This study evaluates hybrid hydrological models combining physics-based and data-driven components, using Information Theory to measure their relative contributions. When testing conceptual models with Long Short-Term Memory (LSTM) networks that adjust parameters over time, we found performance primarily comes from the data-driven component, with physics constraints adding minimal value. We propose a quantitative tool to analyse this behaviour and suggest a workflow for diagnosing hybrid models.

This study presents the first detailed hydrodynamic assessment of the May 2024 floods in southern Brazil. The model accurately reproduced flood levels and extent, validated with in situ and satellite observations. Results indicate limited effectiveness of proposed hydraulic interventions, highlighting constraints of structural solutions. The findings improve understanding of complex river–estuary–lagoon dynamics and support better planning for future extreme floods under climate change.

This study presents a novel approach aimed at using biogeochemical data from surface water to decompose streamwater flow into spatial and vertical contributions. A selection of tracers was used in a mixing model to estimate contributions at the outlet of a peri-urban catchment. Results provided a better understanding of hydrological processes in the catchment and demonstrated the potential of biogeochemical data to discriminate spatial contributions according to land use.

Our research presents an improved method to enhance the understanding and prediction of water flows in rivers and streams, focusing on key runoff components: surface flow, baseflow, and total runoff. Using a streamlined model, the MPS model, we analyzed over 600 catchments in China and the US, demonstrating its accuracy in capturing the spatial and temporal variability of these components. This model offers a practical tool for water resource management.

Groundwater level dynamics under freeze-thaw conditions remain unclear. We use interpretable deep learning to simulate water table changes and identify seasonal drivers in seasonally frozen regions. During freeze-thaw, changes in soil water potential cause two-way exchange between soil water and groundwater, while rainfall, runoff, and irrigation dominate in other periods. These insights inform groundwater modeling and management in cold regions.

Water balance data are affected by various errors (bias and noise). To reduce these errors, this study presents a water balance data fusion approach that combines multi-scale data (from satellites and in-situ sensors) for each water balance variable and jointly calibrates them, resulting in consistent, bias-corrected and noise-filtered, water balance estimates, along with uncertainty bands. These estimates are useful for constraining process-based models and informing water management decisions.

This study introduces a new method of detecting how water moves quickly through certain paths in soil, bypassing the usual, slower flow. By analysing natural water markers in soil samples taken at different depths, we identified unusual flow patterns. Our method is simple and non-invasive, and can be used to cover large areas. This helps us to better understand how water travels through the ground, which is important for managing water resources and protecting the environment.

Flooding is a major problem and predicting it accurately over large areas is tough. This study tested a new approach to forecast floods across a large region in the United States. By dividing the area into smaller areas to develop the prediction models and then combining, the method successfully simulated surface water extent for both high and low flow periods. The results were more accurate than existing approaches with similar methods which can improve flood forecasting for larger areas.

Climate change impacts on a hydrologic phenomenon termed the mid-summer drought (MSD), affecting water and food security of smallholder farmers in Central America, depends on how the MSD was defined. We used many climate model projections and MSD definitions to quantify the sources of uncertainty in future MSD characteristics in Nicaragua for varying levels of global warming. We found that the definition can change where the MSD occurs but it does not add substantially to the total uncertainty.

We studied an arid, endorheic basin in the Chilean Altiplano to understand how rainfall and evaporation affect groundwater and water availability. Using a rainfall-runoff model and 40 years of satellite data, we found that much of the water evaporates and less reaches the aquifers than expected. Our results challenge the idea that the basin is fully closed and suggest that current water budget estimates may need revision, an urgent task under a changing climate.

Compound flooding occurs when multiple drivers, such as heavy rain and storm surge, occur simultaneously. Comprehensive compound flood risk assessments require simulating many storm events using flood models, but such historical data are limited. To address this, we developed a statistical framework to generate large numbers of synthetic yet realistic storm events for use in compound flood modeling.

Recent progress in NN (neural network) accelerated improvements in the performance of catchment modeling. Yet flood modeling remains a very difficult task. Focusing on two headwater streams, we developed N-HiTS (Neural Hierarchical Interpolation for Time Series Forecasting) and N-BEATS (Network-Based Expansion Analysis for Interpretable Time Series Forecasting) models and benchmarked them with LSTM (long short-term memory) to predict flooding. N-HiTS and N-BEATS outperformed LSTM for flood predictions. We demonstrated how the proposed models can be augmented with an uncertainty approach to predict flooding that is interpretable without considerable loss in accuracy.

Evapotranspiration is evaporation that occurs from plants, soil, and water bodies, but determining these components is difficult. We developed a new method that uses measurements such as streamflow and rainfall to determine the portion of plant evaporation. We found differences among vegetation types and climate conditions, and showed that plant water use peaks in moderately dry regions. The results improve understanding of plant-water interactions and can help improve water and climate models.

How landscape features affect water storage and release in catchments remains poorly understood. Here we used water stable isotopes in 12 streams to assess the fraction of precipitation reaching streamflow in less than 2 weeks. More recent precipitation was found when streamflow was high and the fraction was linked to the geology (i.e. high when impermeable, low when permeable). Such information is key for better anticipating streamflow responses to a changing climate.

