In this session, we welcome contributions on novel frameworks for model development, evaluation and parametrization across spatio-temporal scales.
Potential contributions could (but are not limited to):
(1) introduce new global and regional data products into the modeling process;
(2) upscale experimental knowledge from smaller to larger scale for better usage in catchment models;
(3) advance seamless modeling of spatial patterns in hydrology and land models using distributed earth observations;
(4) improve model structure by representing often neglected processes in hydrological models such as human impacts, river regulations, irrigation, as well as vegetation dynamics;
(5) provide novel concepts for improving the characterization of internal and external model fluxes and their spatio-temporal dynamics;
(6) introduce new approaches for model calibration and evaluation, especially to improve process representation, and/or to improve model predictions under changing conditions;
(7) develop novel approaches and performance metrics for evaluating and constraining models in space and time
This session is organized as part of the grass-root modelling initiative on "Improving the Theoretical Underpinnings of Hydrologic Models" launched in 2016.
River discharge directly affects the water-food-energy-environment nexus and can have devastating impacts during floods. Floods often occur after extreme precipitation events, which are challenging to forecast accurately, both in time and space. Unresolved small-scale processes and features, including convection and orography, have direct impacts on our ability to accurately simulate precipitation, its partitioning into surface and sub-surface runoff, and consequently hydrological forecast skill. This motivates a spatial resolution increase in Numerical Weather Prediction (NWP) models, including their land and river components, and the revision of parametrizations suitable for those small-scale processes, such as runoff generation over regions with complex orography.
The Destination Earth programme of the European Commission addresses these issues through the global Extremes Digital Twin (C-EDT): globally coupled simulations at spatial resolutions of ~4.4km. These meteorological simulations are used to force ECMWF’s Land Surface Modelling System (ecLand), the land component of the Integrated Forecasting System (IFS), which in turn generates runoff. By effectively 1-way coupling the hydrodynamic Catchment-based Macro-scale Floodplain model (CaMa-Flood) to the IFS, grid-wise generated runoff is routed as streamflow in rivers and to simulate hydrological events.
Here, we investigate the added value of high resolution for global hydrological simulations by comparing the hydrological C-EDT-CaMa-Flood simulations at ~4.4km with a similar configuration at the operational resolution ~9km. Using the same input data, higher model resolution yields a local mean orography which is more consistent with the true conditions. Such an improved mean orography leads to better representation of the terrain, which is particularly important in mountainous regions. This fosters local precipitation maxima of a higher magnitude, due to convective processes. Conversely, a higher resolution also leads to less variability in sub-grid orography, which is a determining variable when modelling the saturated fraction per grid area over which saturation excess surface runoff is generated. Consequently, a change of the variability in sub-grid orography will directly affect the partitioning of precipitation into runoff and infiltrating water, and therefore the downstream streamflow accumulation and locally also the plant water availability. The results presented here explore the impact of model resolution changes on both the precipitation and runoff generation.
Earlier results focus on several streamflow events in the European Alps and first analyses show that, despite slightly higher precipitation totals over complex orographic regions, less surface runoff is generated, and lower peak streamflow values are predicted. As differences between initial soil moisture conditions between the two resolutions were found to be marginal for those events, the reduction in the sub-grid orography is the remaining factor leading to lower surface runoff generation at the higher resolution. Those results suggest that scale-dependent parameterisations for the runoff-generating processes are needed to minimise uncertainty in the streamflow predictions at varying scales.
How to cite: Denissen, J., Arduini, G., Zsoter, E., Wortmann, M., Tanguy, M., Gascón, E., Mazzetti, C., Prudhomme, C., Morrison, O., Düben, P., Sandu, I., Rüdiger, C., and Vannière, B.: The role of sub-grid orography for global high-resolution flood forecasting, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7097, https://doi.org/10.5194/egusphere-egu25-7097, 2025.
HydroBlocks provides a generalized framework for representing subgrid heterogeneity in land surface models. High-resolution environmental data and a hierarchical multivariate clustering scheme are employed to aggregate field-scale grid cells with similar characteristics into coherent Hydrologic Response Units (HRUs), over which the model physics is then simulated individually. This customizable clustering process enables HydroBlocks to approximate fully distributed simulations while maintaining computational efficiency. By preserving the geographic locations of tiles, the model allows for spatially explicit interactions among HRUs. Currently, HRU connectivity has been explored only via subsurface flow, where interactions occur within adjacent height bands. However, connectivity in the lower atmospheric boundary layer is not yet accounted for, despite its critical role in processes like the heat and moisture advection, wildfire spread, and the transport of snow, pollen and dust.
To bridge this gap, we propose three possible approaches for calculating horizontal fluxes between HRUs: the Eulerian Connectivity Matrix (ECM), the Lagrangian Connectivity Matrix (LCM), and the Lagrangian Particle Tracking (LPT). While LPT offers the highest accuracy consistently, it is computationally demanding. In contrast, the performance and computational cost of ECM and LCM are highly dependent on HRU configurations and wind field characteristics. These sensitivities can be brought under control by adjusting the number of HRUs connected, calling for a formal process to determine the optimal parameters for each scheme. To this end, a comprehensive evaluation of ECM and LCM under various HRU configurations and wind conditions is conducted, using LPT as a benchmark. A decision-support model is built accordingly to guide the selection of three approaches and determine appropriate parameter ranges. Incorporating a subgrid horizontal tracer transport scheme into HydroBlocks offers an effective pathway to enhance the representation of spatio-temporal dynamics in land surface modeling.
How to cite: Cai, J., Zorzetto, E., and Chaney, N.: Integrating a subgrid horizontal tracer transport module in HydroBlocks land surface model, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13386, https://doi.org/10.5194/egusphere-egu25-13386, 2025.
Accurate flood forecasting is critical for mitigating risks and safeguarding infrastructure and communities. Usually, we perform these forecasts by using distributed hydrological models, often calibrated at gauged watersheds and extrapolated to ungauged regions. However, depending on the model discretization, forecasts may suffer from significant performance degradation due to inadequate spatial discretization scales (DS), particularly in representing river networks. This study investigates the effects of discretization on watershed geomorphology and hydrological simulations using the Hillslope Link Model (HLM) applied to the Smooky Hills watershed in Kansas, U.S. We analyzed six DS ranging between 0.1 (benchmark DS-BDS and closer to observable network) and 70km2 (closer to USGS HUCs 12 and Hydrosheds). We assessed changes in geomorphological features such as width functions, saturated hydraulic conductivity, slope distributions, and hillslope travel times. We forced a constant runoff HLM formulation with 100 uniform rainfall patterns obtained from MRMS observations for the hydrological simulations. We compared our results at 400 control points distributed along the river network, covering scales between 1 and 50,000 km2 (outlet). The simulations include a version in which all the DS share the same routing parameters and another in which we select the routing parameters with the best performance at each control point and event. Our results show a loss of geomorphological and topological information as we use coarser DSs.Features such as the estimated hillslope travel times and the width function exhibit significant changes compared to the 0.1 km2 DS. Hydrological simulations revealed that coarser DSs result in decreased peak flows and delayed times to peak, highlighting the sensitivity of model performance to spatial scale. Parameter calibration further demonstrated that optimal model parameters vary across discretization scales and locations within the watershed, underscoring the limitations of universal calibration strategies. Our results also suggest a connection between the loss of geomorphological features and the simulations that could eventually be used to explain and overcome the limitations due to inadequate DSs. However, we are still unable to generalize this connection. These findings emphasize the importance of scale-sensitive modeling approaches and caution against the indiscriminate application of coarse discretization in flood forecasting for ungauged basins. By addressing the impacts of spatial resolution on predictive accuracy, this work contributes to advancing process representation in hydrological modeling.
How to cite: Velasquez, N., Rendon, S., and Krajewski, W.: Bridging the Gap: How Watershed Discretization Scale Affects Flood Forecasting Accuracy, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-20593, https://doi.org/10.5194/egusphere-egu25-20593, 2025.
