The methodological heart of the study is a three-dimensional connected-component labeling approach, abbreviated CC3D, borrowed in spirit from image-processing and computer vision. Instead of analyzing drought and heatwave conditions independently, the researchers constructed a three-dimensional data cube in which two dimensions represent space and the third represents time. Within that cube, they identified contiguous clusters of grid cells that simultaneously satisfy drought and heatwave thresholds, treating each cluster as a single coherent object—an event with a birth, a trajectory, a lifetime, and a death. This is the same conceptual machinery used to track thunderstorm cells in radar imagery, applied here to the coupled signature of heat and moisture stress. The approach allows each event to be characterized not only by its intensity and duration but also by its spatial footprint at every time step and by the distance it travels between its origin and its eventual dissipation.

The statistics that emerge from this tracking exercise are striking. Most contiguous compound drought–heatwave events persist for six to twelve days, yet during that short lifetime they can displace by up to roughly 500 kilometers. At their peak, individual events can blanket daily spatial extents of two to five hundred thousand square kilometers—an area comparable to a large European country. Crucially, the analysis reveals a scaling relationship: the larger and longer-lasting an event is, the farther and faster it tends to travel. This coupling between event magnitude and mobility means that the most severe compound extremes are precisely the ones most likely to sweep across new territory, expanding the spatial footprint of combined heat and moisture stress well beyond where the hazard first ignited.

Geographically, the migration signal is strongest in the Northern Hemisphere mid-latitudes, and the dominant direction of travel is eastward. That preference is not accidental. The mid-latitudes sit beneath the storm tracks and the westerly flow of the extratropical circulation, so anomalies embedded in the jet stream—blocking highs, quasi-stationary ridges, and the wave patterns that steer heat domes—tend to propagate downstream from west to east. When a drought–heatwave event is anchored to such a circulation feature, its spatial expression moves with the feature. The study links the observed displacement of compound events to these evolving large-scale atmospheric circulation patterns, providing a physical explanation for why events behave like traveling objects rather than fixed blights on the landscape.

But circulation alone is only half the story. The researchers show that the migration of compound drought–heatwave events is reinforced by downstream drying—a feedback loop in which the moving event itself helps create the conditions it needs to persist. As an event advances, reduced moisture transport leaves the air downwind drier, and strengthened land–atmosphere coupling allows soil-moisture deficits to deepen. In a tightly coupled system, a parched land surface suppresses evaporation, so more of the sun’s energy goes into heating the air rather than into latent heat flux, amplifying temperature extremes. That heat, in turn, accelerates drying. The result is a self-sustaining propagation mechanism: the event carries its own drought precondition eastward, priming each new region it touches for the heat stress that follows. This interplay between moisture transport and land–atmosphere feedbacks is what gives compound events their remarkable coherence as traveling structures.

Perhaps the most consequential finding is the identification of emerging hotspots where compound drought–heatwave migration potential is increasing. The study singles out the European Plains, western North America, the Amazon, the West African Sahel, Central Asia, and northern Australia. These regions span a remarkable range of climates—from the moisture-laden Amazon to the semi-arid Sahel—yet they share conditions that favor mobile compound extremes: extensive flat terrain aligned with prevailing flow, strong land–atmosphere coupling, and sensitivity of the surface water balance to modest shifts in precipitation and temperature. The fact that hotspots are emerging rather than merely persisting suggests a spatial reorganization of risk under a changing climate, with regions that previously experienced isolated droughts or isolated heatwaves increasingly hosting compound events that arrive from upwind.

The agricultural implications are immediate and sobering. The authors highlight that increasing migration potential over agricultural landscapes may expand exposure to compound heat and moisture stress, with particular concern for grazing-dependent regions vulnerable to pastureland-to-scrubland transitions. When a compound event sweeps through rangeland, the combination of depleted soil moisture and extreme heat can push perennial grasses past physiological thresholds, accelerating the conversion of productive pasture into woody scrub—a transition that is difficult and expensive to reverse. Unlike a stationary drought, which degrades conditions in one place over weeks or months, a migrating compound event can deliver acute stress to a sequence of grazing regions within days, leaving herders and livestock managers little time to move animals or secure fodder. Similar logic applies to rain-fed croplands, where a traveling event can strike several breadbaskets in a single growing season.

The study also carries a methodological message for the hazard-modeling community. Risk assessments, insurance products, and early-warning systems frequently treat droughts and heatwaves as spatially fixed hazards, compiling statistics of exceedance for defined administrative or hydrological units. If compound events systematically displace by hundreds of kilometers, such frameworks will systematically misallocate risk—underestimating exposure in regions that lie downstream of current hotspots and overestimating it where events merely originate. Tracking-based approaches like CC3D offer a path forward: by treating extremes as objects with trajectories, forecasters can in principle anticipate not only where a compound event will form but where it is likely to travel, buying lead time for the regions in its path. Embedding this object-based perspective into seasonal prediction systems and coupled land–atmosphere models is an obvious next step for the field.

