Non-parametric multidimensional uncertainty estimation employing a hybrid approach

Understanding the response of nitrogen fluxes to changes in land use and agricultural practices is crucial for improving the instream water quality prediction. In central Germany, the expansion of bioenergy crops during the last decade led to an increase in fertiliser application rates. The purpose of this study is to investigate the effect of agricultural management changes on the stream nitrogen load of a drinking water reservoir catchment (Weida, 99.5 km2) using a hydrological water quality model. The semi-distributed hydrological water quality model—the HYdrological Predictions for the Environment (HYPE)—was calibrated and validated successfully for discharge and nitrate-N concentrations during the period 1997–2003 (the lowest discharge Nash-Sutcliffe efficiency (NSE) was 0.78). Subsequently, stream nitrogen load of six different land use scenarios and their associated agricultural practice changes were compared to the baseline simulations of the period 2006–2009. Some of these scenarios were designed considering the increased cultivation of bioenergy crops. Results revealed that an increase in mineral fertiliser by 20 % for all crops augmented an increase of monthly stream nitrogen loads in the range of 2–6 % compared to the baseline simulations. Also, it was found that stream nitrogen load increased in scenarios where all or some crop areas were converted to maize and rape, which are the established bioenergy crops in Germany. The increase of nitrogen load resulting from these scenarios differed in terms of magnitude and their temporal patterns, reflecting the importance of timing, the amount of fertiliser applications, and harvesting periods. However, results showed that nitrogen load was reduced in situations when only organic farming or summer barley was used and when rape and maize cropping areas were converted to winter wheat. In this intensively used agricultural catchment, the simulated stream nitrate-N loads quickly responded to fertiliser application changes (increase/decrease). This rapid response could be explained by short residence time of the interflow and baseflow runoff components because of the hardrock geological properties of the catchment.

Spatial validation of a semi-distributed hydrological nutrient transport model

As a result of no direct measurement data, accurate stream flow prediction in ungauged basins has remained a challenge in water resource management and flood forecasting. To meet this challenge, advanced hydrological models can be integrated with remote sensing and machine learning techniques. Hydrological models are used to understand the dynamics of streamflow patterns while remotely sensed information provides real-time information on land surface conditions as well as weather parameters thus enhancing the accuracy of these models. Using classical approaches, machine learning methods enable the analysis of complex datasets to detect interrelations or patterns that may not be obvious. Therefore, an integrated approach is crucial for improving flow predictions in ungauged basins that ultimately support effective water management and flood mitigation strategies. This can become even more complicated with climate change as shown by Morocco's high sensitivity to irregular rainfall patterns brought about by climate variations. One such place is the Ouzoud watershed, which stretches 315.6 km 2 across Central High Atlas Mountains and is characterized by frequent flooding leading to socio-economic impacts. This study aims at assessing streamflow prediction methods particularly the Quantile-Quantile Plot (QPPQ) method endorsed by USGS for its applicability in predicting streamflow at ungauged locations. The study area includes both gauged and ungauged watersheds such as the Bernat River, Lakhdar River, Ghassaf River, and Tassaouet River, with Ouzoud River being the only ungauged basin. The geological setting is characterized by carbonate rocks (limestone) throughout the catchment. Mediterranean climate mountainous conditions prevail in this region that experience significant rainfall variations. Other watersheds within the survey site are those of Tassaout and Bernat which have different geological and climatic influences on streamflow patterns. The QPPQ method entails estimating the Flow Duration Curve (FDC) for the catchment of interest, identifying matching donor sites in terms of hydrological characteristics, transferring non-excessive probabilities from these donor sites and mapping these probabilities to streamflow estimates. The performance of the QPPQ method is assessed based on several statistical criteria including Correlation Coefficient (CC), Coefficient of Determination (R 2 ), Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), Nash-Sutcliffe Efficiency (NSE), and Kling-Gupta Efficiency (KGE). For instance, the CC value of 0.55 at Bernat River Sgatt station means that there exists a moderate linear relationship between predicted and observed streamflows while R 2 value at 0.36 indicates that about 36 % of the explained variation in observed flow is covered by predicted ones. Meanwhile, MAE=22.50 mirrors an average absolute prediction error while RMSE equals 55.01 in this case refers to the square root of mean squared prediction errors. Furthermore, NSE gives information concerning how far off the model's results are from the actual data with NSE=0.36 denoting a moderate fit and KGE which equals 0.34 showing good agreement in overall terms. Together, these measures expose the strengths and weaknesses of QPPQ as a streamflow prediction technique. Although the method gives some useful indications, it should be improved further by integrating it with more data sources and processes to increase the accuracy of its predictions and make them more reliable for ungauged basins. Graphical Summary of the Study Using the QPPQ Method Applied to a Gauged Basin (Bernat Basin) and an Ungauged Basin (Ouzoud Basin) • Improved QPPQ method enhances streamflow prediction in ungauged basins. • Seasonal analysis shows best QPPQ performance in winter, weakest in autumn. • Hybrid QPPQ with LSTM and RF improves prediction of extreme streamflows. • QPPQ shows strong results in Ouzoud Basin; basin-specific calibration is crucial. • Integrating remote sensing and machine learning supports water management.