Spatial scaling and the reconditioning of Digital Elevation Models (DEMs) are fundamental to hydrological modeling, as they directly affect the accuracy of geomorphological parameters and runoff simulation results. This study analyzes two basins, the Po River (Europe) and the Tugela River (Africa), using base DEMs with a resolution of 30 m, scaled to resolutions of 200, 500, 1000, 2500, and 5000 m. The DEMs were reconditioned using the AGREE method (both locally and globally) to evaluate variations in parameters such as flow direction, flow accumulation, slope, and hillslope and river network flow velocity. These variations were analyzed in the hydrological modeling of a precipitation event using TETIS v9.1 software, under the assumption of impermeable soil and reproducing the unit hydrograph principle. In addition, an exploratory analysis of the fractal dimension (FD) was conducted. To identify patterns in the scalar interactions of the results, primarily using the Box Counting methodology. Recognizing the notion that fractal dimension can be mathematically interpreted as regressions of data sets, FD was estimated by clustering sets of 2, 3, 4, 5, and 6 data points for each of the study scenarios, focusing on variables such as the total basin area, the area covered by the drainage network, the network length, and the drainage density.
The results indicate that the total area of the basins increases with scaling: from 28,955 km² to 31,225 km² for Tugela, and from 67,021 km² to 95,925 km² for Po. Flow direction alterations were observed at intermediate scales (1000 and 2500 m) reaching up to 60%, while the percentage of unaltered flow velocity decreased to 0% for scales between 500 and 1000 m . The slope exhibited a substantial decrease, from mean values of 1.84 and 3.16 to 0.07 and 0.11 for Tugela and Po, respectively. In the modeling, scale variations could amplify simulated peak flows by up to 10% or reduce them by up to 26%, while simulated peak times could be delayed by up to 12% or advanced by up to 20%. Regarding FD, it was observed that the variable "area covered by the drainage network" exhibited a tendency to converge at a value of 1.0 when the dataset corresponded to two fine scales. Conversely, when the dataset corresponded to two coarse scales, the results exhibited a tendency to approach a value of 2.0. Finally, the analysis indicated that clustering 6 data points minimizes uncertainty in the regressions. For instance, the mean values for the "area covered by the drainage network" converged to 1.17, while for total area, network length, and drainage density, the mean values were 1.97, 0.14, and 0.17, respectively.
In conclusion, each spatial scale requires specific adjustments to achieve precise calibration in hydrological modeling. These adjustments are essential to ensure that the results are consistent and reliable, allowing the model to accurately reflect the actual basin conditions and flow dynamics.
How to cite: Cortes Torres, N., Salazar-Galán, S., and Francés, F.: Fractal Dimension and Multiscale Analysis in Geomorphological Parameter Assessment and Hydrological Modeling , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11047, https://doi.org/10.5194/egusphere-egu25-11047, 2025.
Agriculture has always had to face up climate variability, especially rainfall variability, as water stress can damage crops. In many regions of the world, infrastructures to store runoff and stream water such as small reservoirs are seen as a solution to secure food production. The presence of multiple reservoirs in one catchment leads to cumulative impacts which are not necessarily the sum of the individual impacts. In the literature, their impacts are generally studied through modeling. However, the large size of the studied catchments and the aggregated representation of reservoirs in models do not allow to precisely study their cumulative impacts (Habets et al.,2018, https://doi.org/10.1016/j.scitotenv.2018.06.188).
In this work, the effect of small reservoir spatial distribution on their hydrological impacts in a small catchment is investigated by a modeling approach using the distributed agro-hydrological model MHYDAS-small-reservoir (Lebon et al., 2022, https://doi.org/10.1016/j.envsoft.2022.105409). This model features physically-based modeling of surface hydrological processes coupled with a soil-crop model, a reservoir model, a groundwater model, and a decision model for agricultural practices. Space discretization is done following parcel shapes and topography. The specificity of the model is thus the spatially explicit representation of reservoirs and associated processes such as irrigation. We constructed eightteen situations with contrasted spatial distribution of reservoirs along the stream network, namely i) upstream, ii) balanced, and iii) downstream, and with varying values of reservoir densities and cumulated reservoir volume.
The Gélon catchment (20km², France), for which MHYDAS-small-reservoir had been previously tested and validated, was used as basis for the numerical experiment. We performed the simulations on 20 years at a hourly time step, with multiple repetitions for each situation. We considered a reference situation without any reservoir, and with the spatial crop distribution of the year 2015. For each situation, reservoirs were randomly positioned along the stream network, with different probability distributions for the upstream, balanced, and downstream modalities. Nearby crops were modified compared to reference and connected to the reservoirs to reach a total irrigated area of 1 km² in all tested situations. The impact of reservoirs is thus due to the infrastructure itself and the associated nearby modifications of cultivated species and practices. Impacts are quantified relatively to the reference situation, and based on stream discharges and crop yields at different time horizons.
The first analysis of the results revealed that the mean interannual outlet discharge decreased in all the simulations, with high interannual and seasonal variability. Higher reservoir number, higher total stored volume, and downstream distributions generally led to higher hydrological impacts, with interactive effects of these factors. The main driver for these impacts was found to be the water withdrawals in reservoirs, which depends on irrigation needs and water availability. The spatial distribution of reservoirs thus appears as an important factor to consider in models to evaluate their impacts.
How to cite: Lechevallier, H., Molénat, J., Dagès, C., and Burger-Leenhardt, D.: The effect of spatial distribution of small farm reservoirs on their cumulative hydrological impacts in a small agricultural catchment: a modeling exploration, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-6876, https://doi.org/10.5194/egusphere-egu25-6876, 2025.
A correct representation of reservoirs in a large-scale hydrological model is crucial for simulating reliable streamflow, particularly in strongly managed catchments. The challenge lies in developing a simple, universally applicable routine for reservoirs worldwide, regardless of their size, purpose or climate. However, a thorough analysis of reservoir routines is hindered by the limited availability of in situ observations.
To address this limitation, we created a harmonised dataset of reservoir operations —including inflow, storage, and release— using observations in the US (Steyaert et al., 2022), Mexico, Brazil and Spain. The dataset also includes meteorological time series and static attributes —such as reservoir and dam characteristics, water use, and climate indices— that can be used to train data-driven models or regionalise model parameters.
Using that dataset, we compared five reservoir routines from the literature: a simple linear reservoir, the routines implemented in the hydrological models LISFLOOD-OS (Burek et al., 2013), CaMa-Flood (Hanazaki et al., 2022), and mHM (Shrestha et al., 2024), as well as the Starfit routine (Turner et al., 2021). Our comparison analysed the ability of these routines to model both reservoir storage and release, as well as their potential for implementation in continental or global operational systems.
Our results indicate that the Hanazaki routine strikes the best balance between storage-release performance and complexity, as it minimises the number of parameters to be calibrated and has limited data requirements. Consequently, we have implemented the Hanazaki reservoir routine in the LISFLOOD-OS v5 model, which will be used in future versions of both the European Flood Awareness System (EFAS) and Global Flood Awareness System (GloFAS).
How to cite: Casado-Rodríguez, J., Disperati, J., Grimaldi, S., and Salamon, P.: Comparison of reservoir routines for large-scale hydrological models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-16223, https://doi.org/10.5194/egusphere-egu25-16223, 2025.
Modelling the streamflow of flat, and pothole-dominated prairie or arctic regions presents challenges due to the influence of variable Non-Contributing Areas (NCAs) on converting local runoff to streamflow. Various models have been developed to represent these NCAs and their impact on streamflow prediction. However, these models may not adequately capture NCAs dynamics, rely heavily on calibration, are not applicable to large-scale basins, or are not model agnostic. In response, we introduce an open-source and model-agnostic version of a revised Hysteretic Depressional Storage (HDS) model. This model incorporates an improved numerical solution that accurately captures the hysteretic relationships of prairie potholes and NCAs, and their effect on streamflow generation. The revised HDS model is implemented and tested in three hydrological models (HYPE, MESH, and SUMMA) on a prairie pothole basin in Canada. Results demonstrate enhanced simulations of streamflow responses in the tested basins. Notably, the modified models successfully replicate the known hysteretic relationships between depressional storage and contributing areas in the region. The open-source HDS implementation approach facilitates integration into hydrologic or land surface modelling systems, enabling improvements in simulating complex hydrology and streamflow patterns globally.
How to cite: Ahmed, M. I., Clark, M., Pietroniro, A., and Stadnyk, T.: From Unmodelable to Understandable: A Model Agnostic Approach in Prairie Pothole Hydrology, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11874, https://doi.org/10.5194/egusphere-egu25-11874, 2025.
