ABSTRACT Reliable hydrological drought monitoring in data-scarce regions such as India is constrained by limited long-term streamflow records. Global reanalysis products like GloFAS offer a potential alternative, but their applicability remains insufficiently evaluated. This study presents the first nationwide assessment of GloFAS for streamflow representation and hydrological drought monitoring using discharge from 109 gauging stations (>25 years). Model performance was evaluated across daily to annual timescales, monsoon and non-monsoon periods, and different flow conditions, and SDI (1, 3, 6, 12 months) was benchmarked against GloFAS. Performance improved with aggregation: correlations increased from ρ = 0.50 (daily) to 0.78 (annual), and from ρ = 0.43 (SDI-1) to 0.58 (SDI-12). GloFAS reproduced medium-to-high flows well but underestimated low flows, limiting drought onset detection. At SDI-12, detection skill was moderate (POD = 0.73, CSI = 0.60). GloFAS can reliably detect drought occurrence but is less effective in distinguishing drought severity, with potential for application in ungauged basins.

Tidal-dominated coastal urban areas, such as the Dong Nai-Sai Gon River Basin, face escalating flood risks from heavy rainfall, reservoir releases, and tides. Accurate daily peak water level (Zmax ) forecasting is vital for flood and traffic management. Conventional harmonic analysis and standard machine learning (ML) methods often overlook future tidal data, limiting predictive accuracy. This study proposes a hybrid framework (M3) integrating 24-hour forecasted tides, derived via Fourier analysis and a multilayer perceptron, with hydrometeorological predictors using ensemble ML. Using data from four stations between 2019 and 2022, the optimal M3-cbr model achieved strong performance, yielding RMSE values from 0.0581 to 0.1103 m, KGE from 0.8499 to 0.9414, WI from 0.9509 to 0.9719, and |PBIAS| < 2.4325%. It improved accuracy by 9% to 43% over M2 and 45% to 86% over M1 during flood seasons. This cost-effective framework supports operational early warning and sustainable flood management.

ABSTRACT 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 Floods, major natural disasters, are shaped by climatic phenomena and regional dynamics. This study explores factors such as event cause, magnitude, seasonality, and the influence of climatic phenomena such as the El Niño Southern Oscillation. The study area is Rio Grande do Sul, southern Brazil, a region that suffers recurrently from floods and faced severe events in 2023 and 2024. The analysis revealed an annual number of 1.6 flood events per station. Excessive rainfall on saturated soils is the primary flood cause, accounting for 67% of events. Larger basins experienced more frequent excessive rainfall floods, while smaller ones were vulnerable to long and short rainfall events. The research revealed no seasonality. El Niño years showed more than twice the frequency of flood events than La Niña and neutral years. The findings align with trends of increasing flood events, emphasizing the need for robust flood risk management strategies in the face of climate change.

ABSTRACT The traditional stability assumption of dependence structure constructed by diverse hydro-meteorological variables has been altered, and the non-stationarity characteristic and its driving factors recognition of dependency structures are crucial for drought disaster prevention analysis. In this study, we utilized the copula-based likelihood ratio (CLR) method to reveal the non-stationarity features of the dependency structure between precipitation (P) and its influencing factors including temperature (T), relative humidity (RH) and soil moisture (SM). Also, the driving factors of dependency structures P-T, P-RH and P-SM series were illustrated using an Random Forest (RF) approach. The application results in Anhui Province, China, indicated: (1) evident non-stationarity features were identified for the evolution of the P-RH series in seven cities; (2) SM, T, SD and ENSO are the primary driving factors both before and after the variability points of P-RH series. These findings will be favourable for understanding the occurrence mechanism of non-stationarity characteristics of hydrological variables.

ABSTRACT We propose a robust and parametric–parsimonious framework to design the agricultural water balance at the planning level. The methodology is based on Melisenda’s climate index, Benfratello’s water balance and Budyko’s curve, and deploys standard crop scenarios. It allows us to estimate the annual irrigation deficit and to assess the hydrological sustainability of the agroecosystems. We tested the method on two Mediterranean semiarid case studies, the Capitanata irrigation district (Southern Italy) and the Mygdonia basin (Northern Greece). Results show how the methodology effectively depicts (i) the crop suitability to the local climatology, (ii) the noteworthy effect of field capacity on the irrigation deficit and (iii) the irrigation water footprint of each crop scenario, thus providing a guideline for stakeholders to address a hydrologically-sustainable agricultural development. The methodology moreover highlights the field capacity as a key hydrological soil–health parameter and allows us to explicitly estimate its effect on the annual water balance.