What makes the work resonate beyond hydrology is the way it captures a shift in how scientists think about climate extremes. For decades, the discipline has moved from analyzing single hazards to compound ones, from fixed locations to connected systems, and from static climatologies to evolving risk landscapes. This study pushes that evolution another step: the hazard itself is in motion. As global warming intensifies both drought and heat extremes and alters the circulation patterns that steer them, the corridors along which compound events travel may widen and shift, redrawing the map of vulnerability. For the millions of people who depend on the rain-fed pastures and croplands of the European Plains, the Sahel, Central Asia, and beyond, the message is that the next compound drought–heatwave may not come to them—it may already be on its way from somewhere else.

Subject of Research: Migration and expanding spatial footprint of compound drought–heatwave events across global land regions

Article Title: Global expansion of compound drought–heatwave risk

Article References: Sabut, A., Mishra, A., Leung, R., & Entekhabi, D. (2026). Global expansion of compound drought–heatwave risk. Nature Communications. https://doi.org/10.1038/s41467-026-78045-3

Image Credits: AI Generated

DOI: 10.1038/s41467-026-78045-3

Keywords: compound events, drought, heatwave, climate change, land–atmosphere coupling, hydrology, natural hazards, agriculture, water security, atmospheric circulation, CC3D tracking, ecosystems

The team focused on the Standardized Precipitation Evapotranspiration Index, or SPEI, a drought metric that captures both water supply through precipitation and water demand through potential evapotranspiration, which rises as temperatures climb. Drawing on independent datasets including the CRU TS4.09 gridded climate records, NOAA’s PREC/L precipitation analysis, TerraClimate water-balance data, and the ERA5 reanalysis, the researchers tracked summer conditions across the steppe region bounded roughly by 40 to 55 degrees north and 30 to 65 degrees east. Using a sequential regime-shift detection algorithm known as STARS, they identified a statistically significant break point in 2006, surpassing the 95 percent confidence level, after which the mean summer SPEI dropped sharply and showed no sign of recovery through the end of the observational period in 2024.

What makes this finding remarkable is not simply that the steppe has dried, but how the drying unfolded. The strongest downward tendency in the 7-year running trends occurred between 2004 and 2010, and the post-shift mean remains far below anything seen in the preceding decades. When the researchers compared the observed post-regime conditions against pre-industrial control simulations from 14 CMIP6 models, which represent only natural variability with no human-induced forcing, the observed drying fell outside the 5th to 95th percentile range of what nature alone could produce. In other words, the new climate state of the Western Eurasian Steppe is unprecedented in the context of natural climate variability over the past century.

Attributing the shift required separating the slow, forced hand of greenhouse warming from the capricious hand of internal climate variability. The researchers turned to the Max Planck Institute Grand Ensemble, a collection of 100 simulations with identical forcing but different initial conditions, along with 50 members from the MIROC6 model and 18 models from the CMIP6 archive. The ensemble means of all three archives show a background drying trend over the steppe during 1992 to 2010, indicating that external forcing provided the underlying push toward aridification. Yet the magnitude of the observed drying exceeds the forced signal alone, and comparisons between the most and least drying ensemble members reveal that internal variability substantially amplified the regional trend since the early 1990s. Signal-to-noise analysis of the Grand Ensemble confirms that the forced component alone cannot explain the full depth of the transition.

The mechanism that turned a gradual trend into an abrupt regime shift lies in the feedback loop between the land surface and the atmosphere. As soil moisture declined across the steppe, a threshold was crossed beyond which evaporative cooling weakened and the surface energy budget tilted toward sensible heat, the form of energy that directly warms the air. The study documents strengthening trends in the contribution of soil moisture deficit to surface sensible heat flux during 1992 to 2010, alongside rising summer surface air temperature anomalies. Drier soils mean less water available for plants and the ground to evaporate, so more of the sun’s energy goes into heating the air rather than converting moisture into vapor. That hotter air increases evaporative demand, pulling yet more moisture from an already stressed landscape and deepening the hydrological deficit in a self-reinforcing spiral.

The consequences of this feedback are visible in the extreme-temperature record. The researchers found that heatwave frequency trends across the steppe intensified in step with the soil moisture decline, and that the strength of land-atmosphere coupling itself underwent a regime shift detected by the same statistical method applied to the drought index. Independent soil moisture products, including ERA5, the GLEAM4.2 dataset, and the ESA CCI satellite-based soil moisture record, all corroborate the abrupt transition. Simulations from the CMIP6 Land Surface, Snow and Soil Moisture Model Intercomparison Project, known as LS3MIP, provide the clinching evidence: in fully coupled historical and future-scenario runs, five of seven models exhibit significant post-1990 regime shifts in summer SPEI, but when soil moisture is prescribed to its climatological seasonal cycle and the interactive feedback is suppressed, none of the corresponding simulations shows such a shift. The land-atmosphere coupling, in other words, is not a bystander but a necessary ingredient for the abruptness.

The ecological toll is already measurable. Satellite-derived land cover data from the MODIS record show a declining areal fraction of grassland and an expanding share of barren land within the steppe relative to the pre-2010 mean, a signature of widespread land degradation. Gross primary productivity, the rate at which plants convert carbon dioxide into biomass, has been reduced across the region, as confirmed by multiple products including the FLUXCOM flux synthesis and a near-infrared reflectance based estimate of vegetation activity. Rising atmospheric vapor pressure deficit, the thirst of the air for moisture, suppresses plant growth even where water is not yet fully exhausted, and the combination of heat and drought stress has pushed the steppe’s vegetation past a point where recovery during the observational record has been absent.