Dissolved organic carbon (DOC) concentrations in forested headwater streams have shown critical upward trends in the last decades with potentially harmful environmental consequences and potential impacts on drinking water production from downstream reservoirs. Using chemical fingerprints of DOC in the riparian zone (RZ) of a temperate headwater catchment in the Harz Mountains in Germany and a hydrologic process model for the riparian corridor, we could identify dominant stream flow generation processes and DOC source zones for a representative river reach (Werner et al. 2021). The gained local process understanding was used to adapt a parsimonious, box-type, hydrology-biogeochemistry model for the entire headwater catchment to reflect the dominant runoff generation and DOC mobilization processes using a threshold-controlled, surface flux mechanism for the RZ. The model was used to simulate DOC export dynamics to a downstream drinking water reservoir. A multi-objective calibration on stream flow and instream DOC concentration (Kling-Gupta efficiencies of 0.79 and 0.73 for the hydrological and biogeochemical modules, respectively) yielded reasonable riparian zone water and DOC dynamics as well as stream DOC exports, which were in line with observations. Fast, surficial water flow components from the RZ accounted for the largest fraction of total DOC export.
Calibrating the hydrological module of the model to discharge first, followed by a consecutive calibration of the biogeochemical model to DOC flux, produced unrealistic groundwater (GW) dynamics and GW DOC concentrations, despite a reasonable match with observed discharge and stream DOC concentrations. In contrast, the multi-objective simultaneous calibration of both, the hydrologic and biogeochemical modules, yielded an internally consistent model with adequately simulated discharge and DOC at the catchment outlet. This highlights the strong coupling between catchment internal water partitioning and DOC mobilization and export, which cannot be captured, when calibrating water and solute fluxes separately.
References:
Werner, B.J., Musolff, A., Lechtenfeld, O.J., de Rooij, G.H., Oosterwoud, M.R., Fleckenstein, J.H. (2021) Small-scale topography explains patterns and dynamics of dissolved organic carbon exports from the riparian zone of a temperate, forested catchment, Hydrology and Earth System Sciences, 25, 6067–6086, https://doi.org/10.5194/hess-25-6067-2021
How to cite: Fleckenstein, J., Werner, B. J., Schauer, L. S., Musolff, A., Lechtenfeld, O., and Birkel, C.: Using chemical fingerprints and process modeling to inform a parsimonious model for DOC export from a temperate headwater catchment, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-20917, https://doi.org/10.5194/egusphere-egu25-20917, 2025.
It has now been almost five years since Grey Nearing and his colleagues published their provocative commentary “What Role Does Hydrological Science Play in the Age of Machine Learning?”. Nearing et al. reviewed experiments that use deep learning to simulate time series of streamflow, emphasizing results that show there is substantially more information in large‐domain hydrological data sets than hydrologists have been able to translate into theory or models. In their commentary, Nearing et al. encouraged the hydrology community “to focus on developing a quantitative understanding of where and when hydrological process understanding is valuable in a modeling discipline [that is] increasingly dominated by machine learning.”
This presentation will summarize advances in process-based hydrological modeling in our research group in the five years since Nearing et al. published their controversial commentary. To bridge the gap between process-based modeling and machine learning, we depart from the focus of Nearing et al. where machine learning has a central role in the modeling ecosystem – instead, we ask how machine learning can enable and accelerate the development of process-based hydrological models. We will emphasize the components of the model ecosystem where we use machine learning and artificial intelligence, and the ecosystem components where we do not. We will discuss our advances in generating ensemble spatial meteorological fields, the numerical implementation of process-based models, process-based parameter estimation, multi-model combinations, and reproducible and transparent workflows. We will demonstrate tangible progress in closing the gap between the predictive performance of (hybrid) process-based models and pure machine learning algorithms for hydrological predictions across large geographical domains. We also demonstrate prototype workflows that use artificial intelligence to support the hydrological modelling exercise from A-Z, including the configuration, running, optimisation and interpretation of complex process-based models. We consider the community value and dangers of using AI to assist in different aspects of the process of scientific discovery.
We will end the presentation by returning to the question posed by Nearing et al. – What Role Does Hydrological Science Play in the Age of Machine Learning? We will argue that the appropriate use of machine learning and artificial intelligence is beginning to enable the development of process-based models that effectively use the information in large-domain hydrological datasets, while maintaining the interpretability and transparency of physically grounded simulations. We will suggest a path forward for the discipline where machine learning and artificial intelligence are essential to develop the next generation of hydrological prediction systems.
How to cite: Clark, M. P., Thebault, C., Eythorsson, D., Vasquez, N., Knoben, W., and Wood, A.: What is the role of machine learning when we want to simulate hydrological processes?, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12547, https://doi.org/10.5194/egusphere-egu25-12547, 2025.
As the primary source of freshwater, groundwater plays a critical role in supporting agriculture, sustaining ecosystems, and enhancing climate resilience. In the context of climate change, understanding the processes that influence groundwater is essential to predict how shifting precipitation patterns, altered evaporation rates, and the increasing frequency and intensity of droughts and floods will impact aquifers. However, assessing these processes is complicated by the limited availability of direct observations.
In this work, we aim to investigate the response times of groundwater systems and identify the key processes driving their evolution over time. Understanding the critical time of groundwater is crucial for predicting how groundwater systems will react to climate change, especially in terms of recharge and discharge dynamics.
First, we start by assessing the critical times of groundwater. For this purpose, the spectro-analysis method, whose efficiency has been demonstrated by Houben et al. (2022), is extending to discharge data collected across Europe. Our data spans from 1950 to the present, allowing us to identify critical times across a wide range of scales.
Second, we assess how critical times are influenced by rheological factors, climate conditions, and soil properties. In particular, we describe the relationships between critical times and various parameters selected to characterize the rheology, climate, and soil. This information enables us to identify the aquifers that are primarily controlled by climatic conditions and that may therefore be more vulnerable to climate change in the future.
Finally, we develop a machine learning model to predict the evolution of critical times with the variations in precipitation and evapotranspiration induced by climate change. We are therefore able to bring new constrains regarding the response time of groundwater to climate change for the years to come.
Our work will provide new insights regarding the impact of climate change on groundwater.By identifying the key parameters that control the critical times of groundwater, we can pinpoint the aquifers most vulnerable to climate-related changes. This knowledge allows us to focus on regions where the development of targeted adaptation strategies will be beneficial. These plans could help mitigate the potential risks of groundwater depletion and ensure the resilience of water resources in the face of climate change
How to cite: Ibourichene, A. and Attinger, S.: Assessment of the groundwater critical times to predict the impact of climate change on aquifers, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19178, https://doi.org/10.5194/egusphere-egu25-19178, 2025.
Remote sensing data are crucial for informing earth science models with key hydrological variable, such as evapotranspiration, soil moisture, or terrestrial water storage. However, gaps in historical data, especially pre-2000, hinder long-term hydrological modelling efforts. To address this, we propose a novel method to generate synthetic data, bridging temporal gaps and extending existing datasets for water resource modelling and climate impact studies.
Unlike existing approaches that resample and interpolate historical data as cohesive wholes, the proposed method adopts a pixel-wise perspective. Each pixel’s associated climate time series are analysed independently using a k-Nearest Neighbour (kNN) algorithm paired with a process-specific similarity metric. This allows the identification of pixel-specific analogues based on climate reanalysis data. The selected pixel-wise analogues are then combined to create “compound synthetic images,” preserving spatial and temporal heterogeneity often lost when using a domain-wise approach.
To enhance variability and assess uncertainty, the proposed method integrates stochastic sampling within the analogue selection process. This generates ensembles of synthetic data, enabling quantification of pixel-level uncertainty on any given day.
The proposed approach is tested in the Volta River Basin, a West-African region with strong climate variability and affected by data scarcity. The synthetic data are applied to a spatially distributed hydrological model and evaluated based on their ability to reproduce observed streamflow patterns. Additionally, the model is calibrated separately with real and synthetic data, and the resulting evapotranspiration outputs are compared to assess their closeness.
Preliminary results show that the hydrological model performs equally well in terms of streamflow and evapotranspiration when using either real or synthetic data. This demonstrates the reliability of the synthetic data generation and its suitability for modelling unobserved processes.
