Hydrological extremes

Climate variations affect access to recreational opportunities through different mechanisms that influence the provisioning of water and ecosystem services. From 2000 to 2022, Northern and Central California experienced extreme climate fluctuations, with alternating multi‐year dry spells and wet years causing highly variable runoff from the Sierra Nevada into the Sacramento‐San Joaquin Delta. This study assesses how these hydroclimatic variations influence regional recreation by analyzing summer activities (boating, day use) alongside winter activities (birding, waterfowl hunting) at state parks and wildlife areas. The objective is to evaluate how monthly visitor volume and timing shift during drought versus wet conditions, distinguishing between activities directly or indirectly dependent on water resources. During summer droughts with critical low runoff, recreational activity declines sharply; boating drops by 65.8% and day use decreases by 13.8% at state park units. Conversely, winter droughts correlate with increased recreational engagement. Critical low runoff is associated with a 23.4% increase in birding observations and a 50.2% increase in waterfowl hunting takes. This winter increase occurs because restricted wetland resources concentrate waterfowl populations, making them more accessible to visitors. Activities directly dependent on water resources, including recreational fisheries, show the highest sensitivity to runoff fluctuations. We also find that visitor spending is highest during critical drought and wet runoff extremes, reflecting winter demand in drought years, and summer demand in the wettest years. These results highlight the vulnerability of water‐dependent recreation and its economic contributions as climate change accelerates the frequency and magnitude of extreme wet and dry cycles.

Nearly half of Earth’s seasonal freshwater storage variability occurs in human‐managed reservoirs, resulting in large areas of reservoir‐bed that are episodically inundated or subaerially exposed. Here we use hyperspectral data from NASA’s Earth Surface Mineral Dust Source Investigation (EMIT) mission to map calcite mineral assemblages across exposed surfaces in two large U.S. reservoirs, Lake Mead and Lake Powell. Results show distinct patterns in the occurrence of carbonate, clay, and mica minerals. Reservoir‐deposited sediments exhibited significantly higher calcite proportions than adjacent, lithologically similar, non‐inundated terrestrial surfaces. These differences likely reflect hydrologically‐driven calcite precipitation, wherein warm, clear, productive reservoir waters drive calcium carbonate out of solution. We also detect reductions in calcite proportions in areas of the shoreline disturbed by human recreation. EMIT's imaging capabilities enable rapid, high‐resolution mineralogical inventories across vast areas where field sampling is infeasible. We find EMIT is a scalable tool for monitoring geochemical changes associated with large water level fluctuations.

Reliable flood change projections under climate change are critical for water resources planning yet remain challenging due to uncertainty in climate‐informed statistical models. Conventional approaches relate flood distribution parameters to climate variables using log‐linear formulations, which can produce unrealistically large flood estimates. In this study, we develop a semi‐parametric statistical framework that constrains flood variability using a bounded logistic link function and estimates parameters from hydroclimatically analogous samples. Using 548 natural basins across the coterminous United States, we evaluate the proposed approach against log‐linear formulations from the entire sample. Model performances are assessed based on both the accuracy and sensitivity of flood quantiles to climate forcings. Results show that the proposed framework yields more stable and realistic flood quantile estimates, particularly under high precipitation conditions, and better preserves flood sensitivity. These findings highlight the value of bounded, locally informed statistical models for robust flood risk assessment under climate change.

This study investigated the roles of two different ensemble-boosting approaches in numerical weather modeling (i.e., storm transposition and time-lagged approaches) for estimating worst-case precipitation scenarios associated with mesoscale convective systems (MCSs). As a case study, we focus on a record-breaking extreme precipitation event that occurred in Kumamoto, central Kyushu, Japan, on 10 August 2025. All numerical experiments were conducted using the Cloud Resolving Storm Simulator (CReSS). Time-lagged ensembles were generated by varying initialization times, while storm transposition ensembles were produced by meridionally shifting the background atmospheric fields in the control (CTL) runs at 0.05° increments. Our analysis revealed that both the storm transposition and time-lagged ensemble approaches produced significantly intensified maximum precipitation scenarios over the target watersheds. However, these scenarios exhibited distinct differences in precipitation characteristics and environmental drivers, reflecting the fundamental conceptual divergence between the two methods. Storm transposition not only shifted the spatial distribution of rainbands but also enhanced low-level convergence through interactions with the complex topography of western Kyushu. In contrast, time-lagged ensembles produced diverse spatiotemporal patterns by sampling initial condition uncertainties, leading to variations in convection initiation and rainfall duration. These results indicate that worst-case precipitation estimates are highly sensitive to the choice of ensemble generation. While storm transposition captures deterministic shifts in meso- α -scale features, time-lagged ensembles represent internal chaotic evolution. Therefore, considering both spatial displacement and initial condition uncertainty is important for a more comprehensive understanding of the upper bounds of extreme precipitation.

Concurrent floods and concurrent droughts in nearby catchments pose challenges to risk assessment and water management. Climate change is affecting extremely high and low discharge, but the complex interplay between changes in individual catchments, and in the dependence across catchments, make it difficult to provide accurate assessments of the occurrence probabilities of concurrent extremes. In this work, we use a contemporary statistical deep learning model to capture concurrent river floods and droughts in four catchments in the Upper Danube basin, based on discharge simulated by a hydrological model driven with large ensemble climate model output. The statistical model is able to accurately capture the tail behavior of the simulated discharge, which we assess by making use of the large available sample. We subsequently use our statistical model to study changes in joint tail behavior of discharge over time, finding that both compound flooding and drought‐like conditions are becoming increasingly likely toward the end of the 21st century under a high‐emission scenario. In particular, our results highlight that changes in the dependence structure of extremes strongly contribute to the detected changes, an aspect that would be difficult to capture with traditional approaches. This work paves the way for highly flexible, general inference on compound extremes in hydrological applications, and demonstrates key advantages of using statistical deep learning in this setting.

Hydrological extremes display long‐term trends and natural oscillations in response to climatic and socioeconomic factors. Knowing how and why these extremes are shifting is crucial for increasing water resilience in Europe. This work analyses trends in floods and droughts in European catchments between 1951 and 2020. Using the OS LISFLOOD hydrological model and non‐stationary extreme value analysis, we disentangle the effects of dynamics in climate, encompassing climate variability and climate change, land use, water abstraction and reservoirs on river flow extremes. Our results reveal spatial and temporal variations in magnitude of hydrological extremes. Land use changes have generally increased flood and drought hazards, while water demand primarily intensified droughts. Reservoirs, conversely, have decelerated the hydrological cycle. Our work delivers insights into the intricate connections between climate, society, and water in European river basins, enabling the development of effective strategies for enhancing resilience to extreme water events under global change.

The Amazon carbon sink is weakening as deforestation 1–5 , rising tree mortality 6–8 and climate change disrupt the climatic conditions that sustain a humid tropical forest 9–11 . Forest loss reduces precipitation, increases temperatures and lengthens the dry season, driving higher carbon emissions 1–4,11,12 . Climate stress is stronger in more deforested regions, rendering forests progressively more flammable 1,9,13 . Large-scale climate events, such as El Niño, further amplify regional climate impacts 9,14–17 . Here we present an analysis of 15 years of Amazonian carbon fluxes derived from atmospheric CO 2 measurements. We determine a carbon balance of 0.32 ± 0.08 PgC yr −1 , with fires acting as the dominant source (0.40 ± 0.07 PgC yr −1 ) and forest uptake becoming increasingly unable to offset emissions. Furthermore, we investigate key drivers of the carbon emissions of Amazonia and identify three distinct year groups. Group 1 (2011–2014, 2017 and 2018) represents the role of the Amazon in the global carbon balance (0.11 ± 0.08 PgC yr −1 ) during wetter years with low human activity, such as deforestation, biomass burning and degradation. Group 2 (2019–2022) represents Amazon carbon emissions (0.43 ± 0.13 PgC yr −1 ) from human activities (mainly deforestation and degradation) during years with low climatic stress. Group 3 (2010, 2015, 2016, 2023 and 2024) has the highest carbon emissions (0.49 ± 0.07 PgC yr −1 ) and comprises warm and dry years (high climate stress), such as El Niño and the warming of the Tropical North Atlantic Ocean, associated with biomass burning.

The French Alps, characterised by complex topography and diverse climatic influences, are highly vulnerable to extreme precipitation events, which can trigger floods and landslides with significant socio-economic impacts. While local stakeholders report a perceived increase in sub-daily extreme precipitation events, such a trend has not yet been statistically confirmed based on observations. In this study, we apply extreme value theory to long time series of in situ observations of daily and hourly precipitation to provide an updated climatology and trend analysis of precipitation extremes across the region. We focus on 20-year return levels to analyse rare, potentially high-impact events and show that precipitation extremes exhibit distinct spatial patterns with marked differences between daily and hourly timescales. By examining trends at monthly and seasonal levels and making use of an existing weather pattern classification, we gain a deeper understanding of the dynamics underlying these trends. For extremes of daily precipitation intensity, we observe strong positive trends in the southern French Alps and negative trends in the north, primarily driven by a stronger influence of southerly (Mediterranean) flows in autumn. The spatial patterns for hourly extremes show more spatially variable trends, and seem to be primarily influenced by local thermodynamic effects potentially linked to e.g. changing moisture availability or changes in surface albedo due to shorter snow cover persistence. This detailed analysis provides important insights for stakeholders who are working on prevention and protection from precipitation-related hazards, underscoring the need for adaptive strategies in response to changing precipitation patterns. It also highlights the importance of region-specific and timescale-specific approaches in future studies of precipitation extremes.

Extreme floods rank among the most destructive natural hazards, disrupting economic development, societal stability, and river ecosystems. Although macro‐scale flood risk disparities across poverty‐classified regions have drawn increasing attention, they remain critically underexplored in China, the world's most flood‐vulnerable country, where flood‐induced human displacement accounts for one‐third of the global total. Here, we develop a rigorously validated, macro‐scale framework to estimate flood‐induced economic losses across China and to quantify the disparity of these impacts across distinct poverty strata. We find that impoverished regions experience disproportionately severe relative impacts, with their flood losses escalating by at least 22% for each 0.5°C of global warming, outpacing non‐poor regions by up to seven percentage points. Furthermore, economic growth within projected flood‐prone areas markedly exacerbates these losses. Although enhancing flood‐defense standards can mitigate damages, maintaining historical loss levels requires ambitious defense standard upgrades of 80% (under 1.5°C of warming) and 95% (under 2.0°C), corresponding to an average 16‐percentage‐point improvement per 0.5°C of warming. Our findings reveal the accelerating flood risk and macro‐scale disparity faced by impoverished regions in China, a dynamic long overlooked, and challenge the adequacy of historical flood risk management strategies. As nations work toward the United Nations Sustainable Development Goals (SDGs), particularly SDG 10 (Reduced Inequalities), which intrinsically requires a foundational understanding of risk disparities, our study provides critical baseline evidence for national adaptation planning and offers globally transferable methodologies.

Atmospheric rivers (ARs) play a crucial role in the global hydrological cycle, yet how the intrinsic midlatitude atmospheric variability modulates their activity on subseasonal timescales remains poorly understood. Here, we show that AR frequency and landfall in both hemispheres vary strongly with the strength and phase of the Baroclinic Annular Mode (BAM). During its positive phase, BAM increases AR occurrence by up to ∼50% and AR‐driven precipitation extremes by up to ∼40% relative to climatology. Stronger and more frequent ARs during the positive phase of the BAM are primarily driven by enhanced potential‐to‐kinetic energy conversion, which favors AR growth and intensification, particularly in the storm‐track regions. This response is consistent with dynamically enhanced poleward moisture transport due to a more baroclinically unstable and stormier environment. The results underscore the importance of BAM‐related intrinsic midlatitude storm‐track variability in modulating global AR activity and its potential to improve extended‐range AR prediction on subseasonal timescales.

Deterministic model predictions can struggle to adequately capture extreme events such as floods and droughts, which are of particular importance in hydrology. This limitation arises because deterministic models collapse the conditional discharge distribution to a single point estimate. Probabilistic modeling provides a way to address this issue by explicitly representing uncertainty and assigning non-zero probabilities to all possible outcomes, including extreme events, thereby capturing the entire conditional discharge distribution. Motivated by this perspective, we examine whether probabilistic Long Short-Term Memory (LSTM) models improve the representation of extreme events in rainfall–runoff simulations across Switzerland. The differing degree of flexibility in the models leads to different degrees and distinct patterns of miscalibration in the predictive distributions. For the extreme discharges (top 0.05 % of the marginal discharge distribution), the deterministic LSTM systematically underestimates 97 % of these events, whereas 83 % of these fall within the probabilistic model's 99th percentile prediction interval. Given the rarity of these events, representing them as outcomes near the upper tail of the predictive distribution is both expected and statistically consistent. Building on the additional information provided by probabilistic runoff predictions, we further show how they can be translated into actionable flood warnings using reinforcement learning. To this end, we introduce a Flood Risk Communication Agent (FRiCA) that operates on probabilistic runoff predictions and learns risk-aware decision rules for issuing warnings of varying intensity. Results indicate that the FRiCA outperforms simple fixed heuristics, such as issuing warnings based on the deterministic LSTM predictions or a fixed high quantile (e.g., the 99th percentile). The learnt behavior demonstrates the potential of reinforcement learning as a decision layer for improved flood risk communication, and motivates further exploration of better reward design and policy network optimization for context-dependent decision policies that adapt to varying hydrological and societal contexts.

Tribal nations across the United States face disproportionate flood impacts yet remain significantly underrepresented in national‐scale assessments, creating a critical equity gap in climate adaptation. We address this deficiency through a climate‐to‐local‐inundation framework for the Santee Sioux Nation, Nebraska, by forcing 10‐m hydrodynamic simulations with a 42‐year paired historical and pseudo‐global‐warming CONUS404 runoff event set. Results show that projected changes in inundation frequency and flood depth are limited, with no evidence of widespread intensification of flood extremes. Instead, changes manifest as localized adjustments in floodplain activation and tributary connectivity. Pixel‐wise return‐period analysis shows small changes in flood depth and only minor deviations in the statistical structure of extremes. These findings demonstrate a transferable pathway for translating climate‐driven hydrologic forcing into locally resolved inundation information for under‐mapped Tribal lands.

Flood peaks can develop and recede within a day, making them difficult to capture with satellites with multi‐day revisit intervals. Here, we exploit daily observations during the Calibration and Validation phase of the SWOT mission to examine a short‐lived flood in May 2023 along the Du Gouffre River in Canada. Despite channel widths below 100 m, SWOT tracked day‐to‐day changes in water surface elevation, capturing a rapid rise of approximately 5 m consistent in timing and magnitude with in situ measurements. We derived flood extent based on classification and confidence information. The results show partial spatial overlap with an independent post‐event flood hazard map, with differences primarily associated with reduced classification confidence in vegetated floodplains and narrow reaches. Some areas not identified as water may reflect erosion, sediment deposition, or post‐peak conditions. These results highlight the value of high‐frequency wide‐swath radar altimetry for monitoring transient floods.