The Western Eurasian Steppe is not the first mid-latitude dryland to sound this alarm. A 2020 study in Science documented an abrupt shift to hotter and drier conditions over inner East Asia beyond a tipping point, and the authors of the new study explicitly frame their region as an early warning signal for other Northern Hemisphere mid-latitude drylands where similar land-atmosphere feedbacks operate. The steppe’s position downstream of the westerly jet stream, which acts as a waveguide for circumglobal teleconnection patterns, adds an atmospheric dimension to its vulnerability, linking regional soil desiccation to the large-scale circulation anomalies that have been implicated in persistent heatwaves across Eurasian breadbasket regions. Multi-model projections extending to 2080 under different future scenarios suggest that the ensemble-mean drying continues, with the range of outcomes widening as scenario uncertainty grows over time.

Methodologically, the study is notable for its redundancy. The regime shift appears in SPEI computed from three different combinations of precipitation and potential evapotranspiration datasets, in soil moisture from three independent products, and in the coupling-strength metric derived from reanalysis. Trend significance was assessed with Mann-Kendall tests, piecewise regression helped locate the turning point, and the Theil-Sen estimator provided robust slope estimates resistant to outliers. The pre-industrial control comparison rules out the possibility that the observed transition is simply a rare draw from unforced natural variability, while the LS3MIP experiment isolates the causal role of interactive soil moisture. Together, these lines of evidence converge on a coherent story: external forcing set the stage, internal variability pulled the trigger, and land-atmosphere feedback turned the bullet into a lasting change of climate state.

For the tens of millions of people who depend on the Eurasian steppe for grazing, agriculture, and the ecosystem services of one of the world’s largest contiguous grasslands, the finding carries sobering implications. A regime shift, unlike a gradual trend, offers little time for adaptation; the baseline against which farmers and herders plan their seasons has already moved. The study also sharpens the scientific debate over climate tipping elements, suggesting that semiarid drylands with strong land-atmosphere coupling may harbor thresholds that sustained warming can breach, producing transitions that are abrupt, persistent, and effectively irreversible on societal timescales. As global temperatures continue to rise, the Western Eurasian Steppe stands as a real-time demonstration that climate change does not always arrive as a slow creep. Sometimes it arrives as a door that swings shut.

Subject of Research: Abrupt hydroclimatic regime shift and land-atmosphere feedbacks in the Western Eurasian Steppe

Article Title: An abrupt shift towards a hotter and drier climate regime in the Western Eurasian Steppe

Article References: Piao, J., Kug, J.-S., Chen, W., Sung, M.-K., Chen, S., Wang, Z., Cai, Q., & Lee, D. (2026). An abrupt shift towards a hotter and drier climate regime in the Western Eurasian Steppe. Nature Geoscience. https://doi.org/10.1038/s41561-026-02125-8

Image Credits: AI Generated

DOI: 10.1038/s41561-026-02125-8

Keywords: Western Eurasian Steppe, climate regime shift, drought, soil moisture, land-atmosphere coupling, heatwaves, SPEI, CMIP6, internal climate variability, land degradation, gross primary productivity, tipping point

The instrument at the heart of the effort is the Ka-band cloud radar, which transmits millimeter-wavelength pulses exquisitely sensitive to the tiny droplets and ice crystals that make up non-precipitating and weakly precipitating clouds. Unlike conventional weather radars tuned to rain, these systems can peer through thick, multi-layered cloud decks and return continuous vertical profiles up to 18 kilometers above the ground. Of roughly 120 radars operating nationwide in 2024, the researchers retained 80 stations that logged at least 100 days of valid observations each, a quality threshold that made a robust national analysis possible for the first time. Before 2024, fewer than 30 such radars were in operation, far too sparse for a coherent countrywide picture.

Raw radar echoes, however, are notoriously contaminated. Near-surface dust, insect swarms, turbulence, and instrumental noise can masquerade as cloud, while a radar’s sensitivity to faint high clouds fades with distance. The team addressed both problems with an improved retrieval algorithm that replaces the traditional fixed reflectivity cutoff with a height-dependent noise-equivalent threshold, so that weak echoes aloft are judged against a realistic sensitivity curve rather than a single number. A second, variance-based filter exploits the fact that true cloud layers are temporally coherent, whereas clutter flickers erratically; a three-minute sliding window separates the two. Validation against collocated radiosondes showed discrepancies of well under half a kilometer for thin low and mid-level clouds and no more than about a kilometer for deep precipitating systems.

The headline numbers are striking. Across the network, clouds were present 56.7 percent of the time on average, and single-layer clouds dominated, accounting for 34.7 percent of all observations, roughly 62 percent of cloudy moments. Two-layer clouds occurred 14.7 percent of the time and multi-layer systems, with three to five distinct decks, just 7.1 percent. Geography mattered enormously. Single-layer clouds prevailed over the arid basins and plateaus of northwestern China, where dry, stable air suppresses vertical motion, while multi-layer clouds clustered over the humid southeast, where monsoon moisture and strong solar heating fuel deep, vertically complex convection. The highest overall cloud frequencies appeared over the southeastern Tibetan Plateau and the Sichuan Basin, where topographic lifting and moisture convergence conspire to fill the sky.