How to cite: Gerber, L. and Mariéthoz, G.: Pixel-wise Synthetic Hydrological Data for Long-term Modelling: A Novel Approach for Bridging Spatiotemporal Data Gaps, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-6811, https://doi.org/10.5194/egusphere-egu25-6811, 2025.
At its core, a hydrological model is comprised of the conservation of energy and mass for a myriad of modeled processes across a suite of spatio-temporal scales. Simulations over North America with the SUMMA hydrological model show that the form of the energy equation most commonly used in land models produces both large violations in energy conservation (especially in cold regions) as well as larger numerical errors in soil temperature and soil water content than is possible with more robust solvers. These numerical issues sabotage the success of efforts to improve process-representation. We present improved energy-conserving solutions for land models, testing five approaches over North America with the SUMMA model and evaluating tradeoffs between strict energy conservation and numerical errors in the energy equation. We include approaches that do not use time integration methods with rigorous error control (as is common in hydrological models) as well as approaches that do. The mixed form of the energy equation is discretized to conserve energy to within machine precision. Alternatively, the direct solution of the energy equation (i.e., using enthalpy as a primary variable) yields the smallest numerical errors because it allows error control to be placed on the inherent state variable. In the spirit of advancing process-representation, we illustrate the importance of accurate energy balance solutions for simulations of partially frozen soils, permafrost, and glaciers. In one prominent example, we demonstrate that debris-covered glaciers have substantially dissimilar runoff contributions when evolved using different solutions to the energy equation. The capability to accurately simulate the energy balance of terrestrial systems is essential to improve the theoretical underpinnings of process-based hydrologic models.
How to cite: Van Beusekom, A., Spiteri, R., and Clark, M.: Improving the numerical solution of the energy equation in land models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10329, https://doi.org/10.5194/egusphere-egu25-10329, 2025.
When starting a new modelling task, one of the very first decisions the modeller has to make is the choice of which model(s) to use. That this selection can have a significant impact on our model results has been shown through numerous studies (e.g. Melsen et al. [2018], Mendoza et al. [2015]). That it is often based on legacy (habit, practicality, convenience) rather than adequacy (fitness for purpose) has also been recognized (Addor&Melsen [2019]). We present the results of a literature review on previous model selection practices to better understand what modellers have considered important when choosing a model for a particular purpose.
We analyze more than 250 studies discussing model selection, model intercomparison or multi-model studies with a focus on conceptual hydrologic models. We identify the criteria used to determine which models were considered “fit for purpose” and why. The analyzed studies compare between two and 7488 model structures in two to 1013 basins. We aggregate information on the evaluation criteria used for different modelling purposes and in different locations and identify common model selection strategies. We monitor the range of model performance in individual comparisons and document both, the progress made and the challenges faced during previous model comparisons.
Our analysis shows a strong dependency on aggregated statistical metrics and a tendency for simplified calibration approaches that were meant to support a broad range of evaluation practices. This often led to a lack of clear answers on which models to prefer. The reasons that were given for (not) selecting a specific model structure seem to indicate a mismatch between the perceptions of when model adequacy is reached. We therefore conclude with a critical discussion of previous model selection strategies and call for a more nuanced approach to model evaluation as well as standards for reporting modelling practices and results.
How to cite: Spieler, D. and Stadnyk, T.: Fantastic Models and How to Find Them: A Literature Review on Model Selection Practices, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13119, https://doi.org/10.5194/egusphere-egu25-13119, 2025.
This study provides a comparison of water balance components and model performance, using the LSM/HMs models: TETIS, mesoscale Hydrologic Model (mHM), PCRaster Global Water Balance (PCR-GLOBWB) and Community Land Model (CLM), between three different experiments. The first one is a calibrated model using EMO1 [1] precipitation as the meteorological forcing, without explicit irrigation representation. The second experiment is a simulation with EO irrigation [2] added to the previous precipitation as a rainfall input. The resulting discharges are adjusted during post-processing stage to account for irrigation abstraction [3] from surface waters. Finally, the calibration of the latter is based on a naturalised discharge dataset, estimated as the sum of observed flow series and irrigation water abstractions [3] at several stations on the Po river. Simulations, calibrations and comparisons are carried out at two spatial scales, 5 km and 1 km, to take into account possible scale effects on the water balance and model performance.
The results show that, using the initial experiment as a baseline, there is an increase in evapotranspiration at both scales due to the additional irrigation. However, the streamflow may fluctuate between the second and first experiments depending on the model employed, with the difference being corrected through calibration in the third experiment. In terms of performance metrics, the Kling-Gupta Efficiency (KGE) decreased, thus the third experiment was conducted to improve the metrics on both scales, besides the representation of basin fluxes and storage.
[1] Joint Research Centre, «EMO: A high-resolution multi-variable gridded meteorological data set for Europe [Data set],» European Commission, Joint Research Centre., 2020. [En línea]. Available: http://data.europa.eu/89h/0bd84be4-cec8-4180-97a6-8b3adaac4d26.
[2] J. Dari, L. Brocca, S. Modanesi, C. Massari, A. Tarpanelli, S. Barbetta, R. Quast, M. Vreugdenhil, V. Freeman, A. Barella-Ortiz, P. Quintana-Seguí, D. Bretreger y E. Volden, «Regional data sets of high-resolution (1 and 6 km) irrigation estimates from space,» Earth System Science Data,15-4, pp. 1555--1575, 2023.
[3] Autorità di Bacino del Fiume Po, «Piano del Bilancio Idrico per il Distretto del fiume Po. Bilancio idrico dell'asta del fiume Po,» 2016.
How to cite: Güiza-Villa, N., Cortes-Torres, N., Frances, F., Shrestha, P. K., Modiri, E., Rakovec, O., Droppers, B., Wanders, N., Ávila, L., and Kollet, S.: Impact of EO irrigation on LSM/HMs modelling: comparing water balance and model performance in the Po river basin. , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-14499, https://doi.org/10.5194/egusphere-egu25-14499, 2025.
Accurate representation of terrestrial Essential Climate Variables (tECVs) is crucial for practically understanding the Earth's climate system and supporting policy decisions. This study initiates benchmarking practices within the Land Surface/Hydrologic Model (LSM/HM) communities by integrating high-resolution data with hyper-resolution hydrological modelling. The European Space Agency (ESA)-funded 4DHydro project employs six advanced LSM/HMs: Community Land Model (CLM), GEOfram, mesoscale Hydrologic Model (mHM), PCRaster Global Water Balance (PCR-GLOBWB), TETIS, and wflow_sbm.
We benchmark, calibrate, and analyze scalability using consistent EMO1 precipitation forcings, focusing on 1 km spatial resolution. We introduce a novel multi-basin (MB) calibration technique based on streamflow data from the Po, Rhine, and Tugela River basins, highlighting its impact on model performance. Scalability analysis evaluates computational trade-offs and performance improvements at higher resolutions while ensuring flux matching. The study includes 34 simulations addressing water balance closure to enhance tECVs.
Key findings explore the advantages of high-resolution modelling, introducing a reference benchmark dataset of 1 km hydrological simulations, optimal gauge selection for MB calibration, and comparative performance of different LSMs and HMs in flux matching across spatial scales. These insights contribute to advancing the integration of high-resolution data with hydrological modelling, promoting consistent and accurate terrestrial ECVs at regional and continental scales.
How to cite: Modiri, E., Rakovec, O., Shrestha, P. K., Droppers, B., Avila, L., Azimi, S., Salehi, H., Cortes-Torres, N., Güiza-Villa, N., Imhoff, R., Frances, F., Kollet, S., Rigon, R., Weerts, A., García-García, A., and Samaniego, L.: Advancing Terrestrial ECVs through High-Resolution Hydrological Modeling: Insights from the 4DHydro Project, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13751, https://doi.org/10.5194/egusphere-egu25-13751, 2025.
Accurate streamflow simulations are needed to manage water resources, evaluate flooding risks, and support agriculture and industry. Traditional ensemble approaches are usually based on meteorological ensemble but rarely consider hydrological ensemble. However, hydrological forecasts based on a single model often fail to capture the dynamic nature of hydrological systems. Addressing this gap, we present a novel dynamic combination method that adaptively leverages hydrological ensemble diversity to enhance streamflow simulations.