Understanding changes in extreme precipitation is critical for enhancing climate resilience, particularly in regions such as West Africa, where communities are highly vulnerable to hydroclimatic extremes. This study examines the spatial and temporal variations in extreme precipitation across West Africa over 1998–2024 using high-resolution satellite data from NASA’s Integrated Multi-satellitE Retrievals for GPM (IMERG) V07B Final Run product. Extreme precipitation is analyzed using daily accumulated precipitation indices, including absolute and threshold-based indices. The Theil-Sen slope estimator and the Mann-Kendall test are applied to identify statistically significant trends at three significance levels. To assess the robustness of the detected trends, results from IMERG are compared with those from the Global Precipitation Climatology Project (GPCP) Version 3.3 Climate Data Record over the same period. The comparison shows generally similar spatial patterns and directions of change between the two datasets across West Africa. Results reveal distinct regional patterns of change. The Guineo-Congolian and Guinean regions, particularly Nigeria’s eastern coast, show significant declines in annual precipitation, along with significant decreases in the frequency and intensity of extreme events. In contrast, the western Sudanian and Sahelian regions, notably southern Mali and Burkina Faso, exhibit increasing trends in both total precipitation and extreme precipitation indices, indicating a shift toward wetter conditions driven by more frequent and intense extreme events. Furthermore, the contribution of extreme precipitation to annual totals varies across climatic zones, with smaller contribution from extreme days at lower latitudes. These findings highlight the importance of regionally tailored climate adaptation strategies for effective flood risk management and sustainable water resources.

Sub-daily quantitative precipitation forecasts (QPF) are a primary input to flood early warning, yet over Brazil they are hampered by convective, intermittent rainfall and a sparse observing network. We present a machine-learning post-processing system (V2.0C) producing 3-hourly QPF on a 0.25° grid by coupling a LightGBM precipitation-occurrence classifier with a Tweedie amount regressor and a dual-gate routing scheme with a physically guarded heavy-rain (≥5 mm) hard route. Its 71-dimensional feature set is dominated by GEFS ensemble statistics. The system is verified over two independent windows — April–May 2026 and the complete austral-summer convective season (October 2025– March 2026) — against the IMERG satellite product (areal) and the INMET automatic-station network (point). Against IMERG, V2.0C attains the highest correlation (0.324 and 0.286) and a heavy-rain CSI (≥10 mm) of 0.063–0.071, 3.3–4.4 times raw GFS, with stable month-to-month skill through the SACZ peak. A common mask analysis localizes the 60–65% ensemble-driven correlation gain to the southern third of the domain, where GEFS statistics are the only direct precipitation predictors available. Controls show that the guarded routing makes a high-recall heavy-rain expert safe to deploy: it adds 6–10% heavy-rain detection over a no-expert control, whereas unguarded use halves heavy-rain CSI. Gauge verification reproduces the heavy-rain advantage (CSI ≥10 mm an order of magnitude above raw GFS) and the wet bias, and the point–area correlation gap is explained by sub-grid representativeness. The detection-oriented bias is removable a posteriori: leakage-free quantile mapping eliminates it (bias +0.02 mm, MAE −69%) without reducing detection, supporting a dual-product strategy for flood-warning and hydrological applications.

The zenith total delay (ZTD) retrieved from 5G communication base stations represents a novel observation source for numerical weather prediction (NWP), leveraging the extensive global deployment of 5G infrastructure. In this study, the direct assimilation of 5G‐ZTD observations into a high‐resolution NWP system was accomplished for the first time. Two precipitation events were selected to evaluate the impact of 5G‐ZTD assimilation on short‐term precipitation forecasting. Assimilating 5G‐ZTD data corrected the dry bias in the initial water vapor field and improved precipitation forecasts in both cases, demonstrating its practical utility for short‐term heavy rainfall forecasts and operational potential given its low cost and extensive coverage.

We combine CONUS404 convection‐permitting climate simulations with the physics‐based WRF‐Hydro model to assess how climate warming may influence flood frequency and seasonality across the Midwestern United States. This framework captures interactions among precipitation, soil moisture, and snowpack that shape flood generation. Both simulated warming and historical records show a regional shift toward more extreme warm‐season, rainfall‐driven floods. This regional signal, however, does not translate uniformly to individual watersheds, where flood responses reflect not only warming but also the influence of subregional climate variability and sampling uncertainty. Local flood‐frequency changes thus remain highly uncertain despite the thermodynamic tendency toward stronger convective precipitation. Ensembles of convection‐permitting simulations are needed to constrain plausible local outcomes. These findings suggest that flood risk communication under climate warming should distinguish broad climate‐driven tendencies from uncertain local responses.

Floods are among the most frequent and destructive natural disasters, yet future flood risk assessments have largely relied on precipitation indices, overlooking runoff depth—a variable that more directly captures flood hazard under diverse regional hydrological conditions. Moreover, the interactions among risk components (hazard, exposure, vulnerability) remain underexplored. Here we integrate 14 multidimensional sub‐indicators, including both precipitation‐ and runoff‐based indices, to assess global rainstorm flood risk from 1991 to 2100 under four SSP scenarios (SSP1‐2.6, SSP2‐4.5, SSP3‐7.0, SSP5‐8.5). Comparative experiments reveal that omitting runoff depth systematically overestimates high‐risk areas and brings forward the time at which risk projections begin to diverge across scenarios. Global risk increases persistently, with SSP5‐8.5 showing the highest rise by 2100, and high‐risk regions are concentrated in the Americas, Africa, South Asia, East Asia, and Southeast Asia. Exposure dominates risk, while hazard and vulnerability play regionally distinct modulating roles. Alarmingly, by 2100, half of the global population (up to 6 billion under SSP3‐7.0) and half of global GDP (under SSP5‐8.5) will face the highest risk level, with South and East Asia accounting for half of that population (or GDP) due to dense populations and infrastructure. This study advances flood risk assessment by reconciling hydrological complexity with socioeconomic pathways and highlights that mitigation strategies must prioritize emission reductions and the relocation of people and assets from flood‐prone areas.

Groundwater plays a critical role in sustaining the water supply in the Colorado River Basin (CRB), a region experiencing persistent drought, population growth, and increasing water demand. Using GRACE/GRACE‐FO satellite observations combined with hydrological data sets, we analyze basin‐scale variations and long‐term trends in terrestrial and groundwater storage across the CRB from water year 2004–2023. We further apply a joint framework integrating copula‐based statistical modeling with statistical inference to quantify probabilistic relationships among storage‐based drought indicators and to characterize groundwater–surface water interactions. Results reveal sustained basin‐wide groundwater depletion over the past two decades, with accelerated losses since 2019 and during the water year 2020–2022 drought. Notably, during this recent drought period, groundwater losses were larger in the typically more resilient Upper CRB (17.5 ± 3.5 km 3 ) than in the Lower CRB (9.4 ± 6.5 km 3 ), indicating increasing vulnerability across both subbasins. Copula‐based analysis shows strong coupling between surface‐water deficits and groundwater drought probability in the Upper CRB, whereas parts of the Lower CRB exhibit weaker statistical dependence, consistent with groundwater extraction operating partly independently of surface water availability under chronically limited supply conditions. These findings provide a basin‐wide observational synthesis of groundwater–surface water dynamics and demonstrate the value of satellite‐based monitoring for diagnosing hydroclimatic stress in large managed river basins. The results highlight emerging shifts in groundwater vulnerability across the CRB and underscore the need for integrated water‐management strategies that jointly consider climatic pressures and sustained anthropogenic demand.

Wetlands play a crucial role in maintaining the global ecosystem balance, and determining the inundation dynamics is of great significance for understanding wetland ecosystems. Hydrodynamic models are effective tools for mapping the inundation dynamics; however, they are affected by uncertainties arising from environmental changes, including anthropogenic pressures and land cover modifications, especially in large wetlands. To overcome hydrodynamic model limitations, this study integrated a hydrodynamic module, a remote sensing module, and a deep learning fusion module to develop an efficient approach for mapping the inundation dynamics. This approach was applied in the Poyang Lake Wetland in China, a vital conservation wetland in the Ramsar Convention. The effectiveness of this approach was verified by comparing the simulated inundation dynamics of the hydrodynamic model during the dry season. We found that the performance of the hydrodynamic model was relatively poor. In particular, the performance in the sub‐lake area was significantly inferior than the other zone, with an overall accuracy (OA, dimensionless, ranging from 0 to 1) of only 0.63–0.66. The proposed approach in this study exhibited better performance, especially in the sub‐lake area, with an OA increasing to 0.82–0.92. Furthermore, the hydrodynamic model tended to overestimate the inundation time, with the median overestimated time reaching 42–50 days. Our study provides a transferable framework that can improve the accuracy of wetland hydrological monitoring under changing environment, contribute to maintaining the ecological integrity and stability.

Flood depth estimation from social media images provides timely and cost‐effective observations, offering significant potential for flood monitoring and emergency response. Existing image‐based methods typically rely on humans, vehicles, or traffic signs as scale references, however, such objects are frequently absent in real flood scenes and often yield unreliable scale information. In contrast, urban 3D models offer consistent and accurate metric references across large spatial extents. In this paper, we propose a flood depth estimation method based on urban 3D models and develop an Object‐level Coarse‐to‐Fine Matching Model that transfers metric scale from 3D models to social media images. The results demonstrate that our model significantly outperforms the representative 2D3D‐MATR on an urban public benchmark, increasing the inlier ratio by approximately 30% and improving the registration recall by about 16%. Building upon the matching capability, our 3D‐based flood depth estimation method generalizes across diverse flood scenarios and achieves improved accuracy, with a Mean Absolute Error (MAE) of 0.164 m and a Root Mean Square Error (RMSE) of 0.212 m. By transforming discrete social media data into spatially explicit, high‐accuracy flood depth maps, our method not only provides reliable support for rapid flood assessment and emergency decision‐making during events but also offers valuable data for calibrating hydraulic models post‐disaster.

Global warming is increasing the frequency and intensity of extreme droughts, threatening terrestrial ecosystem structure and function. However, existing indices often fail to provide a comprehensive representation of the water cycle, and high-resolution datasets for characterizing vegetation responses to drought remain limited. To address this gap, we first developed a monthly Vertical Composite Drought Index (VCDI) by integrating precipitation, surface runoff, soil moisture, and groundwater storage anomalies. This index summarizes multidimensional water constraints across the “atmosphere-surface-subsurface” continuum and captures cumulative effects of drought across multiple timescales. Furthermore, VCDI, land surface temperature, and vegetation greenness were combined to produce a monthly Vegetation Drought Resistance Index (VDRI) dataset for China at 250 m resolution from 2000 to 2022. Multi-source evaluations indicate that VDRI provides spatially detailed and temporally consistent information on vegetation conditions under concurrent hydroclimatic and thermal stress. The datasets support ecosystem assessment, ecological drought monitoring, wildfire risk analysis, and benchmarking for drought-impact studies and related ecological models.

Monsoon rainy seasons shape water supply and flood hazard across northern East Asia, but their long-term variability remains poorly understood because few daily rainfall records extend beyond the instrumental era. Here, we reconstruct rainy-season onset, retreat, and duration from a 256-year daily precipitation record for Seoul (1770–2025), providing an unusually long view of seasonal monsoon behavior. The reconstruction reveals structured quasi-decadal to multidecadal variability in rainy-season duration, indicating that recent rainy-season lengthening over northern East Asia is consistent with the latest positive excursion of a recurrent multidecadal modulation rather than a purely isolated modern anomaly. Coupled with warming-associated intensification of extreme rainfall, this prolonged-duration phase has elevated the compound-hazard probability, with recent probability peaks exceeding the long-term mean of ∼25% in Seoul. Exploratory near-term outlooks suggest that rainy-season duration in northern East Asia is likely to remain prolonged toward mid-century, sustaining heightened compound hydroclimatic hazard when seasonally extended overlaps with intensified event-scale rainfall.

Global warming has intensified extreme precipitation, a trend projected to continue as warming persists. However, how extreme precipitation characteristics respond to global warming mitigation strategies remains poorly understood. This study examines East Asian summer extreme precipitation responses under a net-zero anthropogenic carbon dioxide (CO 2 ) emission scenario. Despite substantial cooling over East Asia following atmospheric CO 2 concentration decline, the intensity of extreme precipitation events continues to increase for about 30 years and does not subsequently weaken. This counterintuitive result arises mainly from intensified upward motion during extreme precipitation events, driven by enhanced baroclinicity and strengthening of the upper-tropospheric jet, which together overcompensate the cooling effect. These are essentially due to the hysteresis in the Atlantic (i.e., Atlantic Meridional Overturning Circulation slowdown), Southern (i.e., sustained warming), and Pacific (i.e., El Niño–like warming) Oceans. Our findings highlight that achieving net-zero anthropogenic CO 2 emissions is insufficient to mitigate intensifying extreme rainfall events in the densely populated East Asian countries.

Water bodies such as lakes and reservoirs can play a crucial role in reducing estimated flood peaks, at both daily and hourly time resolution. While the effect of water bodies on flood peaks has been demonstrated in the past at different time resolutions, it remains unclear how water bodies affect the ratio between estimated daily and hourly quantiles, that is, to which degree their attenuation effect is scale dependent. Here, we analyse how water bodies influence this ratio using two approaches: (1) by comparing estimated daily and hourly flood peak quantiles upstream and downstream of reservoirs of four local case studies, and (2) by comparing daily and hourly flood peak quantiles across a large sample of catchments in Switzerland with various degrees of water-body influence. Our results show that reservoirs dampen hourly peak discharge much more strongly than daily peak discharge, which leads to similar daily and hourly flood peaks downstream of reservoirs. Specifically, our case study analysis highlights that (sub-)hourly peak discharge is attenuated by up to 70 % downstream of reservoirs during flood events with a 10-year return period. We also find that the attenuation effect is particularly pronounced in catchments that are heavily influenced by water bodies, i.e. those catchments where more than 60 % of the area contributes to water body inflow. This implies that daily flood peaks are a good proxy for hourly peaks in strongly regulated catchments in contrast to natural ones. We conclude that considering water body influence on flood peaks is crucial to understand the similarity between daily and hourly flood peaks downstream of water bodies. This suggests that water body influence should also be considered in large-sample studies for which the exact influence may be unknown, and quantified using suitable metrics such as the fraction of the total catchment above the water body.

Droughts and dry periods represent a novel and emerging threat to agriculture in historically water rich, temperate regions, such as the Northeast United States. Water governance institutions to manage water quantity in these regions are underdeveloped in the context of current needs. In this study, we examine the state of Maine as a case study in drought institutions in a historically water‐rich region. Specifically, we review the institutional landscape for drought resilience and resistance for Maine agriculture and identify the gaps that may be hindering agricultural adaptation. Through focus groups with farmers and agricultural advisors, we find: (a) There are few quasi‐governmental organizations and programs dedicated to agricultural drought‐adaptation; rather farmers rely upon mostly federal, state, and non‐governmental regional programs and policies that are non‐drought specific in nature; and (b) there is a need for articulated and persistent coordination between federal‐state‐local organizations and programs, to ensure vertical and horizontal integration and complementarity. In the absence of drought‐specific institutions that coordinate across scales, there is the potential for ineffective drought adaptation and maladaptation in Maine and other temperate regions. This could result in wasted public and private resources, ineffective use of natural resources, and inequitable or otherwise socially undesirable outcomes. Our research adds to the growing body of evidence showing that, even in temperate and historically water‐rich regions, drought is a threat to agricultural viability. Here, we outline the need for institutional coordination around drought‐adaptation to facilitate strategic program development.

