We define fire weather waves as persistent extreme fire weather, which can intensify fire activity by sustaining exceptionally warm, dry, and windy conditions. Here, we use daily fire weather index, fire activity, and meteorological data to examine the impacts of fire weather waves on fires, as well as their patterns and trends across global terrestrial ecoregions. Fire weather waves account for only 4% of days but coincide with 26% of the area burned and nearly half of the top 1% most energetic fires in forested ecoregions. Compared with grassland and shrubland fires, forest fires exhibit a larger and more persistent increase in daily burned area in response to fire weather waves, particularly in Mediterranean forests. Fire weather wave frequency has significantly increased across most burnable lands during 1979–2024. Climate projections indicate that fire weather waves will increase throughout the 21st century. These findings underscore fire weather waves as an essential component of early warning systems to strengthen preparedness for extreme fires. Fire weather waves are strongly linked to both overall fire activity and the most extreme wildfire events, highlighting their value as a predictor of fire risk, based on an analysis of fire, fire weather index, and meteorological data.

Wildfires pose severe threats to lives, critical infrastructure, and ecosystems, particularly in urban-wildland interface settings where rapid situational awareness is essential for response and early recovery. This study presents a comparative multisensor satellite workflow for rapid wildfire damage assessment, demonstrated using the 2025 Pacific Palisades Fire (California) as a case study. We integrate Sentinel-2 optical imagery with ALOS-2 (L-band) and Sentinel-1 (C-band) synthetic aperture radar (SAR) data to map burned areas and characterize structural impacts. Burned regions are delineated using optical spectral indices, while SAR coherence and texture features support urban damage characterization, including conditions where optical observations are hindered by smoke or clouds. Results indicate that ALOS-2 achieved higher classification performance (84.64% accuracy, Kappa = 0.73) compared with Sentinel-1 (78.31% accuracy, Kappa = 0.61), underscoring the added value of Lband SAR for post-fire impact mapping. The event affected 92.99 km2, including 92.05 km2 of forest and 0.94 km2 of residential areas. Beyond technical performance, the comparison highlights operational trade-offs relevant to disaster response, where higher-accuracy products must be balanced against data accessibility and acquisition frequency. Overall, the proposed multisensor analysis supports rapid post-fire assessment and provides actionable information for emergency response and evidence-based recovery planning in fire-prone regions.

Concurrent extreme fire weather creates favorable conditions for widespread large fires, which can complicate the coordination of fire suppression resources and degrade regional air quality. Here, we examine the patterns and trends of intra- and interregional synchronous fire weather (SFW) and explore their links to climate variability and air quality impacts. We find climatologically elevated intraregional SFW in boreal regions, as well as interregional synchronicity among northern temperate and boreal regions. Significant increases in SFW occurred during 1979 to 2024, with more than a twofold increase observed in most regions. We estimate that over half of the observed increase is attributable to anthropogenic climate change. Internal modes of climate variability strongly influence SFW in several regions, including Equatorial Asia, which experiences 43 additional intraregional SFW days during El Niño years. Furthermore, SFW is strongly correlated with regional fire-sourced PM 2.5 in multiple regions globally. These findings highlight the growing challenges posed by SFW for firefighting coordination and human health.

It is widely perceived that wildfire activity has increased across the western United States (WUS), with studies generally focusing on large wildfires. This study examines changes in the number of wildfire ignitions and burned area in the WUS using a comprehensive record of more than 750 000 wildfire incidents from 1992 to 2020 across biophysical gradients, seasons, and ignition causes. It further analyzes the environmental conditions associated with elevated fire activity. Our analysis comparing the periods 1992-2006 and 2007-2020 indicates that the annual number of recorded wildfires (>0.04 ha) in the WUS declined by 31% despite a 40% increase in burned area. This finding denotes that although wildfire prevention and mitigation strategies have decreased the overall number of fires, environmental conditions promoted larger fire sizes. Burned area increased in both forested and non-forested areas and across human- and naturally-ignited fires, with the greatest increase (84%) observed in lightning-caused forest fires. The seasonality of natural ignitions remained largely unchanged from 1992 to 2020, whereas the average day of year of ignition for human-caused wildfires shifted >12 d earlier (statistically significant), driven by an increasing number of springtime ignitions across the WUS. Ignitions increasingly occurred on days with abnormally drier-hotter weather compared to their climatology. Our analysis also showed that the median energy release component activation threshold associated with wildfire ignitions was similar to 42 (interquartile range, IQR: 34-50 across ecoregions). Activation threshold refers to the fire danger level beyond which the number of ignitions increases markedly. This threshold for fire size area was similar to 50 (IQR: 43-60). Findings of this study provide critical information for the development of wildfire prevention and mitigation strategies.

