite relatively sparse 1945 observations.

ABSTRACT Upland farming in Europe is frequently assumed to have been limited and precarious, despite relict spade‐dug ridges providing firm evidence of past crop growing. Direct evidence for crop types in post‑Medieval to Modern (~since 1500 CE) uplands is rare; yet, in Ireland, relict spade‐dug ridges that date to this period are commonly interpreted as potato beds, generally assumed to have been abandoned during the Great Irish Famine of 1845–1852 CE. This assumption has seldom been tested and reinforces narratives of upland marginality and climatic vulnerability. Here, we apply palynology to cultivation ridge soils in three relict field systems associated with disused cottages in the foothills of the Mourne Mountains, Northern Ireland. We find that cereals—not potatoes—were grown in these upland fields, extending the history of cereal growing in the Mournes from the mid‐12th until the 20th century CE. These results show that permanent occupation persisted through a period when uplands were abandoned elsewhere in Ireland, demonstrating the resilience of upland populations in what is widely regarded as a marginal environment. Our findings challenge entrenched assumptions about historic upland agriculture and highlight the importance of empirically testing the nature of past land use.

Geophysical Research Letters, Volume 53, Issue 16, 28 August 2026.