Drought assessment is an extremely complex task that relies on high-quality datasets, posing a major challenge for scientists and water managers. Here, we present the Global Multi-Index Drought (GMID), a multiscalar high-resolution dataset that includes five standardised drought indices: the Standardised Precipitation Index (SPI), the Evaporative Drought Demand Index (EDDI), the Standardised Precipitation Evapotranspiration Index (SPEI), the Standardised Evapotranspiration Deficit Index (SEDI), and the Standardised Vapor Pressure Deficit Index (SVPDI). This dataset is available globally at multiple time scales (1- to 12-, 18-, 24-, 48-month) for the period 1980–2024, with high spatial (0.1°) and temporal (weekly and monthly) resolution. The GMID dataset contains two versions of the drought indices; one generated using a traditional single-distribution approach, and another using a new flexible-distribution approach allowing the selection of probability distributions providing the best-fitting normalised series in each case. GMID adds significant value to previous drought data products through its high resolution and flexible multi-index nature, representing a comprehensive and robust long-term dataset for drought (and flash drought) assessment at the global scale.

Hydrological extremes are projected to intensify under climate change, yet their implications for water scarcity remain poorly understood. We use a multi‐model ensemble to assess how changes in hydrological low and high extremes will constrain renewable water for human use across 520 major river basins worldwide. We evaluate trends in two metrics of water availability: (a) annual lowest monthly runoff, and (b) annual accessible runoff, a conceptual indicator defined as total runoff excluding extreme high flows typically released rapidly for flood control. More than 40% of river basins are projected to experience declines in both metrics during 2015–2100. Mediterranean and several scattered basins are identified as hotspots where runoff reductions compound existing water scarcity. Scenario comparison shows that low emission mitigates, though do not fully eliminate, these declines, highlighting hydrological co‐benefits of climate mitigation. The novelty lies in explicitly incorporating hydrological extremes into water availability assessment.

Tropical forests are an important component of the global carbon cycle. Increasingly frequent droughts and heatwaves are undermining their resilience, with far‐reaching impacts on local livelihoods and on regional and global hydrological cycles. El Niño events, recurring climate patterns that lead to dry and hot conditions across multiple tropical regions, have always influenced these dynamics, but the 2023–2024 El Niño produced extreme droughts and heatwaves that exceeded the typical El Niño footprint. Here, we place the 2023–2024 El Niño in both past (2000–2024) and future (2025–2100) climatic contexts across pantropical forests. We assess its climatic magnitude using ensembles of reanalyses, satellite climate products, and Earth System Models. Climate anomalies over 2000–2024 show that three quarters of pantropical forests experienced, on average, 6 months of compound droughts and heatwaves during 2023–2024. Based on future climate projections, most tropical forest regions will reach high probabilities of experiencing 2023–2024‐like temperatures on an annual basis, while the probability of similar droughts is expected to increase mostly in Amazonian forests. These results indicate that the 2023–2024 El Niño may have presented a preview of future tropical forest conditions, which can only be prevented by rapid and drastic reductions in global greenhouse gas emissions.

Accurate estimation of regional-scale terrestrial carbon budgets is of great importance but remains challenging. With particular advantages, the Long Short-Term Memory (LSTM) network method shows potential for improving regional carbon budget upscaling estimations. Here, using LSTM, we upscaled regional net ecosystem carbon exchange (NEE) with available flux tower measurements and satellite land surface observations in North America. With well-established ecosystem-specific LSTMs, we produced monthly NEE at a spatial resolution of 0.1 ° ×0.1 ° over 2001–2021 (labelled as MemoryFlux). Unlike existing upscaling estimates, our dataset properly identified the Midwest Corn Belt as a region of large seasonal carbon uptake during the peak growing season, a feature revealed by previous top-down studies and recognized as a model benchmark. Moreover, the seasonal variations in NEE estimated by MemoryFlux were strongly correlated with those from independent atmospheric inversions, including the ensemble mean of the Orbiting Carbon Observatory-2 Model Intercomparison Project (OCO-2 v10 MIP; r =0.96, p <0.001) and CarbonTracker2022 (CT2022) ( r =0.97, p <0.001). The mean annual NEE was estimated at −1.27 ± 0.12 Pg C yr −1 , which was closer in magnitude to inversions (−0.83 to −0.70 Pg C yr −1 ) than existing upscaling estimates (−3.30 to −1.68 Pg C yr −1 ). In addition, MemoryFlux captured spatial NEE anomaly patterns associated with six selected severe drought and flood events. We further found that explicitly incorporating historical predictor information improved the representation of NEE interannual variability and spatial anomalies associated with climate extremes. MemoryFlux provides an improved bottom-up estimation of North American NEE and shows greater consistency in magnitude with independent atmospheric inversion estimates than several existing EC-based upscaling products. The MemoryFlux dataset is available at https://doi.org/10.5281/zenodo.20482274 (Huang and He, 2026).

One of the most important skills of hydrological models is to simulate timing and magnitude of flood events. Long Short-Term Memory (LSTM) networks are currently among the most successful models for streamflow and flood prediction over large regions. In snow-influenced catchments, which typically comprise a minority in large-scale studies, floods are generated by two distinctly different processes, snowmelt and rainfall. The applicability of hydrological models in such regions is therefore dependent on their ability to represent both types of floods. Nevertheless, flood evaluations of LSTM taking different flood-generating processes into account are currently lacking. This study fills this gap by evaluating the ability of LSTM to model flood peak characteristics separately for snowmelt and rainfall generated floods. The trained LSTM model successfully simulated streamflow time series across the 103 evaluated catchments, with average NSE of 0.84 and average KGE of 0.86 over the unseen evaluation period. LSTM exhibited better performance in the majority of the catchments in terms of flood peak timing and magnitude for both rainfall and snowmelt generated floods when compared to the operational hydrological model in the region (HBV) used as a benchmark. LSTM had a 27 pp higher percentage of correctly simulated peak days for rainfall generated floods as compared to snowmelt generated floods, similar to what was found for HBV (29 pp). LSTM's ability to simulate flood peak magnitudes was similar for the different flood types, with percent errors within 20 % for close to half of the events. False alarm rate and probability of detection results were similar for snowmelt and rainfall generated events, both slightly better than those of mixed events. LSTM outperformed HBV in 70 %–91 % of the catchments in terms of flood peak timing and magnitude for snowmelt and rainfall generated floods. The largest improvements in flood peak magnitudes were found for rainfall generated events, in particular for catchments where HBV exhibited high (>40 % ) absolute errors. Overall, our findings bring confidence that LSTM can improve hydrological services in regions subject to both snowmelt and rainfall generated floods.

Amazon droughts have historically been driven by precipitation deficits, but recent escalating drought events highlight a regime shift toward “hot droughts.” By decomposing historical drought drivers (1979–2024), we show that rising evaporative demand (represented by potential evapotranspiration, PET) now acts as a primary driver of drought intensification. While rainfall deficits remain the dominant contributor to the record‐breaking 2024 drought, escalating atmospheric thirst almost doubled its severity. Rising PET also expands the total drought footprint, disproportionately amplifying exceptional (D4) droughts and nearly tripling their affected area in 2024 compared to precipitation deficits alone. This PET increase is forced comparably by background warming and increased shortwave radiation; neutralizing either factor during 2024 reduces D4 drought coverage by 25%–33%. Linked to observationally confirmed declines in dry‐season cloud cover, these compounding thermodynamic and radiative feedbacks are driving a shift toward increasingly extreme hot droughts across the Amazon.

Water allocation has long been framed as a general optimization problem in which users compete for a finite resource, leading much of the existing literature to adopt a top‐down perspective on how water should be distributed among users. However, in practice, water withdrawals may also be influenced by physical access and operational arrangements, such as direct abstraction from river channels through intake facilities for agricultural and industrial use. In such systems, disparities may arise between institutional allocation principles and effective access to water during shortage conditions, potentially creating asymmetric impacts among upstream, midstream, and downstream stakeholders. This study quantifies these impact asymmetries across river locations and evaluates stage‐based intake restriction strategies designed to mitigate inequitable outcomes during droughts. The proposed framework is applied to the Yeongsan River Basin in southwestern Korea through computational experiments to demonstrate how drought‐stage‐based intake regulations can improve both net profit and equity among stakeholders. A game‐theoretic approach is further employed to assess the potential for compensation across Pareto‐optimal solutions by examining whether the additional gains generated under cooperation are sufficient to offset stakeholder losses relative to the non‐cooperative baseline. The analysis is extended to projected scenarios of intensified climatic and socio‐economic droughts to assess the robustness of the proposed policies under future stress conditions. Through this approach, the study introduces a generalized framework for identifying water allocation policies that reconcile efficiency, equity, and sustainability under institutional and hydrologic constraints.

Using satellite‐derived dust optical depth (DOD), reanalysis meteorology, and a global topsoil flash drought (TSFD) inventory for 2003–2022, we find that TSFD events are associated with increased dust loading at the global scale. During TSFD events, 44.7% of land pixels exhibited positive DOD anomalies. The net global DOD anomaly increased significantly over the study period; however, this trend cannot be attributed solely to TSFD‐related effects, as changes in background environmental conditions may also have contributed. Compound TSFD–strong dust (TSFD_SD) events occurred most frequently under sparse vegetation and moderately elevated wind conditions. SHapley Additive exPlanation (SHAP) analysis showed that DOD variability during TSFD events was more strongly associated with soil moisture, background aridity, and vegetation state than with wind speed. These relationships are consistent with a mechanism in which rapid topsoil drying and vegetation stress increase surface erodibility and may enhance dust loading, although the observational analyses do not establish causality.

Flood frequency analysis (FFA) forms the basis of many hydrological assessments and practical design calculations. Traditionally, FFA is based on the assumptions of stationary distributions and homogeneous samples. These assumptions are questionable in light of climate change and the wide variety of hydrological processes that lead to flooding. Furthermore, these assumptions can obscure important insights, including process‐specific changes. In this study, we develop a process‐aware, non‐stationary mixture model that combines type‐specific distributions with time‐varying parameters. An application to more than 600 catchments across Europe shows that the various flood types change differently over time. Heavy rainfall floods at many locations in Europe (e.g., the Alps and UK) are increasing significantly, whereas snowmelt floods are decreasing. In conventional models, these findings are obscured by the combined analysis of opposing trends, and trends are often not recognised. This causes design quantiles, for example, for a 100‐year return period, to be underestimated by about 15% on average in many regions of Europe when non‐stationary, type‐specific distributions are not used. Our results highlight the importance of incorporating process understanding into FFA to improve design‐flood estimation and our understanding of changing flood‐generating processes.

Increasing droughts are exacerbating the imbalance between water supply and demand, complicating efforts to secure reliable water and food supplies for key sectors. However, limited understanding of how water use responds to drought constrains effective risk management and climate adaptation. Using monthly water use and meteorological data across mainland China from 1981 to 2020, we developed an event‐scale framework integrating time series decomposition, deep learning, and event coincidence analysis to quantify sector‐specific responses to meteorological droughts. Droughts impacted a relatively small portion of land (14.9% of China mainland) but accounted for a disproportionately large share of water use (47.4%). Northern basins were particularly sensitive, with water use rising by 5.67% (4.49%–6.37%) during drought events. Irrigation emerged as the primary driver of water use changes, with demand increasing as drought intensity rises. Projected intensification of droughts suggests that by 2100, more than 10% of affected regions could face high water scarcity risks under SSP1‐2.6, which may increase to 25.5% under SSP5‐8.5. These findings highlight pronounced spatial disparities in drought responses and emphasize the need to incorporate water use dynamics into climate risk assessments.

In May 2019, unprecedented extreme rainfall over the central United States (US) facilitated severe flooding along the Arkansas River. The well‐documented relationship between atmospheric rivers (ARs) and central US heavy precipitation motivated investigation into AR relevance to this event and its predictability on extended timescales. AR conditions coinciding with the extreme precipitation indicate ARs contributed to the Arkansas River flooding, while poisson weighted forecast probabilities of SHiELD ensemble hindcasts show both the ARs and extreme precipitation were predictable at 2 week leads. Comparison of precursory atmospheric conditions between the top and bottom five performing ensemble members reveal the top performers developed a Rossby wave train over the North Pacific, which promoted AR‐favorable conditions downstream and facilitated better forecasts of precipitation than the bottom performers. Results suggest cross‐Pacific Rossby wave trains can provide a forecast of opportunity for extended predictability of ARs and extreme precipitation over the central US.

The resilience of human society to multiple climate disasters is crucial for supporting sustainable development progress and promoting human well‐being in an era of climate change. Yet little research has attempted to assess global spatial‐explicit capacity of communities to deal with compound extreme climate disasters simultaneously, and the relationship between resilience to different climate extremes remains unknown. Here we presented a high‐resolution assessment of community resilience to compound climate extremes simultaneously by developing the Community‐Climate Resilience Index (CCRI) using the multisource satellite remote sensing data sets, the geocoded disaster records and fuzzy logic model. We further explored interrelationships among resilience to major climate extremes, that is, resilience to floods (Resilience flood ), resilience to extreme temperature (Resilience ET ) and resilience to droughts (Resilience drought ). The results indicated western Europe showed higher CCRI while South America, Africa and Oceania showed lower CCRI. Urban areas showed higher CCRI but lower scaling effects comparing with rural areas with lower CCRI level and higher scaling effects. Most of the regions showed positive associations among resilience to different climate extremes, but there were also negative associations between resilience in some areas, especially located in suburban and rural Sub‐Saharan Africa. Our analysis suggested leveraging integrated and coordinated risk management for resilience capacity building to multiple climate extremes in vulnerable areas and on a local and global scale.

Streamflow estimates derived from stage–discharge rating curves (HQRC) inadequately represent the spatiotemporal complexity of unsteady gradually varied flows (UGVF) and are commonly reported without uncertainty, encouraging their treatment as deterministic. These limitations are increasingly problematic as data users demand defensible confidence in streamflow estimates for cross‐agency comparisons and for scientific applications. One of the dominant yet poorly quantified sources of HQRC error is hysteresis arising from ubiquitous unsteady and nonuniform flow conditions characteristic of UGVFs. Hydrometric agencies attempt to mitigate these effects through using alternative approaches such as HQRC corrections or use of stand‐alone multivariable methods that augment stage with velocity or free‐surface slope measurements. However, evaluating the effectiveness of these alternatives remains challenging, as HQRC corrections rely on simplified flood‐wave representations, while multivariable methods retain the semi‐empirical underpinnings of traditional HQRC formulations. To address this evaluation gap, we introduce a hysteresis‐centric performance metric that quantifies departures between bivariate relationships obtained using HQRC correction and multivariable methods and those derived from conventional HQRC ratings. The metric integrates three complementary indicators: (a) flow deviation magnitude, (b) hydrograph peak phasing, and (c) cumulative volumetric differentiation, collectively capturing the hysteresis severity, variable peak timing, and integrated impacts of hysteresis on discharge estimates. The framework is applied to three U.S. Geological Survey gaging stations spanning diverse hydro‐geomorphic settings. Results provide practical guidance for evaluating and selecting alternative monitoring methods in relation to dominant flood‐wave behavior (kinematic vs. non‐kinematic) and are synthesized into a structured decision‐making workflow for monitoring‐system selection under UGVF conditions.