Using the Framework for Understanding Structural Errors (FUSE), we generated 78 hydrological models applied to 559 catchments from the CAMELS dataset across the contiguous United States. Each model was calibrated to optimize both high-flow and low-flow performance, producing a hydrological ensemble of 156 members per catchment. Our dynamic combination approach can be divided in two parts: a conceptual k-nearest neighbor algorithm to identify similar historical conditions and then model predictions at the time step of interest are weighted based on their performance for the k-nearest neighbors.
Results demonstrate that this dynamic combination approach improves upon traditional static methods, particularly in representing diverse streamflow conditions. The method captures temporal variability, reduces trade-offs among objective functions, and provides a model-agnostic framework for enhanced streamflow simulations. While the approach shows promising results, it faces limitations in its reliance on hydrological ensemble and meteorological data quality. Future work could explore machine learning integration for dynamic combination and applications to real-time forecasting and ungauged catchments.
How to cite: Thébault, C., Knoben, W. J. M., Addor, N., and Clark, M. P.: Dynamic Combination of a Multi-Model Ensemble for Improved Streamflow Simulations, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-1916, https://doi.org/10.5194/egusphere-egu25-1916, 2025.
Accurately representing actual evapotranspiration (AET) is crucial for hydrological modelling, as it is a major component of the catchment water balance. However, AET is often neglected when calibrating conceptual rainfall-runoff models to reproduce observed streamflow. This oversight can lead to models with accurate streamflow predictions but flawed internal fluxes, which could become problematic under changing environmental conditions. To address this gap, we systematically evaluated 15 evapotranspiration equations by substituting them into three widely used conceptual hydrological models (GR4J, Simhyd, and VIC). These equations represent diverse process assumptions found across common conceptual rainfall-runoff models by variously converting potential evapotranspiration (PET) and soil moisture into AET. The performance of each model-equation combination was assessed using a multi-objective calibration approach, accounting for simulated streamflow and flux tower-derived AET. We applied this method across seven Australian catchments spanning diverse climatic conditions.
Our analysis reveals that the choice of evapotranspiration equation significantly influences both internal flux accuracy and streamflow predictions. The performance among the tested equations varied extensively. It was found some widely used evapotranspiration equations struggled to replicate observed AET, underscoring potential limitations in their assumptions. Conversely, the better performing equations captured observed evapotranspiration signatures and achieved higher objective function values for both AET and streamflow, suggesting they better represent underlying hydrological processes. One equation consistently performed best across model structures and catchments. This equation incorporated a non-linear relationship between soil moisture and AET, while limiting AET to below potential evapotranspiration (PET).
Our findings underscore the need to improve the realism of evapotranspiration processes in conceptual hydrological models, particularly in relation to vegetation dynamics and their interactions with soil and atmosphere. By incorporating flux tower observations into model calibration and evaluation, our study bridges the gap between experimental data and catchment-scale modelling. We recommend that similar systematic reviews be undertaken on other continents to assess global patterns and differences. Enhancing the representation of these processes could improve model reliability across temporal and spatial scales, especially under changing climatic and environmental conditions.
How to cite: Burns, G., Fowler, K., Stephens, C., and Peel, M.: Leveraging flux tower data to systematically evaluate evapotranspiration formulas in conceptual hydrological models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8096, https://doi.org/10.5194/egusphere-egu25-8096, 2025.
Hydrological models are typically calibrated against discharge data. However, the resulting parameterization does not necessarily lead to a realistic representation of other simulated variables, such as actual evapotranspiration, soil water storage, or total water storage. Since a variety of hydrological variables are now freely available globally, multivariate model calibration has become a popular method to overcome the aforementioned limitation of a discharge-based calibration. Given the improved process representation after multivariate calibration, it seems reasonable to expect that such a calibration also leads to reduced hydrograph uncertainty, associated with more constrained flux maps (i.e., combinations of streamflow generation mechanisms). However, this expectation assumes that an intersection exists within the parameter space between the separate behavioural clouds of the two or more variables considered in multivariate calibration. Here, we tested this assumption in twelve Australian catchments located in five different climate zones. We calibrated the SIMHYD model using a Monte Carlo-based approach in which an initially large sample of parameter sets was constrained using discharge only, actual evapotranspiration only, and a combination of both variables (combined into a single objective function). As could be expected, considering both variables in model calibration resulted in the best overall model performance in all catchments. However, adding actual evapotranspiration to a discharge-based calibration caused hydrograph uncertainty to increase for 11 of the 12 study sites, whereby increases tended to be larger for low flows than high flows. Similarly, flux map areas increased on average by 27% as a result of less constrained streamflow generation mechanisms under multivariate calibration relative to univariate calibration. Analysis of behavioural parameter sets suggests that these symptoms could be caused by non-overlapping behavioural parameter distributions among the different variables. By separately considering both locally observed and remote sensing-based evapotranspiration in the analysis, we could demonstrate that the source of the information did not affect our findings. This has implications both for model parameterization and model selection, emphasising that the value of non-discharge data for improving process representation through calibration is contingent on the suitability of the model structure.
How to cite: Pool, S., Fowler, K., Gardiya Weligamage, H., and Peel, M.: Multivariate calibration can increase simulated discharge uncertainty, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-1371, https://doi.org/10.5194/egusphere-egu25-1371, 2025.
Hydrological models often struggle to accurately represent subsurface processes, which are crucial for understanding groundwater dynamics and recharge, particularly in snow-dominated catchments. Traditional calibration methods, primarily focused on streamflow, can produce models that perform well for discharge prediction but inadequately capture internal hydrological processes such as groundwater recharge, baseflow, and soil moisture. This study explores how incorporating internal state variables into the calibration process can improve model realism and better reflect the complex interactions within the hydrological cycle.
Using the physically based Water Balance Simulation Model (WaSiM), we implement and compare three model configurations across 34 catchments in southern Quebec, a region characterized by diverse hydrological conditions and significant seasonal snowmelt. The first configuration (Baseline, BL) employs a conventional calibration approach, focusing on streamflow while relying on conceptual methods to simulate groundwater flow. In the second configuration (Groundwater, GW), we enable the groundwater module, which uses physically based equations to model subsurface processes. The third configuration (Groundwater with Recharge Calibration, GW-RC) further refines the model by incorporating groundwater recharge as a constraint in the calibration process.
Our results show that while the BL and GW configurations achieve high Kling-Gupta Efficiency (KGE) scores for streamflow predictions, they underperform in representing other critical hydrological processes, such as groundwater recharge and baseflow variability. The GW-RC configuration, despite a modest reduction in streamflow performance, significantly improves the representation of subsurface processes, particularly during snowmelt periods. This enhancement is achieved by including internal state variables such as recharge in the objective function during calibration. As a result, GW-RC offers a more comprehensive understanding of watershed dynamics and provides insights that are crucial for water resource management and climate adaptation strategies.
The study highlights the value of multi-variable calibration frameworks, which move beyond streamflow optimization to incorporate additional hydrological data. Such frameworks offer a more accurate depiction of watershed processes, especially in the context of climate change. The GW-RC approach demonstrates that even small adjustments in the calibration process, such as the inclusion of recharge as a constraint, can lead to substantial improvements in model realism without sacrificing overall model stability.
The results underscore the importance of developing robust hydrological models capable of simulating both surface and subsurface processes, which are essential for adapting to future hydrological shifts. This study provides a framework for improving hydrological model calibration and offers valuable contributions to the fields of water resource management and climate adaptation.
How to cite: Talbot, F., Sylvain, J.-D., Drolet, G., Poulin, A., and Arsenault, R.: Enhancing physically based and distributed hydrological model calibration through internal state variable constraints, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10, https://doi.org/10.5194/egusphere-egu25-10, 2025.