Cloud base height told a vivid seasonal story. Nationally averaged bases climbed from 2.15 kilometers in winter to 3.24 kilometers in spring, 3.95 kilometers in summer, and back down to 2.78 kilometers in autumn, a progression the authors attribute to the interplay of surface heating, boundary-layer depth, and the lifting condensation level. Counterintuitively, cloud bases over the Tibetan Plateau sat distinctly lower than over surrounding lowlands during spring and summer, because the Plateau’s cold ambient air lowers the condensation level for a given humidity even as intense surface heating deepens mixing. Cloud tops followed a similar seasonal rhythm, averaging 7.51 kilometers in summer against 4.70 kilometers in winter, and vertically the distribution of cloud tops was bimodal, with one population peaking between 1 and 3 kilometers and a second between 6 and 12 kilometers, the signature of coexisting shallow stratiform and deep convective regimes.

The radar network’s minute-by-minute sampling also exposed a diurnal drama that twice-daily radiosondes have long missed. In summer, cloud occurrence between 3 and 9 kilometers rose from an early-morning minimum to a pronounced late-afternoon and evening peak between 16:00 and 20:00 local time, as solar heating destabilized the boundary layer and ignited moist convection that punched into the upper troposphere. Winter flipped the script: clouds stayed largely below 3 kilometers and peaked near sunrise, when minimum temperatures under a stable nocturnal boundary layer pushed relative humidity high enough to condense shallow stratiform decks. Spring and autumn traced transitional patterns in between, with afternoon mid-tropospheric peaks in spring and increasingly even, low-topped cloudiness in autumn.

What drives all this variability? The team matched the radar retrievals hourly against ERA5 reanalysis fields and wind-profiler measurements to isolate three candidate controls: near-surface specific humidity, lower-tropospheric stability, and vertical wind shear. Moisture emerged as the dominant driver. Specific humidity correlated positively with both cloud base and cloud top at more than 93 percent of stations, and with cloud thickness at 71 percent, confirming that abundant low-level moisture sustains the buoyancy needed for deep vertical development. Stability acted as a powerful brake: a terrain-following measure of lower-tropospheric stability correlated negatively with cloud base at 97.5 percent of stations and with cloud top at 93.8 percent, meaning strongly stratified air confines clouds to shallow, low-based layers regardless of how moist the surface is.

Wind shear played a subtler, conditional role. Strong vertical shear correlated negatively with cloud base, top, and thickness at roughly three-quarters of stations, consistent with shear-enhanced entrainment of dry air that erodes buoyancy and shreds organized updrafts. Yet about a quarter of stations, mainly in the northwest, showed the opposite sign, echoing the well-known optimal shear framework in which moderate shear organizes convection rather than destroying it. Joint analyses made the hierarchy explicit: the tallest, thickest clouds required both high humidity and low stability, while under stable conditions cloud bases stayed low no matter the moisture or shear. Moisture was the engine; stability and shear were the governors.

Perhaps the most novel finding concerns the ground itself. Comparing cloud base heights across land-cover types mapped by MODIS, the researchers found that low clouds rooted lowest over densely vegetated surfaces, where evapotranspiration moistens the boundary layer and drops the condensation level, and highest over barren land, where moisture scarcity lifts it. The contrast sharpened dramatically in the afternoon. Single-layer low clouds observed between 14:00 and 18:00 local time had bases averaging 21.7 percent higher than the all-hour climatology, with the enhancement reaching 41.7 percent over evergreen broadleaf forests and 33.3 percent over croplands, but only a few percent over water bodies whose thermal inertia damps the diurnal boundary-layer cycle. The result is a crisp observational fingerprint of land-atmosphere coupling, showing that what grows on the surface measurably shapes where clouds form above it.

For modelers, the implications are immediate. The observed sensitivities of cloud base and top to humidity and stability provide quantitative benchmarks for calibrating the boundary-layer and convective parameterizations that dominate cloud-related uncertainty in weather and climate prediction, while the land-cover contrasts offer a test of how faithfully land surface models capture surface-atmosphere feedbacks. The authors caution that a single year of data cannot establish trends, and they plan to extend the analysis to multi-year records to probe interannual variability and the role of cloud structure in extreme weather. But as a first national snapshot, the study demonstrates what a coordinated ground-based radar network can do: turn one of climate science’s murkiest uncertainties into a measured, mappable, and model-testable quantity, one kilometer of sky at a time.