Synchronous flooding across distant regions can amplify socioeconomic impacts. As the hydrological cycle intensifies, flood synchronization appears to be expanding, yet predictive understanding beyond gauged basins remains limited. We analyze the global synchronization of annual peak runoff (a proxy for flooding) using gridded data sets and complex network approach, and evaluate 13 CMIP6 (Coupled Model Intercomparison Project Phase 6) models in reproducing spatial synchronization patterns. CMIP6 models reproduce major synchronization hotspots but systematically overestimate connectivity and local clustering (in 90% and 85% regions, respectively) while underestimating average link length (in 70% of regions), yielding overly connected networks disproportionately dominated by short‐range synchronization. We show that exaggerated temporal dependence driven by overly concentrated and shifted peak timing is the primary source of this structural bias. These emergent errors suggest a potential overestimation of spatially compound flood risk and highlight deficiencies of climate models in reproducing global hydrological seasonality, particularly in snow‐dominated regions.

Flood‐Regulating Ecosystem Services (FRES) are widely used to assess the capacity of ecosystems in retaining water during a storm and mitigating hydrological extremes. Agriculture can play a role in the regulation of flood hazard, but the role of specific crop types in influencing flood regulation in small agricultural catchments remains largely unexplored. Indeed, FRES are typically evaluated by comparing the runoff generated in current conditions and a barren land scenario, treating agricultural land use uniformly in space (without distinguishing between crops) and time (without considering that agriculture has a seasonality). This study presents an approach to assess the seasonality of crop‐specific FLOod Regulating Agro‐Ecosystem Services (FLORAES). The proposed approach is based on coupling two fully distributed hydrological models: WATNEEDS, an agro‐hydrological model that estimates daily soil moisture based on agricultural water demand, and MOBIDIC, which uses the soil moisture conditions derived from WATNEEDS as the initial state to simulate flood hazards during extreme storm events. As a proof of concept for the methodology, we analyzed the upper part of the Ombrone Grossetano river basin (Italy), where agriculture plays a huge role in the local economy. Our results reveal that, under identical climatic conditions, different crops generate distinct initial soil moisture conditions, which can affect flood peak discharges. FLORAES don't contribute to flood mitigation in winter, whereas their contribution at the end of the growing season of each crop is not negligible, reaching the 20% of the peak discharge depending on the crop and the river basin size.

Floods are among the most damaging disasters worldwide, necessitating high-resolution modeling of where terrain and environmental conditions predispose societies to flooding. While geospatial data are invaluable for flood susceptibility mapping, comprehensive and globally consistent datasets remain scarce. We present the Global Flood Susceptibility Map (GFSM v1), a globally harmonized 30 m flood susceptibility dataset produced by incorporating multi-source flood conditioning factors accounting for topographic, hydrological, meteorological, and anthropogenic characteristics within a machine learning framework. A gradient-boosted tree model was trained on ~17,000 curated grid tiles with ~30.45 million training samples, using a global flood inundation dataset as training labels and nine conditioning factors (elevation, slope, aspect, wetness index, vegetation index, height above nearest drainage, water/roads proximity, and rainfall frequency). The model was deployed based on geopolitical and climate strata (units n = 192) to generate global pixel-level susceptibility. Rigorous cross-validation indicates strong model predictive accuracy (median area-under-curve ~0.95), and the output captures localized flood-prone regions worldwide. GFSM v1 offers a global open-source, high-resolution resource to support informed flood risk management, particularly in data-sparse regions.

Drought is a highly consequential natural disaster that may likely increase in both severity and extent across the conterminous United States (CONUS). The mechanisms affecting the propagation of drought from the atmosphere to streamflow are complex and interactive, making the prediction of streamflow drought difficult with current modeling approaches. Machine learning is an emerging tool in the field of hydrology that may be well-suited to prediction of streamflow drought across large and topographically diverse areas. Here, we train and analyze 3198 random forest models at U.S. Geological Survey streamgages to understand common meteorological drivers of streamflow drought and to define physiographic characteristics of basins sensitive to these drivers. We also develop a novel dynamic regionalization approach using donor gages to predict daily streamflow drought at pseudo-ungaged locations. Our results show that teleconnections, temperature, evaporative demand, and snow–water equivalent are important drivers of streamflow drought in the West, Southwest, and Northern Rocky Mountains (Northern Rockies) regions of the United States, and precipitation and soil moisture are primary drivers of streamflow drought in the Northeast, Southeast, and the Northwest regions. Prediction using dynamic regionalization shows comparable performance to at-site models. This method may be applicable to other hydrologic prediction problems requiring transferability to ungaged locations.

Rainfall-runoff events are widely used in hydrological applications, from flood estimation and flood forecasting, to understanding catchment responses in a climate/anthropogenic affected world. The majority of methods used to identify rainfall-runoff events are statistical in nature, relying on subjective, user-defined “rules” (i.e., parameters) that define a rainfall-runoff event. Since no ground-truth information is available to confirm the exact beginning and end of rainfall-runoff events, there is noticeable inconsistency (i.e. uncertainty) within the results. In this study, we propose the Robust Variance-based Event Identification Method (RVEIM), a new parsimonious rainfall-runoff event identification method which uses fewer parameters and better mimics the natural runoff generation process; decreasing the uncertainty in rainfall-runoff event identification. RVEIM detects runoff events by focusing on changes of streamflow variance and pairs to the corresponding rainfall event(s) simultaneously. RVEIM was compared to two benchmarking event identification methods in 8 representative catchments in Australia. A sensitivity analysis was performed using a comprehensive set of plausible event identification parameter values for all methods. Results revealed that the variation of rainfall-runoff events characteristics – including annual number, length, and the mean of paired runoff events – showed limited uncertainty from the RVEIM (standard deviation within ±15 % of the mean across 8 representative Australian catchments). In contrast, the two benchmarking methods exhibited substantially higher uncertainties (43 % to 93 %). Using RVEIM, we present the first comprehensive summary of rainfall-runoff event characteristics across 467 Australian catchments. The distribution of event-scale runoff coefficients for individual catchments shows a strong climate gradient. Such systematic shifts in the distribution of runoff coefficient across climate regions indicate that climate variables play an important role in catchment response and point to potential contrasts in dominating runoff generation mechanisms across climate zones.

Riparian landscapes support biodiverse and productive ecosystems that in turn provide valuable services like water filtration and flood attenuation. Extensive anthropogenic modifications to riparian environments have altered natural water connectivity patterns important for ecosystem health, leading to an increased effort to understand and restore natural floodplain connectivity. It is well known that secondary floodplain channels facilitate exchange at the river‐floodplain boundary with increasing discharge, but comparatively little is known about connectivity of floodplain waterbodies that are productivity hotspots in riparian ecosystems. In this study we investigate patterns of floodplain and floodplain waterbody connectivity in the coastal meandering Mission River, TX, USA. We use a modeling approach to examine water movement and connectivity, and find that water exchange generally increases with discharge, while the number of active floodplain connections is greatest at moderate flows. We also observe that the number of effective connections, a measure of connection prominence, is maximized at moderate discharges and is typically one‐third to half the number of active connections, indicating that few connections dominate river‐floodplain exchange. Counterintuitively, we find that exchange at the river‐floodplain interface is most uneven at the highest discharges, despite complete bank inundation. We interpret a process‐based transition in connectivity from being topographically controlled at lower discharges to hydraulically controlled at high discharges, and identify a spatial association to this transition. Finally, we consider that exchange and exposure time in each floodplain region can describe nutrient processing potential in a framework that reflects complexity inherent in natural floodplains.

Bedforms in lowland sand‐bed rivers, which contribute substantially to total flow resistance, are known to adapt to changing flow conditions. While the growth of subaqueous dunes with increasing bedload transport and their wash‐out at higher transport stages are well documented in laboratory experiments, their possible evanescence in lowland rivers characterized by very low Froude numbers remains poorly observed, as do the effects on flood propagation. In this study, a 2D depth‐averaged fixed‐bed hydrodynamic model is enhanced to account for bedform evolution and applied in a real river setting to analyze conveyance shifts caused by dune evanescence during floods. Bed resistance is dynamically updated based on local flow conditions using different roughness predictors for dune dimensions and complemented by additional dissipative components. The model is applied to the lower 200 km reach of the Po River (Italy), where traditional fixed‐resistance approaches misrepresent the stage‐discharge rating curve, overestimating peak water levels by more than 2 m (~25% of the water depth). The results confirm a marked resistance reduction at flood stage due to dune wash‐out into upper‐stage plane bed, occurring at very low Froude numbers (∼0.2) and initiating at approximately 60% of the maximum flow capacity. Enhanced with the Van Rijn roughness predictor, the model enables accurate water level predictions under all flow conditions. In real river settings, flow‐bedform interactions are complicated by irregular morphology and mutual feedback mechanisms. This study demonstrates that an appropriately tuned roughness predictor may substantially improve flood prediction and flow capacity assessment in lowland alluvial rivers.

Dynamic imbalances in model initialization severely constrain the predictability of the rapid intensification (RI) and abrupt track recurvature of tropical cyclones (TCs). In this study, the effectiveness of an incremental analysis update (IAU)‐based four‐dimensional variational (4DVar) data assimilation scheme in forecasting Typhoons Doksuri and Khanun in 2023 was investigated. By reducing spin‐up impact, the mean track errors obtained with the IAU scheme stabilized below 60 km, and the intensity forecast errors were consistently reduced compared to the control experiment. These improvements can be mechanistically attributed to a superior representation of the fractured subtropical high and the internal asymmetric dynamic structure of TCs, which resulted in critical steering flow correction. Consequently, this dynamical enhancement yielded a 60% relative improvement in forecast skill for heavy rainfall (≥50 mm) at the 72‐hr lead time, indicating the potential benefit of improved TC initialization for heavy precipitation prediction.

Atmospheric Rivers (ARs) often co‐occur with other weather systems, including fronts, low‐pressure systems (LPSs), and mesoscale convective systems (MCSs), which can amplify precipitation and flood risk. While extreme rainfall is frequently linked to multiple interacting features, how these systems co‐occur with AR intensity scales and regions remains unclear. Using ERA5‐derived feature masks and satellite data, we quantify spatiotemporal overlaps between ARs and other systems. AR + Front combinations dominate across much of the conterminous U.S. In higher AR scales (4–5), more complex MCS‐related combinations become increasingly common, particularly in DJF and MAM and along the West Coast, while limited LPS‐related co‐occurrences appear mainly in JJA and SON and are confined to regions such as the Gulf of Mexico and offshore Pacific genesis regions. In contrast, AR‐only occurrences dominate in some inland western regions. Extreme precipitation exhibits seasonal variation, with distinct co‐occurring systems contributing across seasons.

The decoupling coefficient (Ω), a dimensionless metric that quantifies the balance between stomatal conductance ( G s ) and aerodynamic conductance ( G a ) in regulating evapotranspiration (ET), is key to understanding vegetation‐atmosphere interactions, which are fundamental to ecohydrological cycles and ecosystem functioning. However, the spatiotemporal patterns of Ω across different plant functional types (PFTs) and climatic gradients remain poorly understood, particularly in heterogeneous land surfaces under climate change. Here, we quantified Ω at 52 AmeriFlux sites across four PFTs in the contiguous U.S. (CONUS) from 2001 to 2024 and developed a high‐resolution Ω modeling framework that integrates eXtreme Gradient Boosting (XGBoost) with satellite observations and reanalysis data to produce spatially explicit 8‐day Ω maps at 0.1° resolution. This framework reveals a significant increasing trend in continental‐mean Ω (1.0 × 10 −3 yr −1 , p < 0.05), indicating progressively weaker vegetation‐atmosphere coupling, whereby ET becomes relatively insensitive to G s and increasingly constrained by available energy. We also found that Ω in water‐limited regions show highly volatile and asymmetric responses to drought, shifting toward stronger coupling in dry periods and toward stronger decoupling in wet periods, whereas Ω in energy‐limited regions remain strongly buffered with a stable coupling state. Our study reveals the spatiotemporal variability of Ω in CONUS, shedding light on the pivotal role of vegetation in modulating water fluxes under shifting climatic regimes and drought stress.

Climate change is intensifying the global water cycle, with extreme precipitation events increasing in frequency and intensity at the global scale. While trends in daily precipitation extremes are well-documented, sub-daily extremes – critical for flash flood risk – remain poorly characterized, due to limited long-term sub-daily observations. Convection-permitting models explicitly resolve deep convection and therefore offer the potential for a substantially improved representation of convective processes and short-duration precipitation extremes. This study evaluates the ability of the convection-permitting regional climate model AROME (2.5 km resolution, 1959–2022), forced by ERA5 reanalysis, to reproduce precipitation extremes and their trends, at daily and hourly scales, using a dense network of Météo-France stations. Using extreme value theory (GEV modeling), we analyze trends in 10-year return levels for both daily (1959–2022) and hourly (1990–2022) extremes. At the daily scale, AROME reproduces observed positive trends in southeastern France, consistent with previous studies. Hourly trends are more heterogeneous and less robust, with high spatial variability and low model-observation correlation. Overall, the results demonstrate the value of the explicit-convection model for studying extreme precipitation, while also highlighting its limitations for convective extremes.

Groundwater drought reflects the combined influence of meteorological forcing, aquifer memory, hydraulic connectivity, and anthropogenic modification. This study develops a hydrogeologically informed statistical framework that integrates temporal groundwater-drought characteristics with spatial indicators of Exposure, Pressure, and environmental Sensitivity. Standardized Precipitation and Groundwater Level Indices were analysed at 3-, 6-, and 12-month aggregation scales for 16 piezometers grouped into shallow unconfined, intermediate confined, and deep confined aquifer systems. The shallow unconfined aquifer showed the strongest SPI–SGI association, whereas the deep confined aquifer exhibited weaker correlations, longer drought persistence, and no observed recovery within the available record. These contrasts are consistent with differences in aquifer memory and hydraulic conditions, but conclusions concerning the deep system remain exploratory because it is represented by only two monitoring wells and may also be affected by sustained anthropogenic pressure. The Drought Impact Potential Index (DIPI) combines piezometer-level SGI-12 drought-response metrics with proxy indicators of groundwater abstraction, mining influence, groundwater-dependent ecosystems, and groundwater quality. Weighted and equally weighted formulations produced broadly similar regional patterns, although leave-one-out cross-validation indicated substantial interpolation uncertainty. The resulting maps should therefore be interpreted as exploratory screening products rather than predictive vulnerability maps. The E − P contrast describes only the relative mapped magnitudes of Exposure and Pressure and does not provide causal attribution of climatic and anthropogenic controls.

Snow droughts, defined as periods with below-normal Snow Water Equivalent, have recently received substantial attention as an emerging hazard in a warming world, but their impacts are still poorly understood. Here, we shed light on these impacts across the socio-ecohydrologic spectrum, by leveraging heterogeneous data sources from 38 catchments in Italy: 13 years of snow and runoff data, remote-sensing and in-situ measurements of Gross Primary Production, an inventory of emergency water restrictions obtained via a web-scraping tool and direct consultation of national to local regulations, and a survey among 113 mountain huts. We found that the majority of snow droughts in our sample were warm-dry (53 %), that is, the combination of higher-than-usual temperatures and low precipitation, followed by warm-wet snow droughts (22 %). These events resulted in the snow season being shortened by as much as one month across all elevations, more melt-out events compared to non-snow-drought years, approximately −50 % summer runoff, and a decline in runoff ratio. Notably, growing-season Gross Primary Production of vegetation after a snow drought was up to 10 % higher than after a non-snow-drought winter, particularly above 1500 m, which ground-based data suggest was due to an earlier meltout of snow leading to an earlier-than-usual greening date. By focusing on the recent 2022 and 2023 snow droughts, we also found that water-supply restrictions were issued at all elevations, but particularly across foothills rather than floodplain regions. Meanwhile, about 70 % of hut managers at elevations above 2000 m reported water-supply impacts, with 28 % of them even reporting earlier closing dates. Overall, snow droughts emerge as a deeply multi-sectorial and multi-elevation risk, interconnecting the cryosphere with hydrology, ecology, and society.