The major environmental and engineering applications of the self-potential method have been investigations of subsurface water movement. Specific uses have included mapping of seepage flow associated with dams, dikes, reservoir floors, and other containment structures; studies of general groundwater movement; and delineation of flow patterns in the vicinity of landslides, wells, faults, drainage structures, shafts, tunnels, and sinkholes. The method also has been used to a lesser extent for subsurface thermal investigations such as mapping of steam flood fronts and coal mine fires, and preliminary studies indicate that it could be used to help delineate chemical concentration gradients associated with subsurface contaminants. Self-potential anomaly amplitudes generated by such environmental and engineering sources tend to be lower than those seen in mineral and geothermal exploration, and spurious anomalies associated with the presence of man-made structures at most field sites often create very high noise levels. Therefore extreme care in the acquisition and reduction of field data, and recognition of characteristic fields associated with artificial and natural noise sources, are critical to the success of environmental self-potential investigations. Interpretation of self-potential data often is qualitative, using visual correlation between observed profile or contour patterns and known or suspected seepage flow paths or other sources. The use of available geometric source models, including charged points, lines, spheres, cylinders, and sheets, can provide useful first estimates of source depth and configuration. Computer programs based on analytical modeling techniques can be used to calculate self-potentials generated by subsurface gradients of pressure, temperature, or chemical concentration. If in-situ physical property parameters such as cross-coupling coefficients and hydraulic, thermal, and electrical conductivity are known or can be reasonably estimated, these techniques can provide useful information about the nature, location, and strength of self-potential sources.

This paper applies recent results of risk bounds for time series forecasting to identify optimal complexity of conceptual models and complexity regularised streamflow predictions based on Vapnik-Chervonenkis generalization theory. Earlier reported similar study with SAC-SMA and SIXPAR conceptual models but on two large regions from CAMELS data set is extended to more basins in CONUS to demonstrate the effect of regularizing hydrological model calibration on streamflow prediction over unseen data. SAC-SMA and SIXPAR (conceptual simplification of SAC-SMA) are used as model examples. Results show that the effect of complexity regularization more visible on SIXPAR than SAC-SMA. Results further suggest that when basins itself are complex, regularizing complexity of models does not help and depends on hydrological characteristics of the basins. The benefits of complexity regularization are more evident when assessed based on variance based performance metrices such as correlation coefficient and the slope of observed vs predicted fit than bias and mean absolute error metrices. The paper therefore offers a novel, though computationally intense, method to calibrate conceptual models while controlling for their model complexity.

