Accurate groundwater level predictions are crucial for sustainable management. This study applies two machine learning models – Neural Hierarchical Interpolation for Time Series Forecasting (N-HiTS) and the Temporal Fusion Transformer (TFT) – to forecast seasonal groundwater levels for 5288 wells across Germany. N-HiTS outperformed TFT, with both models performing well in diverse hydrogeological settings, particularly in lowlands with distinct seasonal dynamics.

To better utilize a given set of predictions, identifying “forecasts of opportunity” is valuable as this helps anticipate when prediction skill will be higher. This study shows that when strong land–atmosphere (L–A) coupling is detected 3–4 weeks into a forecast, the surface air temperature prediction skill at this lead time increases across the Midwest and northern Great Plains. Regions experiencing strong L–A coupling exhibit warm and dry anomalies, enhancing predictions of abnormally warm events.

The review highlights the need to integrate urban land surface and hydrological models to better predict and manage compound climate events in cities. We find that inadequate representation of water surfaces, hydraulic systems and detailed building representations are key areas for improvement in future models. Coupled models show promise but face challenges at regional and neighbourhood scales. Interdisciplinary communication is crucial to enhance urban hydrometeorological simulations.

As the soil thaws in summer, the land subsides, owing to the greater volume of ice than of water. This deformation helps monitor water storage because the subsidence magnitude is proportional to water volume. In this study, the  interferometric synthetic aperture radar (InSAR) technique was used to map subsidence around Toolik Lake, Arctic Alaska. Both InSAR and field observations suggest that soil water storage ranges from 0 to 75 cm, with small errors, and that the spatial distribution of soil water correlates strongly with topography and vegetation.

This study presents a process-behavioural calibration approach for water balance models. The different calibration steps aim at calibrating different hydrological processes: evapotranspiration, the runoff partitioning into surface runoff, interflow, and groundwater recharge, as well as the groundwater behaviour. This allows for selection of a model parameterisation that correctly predicts the discharge at the catchment outlet and simultaneously correctly depicts the underlying hydrological processes.

This study evaluates two reanalysis datasets, which are critical in climate, weather research, and water resources analysis, for the Australian region in terms of simulating daily mean precipitation and six other selected precipitation extremes. While spatial patterns of mean precipitation are well reproduced, substantial biases exist in precipitation variability, trends, and extremes. Caution in applying these datasets is thus advised in terms of the latter aspects.

This research examines how future climate changes impact root zone storage, a key hydrological model parameter. Root zone storage – the soil water accessible to plants – adapts to climate but is often kept constant in models. We estimated climate-adapted storage in six Austrian Alps catchments. While storage increased, streamflow projections showed minimal change, which suggests that dynamic root zone representation is less critical in humid regions but warrants further study in arid areas.

The Aral Sea is both an example of large-scale environmental degradation caused by human activity and a message of hope through its partial restoration. Our study shows that the restored part of the Aral Sea is now healthy in terms of vertical mixing and oxygenation. However, small perturbations of water level or transparency could significantly alter the entire ecosystem. The results contribute to understanding the consequences of large-scale lake management worldwide.

Understanding and modeling flash-flood-prone areas remains challenging due to limited data and scale-relevant hydrological theory. While machine learning shows promise, its integration with process-based models is difficult. We present an approach incorporating machine learning into a high-resolution hydrological model to correct internal fluxes and transfer parameters between watersheds. Results show improved accuracy, advancing the development of learnable and interpretable process-based models.

Rating curves are used to estimate flow from recorded stage at hydrometric stations. Uncertainties in these flows are generally not estimated or reported. This paper proposed an approach to develop rating curves and to assess the corresponding uncertainties in estimated flow that can be readily applied to large-scale hydrometric networks. The mean standard deviations of estimated flows are of the order of 6.5 %. Uncertainties increase significantly for low stages and can reach values larger than 20 %.

This study uses a multi-model approach to assess future changes in river flow intermittency across France under climate change. Combining projections from the Explore2 project with historical flow observations, logistic regressions estimate the daily probability of flow intermittency (PFI) under RCP2.6, RCP4.5, and RCP8.5 scenarios. Results suggest intensifying and prolonged dry spells throughout the 21st century, with southern France more affected, while uncertainty remains higher in northern regions.

This study explores how streams react to rain and how water travels through the landscape to reach them, two processes rarely studied together. Using detailed data from two temperate areas, we show that streams respond to rain much faster than rainwater travels to them. Wetter conditions lead to stronger runoff by releasing older stored water, while heavy rainfall moves newer rainwater to streams faster. These findings offer new insights into how water moves through the environment.

Our study focuses on filling in missing precipitation data using an advanced neural network model. Traditional methods for estimating missing climate information often struggle in large regions where data are scarce. Our solution, which incorporates recent advances in machine learning, captures the intricate patterns of precipitation over time, especially during extreme weather events. Our model shows good performance in reconstructing large regions of missing rainfall radar data.

This study enhanced a popular water flow model by adding two components: one for snow melting and another for frozen ground cycles. Tested with satellite data and streamflow, the updated model improved accuracy, especially in winter. Frozen ground delays soil drainage, boosting spring runoff by 39 %–77 % and cutting evaporation by 85 %. These findings reveal that frozen ground drives seasonal water patterns. 

Accurate river flow measurements are essential for understanding river processes. This study evaluates the freely available software KLT-IV for automatic river surface velocity measurement. Analysing over 11 000 videos and comparing them with 274 traditional flow measurements, we find strong correlations (r² = 0.95–0.97) between KLT-IV and traditional methods. KLT-IV effectively estimates river flow with high accuracy, making it a valuable tool for autonomous water resource management.

This paper reformulates the model equations of the distributed Xinanjiang hydrological model in fully differential form and incorporates two-dimensional slope runoff routing methods, which are solved using appropriate numerical techniques. The main contribution is to provide an approach for distributed hydrological models that have evolved from lumped counterparts to reduce inherited numerical errors and to improve terrain representation, thereby enhancing their physical realism and simulation accuracy.

Our study examines how drought and extreme sea-level events affect saltwater intrusion in Thailand's Lower Chao Phraya River. We identified key drivers influencing this process and developed a numerical model to understand their effects and interactions. The model helps pinpoint effective ways to reduce saltwater intrusion impacts, providing practical guidance for managing freshwater resources sustainably in similar coastal areas facing changing climate conditions.

This study examines the timing and topics of newspaper coverage of droughts in England. Media attention correlated with drought-prone hydroclimatic conditions, particularly low precipitation and low groundwater levels, but also showed a seasonality bias, with more coverage in spring and summer, as exemplified by the 2022 summer drought. The findings reveal complex media dynamics in science communication, suggesting potential gaps in how droughts are framed by scientists versus the media.

Through a survey involving six water-dependent sectors, the relevance and impact of drought vulnerability factors for sectors and societies in forested cold climates were studied. Results show that the relevance and impact of vulnerability factors differ across sectors and how governance processes and policies are important for drought risk. Results offer unique insights into the dynamics of drought vulnerability that are valuable for risk assessment, drought plans, and increasing resilience.