Drought is increasingly reshaping terrestrial ecosystems under climate change, yet how aridity modulates the climatic sensitivities of satellite‐derived green‐up rate (GR) and senescence rate (SR) in natural ecosystems remains poorly understood at the global scale. Here, we combine two decades of satellite observations and climate data sets to identify the primary climatic drivers (temperature, precipitation, solar radiation, soil moisture, and vapor pressure deficit) of maximum GR and SR across 764 global ecoregions and to quantify their sensitivities to interannual variability in climate‐forcing accumulation rates of these drivers. We reveal contrasting temperature‐response patterns along the aridity gradient in arid versus humid ecoregions. In arid ecoregions, decreasing aridity increases the GR sensitivity to temperature accumulation rates during the green‐up phase, whereas in humid ecoregions this sensitivity weakens and can even become negative as conditions become wetter. Similar, though asymmetric, those sensitivity reversal patterns are observed for SR. These findings demonstrate that aridity fundamentally reshapes vegetation‐climate interactions, with important implications for the stability of ecosystem carbon uptake under increasing hydroclimatic variability. By identifying critical transitions in climate sensitivities across aridity gradients, this study provides new insights for improving predictions of ecosystem responses to intensifying drought and for assessing risks to carbon sequestration under future climate change.

Climate extremes are increasingly disrupting livelihoods and shaping patterns of human mobility. Their longer-term and compounding relationships with migration remain understudied, particularly at the subnational scale. In this study, by analysing global subnational net migration data, we show that associations between climate extremes and migration vary by hazard type, timing, consecutive events, income level and pre-existing migration context. Floods are associated with decreased net migration across contexts, with effects lasting several years and appearing stronger in lower-income areas. Droughts, by contrast, show associations with net migration primarily in higher-income settings, with effects generally short-term. Tropical cyclones correspond to decreased net migration primarily in lower-income areas, with effects lasting over several years. For compound exposure, droughts show the clearest patterns, with prolonged exposure linked to reduced migration inflows in population-attracting areas. These results highlight how climate extremes interact with local socioeconomic contexts to shape mobility.

Rainfall during the snow season plays an increasingly important role in flood risk as climate warms and extreme events become more frequent. However, a given sized rain-on-snow (ROS) event can yield outcomes ranging from flooding to no runoff, depending partly on the snowpack's antecedent cold content and capillary retention forces. Here, we analyze the seasonal evolution of the snowpack's physical state over a 72-year period to assess long-term changes in its capacity to buffer runoff from liquid water input. We use ERA-5 Land data to force Alpine3D, a distributed snowpack model that tracks the layer-by-layer development of heat, mass, and structural framework of the snowpack throughout the snow season. We test our approach in a large Columbia River headwaters basin in NW Montana, USA. We evaluate cold content and total capillary retention of the snowpack to determine long term trends in Liquid Water Buffering Capacity (LW bc ) as it evolves throughout the snow season. The LW bc of the snowpack exhibited robust long-term declines across all elevation bands, that is driven by reduced SWE along with reduced cold content in late November through late January across all elevation bands, despite high intra- and interannual variability. The largest declines occurred during the Spring period, trending downward across the historical period by 43 % to 80 % depending on the elevation band. The core five weeks of mid-winter showed no trending change of LW bc , and in fact demonstrated an increase in cold content over the 72 years. Our findings demonstrate that changes in the snowpack's ability to buffer runoff, including dependencies on local basin factors related to snowpack seasonality and elevation, are a key component of evolving ROS risk.

Differentiation in water use strategies is essential for maintaining the function and resilience of dryland ecosystems. Under prolonged drought, dominant shrubs self‐organize into two typical spatial configurations: scattered with separated individuals and clumped with clustered individuals. How the coordination between root water uptake and leaf physiological traits drives water use strategies of self‐organized shrubs for adapting to drought remains poorly understood. To address this gap, soil moisture, morphological and root traits, leaf‐level physiological traits, and stable isotope ( δ 2 H, δ 18 O, and δ 13 C) of scattered and clumped Vitex negundo were observed during 2022–2024 in the semi‐arid Loess Plateau. The scattered shrubs primarily utilized middle and deep soil water (62.3 ± 7.8%), associated with isolated canopies that promote infiltration into deep soil layers. In contrast, clumped shrubs relied on shallow and middle soil water (82.0 ± 6.5%), linked to aggregated canopies and shallow roots. Scattered shrubs showed tight stomatal regulation with high midday leaf water potential and intrinsic water use efficiency (iWUE) during dry season, whereas clumped shrubs exhibited relaxed stomatal regulation, with declines in midday leaf water potential and iWUE. In rainy season, scattered shrubs constrained gas exchange, while clumped shrubs enhanced stomatal conductance and photosynthesis. These findings indicate that scattered shrubs adopt an active deep‐water acquisition with conservative water utilization strategy, enhancing drought resistance, while clumped shrubs exhibit an opportunistic shallow‐water acquisition with acquisitive water utilization strategy, improving water capture efficiency for drought adaptation. This study reveals the ecohydrological processes of self‐organized vegetation in drylands and provides basis for vegetation allocation in ecological restoration.

Understanding how meteorological drought propagates to hydrological drought under natural conditions is essential for improving drought early warning, yet the roles of precipitation frequency and intensity in regulating this process remain poorly understood. Here, we quantified drought propagation time (DPT) and drought response rate (DRR) across mainland China under natural conditions, and further assessed the spatial effects of precipitation intensity and frequency using geographically weighted regression (GWR) and future climate projections. The results show that: (a) DRR declines significantly with aridity, with 22.9% of the study area showing DRR values of ≤0.2, indicating weaker meteorological–hydrological coupling in drylands. (b) GWR results indicate that meteorological controls on DPT are dominated by precipitation, while the effects of precipitation intensity and frequency exhibit strong spatial heterogeneity: total precipitation metrics fluctuate around zero across aridity classes, whereas light‐precipitation intensity and heavy‐precipitation frequency are more strongly associated with DPT in drylands, indicating that drought propagation depends more on precipitation partitioning than on bulk precipitation alone. (c) Under future climate scenarios (2030–2100; SSP2‐4.5 and SSP5‐8.5), future changes in drought propagation are driven less by shifts in mean precipitation intensity than by increasing uncertainty in precipitation‐frequency effects, especially in arid and hyper‐arid regions. These findings highlight the critical role of precipitation partitioning in regulating drought propagation and establish a natural‐condition baseline for drought early warning and climate adaptation in water‐limited regions.

The flood reduction potential of individual reservoirs within a large river network continuum remains poorly understood due to the complex interplay between reservoir characteristics and network properties. Here we investigate whether a collection of relatively small reservoirs can play a significant role in mediating downstream floods and assess how that role may be influenced by reservoir network properties compared to traditionally known reservoir characteristics. Our unique contribution was the simulation of downstream flood inundation maps alongside peak flows for each of the 81 major reservoirs (6 × 10 4 to 8 × 10 9 m 3 ) across 15,000 river reach segments, integrated into a process‐based hydrologic model covering a 415,000 km 2 region in the Texas Gulf Coast, United States. The three key takeaways from our study are as follows. (a) Smaller reservoirs can substantially reduce downstream flooding, suggesting that inclusion of large reservoirs—the traditional approach in flood risk management studies—may present only a partial picture. (b) Flood reduction by smaller reservoirs is more effective upstream, although this phenomenon may be linked with aridity and overall water availability. (c) While reservoir size matters, it is not the primary factor determining its downstream flood reduction potential; the influence of network properties, such as catchment area, the count of upstream reservoirs, the cumulative maximum storage capacity of upstream reservoirs, and Stream Order (i.e., location), is equally and often more important. The broader impact of our findings goes beyond just floods, providing foundational insights for addressing emerging challenges such as aging dams and river connectivity.

We present the globally available dataset of Line-of-sight Gravity Differences (LGD) as a new data product to fill the long-standing gap of investigating sub-monthly surface mass change from the satellite gravimetry along-track perspective. The dataset is generated directly from Gravity Recovery and Climate Experiment Follow-on (GRACE-FO) Level-1B intersatellite ranging observations, positioned conceptually between Level-1B measurements and Level-2 monthly gravity field solutions. Provided as along-track time series, LGD delivers instantaneous gravity change at satellite altitude, thus ensuring the fidelity of measurements by avoiding the complex preprocessing, regularization, and filtering applied to monthly products. The instantaneous LGD measurements enable monitoring of sub-monthly terrestrial water storage variability and provide accurate information on timing and magnitude of water mass changes. GRACE-FO's ground track allows LGD retrievals every 5–6 d in most low and mid latitude regions, while the revisit interval becomes shorter toward higher latitudes. Our LGD data products also contain climatology in the form of along-track gravity variations. We demonstrate its potential through case studies, including along-track diagnosis of flash drought evolution in the southeastern United States and the characterization of sub-monthly hydrological extremes in global hotspots. Results show that the LGD dataset has the potential to resolve the critical timing (e.g., onset and peak) of hydrological extremes, which is unattainable from monthly products. The LGD dataset is available at https://doi.org/10.5281/zenodo.17766177 (Tang et al., 2025a), which is expected to pave the way for advances in sub-monthly hydrological monitoring and can serve as a unique constraint for hydrological and geophysical models, opening new opportunities for Earth system sciences.

Data‐driven rainfall–runoff models have advanced rapidly, yet the majority of large‐scale applications still rely on lumped inputs that smooth out spatial variability in precipitation, temperature, and landscape properties. This simplification can introduce substantial biases in flood peaks, hydrograph timing, and water balance estimates. Here, we develop a deep learning framework that ingests spatially distributed (gridded) meteorological forcings and is evaluated across 921 catchments in North America using the CAMELS‐SPAT data set. We compare two strategies for fusing spatial and temporal information, namely a spatial‐to‐temporal (S2T) architecture and a temporal‐to‐spatial (T2S) architecture, and assess the added value of gridded catchment attributes and finer spatial resolutions. The T2S approach consistently achieves the best performance, increasing median Nash–Sutcliffe Efficiency (NSE) from 0.60 (lumped baseline) to 0.65 and Kling–Gupta Efficiency (KGE) from 0.63 to 0.69, with NSE improvements observed in 75.4% of catchments. Performance gains are consistent across catchment sizes and aridity regimes, with enhanced skill in low‐flow, high‐flow, and peak timing representations. Integrated Gradients (IG) analysis shows that the model learns physically coherent spatial behavior, highlighting grid cells near the outlet and identifying dominant meteorological drivers of peak runoff events. These results demonstrate the practical value of spatially distributed forcings for data‐driven hydrologic modeling and outline a pathway toward interpretable, large‐scale streamflow prediction.

Groundwater‐dependent ecosystems (GDEs) provide critical ecosystem services in arid and semiarid regions and help sustain ecosystems during droughts in humid areas. Accurate mapping of GDEs is essential for improving water resource management and ecosystem conservation. The distribution of GDEs remains largely unknown across contiguous China. In this study, remote sensing data, soil data, and modeled groundwater depth (GWD) data are integrated to map GDEs across both arid and humid regions of contiguous China. The mapping framework utilizes long‐term remote sensing data combined with drought‐related climatic data to characterize vegetation responses under selected hydroclimatic conditions. These conditions are selected to highlight the differences in vegetation responses between GDEs and non‐GDEs. By integrating vegetation temporal dynamics with hydroclimatic indicators, this approach enables systematic mapping of GDEs. The results indicate that high‐, moderate‐, and low‐potential GDEs account for 1.6%, 7.7%, and 22.1% of the mapped area, respectively. Assuming a vegetation rooting depth within 10 m, the GDE map shows a commission error of 13.6% against observed GWD. The reliability of the GDE map is further supported by vegetation responses to drought and the characteristic topographic settings of GDEs.

Tropical cyclone (TC)‐induced extreme precipitation (TCEP) poses heavy risks, yet its seasonal predictability remains limited by insufficient understanding of cross‐scale climatic drivers. Here, we elucidate a stratosphere‐to‐stratosphere coupling pathway where the quasi‐biennial oscillation (QBO) modulates TC activity and subsequent TCEP over the southeastern coast of China, with the second boreal summer intraseasonal oscillation mode (BSISO2) acting as a critical mediator. The westerly QBO phase reshapes the spatial structure of BSISO2 during its Phases 4–6, creating favorable dynamical conditions and moisture convergence that nearly double TCEP over the southeastern China coast relative to the easterly QBO phase. Statistical fitting results further show that incorporating the QBO signal substantially improves seasonal TCEP predictability beyond conventional El Niño‐Southern Oscillation predictors. These findings establish the QBO as a coherent stratospheric driver for TCEP, providing a mechanistic basis for enhanced seasonal TCEP prediction and regional disaster preparedness.

Fire danger can escalate rapidly when vegetation dries over days to weeks. Flash droughts, characterised by rapid root-zone soil moisture depletion under high evaporative demand, may accelerate wildfire spread. Here, we quantify relationships between flash droughts and wildfire spread dynamics across land-cover types in the conterminous United States, using a model-derived root-zone soil-moisture flash-drought metric and contrast these with conventional drought conditions defined by persistently low soil moisture without rapid onset. We find that fires following flash droughts exhibit faster early spread than fires not associated with flash droughts, with model estimates indicating about 1.3–1.8-fold faster early burning across major flammable land covers. Grasslands and croplands show the strongest acceleration, followed by open shrublands, woody savannas, savannas and deciduous broadleaf forests, consistent with faster drying and greater fuel continuity in fine-fuel ecosystems. In contrast, conventional drought conditions produce weaker and less consistent changes in fire spread. Our results demonstrate that rapid soil-moisture drawdown—rather than background dryness—appears to provide a more direct pathway linking drought to accelerated wildfire spread. We recommend that monitoring declines in root-zone soil moisture in the days to weeks preceding ignition could improve early warning of fire risk in ecosystems prone to rapid drying.

We evaluate the performance of the Super-Fast INundation of CoastS (SFINCS) hydrodynamic model for simulating riverine floods, combined with a fully automated open-source data preprocessing pipeline. To do this, we assessed the simulated extent of 499 historic flood events against the satellite derived flood extents using the Critical Success Index (CSI) as a performance metric. We utilised simulated discharges from the Global Flood Awareness System (GloFAS) hydrological model and found that SFINCS performance improved with upstream basin size, with a global mean CSI of 0.42 for basins with large upstream area (>1000 km 2 ) and a CSI of 0.29 for basins with small upstream area (<50 km 2 ). Our results illustrate the importance of accurate discharge data input to flood hazard simulations. When the (globally simulated) GloFAS data is replaced with observed discharge data for ten events in the US, the CSI improved from 0.39 to 0.67. These results suggest that global hydrological model performance limits the accuracy of the flood hazard simulations. Our findings also showed a significant improvement in the CSI (from 0.37 to 0.57) when changing to a higher-resolution elevation input by contrasting a ∼1 m digital elevation model (DEM; 3DEP) with our default ∼30 m global DEM (FABDEM) in six US events. Sensitivity analysis of bathymetric calculations revealed a systematic underestimation of the default 2-year return period estimated by GloFAS discharge, likely driven by underrepresentation of annual block maxima, which resulted in underestimated channel dimensions. All of these factors resulted in a loss of detail, which impacted model performance, especially in smaller headwater rivers. We recommend to improve the estimation of bathymetry, for instance by employing the “gradually varying solver” method or using data from the SWOT mission. Furthermore, incorporating additional validation data which ideally includes flood depth measurements can largely enhance our understanding of the model performance.