Added value of distribution in rainfall-runoff models for the Meuse basin

At the occasion of the 25-year anniversary of EGU Interactive Open Access Publishing, the Executive Editors of Hydrology and Earth Systems Sciences (HESS) invite the EGU community to join an interactive presentation about HESS’s past achievements and its vision for the future.HESS encourages and supports fundamental and applied research that advances the understanding of hydrological systems, their role in providing water for ecosystems and society, and the role of the water cycle in the functioning of the Earth system. A multi-disciplinary approach is encouraged that broadens the hydrological perspective and the advancement of hydrological science through integration with other cognate sciences and cross-fertilization across disciplinary boundaries.HESS was founded in 1997 and in the editorial of Volume 1, Issue 1 the rationale for hydrology as a geoscience was given: “Much has been written over the past decade about the need to establish a strong identity for hydrology as a distinct geoscience alongside the atmospheric, ocean and solid earth sciences. The aims and scope of Hydrology and Earth System Sciences (HESS) have been designed to give full expression to this goal, and have been strongly influenced by 'Opportunities in the Hydrologic Sciences' (National Academy Press,1991). The functioning of the hydrological cycle within an earth system undergoing global change is currently the focus of research by many leading scientists and it is hoped that HESS will become a major forum for the publication and discussion of such research, as well as all new findings which enhance the position of hydrology as a geoscience.” Interestingly, by now hydrology is very well established as a geoscience and the Hydrological Sciences Division is one of the largest divisions of EGU.In 2005 HESS then moved from print issues to fully web-based open access with a transparent and interactive peer-review process. From the editorial at that time (written by Huub Savenije):“The transition from the classical anonymous and hidden peer review processes to the open discussion style follows immensely successful developments in other communities (in particular high-energy physics) and has the potential to greatly improve the communication in our discipline. Another great advantage of this discussion forum is that it will make publication (and participation in the discussions during the review process) more accessible to the wider community, especially to young scientists. HESS-D will act both as a discussion forum and as a new way (i) of contributing to our science, and (ii) for any scientist to become active and be visible as a creative and responsible thinker.” This transparent and inclusive aspect of HESS remains something we value very much and constantly try to improve on.Current and prospective authors, reviewers, current editors and anyone who might be interested in exploring the possibility to join our editorial board will learn more about HESS’s editorial workflow, from preparing a strong submission with high scientific significance to navigating the interactive public discussion phase on the community platform EGUsphere.Anyone interested in HESS’s transparent, community-driven and not-for-profit publication model is welcome to join, ask questions, share ideas, and explore ways to get involved.

Long Short-Term Memory (LSTM) networks have become popular for streamflow prediction in hydrological systems due to their ability to model sequential data. However, their reliance on lumped catchment representation and climate summaries often limits their capacity to capture spatial heterogeneity in rainfall patterns and evolving rainfall trends, both of which are critical for hydrological consistency. This study explores the limitations of LSTM-based streamflow predictions by employing a distributed conceptual hydrological model, SIMHYD, coupled with Muskingum-Cunge routing, to generate synthetic datasets representing diverse hydroclimatic conditions. These datasets are designed to replicate rainfall-runoff dynamics across selected catchments from all 18 ecoregions in CAMELS-US and key Indian river basins, providing a robust testbed for evaluating model performance under controlled conditions. The pre-trained LSTM model is tested against synthetic discharge data, enabling direct comparisons to assess its ability to simulate realistic hydrological responses. Performance is evaluated using multiple metrics, including Nash-Sutcliffe Efficiency (NSE), Kling-Gupta Efficiency (KGE), Percent Bias (PBIAS), and mean peak timing errors, to identify systematic deviations. Results reveal that LSTM models struggle with spatially variable and temporally shifting rainfall patterns, leading to inaccuracies in peak flow timing, magnitude, and overall discharge volumes. These issues highlight vulnerabilities in current LSTM-based flood forecasting systems, particularly in their ability to generalize across diverse climatic conditions and regions. This study also characterizes specific failure pathways, such as underestimation of extreme events and poor temporal coherence in hydrographs, which are critical for operational forecasting. By diagnosing these limitations, the study provides a framework for integrating process-based hydrological knowledge with data-driven techniques to improve model robustness. The findings underscore the importance of using synthetic datasets and diverse diagnostic tools to evaluate and enhance the reliability of LSTM-based models, paving the way for hybrid approaches capable of addressing the complexities of real-world hydrological systems.