We used a random forest approach to produce estimates of monthly precipitation stable isotope variability from 1962–2023, at high resolution across the entire Australian continent. Comprehensive skill and sensitivity testing shows that our random forest models skilfully predict precipitation isotope values in places and times that observations are not available. We make all outputs freely available, facilitating use in fields from ecology and hydrology to archaeology and forensic science.

We applied rainfall and tracers on two 8 m wide runoff plots to better understand runoff generation processes in pre-Alpine catchments with low permeability gleysols. The results highlight the frequent occurrence of infiltration and exfiltration of water, the importance of preferential flow, and the quick runoff response (i.e., high celerity) and transport of solutes (i.e., high velocity). These results help to understand why streams in the study region respond very quickly to rainfall. 

In lowland catchments, flood severity is determined by both the amount of rain and the groundwater depth prior to the rain event. We investigated the trade-off between these two factors and how this affects peaks in the river discharge, for both the current and future climate. We found that with climate change floods will increase in winter and spring, but decrease in fall. The number and severity of floods will increase. This can help water managers to design climate robust water management.

Hydrological simulation requires parameter calibration. This study quantifies uncertainties in future runoff and drought using the Soil and Water Assessment Tool, twenty general circulation models, and shared socioeconomic pathway scenarios. Results show calibration conditions significantly impact low-flow projections. While climate models dominate uncertainty, calibration choices contribute notably, highlighting the need for robust strategies in water resource planning.

River depth is crucial in flood modelling, yet often unavailable. Estimation methods can fill this gap but have errors that can affect flood modelling outputs. Our study quantified flood-prediction uncertainty due to these errors. Between Uniform Flow and Conceptual Multivariate Regression formulas, river depths from the former have higher uncertainty than the latter. Among parameters used in these equations, river width corresponds to the largest uncertainty, followed by the flow and slope.

We propose a multi-machine learning ensemble—integrating Gradient Boosting Machine, K-Nearest Neighbors, and Extremely Randomized Trees (GBM-KNN-ERT)—to improve Topography-Based Subsurface Storm Flow (Top-SSF) parameter regionalization for flood prediction in ungauged catchments. Validated across 80 Chinese catchments, the ensemble achieved a Nash-Sutcliffe Efficiency (NSE) greater than 0.9 for 90 % of catchments, showing superior robustness to climate and donor variability.

Our study investigates global dryland dynamics and aridification under future climate scenarios. By employing the Food and Agriculture Organisation Aridity Index and an ensemble of 13 models from the 6th Coupled Model Intercomparison Project, we provide projections for dryland distribution and aridity index across three shared socio-economic pathways (2-4.5, 3-7.0, and 5-8.5) for the near-term (2030–2059) and for the long-term (2070–2099) future.

In recent years, global and regional weather datasets have emerged, but validation with real-world data is crucial, especially in diverse regions like Brazil. This study compares seven key weather variables from five datasets with measurements from 11 sites across Brazil’s main biomes. Results show varying performance across variables and timescales, with one reanalysis product outperforming others overall. Findings suggest it may be a strong choice for multi-variable studies in Brazil.

This study provides the first assessment of an European meteorological dataset (E-OBS) for hydrological applications. We compared the dataset to meteorological datasets developed at a country level, and tested how the different data influenced the simulation of streamflow with hydrological model. Our findings show that, despite some limitations, the European dataset offers a reasonable basis for hydrological modelling in most river catchments across Europe.

This study assesses how spatial patterns, temporal trends, and record length in hourly precipitation data affect annual maximum intensities estimated with stationary and non-stationary models across a climatically and topographically diverse region. Comparing five gridded datasets, we find consistent spatial patterns but notable differences in intensities, with non-stationary estimates generally slightly lower.

The availability of high-resolution rainfall data is limited. This study presents a method to generate hourly and 1 km gridded rainfall data for detailed hydrodynamic flood modelling purposes, using point measurements and thin-plate spline interpolation. The analysis shows that the proposed dataset outperforms other gridded datasets in representing spatial distributions and daily and hourly variations of rainfall. The data is suitable for any study where high-resolution rainfall data is needed.

Global reservoir construction has improved water supply, but its impact on evaporation losses remains uncertain. Our study examines how reservoir development affects water evaporation in the Loess Plateau of China. We find that evaporation loss has increased due to expanding surface water, particularly from thousands of small- and medium-sized reservoirs. This new finding underscores the need to consider previously overlooked evaporation in the hydrological cycle and future water management.

The summer of 2022 was extremely warm and dry in Europe, severely impacting water availability. We calculated water balance anomalies for 88 glacierized catchments in Switzerland, showing that glaciers played a crucial role in alleviating the drought situation by melting at record rates, partially compensating for the lack of rain and snowmelt. By comparing 2022 with past extreme years, we show that while glacier meltwater remains essential during droughts, its contribution is declining.

Hydrological models are useful for assessing the impact of landscape organization for effective mitigation strategies. However, using these models requires reducing uncertainties in their results, which can be achieved through model–data fusion. We integrate satellite surface moisture images into a water and pesticide transfer model. We compare three methods, studying their performance and exploring various scenarios. This study helps improve decision support in water quality management.