Anthropogenic water withdrawals have increased substantially due to socio-economic development, changes in consumption patterns, and population growth. Despite comprising a significant portion of available water resources and influencing water availability, anthropogenic water use is often not explicitly incorporated in hydrological models due to data limitations or restricted access. Hydrological models are often calibrated with observed discharge data which implicitly corrects for this missing process. However, parameter calibration alone is insufficient to fully address the spatiotemporal variability of anthropogenic water use and its impact on hydrological fluxes and states. In this study, we evaluated hydrological fluxes and states in the Rhine basin as part of the Horizon Europe project STARS4Water. In a previous effort, we derived high resolution irrigation maps for the Rhine basin (Purnamasari et al., 2024). These derived irrigation maps are used in wflow_sbm to assess and quantify agricultural water use in the Rhine river basin. In addition, the wflow_sbm model also accounts for other water use (e.g., domestic, industrial, livestock). We compare the hydrological fluxes and state variables of wflow_sbm with and without water use (including irrigation) against observations, such as discharge, total water storage and water table depth. Initial assessments show an improvement in model performance that is attributed to a reduction in systematic errors of the model. Analysis of hydrological flows also indicates that high flows and low flows are sensitive to assumptions regarding water withdrawals and return flows which has implications for the model predictive capacity for water management. Finally, we provide estimates of agricultural water use for the Rhine basin in comparison to other anthropogenic water use.
Purnamasari, D., Teuling, A. J., and Weerts, A. H.: Identifying irrigated areas using land surface temperature and hydrological modelling: Application to Rhine basin, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2024-1929, 2024.
How to cite: Purnamasari, D., van Verseveld, W., Buitink, J., Sperna Weiland, F., Dalmijn, B., Teuling, A., and Weerts, A.: Implications of incorporating anthropogenic water use in the hydrological model simulations of the Rhine basin, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15753, https://doi.org/10.5194/egusphere-egu25-15753, 2025.
The impacts of floods can be enhanced when they occur shortly after droughts. Although these types of events have been widely studied separately, it is yet unclear to what extent hydrological models can capture these drought-to-flood transitions and what are the most important modeling decisions to achieve accurate simulations. To address this research gap, we calibrated four conceptual bucket-style hydrological models (GR4J, GR5J, GR6J, and TUW) for 63 catchments in Chile and Switzerland. We assessed the relative importance of different methodological choices, including model structure and calibration metric (based on the Kling-Gupta efficiency - KGE), on the model's capability to capture streamflow transitions. Further, we explored the link between the detection of transitions and the representation of different processes (e.g., snow, soil moisture, and evaporation) during these events. Our results show that i) a satisfactory KGE does not guarantee a good performance in terms of detecting streamflow extremes, ii) the choice of model structure and catchment characteristics play a relatively more important role in the model’s capability to capture transitions (compared to calibration metrics), and iii) the detection of streamflow extremes and transitions primarily depends on streamflow timing rather than other hydrological signatures or variables (e.g., evapotranspiration, snow water equivalent, etc.). We conclude that the model’s capability to simulate transitions depends on how well the streamflow response to high snowmelt or precipitation rates is represented. We showed that a model that adequately simulates individual drought and flood events does not necessarily capture observed transitions. Based on our results, we hypothesize that parsimonious models such as GR4J seem to be more suitable for simulating drought-to-flood transitions. Finally, our work highlights the importance of assessing the model’s ability to detect and simulate streamflow extreme transitions, and not purely relying on the overall model performance retrieved from the calibration or verification period.
How to cite: Muñoz-Castro, E., Anderson, B. J., Astagneau, P. C., Mendoza, P. A., Swain, D. L., and Brunner, M. I.: How well can hydrological models simulate drought-to-flood transitions?, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-14340, https://doi.org/10.5194/egusphere-egu25-14340, 2025.
Climate change challenges the availability and allocation of water resources among different human uses, even in large river systems. Water managers are confronted with the risk of water scarcity and conflicts that can vary in time and space across catchments. Hydrological models incorporating different human water uses are being developed to improve our understanding of how such complex systems operate, to make projections of future water resources under climate change, and to test adaptation scenarios for water management.
Three major human uses have been integrated into the process-oriented distributed hydrological model J2000: water abstraction for drinking water (and associated water release through wastewater treatment plants), water abstraction for irrigation, and river regulation through the management of hydroelectric reservoirs. The model has been applied to the Rhône Basin (~ 100,000 km²), covering part of Switzerland and France. The parameterisation of water uses was based on existing models (econometric model for drinking water consumption, crop water demand for irrigation, reconstructed dam influence for reservoir management) and national databases for the location of abstraction points. The model was evaluated against observed discharge from 63 gauging stations throughout the catchment, observed groundwater levels from 107 piezometers and against water abstraction volumes sourced from an independent database. The sensitivity of the model to potential irrigation adaptation scenarios was also assessed.
The results show that although the model gives correct results in terms of discharge, it struggles to reproduce abstraction volumes. The parameterisation of the water use components appears to be the main problem, in particular because of the need to make simplifying assumptions for the selection of abstraction/release points. The water use model also appears to be very sensitive to the quality of the representation of natural hydrological processes, especially precipitation in mountain areas and groundwater storage. Finally, the influence of irrigation scenarios appears to be limited beyond a certain catchment size. This study shows the advantages of using several different variables for model evaluation and the interest of distributed models to analyse simulation results at appropriate spatial scales.
How to cite: Branger, F., Mimeau, L., Crochemore, L., Bonneau, J., Lévêque, B., Pellerin, N., Chaix, E., Kubina, R., Sauquet, E., Montginoul, M., and Rabotin, M.: Integrating human water uses in a regional scale distributed hydrological model : model evaluation and sensitivity to climate change adaptation scenarios, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-16477, https://doi.org/10.5194/egusphere-egu25-16477, 2025.
Reservoirs play a critical role in shaping hydrological regimes within watershed systems, presenting challenges for accurate flood modeling. While many studies have developed reservoir operation schemes to enhance downstream discharge predictions, the impact of reservoir representation on the calibration of distributed hydrological models remains unclear. Moreover, the limited availability of upstream reservoir regulation data complicates the inclusion of dams in flood modeling. This study introduced a synergistic framework designed to improve flood predictions in data-scarce, dam-regulated basins, with the Nandu River Basin in Hainan, China, as a case study. By integrating 30m FABDEM and multi-source satellite altimetry, we reconstructed daily storage variations of the Songtao Reservoir, optimizing reservoir scheme parameters, for incorporation into the DRIVE hydrological model (DRIVE-Dam). Two calibration strategies—reservoir-inclusive and reservoir-excluded—were tested using streamflow data from the basin outlet. Satellite-derived data effectively captured high-frequency reservoir water level and storage dynamics (CC=0.95), enabling long-term simulations without management information. Evaluations incorporating hydro-stations within the watershed demonstrated that the reservoir-enabled calibration produced more accurate event hydrographs, reduced peak flow errors, and yielded realistic spatial patterns of N-year flood thresholds. This strategy also lowered flood false alarm rates (FAR) from 0.42 to 0.20 and improved the critical success index (CSI) from 0.50 to 0.54. In contrast, the reservoir-excluded calibration exhibited an overly active baseflow and subdued runoff during rainfall events, and slow flood recession, leading to overestimation of minor floods. These discrepancies arose from the mismatch between the model's naturalized assumptions and its attempt to fit observed human-influenced flood pulses, resulting in a delayed response throughout the entire basin drainage network. Our findings underscore the shortcomings of traditional calibration paradigms for watershed flood estimation and highlight the strategic value of Earth observation in enhancing hydrological and flood modeling.
How to cite: Li, C., Wu, H., and Alfieri, L.: Leveraging Satellite-Derived Reservoir Data for Enhanced Hydrological Model Calibration: Towards Advanced Flood Prediction in Dam-Regulated Basins, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19614, https://doi.org/10.5194/egusphere-egu25-19614, 2025.
A comprehensive two-dimensional hydrodynamic model has been developed for the Lower Mississippi River (LMR) using Adaptive Hydraulics (AdH). Spanning nearly 1000 miles from the confluence of the Ohio River at Cairo, Illinois, to the Gulf of Mexico, this model integrates the river’s main channel, floodplain, and significant tributaries. The model covers over 22,000 square miles and is composed of approximately 1.3 million nodes, providing high resolution across a vast area. Bathymetric data for the model comes from 2023 multi-beam and single-beam surveys conducted by the US Army Corps of Engineers' Memphis, Vicksburg, and New Orleans districts. The model is validated using data from several gage locations distributed throughout the LMR, as well as continuous water surface profiles. By incorporating a system-wide approach, this model enables large-scale analysis of hydrodynamic behavior, moving beyond the more common reach-by-reach assessments. It provides a more comprehensive understanding of river flow dynamics. This model is set to support a variety of future applications, including the evaluation of batture roughness effects on flowlines and flood attenuation throughout the river. Additionally, it will serve as a foundation for the development of an operational low water model, which will enhance predictions of navigational depths across the LMR. Such capabilities are essential for improving navigation during low water events and for optimizing flood risk management strategies. This model represents a critical tool for advancing hydrodynamic modeling and river system analysis at a large, operational scale.