Reservoir inflow plays a key role in hydrological predictions such as drought and flood predictions as well as in reservoir operations, and thus precise prediction of reservoir inflow is of utmost importance. As reservoir inflow pattern predominantly depends upon the rainfall-runoff process, land use and land cover pattern and climate change within a catchment area, accurate prediction of it, is a difficult task. Koyna dam reservoir is one of the major reservoirs in India which serves the multipurpose functions including hydroelectric power generation and thus for efficient reservoir operation; accurate inflow prediction is the need of the day. Consequently, present study aims in predicting the Koyna reservoir inflow using—a conceptual model, Hydrologic Engineering Centre’s Hydrologic Modelling System (HEC-HMS; version 4.8). The Soil Conservation Service Curve Number (SCS-CN), SCS unit hydrograph, and time interval were selected for each component of the inbuilt processes of loss method, transform method, and channel routing, respectively, while developing the HEC-HMS model. To evaluate the model performance, relative error (%), root mean square error (RMSE), correlation coefficient (r), and Nash–Sutcliffe efficiency (CE-NSE) are used along with simulated and observed hydrographs and scatter plots. From all the results, it can be said that the overall performance of the HEC-HMS model is reasonable and trustworthy and can be applied in similar area of research interest.

Artificial neural network (ANN) models combined with time series decomposition are widely employed to calculate the river flows; however, the influence of the application of diverse decomposing approaches on assessing correctness is inadequately compared and examined. This study investigates the certainty of monthly streamflow by applying ANNs including feed forward back propagation neural network and radial basis function neural network (RBFNN) models integrated with discrete wavelet transform (DWT), at Jinsha River basin in the upper reaches of Yangtze River of China. The effect of the noise factor of the decomposed time series on the prediction correctness has also been argued in this paper. Data have been analyzed by comparing the simulation outputs of the models with the correlation coefficient (R) root mean square errors, mean absolute errors, mean absolute percentage error and Nash–Sutcliffe Efficiency. Results show that time series decomposition technique DWT contributes in improving the accuracy of streamflow prediction, as compared to single ANN’s. The detailed comparative analysis showed that the RBFNN integrated with DWT has better forecasting capabilities as compared to other developed models. Moreover, for high-precision streamflow prediction, the high-frequency section of the original time series is very crucial, which is understandable in flood season.