Key Points The editors thank the 2025 peer reviewers

The chronic water crisis in Iran stems from decades of water-intensive development, fragmented governance and national priorities that sidelined environmental protection. Without major governance reform and reprioritization, technical solutions alone cannot stop worsening water and environmental degradation.

Wildfires have increasingly affected human and natural systems across the western United States (WUS) in recent decades. Given that the majority of ignitions are human-caused and potentially preventable, improving the ability to predict fire occurrence is critical for effective wildfire prevention and risk mitigation. We used over 500,000 wildfire ignition records from 2000 to 2020 to develop machine learning models that predict daily ignition probability across the WUS and incorporate a wide range of physical, biological, social, and administrative variables. A key innovation of this work is development of novel sampling techniques for representing ignition absence. Unlike traditional purely random sampling or hyper-sampling, which does not account for temporally autocorrelated factors (such as droughts, insect outbreaks, and heatwaves) and spatially autocorrelated factors (such as proximity to human settlements, infrastructure presence, and fuel type), we introduce spatially and temporally stratified sampling of ignition absence. By drawing absence samples near the location and time of historical ignitions, we better captured the complex environmental and anthropogenic conditions associated with fire occurrence or lack thereof. Models trained without stratified sampling produced ignition probability maps that consistently overestimated fire risk during high fire danger periods, whereas models incorporating stratified fire absence samples more accurately captured the spatial and temporal variability of fire potential and achieved predictive accuracies exceeding 95%. In addition to operational utility for fire prevention and resource allocation, our approach offers insights into the drivers of wildfire ignitions and highlights the value of incorporating spatial and temporal structure in absence sampling for wildfire modeling.

To enhance hydrologic modeling, the hydrology community has developed benchmark datasets (e.g. Model Parameter Estimation Experiment, MOPEX), providing standardized data for model evaluation and parameter estimation. However, these datasets primarily focus on modeling natural hydrologic processes, leaving a critical gap in understanding the role of human influences. Here, we introduce the Coupled Hydrology-Human Activity Information (CHHAI) dataset, a benchmark dataset that integrates coupled human-water data from regions across all continents, excluding Antarctica. CHHAI incorporates data from 25 regions that cover various human impacts such as reservoir management, flood protection, river management policies, land use changes, and water use. Each basin reflects distinct challenges, providing a diverse and globally representative resource for researchers studying these processes. By offering standardized datasets for modeling and analysis, CHHAI aims to enhance our understanding of interactions between people and water and to support the development of improved strategies for managing coupled human-water systems.

Abstract Wildfire impacts on United States (US) communities have escalated in recent decades, highlighting the need to better understand factors that influence wildfire outcomes. We find that 462,069 homes were exposed to wildfires across the contiguous US during 2001–2020, two‐thirds of which occurred in the western US. Residential structure survivability—the percent of structures within a wildfire perimeter that were not destroyed by the fire—remained stable in the eastern US in the past two decades, but declined by 12% in the West coincident with a sixfold increase in the number of homes exposed to fire. Survivability was explained by structural age, fuels, and fire weather. Survivability to large fires in the West was lower for homes built pre‐1990 (84%) compared to post‐1990 homes (90%). Survivability was lower in forests compared to grasslands and shrublands. Finally, survivability was markedly lower for fires coincident with extreme fire weather. Our results suggest that modern building codes, fuel management, and proactive planning are pathways for strengthening wildfire resilience of the built environment.