How to cite: Chewning, B., Jones, K., Zimmerman, J., McAlpin, T., and Hammack, A.: Development, Validation, and Application of a Large-Scale 2D Hydrodynamic Model for the Lower Mississippi River, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-14466, https://doi.org/10.5194/egusphere-egu25-14466, 2025.
Surface runoff from agricultural hillslopes is one of the most important factors controlling soil erosion, land degradation and stream water contamination. In order to improve the understanding of surface runoff, studying the connectivity of flow paths and the different properties present in them is necessary for a more complete understanding of system behaviour. This study aims to analyze the structural connectivity scale index on an agricultural hillslope based on time-lapse photography. The study is conducted on a 26.8 ha hillslope at the Hydrological Open Air Laboratory (HOAL) experimental catchment in Austria. Using digital camera observations, the temporal dynamics of connectivity are estimated from the time-lapse photography for the period of 2014-2020. In order to study the impact of the saturated areas, directly measured field data (precipitation, soil moisture, discharge), and its change, was analyzed in relation to the connectivity scale timeseries. The main driving factor for the generation of the saturated areas are the antecedent conditions of the soil. It was found that there is a significant correlation (r2: 0.83) between maximum sediment output in the stream and the detected integral connectivity scale values. Furthermore, the 5-minutes timeseries of sediment discharge and connectivity scale were compared, which resulted in a set of unimodal cross-correlations with the peak located in the 0-50 lag domain; physically, this implies a consistent 0-4 h delay in the catchment response, varying through events.
How to cite: Marjanović, D., Parajka, J., Szeles, B., Ressl, C., Strauss, P., and Blöschl, G.: Saturation area connectivity in the Hydrological Open Air Laboratory, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9991, https://doi.org/10.5194/egusphere-egu25-9991, 2025.
The rapid growth of large-scale ground-mounted photovoltaic solar panel installations, commonly known as 'solar farms,' has raised concerns about their impact on hydrologic processes and the need for appropriate management practices. Literature review shows a lack of comprehensive field and modeling research on the hydrological impacts of solar farms, and guidance for stormwater management on solar farms varies substantially across the region and US. We have conducted Modeling and field investigation on soil moisture patterns, runoff generation, and solar radiation at two solar farms in central Pennsylvania, USA that are representative of the complex terrain in the region (e.g., high or variable slopes). Soil moisture monitoring and vegetation surveying has occurred at several key locations relative to the panels. Solar radiation has been collected under the panels to understand changes in evapotranspiration. Both solar farms included engineered infiltration basins or trenches, which were instrumented with water level or soil moisture sensors, allowing us to understand the efficacy of these structural stormwater management features in managing runoff in these sites. We have also developed a modelling framework to represent the unique hydrology of solar farms. We are leveraging a freely available, new tool called OpenHydroQual, since this model allows us to represent unsaturated flow in soil. The observed soil moisture data from the solar farm has been used for calibration and validation of the model at one solar farm site. Additional scenarios are in progress to evaluate changes in runoff compared to pre-development conditions, along with selected design storm scenarios, and selected land management scenarios.
How to cite: Yavari Bajehbaj, R., McPhillips, L., Raj, C., and Massoudieh, A.: Understanding and managing impacts of solar farms on landscape hydrology: insights from field monitoring and modeling , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-701, https://doi.org/10.5194/egusphere-egu25-701, 2025.
Non-perennial streams constitute a significant portion of the global river network and play a critical role in catchment hydrology, influencing both the quantity and quality of riverine water resources. The presence of surface flow in a stream reach depends on the imbalance between water inputs from the contributing catchment and the subsurface water transport capacity of the soil prism beneath the riverbed. Since the incoming water flow fluctuates over time due to precipitation events and seasonal cycles, some portions of the stream network may periodically cease to flow, shaping the hydrological behavior of non-perennial streams. The likelihood of a stream reach to maintain surface flow over time is captured by the local persistency index, which indicates the fraction of time during which water is present at the site. Local persistency varies across the river network, reflecting the spatial variability of hydrological and morphological factors that influence the emergence of surface flow. However, this important characteristic of channel networks is often overlooked in existing hydrological models. Understanding how local persistency varies across river networks can provide valuable insights on several hydrological aspects, including: a) identifying scaling laws for estimating the prevalence of non-perennial streams at large scales, b) improving the representation network expansion/retraction and the associated patterns of local saturation in river basins, and c) providing insights on subsurface water fluxes and their spatio-temporal dynamics.
In this study, we analyze local persistency maps from 20 river networks across Europe and the US, spanning a wide range of climatic and geolithological settings and sizes (including a novel dataset for the Biois creek catchment - 20km2, Northeastern Italy). In most catchments, the proportion of non-perennial streams remains high even at larger scales (e.g., >50% in nearly all case studies, even in those with the largest contributing areas). The shape of the local persistency distribution varies depending on the underlying climate and morphometric features, revealing distinct hydrological behaviors. Right-skewed distributions, where low persistencies are more common, suggest flashy responses to rainfall events potentially driven by surface or shallow fluxes. In contrast, left-skewed distributions indicate networks with more stable networks where flows dry out only occasionally. Our analysis also reveals distinctive spatial patterns in local persistency. Abrupt changes in persistency are often associated with specific hydrogeologic features, such as localized springs, indicating shifts in surface/subsurface water fluxes usually driven by the underlying geology. In some cases, local persistency increases or decreases gradually as one moves downstream along the network, reflecting the growing of contributing area or the presence of a losing riverbed. A progressive increase in persistency often leads to an upstream expansion of the network, while reductions in persistency correspond with longitudinal disconnections, particularly in regions with local morphological changes (e.g., variations in slope or topographic curvature). By analyzing local persistency patterns within and among catchments, this study provides valuable insights on the common characteristics of non-perennial streams, their relationship with the spatio-temporal variability of water fluxes within a catchment, and their potential role in improving the reliability of hydrological models
How to cite: Durighetto, N., Barone, F., and Botter, G.: An empirical analysis of local persistency maps for identifying common wetting/drying patterns in non-perennial streams, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10141, https://doi.org/10.5194/egusphere-egu25-10141, 2025.
Catchment response time is an important parameter in hydrological models, particularly in peak flow prediction in catchment scale. Unit hydrographs are hydrological tools that represent the direct runoff response of a catchment to a unit of effective rainfall distributed evenly over a specified duration. They are used to predict the runoff from rainfall events, aiding in flood forecasting, water resource management, and design of drainage systems.
The Unit Hydrograph time to peak (Tp), as proposed in the Flood Estimation Handbook in the UK, is a time parameter that represents the response time of runoff to rainfall. In the Flood Estimation Handbook, the ReFH model employs the unit hydrograph model to transform the rainfall to the direct runoff. Notably, the time parameter Tp, representing the time from the start to the peak in the unit hydrograph model, cannot be directly observed from rainfall and runoff time series.
In previous studies, the observed values of Tp were obtained through calibration. In practical applications, two common issues are often encountered when determining Tp through calibration:
- Ambiguity in Tp Determination: Due to the sensitivity of the calibration process to input data and parameter assumptions, the Tp value obtained may not be unique, leading to potential inconsistencies.
- Overfitting to Observed Data: The calibration process may result in a Tp value that fits the observed data well but lacks generalizability, especially when applied to events or catchments with a higher baseflow index or lower direct runoff.
This study introduces a new methodology to directly calculate Tp from rainfall and runoff time series. The principle of this method is based on the definition of Tp, thereby addressing the issues associated with calculating Tp through calibration as mentioned above. In this new method, the observed rainfall and runoff intensities are treated as realizations of random variables. The variance of the timing of rainfall and the variance of the timing of runoff are recognized as measures representing the time scales of rainfall and runoff events. Using these variances, the variance of the timing of the Unit Hydrograph can be derived. Since Tp is the only parameter in the ReFH model that influences the variance of time of the UH, it becomes possible to directly calculate Tp from the computed variance of time of the UH. This eliminates the need for calibration and provides a straightforward way to estimate Tp based on the intrinsic relationship between rainfall, runoff, and the unit hydrograph.