The resistance and resilience of vegetation determine its ability to maintain ecosystem functional stability under flash drought stress. However, a research framework that quantifies event‐scale response characteristics of vegetation resistance, resilience, and vulnerability remains lacking. To fill this gap, we developed a high‐temporal‐resolution framework to assess the multi‐dimensional attributes of vegetation response to flash droughts, quantifying the spatiotemporal response characteristics of vegetation resistance, resilience, and vulnerability in China from 1982 to 2019. Our analysis shows that, over the past four decades, vegetation vulnerability to flash droughts in China's monsoon regions has increased by 0.019 per year. On one hand, this is primarily due to the significantly higher average decline rate of vegetation ecosystem resistance (0.074 per year) under flash drought stress, compared to the average increase rate of its resilience (0.010 per year). On the other hand, the trade‐off relationship between vegetation ecosystem resistance and resilience has shifted, particularly in the middle and lower Yangtze River Basin. In both monsoon and non‐monsoon regions, explainable machine learning models indicate that temperature is the most significant stress factor influencing vegetation resistance and recovery processes during flash drought events, with SHAP values contributing between 20% and 50% to the overall model explanation. These findings highlight the urgent need to focus on the impact of increasingly frequent flash droughts on the stability of terrestrial ecosystem carbon sink functions in the context of current global warming.

Considerable progress has been made in forecasting the likelihood, magnitude, and mobility of post-wildfire flash floods and debris flows; however, we lack systematic observations of the physical and financial effects of these hazards, which can be used to close gaps between scientific understanding of the hazard and risk reduction. Here, we introduce a 147-year (1879–2025) inventory of damage and losses associated with post-fire flows in the United States. These observations enable an assessment of the geographic convergence and frequency of damaging post-fire flow events, as well as their multi-asset damage profiles and loss distributions. We find that the long-term (1879–2025) mean annual frequency of damaging flow events is greater than one and that recent (2016–2025) fatality rates and levels of monetary loss appear higher compared to historical periods. Our study provides an empirical baseline that can be used for long-term hazard planning amid global pressures compounded by wildland-urban interface expansion, increased wildfire activity, and rainfall intensification.

Conventional constant‐rate sampling methods are inefficient in low‐permeability, low‐storativity formations, where water from casing storage persistently biases collected samples. The High‐Stress Low‐Flow (HSLF) approach addresses this limitation by switching from a high to a low pumping rate, using the drawdown transition to create a sampling window in which downward movement and mixing of casing storage are minimized. In this study, we investigated the purging process of the HSLF approach through controlled experiments in a physical well model, measuring transient drawdown and solute mixing with pressure transducers and the light‐attenuation (LA) technique. A well storage‐mixing model for bounded confined aquifers was developed to interpret the laboratory results and scale them to field conditions. Results show that when the switch occurs later and the pumping‐rate ratio is sufficiently high, a sampling window forms. In bounded aquifers within the tank, the window length approaches a steady‐state limit, whereas in large field‐scale aquifers it increases linearly with later switch times. Comparisons between the constant‐rate and two‐stage purging methods confirm that, with an optimal switch time and pumping‐rate ratio, the HSLF approach effectively minimizes casing storage effect, accelerates mixing, and substantially reduces purge volume and time across all sampling scenarios. Observed drawdown and mixing behaviors are generally consistent with model predictions, and analytical extension of the aquifer domain suggests the potential applicability of the HSLF approach at the field scale. Nonetheless, a comprehensive field study is required to fully assess and validate its performance under complex hydrogeologic conditions.

In the past decade, the Amazon has experienced multiple severe droughts, raising critical questions about how vegetation stress can be detected using remote sensing indicators such as Solar‐Induced Chlorophyll Fluorescence (SIF). In this study, we compare two anomaly detection approaches applied to TROPOSIF data. First a classical pixel wise z‐score method, and second a 3D Convolutional Autoencoder (CAE) that learns spatiotemporal SIF patterns. We evaluate both methods against two drought products, the global Standardised Precipitation‐Evapotranspiration Index (SPEI) metric and Brazil's Monitor de Secas (MSB). Our results show that the CAE identifies broader, more persistent anomalies than the z‐score method and aligns more closely with drought patterns detected by SPEI and MSB, particularly during the peak dry period in September 2024 and the months of delayed recovery that followed. These findings highlight how methodological choices shape the interpretation of vegetation stress and why anomaly detection methodologies should be carefully selected.

Plants often exhibit a midday depression in water use (i.e., transpiration), reflecting a constraint on their ability to sustain maximum water transport, which may occur at the cost of reduced photosynthesis. Eddy covariance observations and geostationary satellites cannot quantify this widespread phenomenon globally while resolving fine-scale spatial variability. Using evapotranspiration measurements from the ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS) and machine learning, we quantify the global distribution of midday depression in evapotranspiration. Midday depression primarily occurs during peak-growing seasons in temperate zones and dry periods in the tropics, with a morning shift of the peak evapotranspiration time by > 0.5 h. Droughts and heatwaves intensify midday depression, with peak evapotranspiration time shifting earlier, typically by 0.5 to 1.5 h and approaching 2 h during local extremes. By contrast, the coarse resolution upscaled flux products and hourly meteorological reanalysis datasets cannot capture this phenomenon. Midday depression of evapotranspiration detected using ECOSTRESS and machine learning is strongly associated with plant water stress, and helps us better understand how plants use water during drought and heatwaves.

Mesoscale convective systems (MCSs) are powerful contributors to extreme precipitation, often resulting in flash floods. However, the extent to which these systems are influenced by natural modes of climate variability is still not fully understood. In this study, we found that the occurrence of extreme precipitation from MCSs nearly doubles during the intense positive phase of the Baroclinic Annular Mode (BAM). The positive phase of BAM slows down the zonal translation speed of extratropical cyclone (ETC) trajectories, creating a favorable dynamical environment for extreme MCS precipitation. Furthermore, we discovered that when ETCs and MCSs coincide, they account for more than 90% of extreme precipitation during springtime in the U.S. These multiscale interactions between the planetary‐scale BAM, synoptic‐scale extratropical cyclones, and MCSs manifest as significant regional impacts on extreme precipitation.

Siberian wildfires, a major disturbance to ecosystems and carbon cycles, have intensified in recent decades, yet long‐term changes in fire seasonality remain unclear. Here, using a refined fire‐season framework, we identify a pronounced shift in fire season characteristics over eastern Siberia during 2002–2023, marked by substantial lengthening of the fire season and increasingly clustered occurrence of large fires. The fire season extends primarily through delayed termination of 1.48 days/yr into late autumn, associated with the growing persistence of extremely large September fires and more wildfire ignitions in October. We further find that enhanced fire‐weather conditions are the main driver, arising from accumulated drought associated with precipitation deficits in the preceding August–September period. These deficits coincide with persistent high‐pressure anomalies linked to declining Arctic sea ice and subpolar North Atlantic cooling, underscoring the importance of understanding mechanisms and improving early warning of antecedent hydroclimate in shaping high‐latitude fire‐season dynamics.

Increasingly frequent extreme weather events are heightening the chances of mosquito‐borne diseases outbreaks, posing a notable threat to human health. However, how extreme weather influences the transmission dynamics of mosquito‐borne diseases remains poorly understood. Here, we integrate heat waves and effective precipitation as parameterization schemes into an SEI‐SEIR dynamic model to quantify their driving effects on chikungunya transmission. Based on the 2025 chikungunya fever outbreak in Foshan, the model successfully reproduced the observed transmission dynamics. Attribution analysis revealed that heat waves exerted a dual effect on transmission. Mosquito suitability declined by 9.7%, yet accelerated viral replication and altered human behavior generated a positive forcing that drove a net 76% increase in infected cases. Effective precipitation regulated outbreak scale nonlinearly, dispersed rainfall increased mosquito abundance by 205% and amplified transmission by 10.5%. Our research offered new insights into the impact of extreme weather on mosquito‐borne infectious diseases.

Compound drought and heat extremes (CDHEs) under global warming pose significant threats to terrestrial ecosystems. However, the concurrent impacts of heat events and diverse drought types, as well as 3D (longitude, latitude, and time) vegetation recovery patterns, remain poorly understood. Here, we defined compound integrated drought-heat extremes through coupling heat extremes with diverse droughts and assessed basin-scale 3D vegetation recovery under varying hazards. The severity of CDHEs increased globally during 1982-2019, with the affected area expanding by 10 million km². Drought was a primary driver of vegetation loss, leading to longer vegetation recovery periods than either CDHEs or heat events, with approximately 44% of recovery events lasting longer than 10 months. In equatorial and high-latitude basins, CDHEs drove prolonged recovery events, though patterns varied among vegetation types. Spatially, vegetation loss generally kept pace with droughts, with extensive loss dominating low-latitude basins. This was less pronounced for CDHEs, which exhibited limited migration distances and could promote vegetation growth in some regions, including Europe and Central Asia. SHAP analysis revealed that temperature was a key factor shaping vegetation responses to CDHEs. Our study advances understanding of multivariate CDHEs, drought and heat impacts on vegetation responses, informing ecosystem management under climate change.

The White Nile from Lake Victoria through Lakes Kyoga and Albert to the Sudd wetlands forms a complex connected lake-river-wetland system where flood propagation, storage, and attenuation remain poorly quantified. Following unprecedented and persistent flooding across South Sudan in 2022, this study quantified system-scale flood-wave transit time and examined how long it takes a flood wave to travel from Lake Victoria to the Sudd and how upstream storage and connectivity shape multi-year flood behaviour. Using daily lake levels, discharge, CHIRPS rainfall, and MODIS-derived inundation for 2002–2024, we tracked sequential flood peaks through the Victoria–Kyoga–Albert–Sudd cascade and mapped monthly wetland dynamics across five South Sudan sub-catchments. Flood-wave tracking showed a mean system transit time of approximately 17 months (16.84±1.95 months; range 13.0–20.9 months), substantially longer than the commonly inferred four-to-five-month timescale based on seasonal peak alignment. Segmental analysis revealed rapid transmission from Victoria to Kyoga (mean 4.2 months) but strong attenuation through the Albert–Sudd reach (mean 9.3 months), consistent with extensive floodplain storage and backwater control. Correlations between Lake Victoria peaks and downstream wetland extents strengthened markedly after 2019, with r 2 exceeding 0.8 at 9–13-month lags, confirming strong hydraulic coupling and long system memory. These statistical lags complement, but do not represent, physical flood-wave transit times. The 2019–2024 high-water regime was not a series of isolated rainfall events but a multi-year propagation of excess storage initiated by the 2019 positive Indian Ocean Dipole anomaly and consecutive rainfall seasons. Lake Victoria reached exceptional peak levels of 1136.48 m a.s.l. in 2020, 1136.50 m a.s.l. in 2021, and 1136.66 m a.s.l. in 2024, each exceeding the historical 1964 maximum (1136.42 m a.s.l.). Over the same period, the Sudd Wetland exceeded its previous MODIS-era maximum extent (81 496 km 2 in 2016) in every year from 2019 to 2024, reaching annual maxima of 120 680, 111 684, 111 480, 163 475, 122 292, and 116 359 km 2 , respectively. Flood-persistence mapping shows a shift from rainfall-driven activation in the eastern Sudd (Baro-Akobbo-Sobat–White Nile) in 2019–2020 to sustained, inflow-driven inundation across the central and western Sudd…

Anthropogenic warming is projected to enhance future drought risk in the northeastern United States, yet the region has experienced precipitation increases. Here, we investigate how seasonal Northeast drought has changed over the last century in response to hydrologic intensification. To assess drought from 1901 to 2022, we use the Palmer drought severity index calculated with Global Historical Climatology Network station data. We find that Northeast drought across all months is 59% less frequent and 19% less intense over the last 40 years relative to 1901–83. Warm-season drought frequency and intensity decreased by 56% and 16%, respectively. These results are spatially consistent throughout the Northeast. The variance in monthly dry and wet events has not significantly changed, suggesting that the Northeast hydroclimate is not more volatile than in the past. We explore the potential drivers of evolving drought by rerunning our analysis with changes in temperature and precipitation removed, finding that reduced drought is attributable to precipitation increases, which to date have outpaced warming-driven evapotranspiration increases. Collectively, our results show a marked reduction in seasonal drought since 1984 in contrast with the enhanced regional drought risk projected across generations of climate model ensembles under anthropogenic forcing. However, these findings do not preclude the possibility of historical increases in drought at subseasonal time scales or enhanced future drought. Our results inform current and future Northeast water management decisions, as well as efforts to prioritize climate change hazards.

Compound heatwave-extreme precipitation (CHWEP) events, characterized by the rapid succession of heatwaves and extreme precipitation, pose escalating compound and cascading hazards. Yet, global-scale comparisons of their spatiotemporal evolution relative to single extremes remain limited. This study systematically examines the changes in CHWEP and corresponding single extremes from 1980 to 2100 using climate observations and projections under SSP (Shared Socioeconomic Pathway) 2-4.5 and SSP5-8.5 scenarios. We find that CHWEP exhibit higher frequency, stronger precipitation, and longer heatwave duration in mid-to-high latitudes of the Northern Hemisphere, while tropical CHWEP feature more intense heatwaves than single heatwave events. These spatial contrasts persist in future projections. Under both scenarios, CHWEP and single extreme metrics intensify globally by 2056–2100, with post-heatwave precipitation exceeding that of single precipitation extremes, particularly under SSP5-8.5, highlighting sensitivity to greenhouse forcing. Critically, the co-occurrence is non-random, indicating an emerging physical linkage. In the tropics, the likelihood of extreme precipitation following heatwaves increases markedly. Our findings demonstrate that CHWEPs are evolving into a distinct, intensifying hazard class, necessitating their integration into climate resilience, early warning, and adaptation frameworks.

Bankfull discharge, the maximum flow a river can convey before spilling over its banks, is central to modelling flood risk and understanding river-channel evolution. Global flood inundation models assume a 2-year return period for bankfull conditions, but this assumption remains untested globally. Here we use observations and machine learning to estimate bankfull discharge at ~1-km resolution along a recently developed global river network. We show that the widely used 2-year discharge reasonably approximates bankfull-flow magnitude, but bankfull return periods vary substantially within and across climate zones. Bankfull conditions occur more frequently in tropical and temperate regions (median return periods of 1.5 and 1.8 years; interquartile ranges of 2.5 and 3.2 years, respectively) and less frequently in cold and arid regions (2.8 and 4.3 years; interquartile ranges of 4.8 and 6.0 years). This variation across climate zones provides a basis for improving the representation of channel capacity in global flood models.