Effective management of coupled human-water (CHW) systems under uncertainty demands policies that are both robust across a wide range of plausible futures and dynamically resilient to shocks. However, existing optimization approaches typically treat resilience statically or as a secondary constraint, overlooking systems' adaptive recovery dynamics and stakeholder-driven local contexts. Here, we apply a two-stage Resilient-Robust (R2) approach that (i) co-develops five localized Shared Socio-economic Pathways (LSSPs) with stakeholders, (ii) optimizes five dynamic resilience metrics within per-LSSP many-objective searches (Stage A), and (iii) conducts a post-search, cross-scenario robustness evaluation to identify stable policies (Stage B). Applied to Pakistan's Rechna Doab region, our framework reveals that under various future scenarios, groundwater level and economic water use efficiency, i.e. farm income, proved to be highly sensitive and unstable. Despite high spatial heterogeneity across the study region, the system showed a non-linear temporal decline from 2020 to 2050 according to the resilience criteria. LSSP1 shows high resilience, recovering 32% faster in farm income and maintaining 20% more stable groundwater than LSSP5, while LSSP3 lags with longer recovery times and more severe disturbances; similarly, policies optimized under LSSP1 scenarios are notably robust (79.2% mean, with 58% exceeding 80%) compared to LSSP3's lower performance (15% below 50%). Empirically, we find that policies optimized under lower adaptation-challenge narratives (LSSP1) exhibit greater cross-scenario robustness than those from high-challenge narratives (LSSP3). By separating the per-scenario search from the post-search evaluation, our R2 approach provides a transparent, stakeholder-grounded basis for resilient water-policy interventions under uncertainty.

In 2025, global annual drought affected ~30% of the global land surface. Near-record warming substantially increased evaporative demand, triggering widespread meteorological and soil moisture drought, even in regions where precipitation deficits were relatively intermediate.

Abstract The Caspian Sea, the Earth's largest inland water body, faces water level decline, drawing comparisons to the collapse of the Aral Sea. Unlike the Aral Sea, the relative roles of climatic variability, hydrological changes, and anthropogenic pressures on the Caspian Sea remain poorly understood. Here, we integrate satellite observations, in situ hydrological records and reanalysis data to examine recent drivers of the Caspian water loss. We show that total river inflow to the Caspian Sea has declined significantly, primarily due to reduced discharge from the Volga River. At the same time, precipitation over the basin has remained broadly stable, while evaporation over the sea has shown a modest upward trend. These findings point to compound anthropogenic and climatic influences on the regional water balance. We also detect a long‐term increase in chlorophyll‐a concentrations in the shallow Northern Caspian, signaling growing ecological stress associated with ongoing hydrological change. Avoiding further ecological disruption requires coordinated international action and policies to mitigate shrinkage by optimizing water allocation and environmental releases, as well as prioritizing long‐term ecosystem resilience. Without urgent intervention, the Caspian Sea risks following the trajectory of other desiccating inland water bodies, with long‐lasting ecological and socioeconomic consequences.

Hyperspectral change detection (HCD) is a critical remote sensing approach for monitoring land surface changes. Despite notable progress, state-of-the-art HCD methodologies encounter difficulties in modeling the high-dimensional, nonlinear spectral characteristics of hyperspectral imagery when comparing pre- and post-change imagery. To address these limitations, we propose a novel deep learning architecture, termed Cross Kolmogorov–Arnold Network (CrossKAN), for accurate and interpretable HCD. The CrossKAN model is predicated on the functional decomposition theory of Kolmogorov–Arnold Networks, a theoretical framework that enables compact and mathematically grounded modeling of complex spectral relationships. A Siamese architecture is employed to process bi-temporal image patches, enabling robust feature extraction by KAN layers based on Chebyshev polynomials. Next, deep features are fused in the CrossKAN layer, and are fed to the subsequent KAN layers to discriminate between change and no-change locations. CrossKAN's performance was assessed using four benchmark datasets in different geographical locations with divergent context and change classes. CrossKAN outperformed state-of-the-art HCD models, including SSTFormer, DBS3TAN, ML-EDAN, and MSDFFN, and achieved an overall accuracy of >94%. Low missed detection and false alarm rates demonstrate CrossKAN's superior effectiveness and generalization in complex regions.