Subject of Research: Cloud vertical structure across China observed by a national Ka-band cloud radar network and its thermodynamic, dynamical, and land-surface controls

Article Title: Cloud vertical structure across China from a national Ka-band cloud radar network: thermodynamic, dynamical, and land-surface controls

Article References: Xu, H., Guo, J., Deng, J., Yang, R., Meng, D., Zhang, Z., Li, N., Sun, Y., Jiang, S., Chen, T., Chen, J., Zeng, L., Zhou, Y., & Tong, B. (2026). Cloud vertical structure across China from a national Ka-band cloud radar network: thermodynamic, dynamical, and land-surface controls. Atmospheric Chemistry and Physics, 26(19), 14303-14327. https://doi.org/10.5194/acp-26-14303-2026

Image Credits: AI Generated

DOI: 10.5194/acp-26-14303-2026

Keywords: cloud vertical structure, Ka-band cloud radar, cloud base height, cloud top height, lower-tropospheric stability, vertical wind shear, specific humidity, land-atmosphere coupling, Tibetan Plateau, East Asian monsoon, cloud parameterization, China Meteorological Administration

The researchers, led by Arnau Garcia Mesa, focused on six locations chosen to span a wide range of climatic behavior: Córdoba in southern Spain, Marrakech in Morocco, Lyon in France, Hannover in Germany, Belgrade in Serbia, and Stockholm in Sweden. These sites run the gamut from water-limited regimes, where soils are almost always dry in summer, to energy-limited regimes, where abundant moisture means the land surface and atmosphere interact very differently. For each location, the team assembled more than seven decades of data from the ERA5 and ERA5-Land reanalyses, covering daily maximum temperatures, large-scale atmospheric circulation fields, soil moisture at three depth levels, and seasonal carbon dioxide concentrations measured at NOAA’s Mauna Loa Observatory.

The machine learning framework at the heart of the study is a hybrid architecture the authors call the CombinedModel. Atmospheric data, which arrives as spatial maps of geopotential height at 500 and 200 hectopascals and sea-level pressure, is processed by a ConvNeXt convolutional neural network, a modern image-recognition architecture that had not previously been applied to driver quantification in this context. Local land-surface data and carbon dioxide values, which arrive as simple time series, are handled by a multilayer perceptron. The outputs of both branches are merged in a final classification network that decides, day by day, whether conditions resemble a hot extreme, defined as a daily maximum temperature exceeding the 90th percentile of the 1950–2000 climatology. Because extreme days are rare, the team used a weighted loss function to prevent the model from simply defaulting to predicting ordinary weather, and they trained an ensemble of twenty members to tame the randomness of network initialization.

But prediction alone was not the goal. The real innovation lies in the interpretability layer. Using SHAP values, short for SHapley Additive exPlanations, the researchers borrowed a technique from game theory that assigns each input feature a precise, additive share of the model’s output. Positive SHAP values push the model toward predicting an extreme; negative values push it away. Because the values are additive, they can be summed across grid points and variables to produce a global accounting of driver importance. The team computed these explanations only for the fifty percent most confident extreme predictions in the 2014–2024 test period, ensuring the attribution rested on predictions the network itself trusted.

The headline result is strikingly consistent: atmospheric circulation dominates everywhere. Across all six locations, circulation features accounted for 66 to 86 percent of the total explanatory power, with the geopotential at 500 hectopascals, the mid-tropospheric field that meteorologists use to spot blocking highs and ridges, contributing the most by a wide margin. This aligns with a growing body of work linking European heatwaves to persistent double jets over Eurasia, deep North Atlantic depressions, and Saharan warm-air intrusions, and it echoes a 2025 global analysis finding that 500-hectopascal geopotential is the dominant heat-extreme driver over most of the world’s land area. In other words, the large-scale shape of the atmosphere is the primary engine of a hot day, almost regardless of where you are.

The land surface, by contrast, tells a sharply regional story. Soil moisture’s share of the explanation follows a clear northward gradient: a negligible 0.2 percent in Marrakech, a moderate 9.4 percent in Córdoba, and a substantial 15.4 percent in Lyon, with similarly elevated values in the temperate and northern sites. The pattern makes physical sense. In already-arid Marrakech, summer soils are so consistently dry that year-to-year variability is small, leaving little for the land surface to modulate. In wetter transitional climates, however, a dry soil anomaly can dramatically amplify a heatwave by suppressing evaporation and shifting the surface energy balance toward sensible heating. The SHAP analysis captured this mechanism directly: for the shallowest soil layer, negative moisture anomalies were consistently associated with positive contributions to extreme-heat predictions, while deeper layers showed weaker and less robust signals, likely because neighboring soil layers are so strongly correlated that the model cannot cleanly separate their effects.

Carbon dioxide emerged as the third actor, and its role was revealing. The CO2 feature carried more than ten percent of the explanatory weight at most sites, and its importance was highest precisely where the yearly count of extreme days showed the strongest warming trend. By including CO2 explicitly, the framework controlled for the long-term anthropogenic signal without having to detrend the data, allowing a cleaner attribution of the physical drivers operating on daily timescales. The researchers stress that the explainability outputs represent the model’s interpretation rather than a direct measurement of reality, but the fact that the identified contributions reproduce well-established physical relationships lends the approach considerable credibility.

Robustness checks formed a major part of the study. When the team redefined extremes using the 80th percentile threshold, the same architecture recovered the same picture: circulation still dominated at 64 to 84 percent, and the northward gradient in land importance persisted, though soil moisture’s relative share shrank, suggesting that land-atmosphere feedbacks are not fully activated in milder heat events. A 95th-percentile model, hampered by too few extreme samples, proved unreliable and was set aside. The team also swapped reanalysis data for observations, using the E-OBS gridded temperature dataset and drought indices computed from it as soil-moisture proxies. The Standardized Precipitation Evapotranspiration Index, which incorporates evapotranspiration and thus better represents the surface water balance, reproduced the original results closely, while the precipitation-only Standardized Precipitation Index captured less of the land signal, a finding with practical implications for applying the framework to climate models, where soil layer definitions differ.

