Snow is the main hydrological component, and its contribution plays a key role in runoff formation in Uzbekistan, Central Asia. However, climate change has altered snow accumulation dynamics over the past two decades. This study investigated the impacts of seasonal changes in snow on river runoff in the Pskem and Uradarya Basins over the past 22 hydrological years (2001–2023), considering air temperature and precipitation variations. In-situ data from meteorological and hydrological stations in the study areas were used to examine the trend dynamics during the study period. Pearson’s correlation analysis was applied to examine the statistical relationship between the winter snow accumulation and summer discharge. The statistical significance of the trend dynamics was tested using the Mann-Kendall test. The results revealed a significant reduction in snow accumulation over the study period and the number of overall snowfall days, with a marked decline observed in recent years. Consequently, the contribution of seasonal snow to river runoff has markedly decreased, leading to a reduction in the discharge volume during the vegetation period. A reduction in discharge volume in both basins over the past 22-year period (2001–2023) was statistically confirmed based on in-situ data.

Drought events in arid regions can be categorized as traditional, characterized by slow development and long duration, and flash, marked by rapid intensification over short timescales. This study assesses the spatial and temporal characteristics of both drought types in Uzbekistan’s Kashkadarya region from 1990 to 2019 using root-zone soil moisture (SM) data from the SiBUC land surface model, compared to ERA5-Land and CCISM datasets. SM anomalies were identified using percentile thresholds, alongside temperature and precipitation anomalies. Flash droughts typically developed in spring under combined short-term precipitation deficits and elevated temperatures, rapidly depleting SM during critical vegetation growth. Traditional droughts evolve more gradually, due to sustained precipitation deficits, with longer duration and greater spatial extent. Traditional droughts showed more severe SM anomalies often below –0.6, while flash droughts were more localized, with anomalies between –0.5 and –0.1. Model comparisons against CCISM indicate that ERA5-Land strongly overestimates wet-season SM, whereas SiBUC shows a weaker positive bias. Nevertheless, all datasets confirm intensifying drought frequency and severity in recent decades, particularly in dryland and pasture zones. These findings support importance of high-resolution SM data for improving drought monitoring, early warning systems, and targeted adaptation strategies in complex land use and water-stressed arid agricultural regions.

A hydrogeochemical investigation was conducted at Takezaki Spring, located in the Nango Valley within the Aso Caldera – one of the largest caldera volcanoes in the world – where the spring is thought to be formed by the mixing of multiple groundwater flow systems. The study was conducted from October 2024 (wet season) to June 2025 (dry season). Seasonal variations in dissolved ion concentrations and stable isotope ratios of oxygen and hydrogen revealed that, toward the dry season, the contribution of groundwater from the central cone flow system – characterized by a larger and more stable flow – became increasingly dominant in the spring discharge. In addition, a marked rise in the oxygen isotope ratio and an increase in discharge was observed after mid-April, when paddy field irrigation commenced upstream. These observations suggest that irrigation water may have significantly contributed to spring discharge even during the dry season. Analysis using multiple hydrological tracers clarified the seasonal and temporal variability of end-members from the wet to the dry season.

This study examined the usability of agricultural reservoirs as water resources and the sustainability of their maintenance and management across Japan, with a focus on the characteristics of municipalities. Using a nationwide database of agricultural reservoirs, we developed metrics to quantitatively assess the water availability and maintenance difficulty of agricultural reservoirs in each municipality, based on precipitation and population data. These metrics were applied in regional analyses, revealing spatial patterns that reflect both demographic and climatic conditions. Future values of the metrics were also estimated using climate model-based precipitation projections and population forecasts derived from the cohort component method. The results emphasized that the potential and challenges for utilizing and managing agricultural reservoirs vary among municipalities. The analysis further identified scattered municipalities across Japan – including not only western Japan, which has traditionally been recognized as having a large number of agricultural reservoirs, but also other regions – where increased attention to reservoir management and utilization will be required in the future.

Earthquake-induced fissures alter near-surface hydrological processes; however, their effects on vertical water movement in layered volcanic soils remain poorly understood. This study aims to elucidate how such fissures affect infiltration dynamics in stratified volcanic soil profiles. A one-dimensional Richards’ equation model (HYDRUS-1D) was applied to simulate water flow in three soil profiles: a reference profile (R1) representing pre-earthquake conditions, and two fissure-containing profiles (F1 and F2) differing in fissure depth and exposed layers. Model parameters were inversely estimated from field-observed pressure head and volumetric water content in R1, providing a reliable baseline for simulations in F1 and F2. The simulations revealed that fissures allowed water to bypass surface horizons and enhance vertical infiltration. Variations in fissure depths and exposed subsurface layers primarily influenced simulated drying patterns. In particular, simulations showed stronger drying beneath the deeper fissure (F2) due to evaporative loss, whereas field observations indicated slower recession, likely reflecting lateral or vertical drainage constrained by structural transitions in the layered soil. These findings underscore the need for future modeling to incorporate and parameterize water retention and flow behavior under abrupt stratigraphic and structural changes to better capture subsurface water dynamics beneath fissures.

The purpose of this study is to propose a digital twin approach that supports evacuation actions in conjunction with flood analysis. In order for local residents to properly understand evacuation sites and evacuation routes and to take action before floods occur, it is necessary to provide content that enables them to experience the situation near and behind river channels, and to provide content that shows the hydraulic phenomena leading to floods in chronological order. In this study, we used 3D point cloud data obtained by laser survey of the Chiribetsu River in Muroran City, Hokkaido, Japan, to represent river channel topography, Fundamental Geospatial Data for the hinterland, and PLATEAU for buildings, and calculated hydraulic phenomena such as river water level and flood depth to create a time series of flood damage using iRIC Nays2D Flood. This paper proposes the hazard map image reproduced on the digital twin by adding evacuation place and route. We believe that this kind of content, which provides a realistic picture of the actual situation of flood disasters, will help local residents take appropriate evacuation actions.

Crowdsourced observations are incorporated as one of the predictors for statistically postprocessing precipitation nowcasts by a numerical weather prediction (NWP) model using machine learning (ML) in Norway. Specifically, we use neural networks and random forests to predict precipitation with a one-hour lead time. Our study indicates that ML based statistical postprocessing combined with crowdsourced observations can improve the NWP precipitation nowcasts in terms of both reducing errors and decreasing uncertainties of the nowcasting. The most important predictor in our study is the accumulated precipitation one hour before the forecasting time by crowdsourced observations from a dense network of Netatmo personal weather stations (PWS). The quality of the crowdsourced observations of precipitation can influence the predictive skills of the postprocessing. In addition, the skewness of precipitation data due to frequent zeros is a limiting factor of the predictive skills. Overall, our results support that one single ML model trained on sample data from a large geographic region can be generalized to other areas in the region, provided that the areas are covered by crowdsourced observations of high quality.