Wildfire impacts on US communities have escalated in recent decades, highlighting the need to better understand factors that influence wildfire outcomes. We find that 567,000 homes were exposed to wildfires across the contiguous US during 2001-2020, two-thirds of which occurred and increased five-fold in the Western US. While residential structure survivability - the percent of structures within a wildfire perimeter that survive the fire - remained stable in the Eastern US in the past two decades, it declined by 10

Effective wildfire prevention includes actions to deliberately target different wildfire causes. However, the cause of an increasing number of wildfires is unknown, hindering targeted prevention efforts. We developed a machine learning model of wildfire ignition cause across the western United States on the basis of physical, biological, social, and management attributes associated with wildfires. Trained on wildfires from 1992 to 2020 with 12 known causes, the overall accuracy of our model exceeded 70% when applied to out-of-sample test data. Our model more accurately separated wildfires ignited by natural versus human causes (93% accuracy), and discriminated among the 11 classes of human-ignited wildfires with 55% accuracy. Our model attributed the greatest percentage of 150,247 wildfires from 1992 to 2020 for which the ignition source was unknown to equipment and vehicle use (21%), lightning (20%), and arson and incendiarism (18%).

Accurate mapping of flood susceptibility (FSM) is of paramount importance for the effective management and mitigation of this deadly disaster. This study introduces a novel framework based on the Kolmogorov-Arnold Network (KAN) for enhanced FSM, which was applied to two basins in Iran: the Karun and Gorganrud basins. Three KAN-based models were implemented and evaluated. The performance of the Boubaker-KAN, Cheby-KAN, and VietaPell-KAN models was evaluated in comparison to state-of-the-art machine learning techniques, including Random Forest (RF), Light Gradient Boosting Machine (LightGBM), Extreme Gradient Boosting (XGBoost), and Categorical Boosting (CatBoost). All models were trained using a set of conditioning factors related to flooding, including topographical indicators, land cover data, and soil characteristics. The delineation of flood-prone areas was conducted through the identification of historical inundation incidents, as observed in satellite imagery and documented in official reports. The results demonstrate that KAN-based models exhibit superior performance, with an average overall accuracy of 92.5 % across the two basins. Furthermore, the KAN models achieved an average F1 score of 93.90 % and an average Matthews Correlation Coefficient (MCC) of 0.866, demonstrating superior performance in these key metrics in comparison to other techniques. The superior performance of KAN-based models can be attributed to their capacity to capture intricate, non-linear relationships between flood conditioning factors and flood occurrences, as per the Kolmogorov-Arnold representation theorem. A visual comparison of the flood susceptibility maps demonstrates that KAN models effectively capture the subtle topographical and hydrological features that contribute to localized flooding. This research contributes to the advancement of FSM techniques, offering improved tools for flood risk assessment and management. Future work should focus on incorporating additional dynamic variables and exploring hybrid approaches combining KAN architectures with ensemble methods.

The area burned in the western United States during the 2020 fire season was the greatest in the modern era. Here we show that the number of human‐caused fires in 2020 also was elevated, nearly 20% higher than the 1992–2019 average. Although anomalously dry conditions enabled ignitions to spread and contributed to record area burned, these conditions alone do not explain the surge in the number of human‐caused ignitions. We argue that behavioral shifts aimed at curtailing the spread of COVID‐19 altered human‐environment interactions to favor increased ignitions. For example, the number of recreation‐caused wildfires during summer was 36% greater than the 1992–2019 average; this increase was likely a function of increased outdoor recreational activity in response to social distancing measures. We hypothesize that the combination of anomalously dry conditions and COVID‐19 social disruptions contributed to widespread increases in human‐caused ignitions, adding complexity to fire management efforts during the 2020 western US fire season. Knowledge of how social behavior changes indirectly contributed to the increased number of ignitions in the 2020 wildfire season can help inform resource management in an increasingly flammable world.