Two case studies brought the method down to the scale of individual disasters. For the August 2021 heatwave that pushed temperatures in Córdoba province to a scorching 47.6 degrees Celsius, the model attributed the event overwhelmingly to circulation, with positive SHAP values appearing precisely over the Iberian Peninsula where positive 500-hectopascal geopotential anomalies were located, a remarkable demonstration of spatial locality in the network’s reasoning. Intriguingly, CO2 ranked as the second most important feature even for this single event, hinting at a detectable anthropogenic fingerprint in an individual heatwave. The July and August 2018 heatwave in Hannover told a different story: the shallow soil-moisture layer played a much larger role, with SHAP values at times rivaling the circulation signal, consistent with the wet transitional regime of northern Germany, where the exceptional 2018 drought turned abundant soils into amplifiers of extreme heat.

The authors see their framework as more than an attribution exercise. Because it is data-driven yet physically interpretable, it could be applied to climate model output, to subseasonal-to-seasonal prediction systems, or to new regions, helping to diagnose where models capture or miss key feedbacks and to target climate risk assessments where they matter most. Future work, they note, should test whether the results hold across different network architectures and whether adding humidity fluxes or aerosol information sharpens the picture further. For now, the study offers a compelling demonstration that artificial intelligence, when forced to explain itself, can do more than forecast the weather; it can tell us, in numbers, why the weather turned deadly.

Subject of Research: Drivers of summer heat extremes in Europe and North Africa quantified with explainable machine learning

Article Title: Quantifying atmospheric and land drivers of hot temperature extremes through explainable Artificial Intelligence

Article References: Garcia Mesa, A., Palma, L., Donat, M., Materia, S., Gràvalos Talló, B., & Marcos Matamoros, R. (2026). Quantifying atmospheric and land drivers of hot temperature extremes through explainable Artificial Intelligence. Weather and Climate Dynamics, 7(3), 1709-1731. https://doi.org/10.5194/wcd-7-1709-2026

Image Credits: AI Generated

DOI: 10.5194/wcd-7-1709-2026

Keywords: heatwaves, explainable AI, SHAP values, atmospheric circulation, soil moisture, land-atmosphere coupling, climate change, CO2, machine learning, temperature extremes, ERA5, Europe

The central question is deceptively simple. When the ground is hot and dry, does that help storms form, or does it suppress them? The answer, it turns out, depends heavily on the time of year and on which link in the chain you examine. The study finds strong and seasonally varying coupling between land surface temperature and atmospheric moisture, but the relationships between individual thermodynamic or dynamical variables and precipitation are generally weak and nonlinear. In other words, no single number measured at the surface or in the boundary layer can reliably predict rainfall on its own. The atmosphere over India behaves less like a machine with a single trigger and more like a system of conditional gates, several of which must open at once before convection fires.

To untangle these relationships, the research integrates an unusually broad set of observational streams. Land surface temperature came from the MODIS MOD11A2 product, evapotranspiration from MOD16A2, both distributed by NASA. Atmospheric variables, including dew point temperature, convective available potential energy (CAPE), convective inhibition (CIN), and wind fields, were drawn from the ERA5 reanalysis produced by the European Centre for Medium-Range Weather Forecasts. Precipitation estimates came from the TRMM 3B43 and GPM IMERG missions, lightning observations provided an independent proxy for actual convective activity, and land-use and land-cover information came from the India WRIS database. Nearly all of the processing was carried out on the Google Earth Engine platform, allowing pixel-level analysis across the entire subcontinent.

Two diagnostic frameworks sit at the heart of the analysis. The first is a normalized CAPE–CIN instability index, or NCII, which combines the fuel available to rising air parcels (CAPE) with the energetic barrier they must overcome (CIN) into a single measure of thermodynamic readiness. The second is the Convective Triggering Potential–Humidity Index framework, or CTP–HI, which classifies the atmospheric profile above a given location according to whether the boundary layer and free troposphere are configured in a way that favors or suppresses feedbacks between the land surface and convective initiation. Together, these tools allow the study to separate the question of whether the atmosphere is primed for storms from the question of whether storms actually occur.

The NCII results deliver one of the study’s most important cautionary messages. The index shows only weak-to-moderate correspondence with observed lightning activity, which serves as a direct tracer of deep convection. Thermodynamic readiness, in other words, does not necessarily coincide with convective occurrence. A region can be loaded with CAPE and still produce no storms if CIN is too strong, if moisture convergence is absent, or if no dynamical lifting mechanism nudges air parcels past their level of free convection. This finding echoes a long-standing theme in convective meteorology, rooted in ingredients-based methodologies for flash flood forecasting: instability is necessary but never sufficient. For forecasters, the implication is that instability indices alone will systematically misidentify both where storms will form and where they will not.