In June 2023, an extreme hydrometeorological event caused widespread flooding across central Chile (34°–39°S), abruptly interrupting a 13-yr dry period. This study provides a comprehensive analysis of the event, characterizing its atmospheric drivers, hydrological impacts, and the role of rain on snow (ROS) in a representative catchment. A category-4 zonal atmospheric river persisted for 4 days in a warm environment, resulting in a steep orographic precipitation gradient over the foothills of the Andes with rainfall accumulations of up to 600 mm and the freezing level above 3000 m MSL. Many rivers reached their highest flows on record, with peak discharges associated with return periods ranging from 9 to 199 years, resulting in significant damage to critical infrastructure and the general population. We also present a quantitative diagnosis of a ROS event in the Andes Cordillera using a physically based and semidistributed hydrological model in a representative catchment. Simulations reveal that snowmelt plus rainfall from snow-covered areas contributed ∼ 15% to the total flood volume, of which snowmelt was the secondary source driven by net longwave radiation and heat from the warm rainfall. This event serves as an example of a compound hydrometeorological extreme event, where intense and prolonged precipitation, combined with high temperatures, converged to produce a major flood in central Chile. Our findings highlight the region’s vulnerability to extreme events and underscore the need for future studies to assess the potential of an integrated hydrological forecasting system to mitigate the impacts of extreme events. Significance Statement The June 2023 flood in central Chile was a historic event, interrupting a 13-yr-long megadrought and producing significant damage to critical infrastructure. Our study provides a detailed hydrometeorological description of the event. Key features include its persistence over 4 days, rainfall accumulations above 600 mm, a steep orographic precipitation gradient, and the freezing level above 3000 m MSL, increasing the pluvial contributing areas and producing rain on snow in some regions, with many rivers reaching their highest flows on record. This study highlights the vulnerability of the region to extreme warm storms.

Understanding how groundwater–surface water interactions shape streamflow variability is critical for diagnosing low flow behavior and prediction bias in continental scale hydrologic models. We present a process informed framework that links observed baseflow (BF) dynamics, watershed attributes, and National Water Model (NWM) performance across the contiguous United States. Using daily observed streamflow from 797 reference quality streamgages, we developed monthly baseflow index (BFI) signatures using a streamgage specific, calibrated digital filter. Hierarchical clustering of these signatures identified seven distinct BFI regimes capturing regional and seasonal variability. We evaluated NWM v3.0 retrospective streamflow performance within each regime using multiple hydrograph and flow duration curve‐based metrics. Model skill varied systematically across regimes: mixed flow systems were simulated most accurately, while predominantly BF dominated and quickflow dominated regimes exhibited substantially poorer performance. Across nearly all regimes, the NWM underestimated observed BFI magnitude and frequently failed to reproduce seasonal BF patterns, indicating systematic biases in simulated low flow contributions. To relate these regimes to potential process controls, we trained a Random Forest classifier using static watershed attributes and applied Shapley Additive Explanations to identify features most strongly associated with each regime. Results highlight regionally varying influences, including the dominant role of snow fraction and seasonal runoff timing in snow dominated basins and the importance of evapotranspiration and aridity in quickflow dominated systems. Collectively, these findings demonstrate how hydrologic signatures combined with interpretable machine learning can diagnose regime specific model biases and generate process‐based hypotheses about limitations in large scale hydrologic prediction systems.

Flash droughts, defined by their rapid intensification, are typically treated as distinct from conventional, slow-evolving droughts. This study challenges this separation by investigating flash and conventional droughts as a unified continuum. Leveraging a cloud-tracking methodology, we develop an object-based drought inventory containing 868 drought events that occurred over the conterminous United States during the 1979–2023 growing seasons (April–October). These “drought objects” had an average duration of 6 weeks and an average area of 480 000 km 2 . We categorized them as conventional drought dominated (394), mixed (401), or flash drought dominated (73) based on the fractional area experiencing rapid intensification. A Lagrangian tracking analysis demonstrates that while more than half of these objects remain geographically anchored, the south-central United States emerges as a primary corridor for active drought genesis and migratory propagation, exhibiting a strong tendency for northeastward movement. Our analysis reveals a critical finding: Flash-dominated drought objects are associated with more extensive areas and longer durations than their conventional or mixed counterparts. Both the area-averaged intensification rate and the fractional area of flash drought significantly correlate with the object’s duration, affected area, and severity. This suggests that rapid intensification during flash drought is a key mechanism driving drought expansion and persistence. Significance Statement While drought is typically viewed as a slow-moving disaster, “flash droughts” develop quickly, often catching farmers and water managers off guard. In this study, we adapted a cloud-tracking method to follow droughts as moving objects across the United States from 1979 to 2023. By analyzing over 800 individual drought events, we found that flash droughts intensify over a longer period of time and that the resultant drought conditions last longer and cover more territory than typical, slower-developing droughts. These results show that a fast-starting drought often leads to a more persistent and expanding threat to water resources. These insights are critical for improving drought early warning systems and helping communities better prepare for these rapidly developing events.

While excess rainfall is associated with mosquito-borne disease because it supports mosquito breeding, drought may also counterintuitively increase disease transmission by altering mosquito and host behavior. This phenomenon is important to understand because climate change is projected to increase both extreme rainfall and drought. In this study, we investigated the extent to which seasonally driven mosquito and primate behavior drove the first yellow fever virus (YFV) epidemic in an urban area in Brazil in nearly a century, coinciding with an equally rare drought, and assessed the role of interventions in ending the outbreak. We hypothesized that drought-like conditions played a key role in the outbreak dynamics by driving the forest mosquitoes and nonhuman primates toward the city in search of water and that the mosquitoes were biting more frequently to avoid desiccation. A dynamical YFV model supports these hypotheses, showing that both behavioral changes were needed to explain the outbreak timing and incidence. Furthermore, a combination of vector control, conservation measures, and vaccination contributed to ending the outbreak, with the strongest effects from vaccination. Together, these results suggest that drought, likely to become more frequent in this region in the coming decades, can considerably influence mosquito-borne disease transmission and that sustained control will require multiple interventions.

Hot droughts are intensifying under anthropogenic warming, yet whether extreme heat systematically alters the triggering of abrupt drought‐to‐pluvial transitions and their downstream impacts remains poorly understood. Here, we identify global drought‐to‐pluvial abrupt transition (DPAT) events and isolate the influence of antecedent thermal conditions by comparing hot‐drought‐triggered DPATs (DH‐DPATs) with transitions preceded by droughts without concurrent heat. We find that hot drought related transitions display broader hotspots and substantially higher occurrence, with frequency increasing from 66.7% to 316.7% across the interquartile range and a higher conditional transition probability (global mean: 44.89% vs. 38.95%). They also transition faster, sustain longer post drought heavy rainfall, and produce stronger rainfall intensity. Mechanistically, although both event types are associated with enhanced moisture convergence, antecedent heat imposes an additional thermodynamic preconditioning characterized by deeper boundary layers and higher convective instability, making comparable moisture recovery more likely to trigger deep convection. These heat‐mediated transitions amplify ecosystem carbon losses, intensifying reductions in gross primary productivity and solar‐induced chlorophyll fluorescence by about 43.7% and 26.8%, respectively, with the strongest relative impacts in croplands. They also contribute to increasing population and managed‐land exposure. Under future warming, DPATs are projected to shift further toward hot‐drought‐dominated pathways, with increases in both the contribution and transition likelihood of DH‐DPATs. These results highlight a thermodynamic pathway through which hot droughts increase the likelihood and impacts of abrupt drought‐to‐pluvial transitions worldwide.

Predicting changes in urban pluvial flood hazards under climate warming is crucial for risk mitigation and disaster management. A key challenge in simulating future urban flood hazards is the scarcity of high-resolution rainfall projections, particularly at the sub-daily and kilometer scales required for hydrodynamic modeling. We present a cascading process-informed framework that requires minimal observed climatic data, enabling scenario analysis even in data-scarce cities. This framework consists of a distribution‐based spatial quantile mapping (DSQM) method to morph observed rainfall fields conditioned on temperature changes, a stochastic storm transposition (SST) method to account for the spatial variability of urban rainfall, and a rain‐on‐grid hydrodynamic model (AUTOSHED) for efficient simulation of urban pluvial floods at high spatio-temporal resolution. The framework allows the generation of stochastic rainfall fields under different rainfall return levels and regional warming levels. It supports the quantification of changes in future urban flood statistics with detailed hazard maps of inundation depth, duration, and flow velocity. We select the metropolitan area of Beijing (300 km 2 ) as a case study area and utilize gridded hourly and 1 km rainfall data to simulate flood evolution at 5 min and 5 m resolution under regional warming levels of 1, 3, and 5 °C relative to the period 1998–2019. Our results show that with rising temperatures, regional storms tend to become more intense but smaller in spatial extent, which may in turn drive increased local flood depth, accelerated flow velocity, and deeper inundation, collectively elevating pluvial flood risk. Specifically, mean rainfall intensity increases by 6 %, 11 %, and 20 % (respectively with the warming levels), peak flood depth exhibits a nonlinear increase of 4 %, 7 %, and 8 %, due to the complex interactions of reduced storm area, increased storm intensities, and rainfall spatial variability. The proposed DSQM-SST-AUTOSHED framework offers a data-driven, physically grounded, and efficient approach to assess urban flood risk under regional warming. It only requires observed rainfall fields and temperature datasets, which are readily accessible from public sources, making the approach easily extendable to other cities.

This study investigates the application of Physics‐Informed Neural Operators (PINOs) for solving the two‐dimensional shallow water equations (2D SWE) in the context of flood modeling. Unlike Physics‐Informed Neural Networks (PINNs), which require retraining for each new initial or boundary condition (BC), PINOs learn the solution operator itself, enabling fast inference across a range of conditions without retraining. PINNs have moreover been shown to struggle with complex bed topography due to the difficulty of satisfying the well‐balanced property (Tian et al., 2025, https://doi.org/10.1029/2025wr040052 ; Dazzi, 2024, https://doi.org/10.1029/2023wr036589 ), a limitation that PINOs can partially mitigate by incorporating data alongside the physics loss. We evaluate the method on experiments of increasing physical complexity: a radial dam break, constant BCs with and without friction, time‐dependent BCs, and a real‐world test case. The results show that PINOs capture key flood dynamics, particularly water depth, while reducing inference time by up to two orders of magnitude compared to numerical solvers. Relative test errors for water depth ranged from 0.3% for the radial dam break to 10.9% for cases with bottom topography, 7.3% with friction, and 9.0% under time‐dependent BCs. For the real‐world case, an additional data loss term yielded a water depth error of 25.8%. Larger errors were consistently observed for velocity components. The findings establish PINOs as a promising complement to traditional numerical solvers, offering a balance between computational efficiency and solution accuracy. Future work will focus on improving accuracy and extending the framework toward real‐world flood forecasting applications.

Socio‐hydrology models explore interactions between human and hydrological systems, enabling the analysis of dynamic feedbacks between society and flooding. Simplified conceptual models of socio‐hydrology often treat the social or human domain as single homogenous entities, failing to account for disparities among the communities that make up a city, and thus underestimating the true impacts of flooding on vulnerable populations. We propose a disaggregated socio‐hydrology model that incorporates a Resilience domain to capture differences among communities within a single city. Using empirical socioeconomic data from United States metropolitan areas, we inform the model's representation of inequality and its influence on community‐level adaptive capacities. With the further addition of a disaster relief component, our approach captures community‐specific recovery trajectories, linking economic inequalities to urban flood resilience. Results show that a community's adaptive capacity is shaped by a city's overall growth rate and the degree of economic disparities between communities. We find that improving a city's economic growth rate can unintentionally widen disparities in community adaptive capacity, consistent with the Kuznets curve hypothesis. However, simulations show that equitable resource allocation reduces recovery gaps without hindering overall city growth. Our results emphasize the necessity of incorporating socioeconomic heterogeneity into socio‐hydrological models to inform equitable disaster response policies. By grounding resilience in empirical economic data, this study advances socio‐hydrology modeling to provide more actionable insights for disaster response and community‐based risk reduction strategies.

Drought is a dominant factor influencing terrestrial ecosystem water-use efficiency (WUE). However, the coupling relationship between WUE and drought remains insufficiently understood. Currently, the coupling relationship is primarily assessed using correlation coefficients or linear regression slopes. However, the optimal drought timescale at which WUE responds to drought has largely been overlooked. Therefore, this study investigated the spatiotemporal patterns of the WUE – meteorological drought coupling relationship across global terrestrial ecosystems from 1982 to 2018 with satellite- derived and model-simulated WUE, together with the Standardized Precipitation-Evapotranspiration Index (SPEI), and explored the potential causal mechanisms. Within the framework of WUE-SPEI coupling, the maximum correlation coefficient between WUE and SPEI represents the sensitivity of WUE to meteorological drought ( R max ), whereas the corresponding optimal drought timescale represents its resistance time ( T opt ). The results indicated that the sensitivity of WUE to meteorological drought decreased at a rate of −0.0003 yr −1 ( p < 0.01), while the resistance time increased at a rate of 0.0155 month yr −1 ( p < 0.01), indicating a weakening of the coupling between WUE and meteorological drought. Attribution analysis indicated that CO 2 fertilization was the primary factor contributing to the weakening of the coupling relationship. Surface soil moisture was the most critical hydrometeorological driver, exhibiting nearly opposite effects and significant threshold effects on R max and T opt . Peter & Clark Momentary Conditional Independence (PCMCI+) algorithm was further employed to construct a causality diagnosis framework for identifying the relationships between WUE-drought coupling and temperature, precipitation, radiation, wind speed, vapor pressure deficit, and surface and root-zone soil moisture. The results showed that the decrease in the R max had direct negative causal effects on precipitation, temperature, and radiation. In contrast, the increase in the T opt was primarily driven by a negative causal effect of radiation. This study highlights the weakened coupling between WUE and meteorological drought, suggesting that vegetation's carbon-water trade-off is evolving toward drought adaptation, which is crucial for understanding the adaptive strategies of…

Traditional static designs of urban stormwater storage systems are increasingly inadequate for addressing extreme rainfalls. Recent advances highlight optimizing both spatial layout and real‐time control (RTC) of storage facilities, yet most of these decisions are made sequentially, overlooking how layout constraints can affect RTC efficiency and attain additional benefits from spatial‐temporal coordination. This study developed a synergistic optimization framework for coordinating stormwater storage system designs with their operational intelligence. This framework integrates a predictive target flow allocation‐based global control method with the optimization of storage capacity distribution, producing combined solutions of layout designs and control settings. Our methodology was applied to an urban catchment in Shenzhen, China, to evaluate the flood control performance under design and monitored heavy storms. Results show that the global RTC method substantially reduces flood peaks and runoff volumes under critical heavy storms. Compared with fixed‐layout or sequential optimization scenarios, integrating global RTC with synergistic optimization achieves the highest performance improvement over static control, with peak flow reduction rates increased by 11%∼32%. Moreover, the synergistic optimization identifies a subset of strategically located, high‐capacity control nodes contributing more than 98% of the system‐wide benefits, thus avoiding unnecessary RTC retrofitting elsewhere. These findings demonstrate the practical value of the synergistic optimization framework in guiding adaptive and efficient urban stormwater management.

An important challenge with Machine Learning (ML) is its transferability; that is, whether an ML model trained on one set of data can be applied to a second set of data without requiring full retraining of the model. Transfer Learning (TL) addresses this challenge by transferring knowledge learned in the source domain (the data it was trained on) to the target domain (a second set of data that is statistically different but related, which the model was not trained on). This study investigates the use of TL for street‐scale nuisance flood forecasting by exploring whether an ML model trained for one set of streets can effectively forecast flooding for another set of streets in the same city. TL is explored using a Long Short‐Term Memory (LSTM) model trained on data for the flood‐prone streets of Norfolk City, Virginia. The results show that full‐weight retraining proved most effective and minimal retraining of only the output layer was insufficient. The advantage of TL was most pronounced when target data was limited, meaning data collected at the new water depth sensor location included generally less than 18 flood events. As the target data increased beyond 18 flood events, the benefit of TL diminished relative to training an ML model directly on the local flood events. These findings can assist cities as they implement street‐scale flood sensing systems to create accurate forecasts for new sensing locations that do not yet have sufficient data records to train a local ML model.