Abstract. The alteration in river flow patterns, particularly those that originate in the Himalaya, has been caused by the increased temperature and rainfall variability brought on by climate change. Due to the impending intensification of extreme climate events, as predicted by the Intergovernmental Panel on Climate Change (IPCC) in its Sixth Assessment Report, it is more essential than ever to predict changes in streamflow for future periods. Despite the fact that some research has utilised machine-learning- and deep-learning-based models to predict streamflow patterns in response to climate change, very few studies have been undertaken for a mountainous catchment, with the number of studies for the western Himalaya being minimal. This study investigates the capability of five different machine learning (ML) models and one deep learning (DL) model, namely the Gaussian linear regression model (GLM), Gaussian generalised additive model (GAM), multivariate adaptive regression splines (MARSs), artificial neural network (ANN), random forest (RF), and 1D convolutional neural network (1D-CNN), in streamflow prediction over the Sutlej River basin in the western Himalaya during the periods 2041–2070 (2050s) and 2071–2100 (2080s). Bias-corrected data downscaled at a grid resolution of 0.25∘ × 0.25∘ from six general circulation models (GCMs) of the Coupled Model Intercomparison Project Phase 6 GCM framework under two greenhouse gas (GHG) trajectories (SSP245 and SSP585) were used for this purpose. Four different rainfall scenarios (R0, R1, R2, and R3) were applied to the models trained with daily data (1979–2009) at Kasol (the outlet of the basin) in order to better understand how catchment size and the geo-hydromorphological aspects of the basin affect runoff. The predictive power of each model was assessed using six statistical measures, i.e. the coefficient of determination (R2), the ratio of the root mean square error to the standard deviation of the measured data (RSR), the mean absolute error (MAE), the Kling–Gupta efficiency (KGE), the Nash–Sutcliffe efficiency (NSE), and the percent bias (PBIAS). The RF model with rainfall scenario R3, which outperformed other models during the training (R2 = 0.90; RSR = 0.32; KGE = 0.87; NSE = 0.87; PBIAS = 0.03) and testing (R2 = 0.78; RSR = 0.47; KGE = 0.82; NSE = 0.71; PBIAS = −0.31) period, therefore was chosen to simulate streamflow in the Sutlej River in the 2050s and 2080s under the SSP245 and SSP585 scenarios. Bias correction was further applied to the projected daily streamflow in order to generate a reliable times series of the discharge. The mean ensemble of the model results shows that the mean annual streamflow of the Sutlej River is expected to rise between 2050s and 2080s by 0.79 % to 1.43 % for SSP585 and by 0.87 % to 1.10 % for SSP245. In addition, streamflow will increase during the monsoon (9.70 % to 11.41 % and 11.64 % to 12.70 %) in the 2050s and 2080s under both emission scenarios, but it will decrease during the pre-monsoon (−10.36 % to −6.12 % and −10.0 % to −9.13 %), post-monsoon (−1.23 % to −0.22 % and −5.59 % to −2.83 %), and during the winter (−21.87 % to −21.52 % and −21.87 % to −21.11 %). This variability in streamflow is highly correlated with the pattern of precipitation and temperature predicted by CMIP6 GCMs for future emission scenarios and with physical processes operating within the catchment. Predicted declines in the Sutlej River streamflow over the pre-monsoon (April to June) and winter (December to March) seasons might have a significant impact on agriculture downstream of the river, which is already having problems due to water restrictions at this time of year. The present study will therefore assist in strategy planning to ensure the sustainable use of water resources downstream by acquiring knowledge of the nature and causes of unpredictable streamflow patterns.

Flood forecasting plays a vital role in disaster management within water resources engineering. This study evaluates the effectiveness of two data-driven approaches: a Long Short-Term Memory (LSTM) recurrent neural network and a Random Forest (RF) regression model for predicting river discharge using a publicly available hydrological dataset. Historical streamflow data, including lagged flow and precipitation variables, serve as inputs to the models. Performance metrics such as root-mean-square error (RMSE), mean absolute error (MAE), Nash– Sutcliffe efficiency (NSE), and coefficient of determination (R²) are employed for model evaluation. Results indicate that the LSTM model exhibits superior predictive performance, with lower RMSE and MAE and higher NSE and R² values. These findings support the advantage of recurrent neural networks in modelling temporal hydrological patterns.