The CTP–HI analysis adds a spatial and seasonal dimension to that caution. The study finds substantial seasonal variation in the atmospheric thermodynamic environment, with conditions favoring a wet-soil advantage, in which moist land surfaces enhance the likelihood of afternoon convection, covering 20.52 percent of the valid analysis area during the monsoon months of June through September and 14.03 percent during the post-monsoon period of October and November. Strikingly, most of the domain remains outside the defined feedback regimes altogether, meaning that for the majority of India, the classical soil-moisture-precipitation feedback loop described in earlier benchmark studies simply does not apply most of the time. The land-atmosphere feedback story over India is therefore not a single narrative but a patchwork that shifts with the seasons.

Perhaps the most vivid result concerns evapotranspiration. When the study examines pixel-level correlations between evapotranspiration and the NCII instability index, the strength of the relationship swings dramatically across the year. It is strongest during October and November, with a correlation coefficient of 0.63, followed by the winter months of December through February at 0.47 and the pre-monsoon spring at 0.28. During the peak monsoon itself, the relationship actually flips sign, showing a weak negative correlation of minus 0.15. This seasonal modulation is physically interpretable: during the monsoon, when the large-scale circulation and abundant moisture supply dominate, the local land surface’s contribution of water vapor to the column becomes relatively unimportant, and heavy rainfall itself suppresses surface heating. In the transition seasons, by contrast, when large-scale forcing is weaker, evapotranspiration from vegetated, moisture-rich land emerges as a meaningful contributor to the spatial organization of instability.

That spatial organization is consistently tied to what covers the ground. The study finds that patterns of NCII are reliably associated with land-use and land-cover characteristics, with vegetated and moisture-rich regions generally exhibiting higher instability than arid or barren areas. Forests, irrigated agriculture, and wetlands effectively pump moisture into the boundary layer, lowering the lifting condensation level and sharpening local gradients in equivalent potential temperature that can focus convective initiation. This connection between vegetation and instability has implications that extend beyond academic interest, because land-use change across India, from deforestation to agricultural expansion to urbanization, may be quietly reshaping where the country’s storms prefer to form.

The broader significance of the work lies in the framework it endorses: a conditional, scale-dependent view of land-atmosphere interaction. Rather than asking whether soil moisture causes rainfall in some universal sense, the study argues that land-surface processes are associated primarily with the spatial organization of thermodynamic instability, while atmospheric moisture, stability, and dynamical forcing jointly determine whether convection and precipitation actually materialize. This division of labor resolves a persistent tension in the literature, where some studies have reported strong soil moisture-precipitation feedbacks and others have found them elusive. Both camps may be right, just in different seasons, regions, and atmospheric regimes.

For a country whose economy remains deeply sensitive to rainfall variability, the practical stakes are considerable. Understanding that the land surface matters most during the post-monsoon and winter transition seasons, and least during the height of the monsoon, could help refine the interpretation of high-resolution model forecasts and guide where improvements to land-surface initialization would yield the greatest benefit. The study also demonstrates the power of open, cloud-based Earth observation: by stitching together NASA satellites, European reanalysis, precipitation missions, and lightning networks within a single analytical environment, it shows how the pieces needed to decode one of the world’s most consequential weather systems are now freely available to any researcher with a laptop and a good question. As climate change continues to warm the troposphere and alter land-cover patterns across South Asia, the seasonal choreography of land and atmosphere documented here will not stay still, making this kind of process-based monitoring an essential tool for the decades ahead.

Subject of Research: Seasonal land–atmosphere coupling and convective instability over India

Article Title: Seasonal modulation of land–atmosphere coupling and convective instability over India: process-based insights into evapotranspiration controls and conditional convection

Article References: Dutta, D. (2026). Seasonal modulation of land–atmosphere coupling and convective instability over India: process-based insights into evapotranspiration controls and conditional convection. Climate Dynamics, 64(11), Article 442. https://doi.org/10.1007/s00382-026-08411-w

Image Credits: AI Generated

DOI: 10.1007/s00382-026-08411-w

Keywords: land-atmosphere coupling, evapotranspiration, convective instability, CAPE, CIN, CTP-HI, Indian monsoon, lightning, MODIS, ERA5, soil moisture feedback, Climate Dynamics

Soil moisture is not a single reservoir but a layered system. Water near the surface evaporates rapidly and responds almost immediately to rainfall deficits, while deeper layers hold reserves that plants tap during dry spells. In a typical drought, this stratification acts as insurance: shallow soils may desiccate, but deep moisture sustains roots through the lean period. The new research shows that this insurance is being increasingly withdrawn. When moisture deficits propagate downward and affect the entire soil profile simultaneously, ecosystems lose their vertical buffer and the damage to vegetation can escalate sharply.

The research team carried out what they describe as a multi-source, multilayer analysis, cross-validating datasets that capture moisture at different depths. This methodological advance matters because most previous drought assessments relied primarily on surface indicators, such as the top few centimetres of soil or indices derived from precipitation and temperature. Such indicators can miss deep-soil depletion entirely, underestimating the true stress experienced by ecosystems whose roots reach metres below the surface. By incorporating subsurface information, the study uncovered a class of drought events that conventional metrics largely overlook.