ABSTRACT Storage reservoir systems influence meteorological, hydrological, and socioeconomic droughts. However, teasing out the effects of reservoirs on drought propagation is challenging due to the integrated operation of hydropower systems. We examine how meteorological droughts propagate into hydrological and socioeconomic droughts within the Paraná River Basin (890 000 km2) in Brazil, using 25 years of daily data from 61 hydropower plants. We define socioeconomic drought using the ratio of hydropower generation to installed capacity, and use SPI and SSI for meteorological and hydrological droughts. Our results indicate that all drought types have become more frequent over the years. On average, meteorological droughts take 9 months to propagate through the reservoir system and generate a socioeconomic drought. The slow propagation is contrasted by a faster recovery of socioeconomic droughts, which can be attributed to the operation of the entire system. These findings highlight the importance of considering reservoir cascades in predicting drought developments.

ABSTRACT The connectivity of impervious areas to drainage systems, or directly connected impervious area (DCIA), is a key factor in urban runoff and water quality modelling, but is rarely measured in developing countries due to cost and complexity. This study provides new, visually observed field data on connectivity for 272 urban sites in Tshwane, South Africa, to support low-cost DCIA estimation across land use classes. Linear regression, logistic regression, and visual interpretation were used to estimate drainage connectivity. Results show that, contrary to findings from developed countries, most formal and informal residential areas exhibit 0% DCIA. In contrast, industrial and commercial sites show consistently high DCIA, aligning with international trends. These findings offer practical guidance for DCIA estimation in data-scarce urban environments: zero connectivity can be assumed for residential areas with total impervious area below 50%, with 100% DCIA for commercial/industrial sites.

ABSTRACT This study integrates the SWAT and WEAP models to address water allocation challenges in Ethiopia’s Abaya Chamo sub-basin. SWAT simulated water availability, while WEAP allocated water to the domestic, irrigation, industrial, and environmental sectors. Baseline (2020) water demands were analysed, and four scenarios were explored: reference (2021–2035), high population growth, irrigation expansion, and irrigation efficiency improvement. The SWAT model estimated the annual water potential at 4.38 × 10 9 m 3 . In 2020, total water demand was 1.22 × 10 9 m 3 , with an unmet demand of 0.17 × 10 9 m 3 . Projections to 2035 revealed persistent unmet demand across all scenarios. The irrigation expansion scenario resulted in a critical water deficit of 0.65 × 10 9 m 3 . In contrast, improving irrigation efficiency from 50% to 60% alleviated water shortages. The findings underscore the importance of developing water storage infrastructure and promoting efficient water-use practices. This study emphasizes the role of advanced modelling tools in guiding sustainable water management in the region.

ABSTRACT Accurate estimation of actual evapotranspiration (ETₐ) is essential for hydrological modelling and drought assessment. Using 270 489 daily observations from 115 FLUXNET2015 sites (2000–2014), this study benchmarks deep-learning and machine-learning model for daily ETₐ prediction. Predictors from Google Earth Engine (GEE) include radiation, temperature, humidity, precipitation, soil moisture, NDVI, and seasonal indicators. Data-driven architectures (ANN, XGBoost, ANN-ET, LSTM, TFT) are evaluated using a strict leaveone-site-out protocol to ensure full spatial independence. Temporal models markedly outperform static approaches: LSTM and TFT reach mean R2 ≈ 0.82–0.83 and RMSE ≈ 0.50–0.52 mm day—1, versus R2 ≈ 0.30–0.38 for ANN, ANN-ET, and XGBoost. Calibrated physical ET products (MOD16, GLDAS, ERA5-Land) are compared globally after linear adjustment and show lower skill. Monte Carlo perturbations yield narrow 95% prediction intervals and minimal RMSE drift for temporal models. Climate- and biome-stratified analyses show consistently stronger performance of temporal architectures.