Despite its energy benefits, hydropower dam development often causes ecological damages and social disruption, including downstream livelihood impacts, and biodiversity loss. Current methods for analyzing changes in downstream inundation extent due to dam operation typically rely on historical ground or satellite observations, or on coupled hydrological‐hydrodynamic modeling. However, while the former fails to isolate hydropower impacts from climate variations, the latter suffers from extensive input data requirements and high computational burden. This study proposes a novel hybrid framework integrating satellite data‐driven Forecasting Inundation Extents using REOF (Rotated Empirical Orthogonal Function) analysis (FIER), and the process‐based Hydrological Predictions for the Environment (HYPE) model incorporating the Integrated Reservoir Operation Scheme (IROS). The framework enables the isolated assessment of long‐term hydropower impacts on downstream inundation dynamics with computational efficiency and reduced ground data requirements, making it suitable for poorly gauged regions. Applying FIER‐HYPE‐IROS to the Lower Mekong River basin (LMB), a region significantly affected by dam proliferation impacting fisheries and agriculture, we found that dam operations decreased decadal‐average wet season water levels by up to 5% and increased dry season levels by up to 11%. Wet season inundation occurrence decreased by 11 days and the inundated area by 6%, while dry season inundation occurrence extended by 6 days and the surface water area increased by 40%. Although the current framework does not explicitly assess the downstream hydrological modifications, it offers a cost‐effective alternative for evaluating upstream alterations on inundation dynamics, such as dam operations, particularly in poorly gauged regions.

In this analysis, we examine relative contributions from climate change and river discharge regulation to changes in marine conditions in the Hudson Bay Complex using a subset of five atmospheric forcing scenarios from the Coupled Model Intercomparison Project Phase 5 (CMIP5), river discharge data from the Hydrological Predictions for the Environment (HYPE) model, both naturalized (without anthropogenic intervention) and regulated (anthropogenically controlled through diversions, dams, reservoirs), and output from the Nucleus for European Modeling of the Ocean Ice-Ocean model for the 1981–2070 time frame. Investigated in particular are spatiotemporal changes in sea surface temperature, sea ice concentration and thickness, and zonal and meridional sea ice drift in response to (i) climate change through comparison of historical (1981–2010) and future (2021–2050 and 2041–2070) simulations, (ii) regulation through comparison of historical (1981–2010) naturalized and regulated simulations, and (iii) climate change and regulation combined through comparison of future (2021–2050 and 2041–2070) naturalized and regulated simulations. Also investigated is use of the diagnostic known as e-folding time spatial distribution to monitor changes in persistence in these variables in response to changing climate and regulation impacts in the Hudson Bay Complex. Results from this analysis highlight bay-wide and regional reductions in sea ice concentration and thickness in southwest and northeast Hudson Bay in response to a changing climate, and east-west asymmetry in sea ice drift response in support of past studies. Regulation is also shown to amplify or suppress the climate change signal. Specifically, regulation amplifies sea surface temperatures from April to August, suppresses sea ice loss by approximately 30% in March, contributes to enhanced sea ice drift speed by approximately 30%, and reduces meridional circulation by approximately 20% in January due to enhanced zonal drift. Results further suggest that the offshore impacts of regulation are amplified in a changing climate.

High-latitude regions are challenging to model due to their inherent data scarcity. This limitation hampers our ability to gain robust process understanding and forecast how these regions will respond to global warming and land-use changes. Additionally, these regions are undergoing rapid changes driven by melting snow and ice with far-reaching implications for downstream areas.In this study, we demonstrate the value of isotope-aided hydrological modeling in improving process understanding and model reliability. Using data from two well-instrumented high-latitude catchments—the Krycklan Catchment Study and Abisko in Sweden—we developed detailed hydrological models in HYPE (Hydrological Predictions for the Environment). We applied a multi-objective calibration approach that includes stable isotopes of water alongside traditional flow data for model calibration and validation. This approach enhances the robustness of model-internal water source partitioning and provides additional insights beyond flow-only calibration.This work is part of the Water4All project ISOSCAN, which investigates how stable isotopes of water, collected through Citizen Science initiatives, can advance hydrological modeling. By comparing flow-only calibrated models with isotope-aided multi-objective calibrated models, we evaluate the contribution of stable isotopes in improving model performance. We explore the potential of high-information-content data (such as stable isotopes of water) collected by Citizen Scientists to overcome data scarcity challenges and enhance the reliability of hydrological models in high-latitude regions.