We investigated the Central Texas flash flood on July 2025 to understand why forecasts underestimated its severity. We analyzed operational streamflow forecasts and evaluate flood inundation maps against benchmark flood extent from High Water Mark. Results show that peak flooding was missed due to spatial error in rainfall and failures in monitoring stations, leading to an underestimation of impacts. We highlight the need for uncertainty aware flood predictions to support emergency response.

Every year, people are caught in avalanches they themselves triggered, and most had some safety training. This study asked what research has learned about how people make decisions in avalanche terrain. We reviewed 35 recent studies and found growth concentrated in decision strategies and methods, with cognitive bias research notably absent. The findings point toward causal research designs and field-based education that meets recreationists where they actually are in the mountains.

Glacial lake outburst floods (GLOFs) are major concern in high mountain regions across the globe and effective disaster risk management requires predictive modelling which needs various input data including potential flood volume. Building on the analysis of breached channels of largest documented GLOFs originating from moraine dam failures, we present new methodology that estimates maximum breach depth and so potential flood volumes of moraine-dammed lake outbursts.

Using a Lagrangian moisture–temperature framework, we show that the 2022 European mega‑drought featured a strong decoupling between enhanced regional moisture uptake and sharply reduced Atlantic–Mediterranean precipitation contributions. Persistent anticyclonic circulation and subsidence suppressed convection and induced adiabatic warming. The abrupt termination highlights that large‑scale dynamics, not moisture availability, control drought severity.

Long-term changes in rainfall conditions triggering landslides in Calabria (Southern Italy) over 1921–2020 are analysed through a dynamic approach in rainfall threshold definition. Shifts in seasonality and a decreasing trend in triggering rainfall are observed over the years, suggesting an increased propensity of the territory to generate landslides, even with less severe triggering rainfall in recent decades. The work highlights the essential role of landslide data collection and archiving.

Climate change will increase the frequency of extreme weather events like floods, droughts and storms. These hazards have impacts on human health. Some hazards, like storms, cause additional hazards like flooding. Little is known about the compound effects of multiple hazards on health, or how impacts vary between communities and areas. We present ten urgent research priorities related to the impact of multiple hazards on health to guide future decision-making towards climate resilience.

Many studies highlight drought impacts on productive sectors, such as agriculture and navigation, but economic consequences for human health remain understudied. We conducted an economic valuation of drought-related health interventions in Brazil, showing that ensuring groundwater access during severe droughts can avert substantial losses. Estimated benefits from reduced diarrhea hospitalizations total 9.92 % of local GDP (Gross Domestic Product), with avoidable losses reaching USD 1.15 billion at the state level.

Forecasting the potential position of future eruptions is fundamental for managing volcanic hazards. Here, we develop a new approach to identify the most likely positions for future eruptions based on the propagation path of the magma that fed past eruptions. Its application to Campi Flegrei shows probability peaks at 2 and 4 km from the caldera center and in the direction of existing topographic peaks. High probability areas correlate well with caldera’s structure and recent major seismicity.

Debris flows are fast-moving mixtures of rocks, sediment, and water that threaten mountain communities. Their speed is controlled partly by sliding at the flow base, but this has never been directly measured in nature. We built a novel sensor setup in Tyrol, Austria, to close this gap. Two events showed signs of partial occurrence of base sliding, strongest at the flow front. Future work will study more events and refine the approach to improve hazard prediction.