Central to the study is the concept of vertical hydrological buffering. Under normal conditions, water moves through the soil profile in response to gravity, capillarity and plant uptake, creating a redistribution system that smooths out short-term rainfall variability. Deep layers recharge during wet periods and slowly release stored water during dry ones. When droughts become vertically compound, this buffering mechanism collapses: surface layers are dry, so evaporation demand cannot be met; subsurface layers are dry, so roots cannot compensate; and the entire column is locked in deficit. The result is drought stress of a fundamentally different magnitude from that recorded by surface-only measures.

The analysis shows that these vertically compound droughts have intensified across much of the globe in recent decades. Regions that experienced them saw markedly stronger losses of vegetation productivity than areas affected by surface-only droughts of comparable severity. Forests and croplands emerge as particularly vulnerable. Deep-rooted forests depend on subsurface reserves during extended dry periods, and croplands, whose productivity is tied tightly to the water available within the root zone, suffer yield consequences when the entire profile dries out. For global carbon accounting, this is a worrying signal: diminished carbon uptake in forests during compound droughts weakens the terrestrial carbon sink precisely when atmospheric carbon dioxide is climbing.

The mechanistic links between layered soil moisture and ecosystem response are grounded in established land-atmosphere coupling theory. Soil moisture controls how much energy from solar radiation goes into evaporation versus heating the air. When moisture is plentiful, evaporation cools the surface; when it is depleted, more energy converts to sensible heat, raising air temperatures and vapour pressure deficits, which in turn forces plants to close their stomata and curtail photosynthesis. Vertically compound droughts thus feed a feedback loop: drier soils intensify atmospheric heat and dryness, which further stresses vegetation and dries the soil still more. By depleting all layers, these events extend and deepen the loop.

The study builds on a growing body of work on the vertical evolution of soil moisture drought. Recent research has classified soil drought across individual soil layers and reported that human activities can enhance subsurface drought development, while other studies have shown that vigorous vegetation growth in one season can exacerbate soil moisture drought in the next through trans-seasonal land-atmosphere interactions. Advances in observational capability have also been critical: machine-learning-based data fusion now produces high-resolution, seamless multilayer soil moisture estimates, and ambient seismic noise has been used to map large-scale deep soil moisture variations from ground vibrations. Together, these developments have made it possible to track drought not just across space and time, but down through the soil column.

The implications for drought monitoring and risk management are substantial. Indices built on surface conditions alone may declare a drought over when rainfall returns, even as deep moisture remains severely depleted, leaving ecosystems exposed to what the authors’ findings suggest can be amplified damage in subsequent stress events. Effective early-warning systems would need to incorporate subsurface measurements and model the depth profile of moisture, identifying when vertical compounding is occurring or imminent. For water resource managers and agricultural planners, the distinction between a shallow dry spell and a full-profile deficit is the difference between a transient inconvenience and a structural threat to yields and biomass.

Climate change amplifies the concern. Warmer temperatures increase evaporative demand, drawing moisture from the soil faster and pushing deficits deeper into the profile. More intense and irregular precipitation patterns reduce the frequency of the steady, soaking rains that recharge subsurface layers, favouring intense bursts that run off rather than infiltrate. The intensification of vertically compound droughts documented in the study may therefore be a signature of a shifting hydrological regime in which the deep-soil safety net that ecosystems have historically relied upon is progressively eroding. Hotter droughts, in effect, are becoming deeper droughts.

For the world’s forests, the stakes extend beyond the trees themselves. Forest carbon uptake represents a substantial share of humanity’s annual carbon budget, and declines in productivity during drought years measurably weaken the sink. If vertically compound droughts continue to intensify and expand, the study’s findings suggest that global models calibrated on surface drought indicators may systematically underestimate future carbon losses, biasing projections toward optimism. Similarly, cropland vulnerability assessments that ignore the depth dimension of drought risk may leave food production systems inadequately prepared for the full-profile deficits that the analysis identifies as especially damaging.

The researchers emphasise that their global analysis synthesises multiple independent data sources, lending robustness to the pattern they detect, though each dataset carries its own uncertainties in vertical resolution and depth coverage. Continued improvements in satellite missions, ground sensor networks and data assimilation methods are expected to sharpen the picture of how moisture deficits propagate through the soil. What is already clear is that drought science must look down as well as out: the vertical structure of soil moisture is not a technical detail but a first-order control on how severely droughts hit ecosystems. As the planet warms, the research warns, the simultaneous drying of every soil layer may become one of the defining hazards of the coming decades, silently stripping ecosystems of the subsurface reserves that have long sustained them through the driest times.

Subject of Research: The amplification of global ecosystem drought impacts by multilayer soil moisture deficits

Article Title: Multilayer soil moisture deficit amplifies drought impacts on global ecosystems

Article References: Multilayer soil moisture deficit amplifies drought impacts on global ecosystems. (2026). Nature Geoscience, 19(10), 1149–1150. https://doi.org/10.1038/s41561-026-02082-2 (Author names are not included in the publisher’s citation export.)

Image Credits: AI Generated

DOI: 10.1038/s41561-026-02082-2

Keywords: soil moisture, drought, vertically compound drought, hydrology, climate change, forests, croplands, carbon uptake, soil profile, land-atmosphere coupling, ecosystem productivity, Nature Geoscience