Abrupt snowmelt, triggered by rain-on-snow events or “snow-eater heat waves,” can cause flooding, initiate or accelerate snow drought, and affect water availability. However, the characteristics (e.g., area, duration, and frequency), impacts, and trends of snow-eater heat waves have received little attention. To address this gap, we developed a method to identify snow-eater heat waves and estimate their melt potential using 20th Century Reanalysis version 3 air temperature data, the TempestExtremes algorithm, and an operational snowmelt model (SNOW-17) across 1850–2015. Melt season snow-eater heat waves typically last 3 to 5 days, with three to five events, doubling snowmelt rates. Seven of 11 spring superfloods are shown to coincide with snow-eater heat waves. Since the 1850s, snow-eater heat waves have increased in area and frequency, decreased in duration, and shifted earlier in the melt season. Incorporating snow-eater heat-wave impacts into SNOW-17 enhances extreme melt estimates, improving water management support tools.

Tropical regions may experience periodic extreme precipitation and suffer from associated periodic deluges in a warmer climate. Recent studies conducted small-domain (around 100 kilometers by 100 kilometers) atmospheric model simulations and found that precipitation transitions from a steady state to a periodic oscillation state in a hothouse climate when the sea surface temperature reaches 320 to 325 kelvin. Here, we conduct global-scale atmospheric model simulations with varying complexity. We find that tropical precipitation in convective regions already transitions to a O (10-day) periodic oscillation state with a O ( 100 mm day − 1 ) amplitude at 305 to 310 kelvin, about 15 kelvin lower than previously reported and within reach in a century under a high–carbon emission scenario. Using a low-order analytical model, we attribute the onset of the periodic extreme precipitation to the intensification of convectively coupled waves that were not considered in the previous small-domain simulations. Our mechanism highlights the importance of the interaction between convection and atmospheric wave dynamics for climate change.

Extreme drought and heat are driving forest die‐off globally, yet regional‐scale mapping of subsurface water storage capacity, governing forest vulnerability, remains a major challenge. Here, we demonstrate that airborne gamma‐ray spectrometry (GRS) of potassium ( 40 K) decay provides a powerful proxy for subsurface constraints on water availability. In southwestern Australia, we found that spatial patterns of forest die‐off during an extreme 2023–2024 drought‐heat event were strongly predicted by surficial 40 K concentrations. While GRS measures the upper ∼30 cm of soil, plot‐scale electrical resistivity tomography suggests that elevated 40 K signals shallow granite or gneiss bedrock, to depths reaching several tens of meters due to leaching of mobile elements, remnant bedrock minerals, erosional thinning of regolith and biological nutrient recycling from K‐rich saprolite. Airborne GRS can map forest vulnerability across highly leached landscapes, which cover one‐third of Earth's ice‐free land surface. This enables targeted adaptive management of forests for future drought‐heat intensification.

Flood risk is correlated in space and time, challenging insurance systems that rely on diversification across assets. Financial instruments governing flood coverage are typically structured as 1–5‐year contracts, exposing portfolios to climate‐driven risk at interannual‐to‐decadal scales. Yet existing tools address climate risk either through seasonal forecasts extending only months or multidecadal projections misaligned with fiscal horizons, leaving a critical gap in actionable flood risk simulation. We introduce a multisite flood simulation framework combining attention‐based analog retrieval with stochastic generation of multivariate flood frequency, intensity, and duration sequences. Applied to over 100 sites in the Mississippi River Basin, the model produces spatiotemporally coherent flood portfolios conditioned on interannual climate variability. Explainable AI attribution paired with wavelet analysis links simulated clustering to large‐scale climate drivers, yielding physically interpretable flood clusters for portfolio‐scale loss simulation. The framework provides plausible, out‐of‐sample flood risk catalogs for interannual‐to‐decadal insurance risk assessment and financial planning.

Climate change is causing the magnitudes of extreme sub-daily precipitation events to increase. The ability to predict changes to these precipitation extremes is crucial for disaster preparedness. The TENAX model was proposed to predict return levels of sub-daily extreme precipitation under climate change based on the projected temperature shifts. It combines a Weibull distribution with an exponential temperature dependence in the scale parameter, accounting for the Clausius–Clapeyron relation, with an explicit representation of the temperatures during precipitation events. The Weibull distribution's shape parameter could also have a temperature dependence, which would mean that the tail heaviness changes with temperature. This implies that the rarest events may increase at faster rates. However, implementing this dependence increases the number of parameters to be estimated, affecting the model's accuracy. Here, we use hourly data from thousands of rain gauges in Germany, Japan, the UK, and the USA to assess the dependence of the Weibull shape parameter on temperature, exploring how it should be implemented in the TENAX model. We find that there is a significant dependence in many stations and that the magnitude and sign of the dependence have regional patterns. In the majority of stations, the sign is negative, implying that rarer events intensify with temperature at a higher rate. However, Monte Carlo simulations show that including this dependence without careful consideration may lead to overestimation of precipitation return levels and increase the model uncertainty. The dependence should therefore be introduced with caution, in the context of surrounding stations.

Characterizing storms that generate heavy precipitation across different timescales is essential for parameterizing and assessing stochastic weather generators, evaluating climate models, enhancing risk management, and understanding climate change impacts. This study develops a detailed climatology of precipitation and temperature time series for storms producing heavy rainfall at sub‐hourly to daily scales across the Greater Alpine Region. Using a high‐resolution data set from ∼400 climate stations with 15–44 years of records for which a storm typology is available, we identify independent storm events and describe them through multiple characteristics, including peak intensity, lifetime, depth, temporal structure, internal variability, and thermal and geographical attributes. At each climate station, we focus on storms producing the largest 10% of the precipitation intensities at durations from 10 min to 24 hr. Results show that storm characteristics vary systematically with both duration and geography. At sub‐hourly scales, the highest intensities are typically generated by short‐lived storms with front‐loaded temporal profiles, high internal variability, and warmer conditions, especially in lowland regions, and exhibit evident reverse elevation‐intensity relationship. At daily scales, the highest intensities are associated with storms of greater depth, more symmetric temporal profiles, and smaller pre‐to‐within‐storm temperature differences, particularly in the inner Alps. At multi‐hour durations, distinct spatial patterns highlight the varying dominance of convective versus stratiform precipitation mechanisms. This climatology provides new insights into storm behavior in complex terrain and can inform precipitation modeling, climate impact assessments, and hydro‐geomorphological studies.

The increased frequency and intensity of hydrological extremes, including drought, due to anthropogenic climate change will drive the need for enhanced water supply resilience, even in water-rich countries. Previous studies have shown that small reservoirs have considerable potential for expanding water supply for various purposes, including when repurposed from flood-only reservoirs for both flood and drought protection. However, whether these repurposed reservoirs retain the same flood protection ability when operating under forecasts is still unclear, as reservoir operation under forecasts has primarily been researched in the context of large reservoirs. In this study, we investigated potential operating rules under forecasts for 30 small-to-midsize (less than 5 million m 3 ) flood reservoirs to (a) determine if the uncertainty introduced by forecasts degrades the performance of repurposed reservoirs so significantly as to render the concept unusable, (b) identify patterns in the relationship between forecast accuracy and optimal reservoir performance, and (c) identify patterns in optimal reservoir operation rules, under the constraint that flood protection should not be compromised. Performance is determined by the modelled ability to either supplement streamflow to avoid low flows or to provide water for irrigation purposes in the area of the reservoir. 1000 combinations of three operation parameters – the warning threshold at which flood pre-release begins, the rate at which water is released from the reservoir for flood pre-release, and the inflow at which the reservoir begins storing water – were sampled to optimize for both maintenance of flood protection (viability) and benefit for the reservoir's additional uses. While some reservoirs indeed were no longer beneficial when optimized to operate under forecasts, many still maintained benefits above 40 %, with a couple even surpassing their performance under perfect knowledge. Comparing changes in benefit from the perfect knowledge operation to forecast accuracy indicated that high rates of hits, false alarms and misses, and misses (HFM) could explain the largest decreases in performance, while other forecast accuracy metrics were less impactful. However, even if HFM were low but nonzero, a poorly-timed false alarm could drain a reservoir's storage before a spike in demand, causing a noticeable loss…

Compound events, arising from the combinations of multiple interacting underlying physical drivers, have gained increasing attention over the past decade with a wide range of applications, including floods, droughts, and wildfires. This growing body of research has advanced both the conceptual understanding and statistical modeling of compound events, often emphasizing the role of dependence between underlying drivers. Building on this foundation, we recognize the need to further describe hydrological and climate phenomena not only through the overall tendency of contributing physical drivers to interact but also by considering the individual combinations of drivers leading to compound events, regardless of whether these drivers are statistically dependent. Here, we introduce the concepts of structural and transient compoundness as system properties of compound events. Structural compoundness quantifies the overall contribution of each driver to the generative mechanism of the compound variable, while transient compoundness captures the role of each driver in the individual realization of the compound variable. This will ultimately improve risk assessment and mitigation strategies. Indeed, by applying these concepts to two types of compound events, river discharge at confluences and hot and dry events, we show that structural and transient compoundness can inform modeling choices and enhance system understanding, with both the dependence and magnitude of the drivers playing critical roles in shaping compound events. Incorporating structural and transient compoundness enhances scientific modeling and, at the same time, improves risk assessment, guiding adaptation strategies for persistent (structural) and temporary (transient) risks.

Dryland tree survival critically depends on deep soil water (DSW), particularly during dry spells. However, DSW is often limited and can be exhausted during periods of active tree growth, as recharge from precipitation is minimal. The eco‐physiological responses of trees to compound droughts (i.e., combined deep soil and atmospheric droughts) remain poorly understood, particularly the variation of sensitivity to atmospheric drought under deep soil desiccation. This study addresses this gap through a three‐year field experiment (2019–2021) on China's Loess Plateau, a global hotspot for DSW depletion due to extensive afforestation. We found that deep soil drying below 200 cm significantly suppressed daily sap flux density ( J s) by 8%–13%, transpiration rate by 24%–36%, and canopy conductance by 25%–34%, in both normal and dry years. Importantly, deep soil drying heightened the sensitivity of trees to vapor pressure deficit (VPD), as demonstrated by two key physiological responses. First, partitioned trees (i.e., those with physically isolated deep and shallow root systems) exhibited reduced hydraulic buffering capacity, with hysteresis analysis revealing a 16% decrease in hysteresis area compared to controls, indicating stronger J s–VPD coupling. Second, partitioned trees initiated stomatal closure at lower VPD thresholds (1.0–1.4 kPa) compared to controls (1.3–2.8 kPa). These findings underscore the role of deep soil water as ecological insurance against compound droughts, and highlight that its depletion exacerbates tree vulnerability to such stresses. Our results may provide insights into the effect of compound atmospheric and soil drought on the evolution of managed ecosystems.

Traditional methods for assessing levee projects focus solely on single‐year economic data, evaluating only the reduction in flood‐related economic losses while neglecting the benefits of saving lives and the dynamic changes in the economy and population over the project's design life. This study introduces a dynamic evaluation method for flood control and waterlogging drainage benefits of levee projects, incorporating economic and demographic changes. The total benefits are categorized into economic and life‐saving components. Separate hydrodynamic models for flood routing in river channels and rainfall‐waterlogging models in the basin are established. Through comparative analysis (with and without projects), the number of lives saved and the economic benefits generated by flood control and waterlogging drainage can be estimated. The Value of Statistical Life (VSL) is used to monetize life benefits, and calculation methods for flood control and waterlogging drainage benefits that considers both economic and life benefits are proposed. Then, a dynamic model incorporating regional GDP growth rates and population change coefficients is constructed, which can quantify the above‐mentioned benefits over the project's lifespan. A case study demonstrates that annual benefits increase from CNY 10.6 million in the base year to CNY 34.4 million by the end of the design life, with flood control benefits accounting for 59.1% and drainage benefits for 40.9%. It provides a comprehensive tool to assess long‐term project value, supporting investment decisions and improving benefit assessment systems for disaster risk reduction in levee projects.

Historical droughts and projected increases in drought risk have created an impetus for irrigation development across much of sub-Saharan Africa, including Malawi. While irrigation can boost incomes and enhance food security, interactions between drought risk and irrigation planning are insufficiently explored, heightening the risks of expensive irrigation infrastructure and competition for water resources. We use a suite of bias-corrected outputs of 18 global climate models coupled with a Soil and Water Assessment Tool (SWAT) model to examine the dynamics of water availability and irrigation water demand (IWD) in five river basins over Malawi, southern Africa. We find that strong coupling between precipitation and streamflow creates conditions suitable for quick propagation of drought signals typified by short lags between meteorological and hydrological droughts. Consequently, episodic spikes in IWD often coincide with periods of diminished streamflow, thus reducing water supply for irrigation when IWD is most pronounced. Mean IWD is projected to increase across the five river basins (ranging from 20% to 50% for the low-mitigation scenario). The frequency of IWD spikes is also projected to increase. Meteorological–hydrological drought lags are projected to become shorter, increasing the prospect of IWD spikes coinciding with hydrological droughts more often. Having grounded the study on extremes, our results highlight other dimensions of climate risks for irrigation aside changing average climatic conditions. Such information is critical in supporting decisions around sustainable and resilient irrigation landscapes, but prevailing uncertainties demand robust decision-making approaches and model improvements. Significance Statement This study sought to develop a picture of the interaction between climate risk and irrigation development in Malawi. Our findings reveal that climate change and variability create conditions of isolated incidents of amplified irrigation water demand that coincide with periods of water scarcity. Irrigation water demand is consistently projected to increase across the twenty-first century, with the risk of decreasing water supply imposing considerable limits to irrigation development. While views that irrigation presents a key route to navigating climate-related challenges—often centered around droughts—underlie the rationale…

Water labeled with stable isotopes provides a conservative tracer, being neither produced nor consumed, for water flowpaths within soils and root systems. We added a strong, evenly distributed 2 HHO label to one m 2 of soil surface and continuously monitored its passage downward into the soil and upward into the stems of surrounding trees, with the objective of illuminating spatiotemporal lateral root water uptake and overlap. The study was conducted during the historic drought of 2018 in a mature Scots pine ( Pinus sylvestris ) forest growing on sandy soil in northern Sweden. Continuous in situ isotopic measurements of tree xylem water evidenced root system overlap of six trees within the labeled square meter. This result is consistent with previous estimates from labelled nutrient uptake measurements at this site. However, label uptake differed sharply among trees, even within the same radius; 90 % of the label was taken up by one of the two trees closest to the labelled plot. Horizontal transport rates in tree roots averaged 0.17±0.05 m d −1 , meaning that the arrival of label pulse in tree stems was delayed by 6–33 d from first tree to last. Root water uptake by trees appeared restricted to the upper 60 cm of mineral soil, even at the peak of the drought; the volume-weighted average depth of uptake was 30.6±1.2 cm throughout the study period. Label intensity of the mineral soil weakened throughout the drought, consistent with the notion that the label was being dispersed or diluted into pre-event water. Our data provide a daily and three-dimensional description of the passage of a labeled water pulse, highlighting the heterogeneity in horizontal water distribution by roots and the uneven partitioning of label among individual trees in a boreal forest.