In this research, we study the 50 large events in 431 catchments in UK from the CAMELS-GB.
The results indicate that Tp calculated using the new method is comparable to Tp obtained through calibration in catchments with a low baseflow index. However, in catchments with large baseflow index, the new method provides more reliable results. Replacing calibration with the new method to calculate Tp allows the ReFH model to be applied to a broader range of catchments.
How to cite: Yin, Y., Rosolem, R., and Woods, R.: A new method to identify time to peak in Unit Hydrographs, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10992, https://doi.org/10.5194/egusphere-egu25-10992, 2025.
Field boundaries, such as hedgerows, dry-stone walls and fences are used on agricultural fields to control livestock and to separate one parcel of land from another; but these features could also have an important role in modifying hillslope hydrology. Here, we focus on boundary features that we would describe as barrier features, which include walls and hedge banks (i.e. hedges with an underlying earthen or stone bank). These features are effectively impervious barriers to overland flow, so are likely to have a substantial impact on hydrology at the field scale. Surprisingly, mention of these features is mostly absent from the hydrological literature, particularly in relation to catchment-scale modelling, where it is common to use a topographic representation to model hydrological flow pathways, with little or no parameterisation of such man-made features.
In relevant catchments, including these features within the structure of our models could be beneficial, particularly when trying to better characterise hydrological response times. Furthermore, there are potential opportunities with regards to semi-natural flood management. Most field boundaries contain gaps in order to provide access to the land via gateways and these access points are often located at the bottom of the field, allowing runoff to continue downslope relatively unimpeded. If these gaps are removed (e.g. by moving a gateway) then the majority of runoff is likely to infiltrate or pond on the surface, resulting in a delayed response. This type of activity could have implications for flood risk management at source, low-flow hydrology and water quality.
We used the semi-distributed hydrological model, Dynamic TOPMODEL for application to the Tamar catchment in South-West England. Using a combination of land use and high-resolution topographic data, we were able to map barrier features across the catchment; it was determined that the majority of agricultural fields contained these features. To account for these features within the model structure, for hydrological response units that were completely blocked by a hydrological barrier, the overland flow velocity was reduced to ~0, resulting in infiltration and ponding. The model was then calibrated using this new model structure.
We then examined the potential impact of removing gaps in barrier features (i.e. relocating gates). Rather than removing all gaps in barrier features, we focused on removing gaps that intersected major flow pathways in order to focus on modifications that provided the greatest hydrological impact. As such, we explored the impact of removing gaps that drained flow pathways with a drainage area of 1 and 10 ha, in comparison to the current state of the Tamar catchment. For all scenarios, model results indicate that removing gaps in barrier features leads to reductions in flood peaks but also significant increases in baseflow.
How to cite: Jackson, B., Kitch, J., Robinson, M., Waly, M., Peng, Z., Panici, D., and Brazier, R.: Restoring hydrological barriers into hydrological models: enhancing realism and opportunities. , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9918, https://doi.org/10.5194/egusphere-egu25-9918, 2025.
In the field of environmental fluid mechanics, flows over vegetated canopies remain a critical research area. Vegetation covering natural watercourse beds significantly influences flow hydrodynamics, turbulence structures (Nepf & Vivoni 2000, Luhar et al. 2008), sediment transport (Morris et al. 2008), and hydraulic efficiency (Nikora 2008). Furthermore, it impacts aquatic habitats (Wilcock 1999) and water quality (Chambers & Prepas 1994). Despite extensive studies, the influence of highly flexible and elongated vegetation morphologies, such as seagrass and water ranunculus, remains poorly understood.
The primary objective of this study is to investigate the influence of highly interactive vegetation structures on velocity profiles, turbulent stress tensors, and drag, in comparison with rigid, less flexible, or less elongated canopies. A secondary objective is to propose a generalized friction law for this type of canopy.
To model this vegetation, we constructed a synthetic bed using rectangular plastic bands (1 cm wide, 29.8 cm long, 𝐸𝐼=1.5×10−6 MPa). Experiments were conducted in a tiltable flume (4 m long, 40 cm wide) at the IMFT laboratory in Toulouse. A total of 24 uniform and stationary turbulent flows were analyzed under various hydraulic regimes, alongside vegetation image analysis. Accurate flow velocity measurements were obtained for 7 uniform regimes using Particle Image Velocimetry (PIV). Spatio-temporal double-averaged decomposition of the turbulent field (Nikora 2007) was employed to estimate mean turbulent profiles, including Reynolds stress tensors, and to calculate vertical drag profiles from the momentum equation.
The vegetation bands exhibited bending near the bed (wake zone) and flapping motions further up (flapping zone), driven by turbulence. The highly elongated canopy flow demonstrated a bi-layer structure, with distinct vertical distributions of drag and turbulent stresses corresponding to each zone’s characteristic length scale. From the observed vegetation structure and drag profiles, we developed physically based drag laws for both the wake and flapping zones. By coupling these drag laws with the universal logarithmic law for flow for the outer flow, we derived a Darcy friction law for flow over dense, submerged, and highly elongated flexible canopies.
How to cite: Vallier, L., Moulin, F., and Cassan, L.: Flows Over Flexible Vegetation Canopies: Hydrodynamic Impact and Drag Modeling, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-18149, https://doi.org/10.5194/egusphere-egu25-18149, 2025.
Accurately characterizing the spatial variability of tension water storage capacity (TWC) within a catchment is challenging due to limited in-situ hydrologic data availability. Conventional conceptual rainfall-runoff models typically rely on an empirically specified TWC distribution. However, this empirical distribution lacks a physical foundation and fails to effectively redistribute critical hydrologic components, such as local capacity and contributing area, to real-world contexts. To overcome this limitation, the topographic wetness index (TWI) and its generalized form (GTWI) are introduced to bridge local topographic information with hydrologic components. Four TWC distribution curves are contrived based on the empirical parabolic distribution, empirical linear distribution, TWI, and GTWI, respectively. The effects of these alternate TWC distributions on streamflow are investigated within the framework of the HYdrologic MODel (HYMOD) across 460 Australian catchments. The results illustrate that the GTWI-based HYMOD (GTHYMOD) outperforms other models in terms of daily streamflow, with high Kling-Gupta Efficiency (KGE) attained in 74.8% of the study catchments during the validation period. The eastern coast of Australian catchments presents a superior streamflow performance compared to that in the western coast. GTHYMOD demonstrates its superiority in characterizing spatial variability, an aspect HYMOD lacked. This study has the potential to refine the empirical TWC distribution from a physical perspective and advance our comprehension of underlying hydrologic behaviors.
How to cite: Zhou, Y., Sharma, A., and Marshall, L.: Streamflow Response to Tension Water Storage Capacity Distributions in a Large Sample Perspective, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7644, https://doi.org/10.5194/egusphere-egu25-7644, 2025.
With climate change, drought events are more likely to occur in Central Europe. The German Drought Monitor is an established tool in Germany for assessing the severity of soil moisture drought events. Based on soil moisture simulations with the mesoscale hydrological model (mHM), it categorizes the probability of occurrence of the current soil dryness compared to a historical reference period. In this model, the potential access to and uptake of soil water by the roots of the vegetation is static in time and determined by the root fraction in each soil layer. Field estimates of root water uptake suggest that this behavior is unrealistic and limits the impact of the vegetation on the evapotranspiration in the model. Therefore, we compared two schemes with and without root length density constraint, resulting in a more or less dynamic representation of root water uptake in the German Drought Monitor. We found that when root length density was removed, the model yielded a more dynamic uptake pattern, which was also comparable to observations. We investigate whether and which parameters and target variables (actual evapotranspiration, soil moisture, soil moisture index indicating drought) are sensitive to this adaptation.
How to cite: Westermann, S., Boeing, F., Peter, C., Schlaak, J., Marx, A., Thober, S., and Hildebrandt, A.: Effect of dynamic representation of root water uptake in a hydrological model on the classification of drought events, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-12583, https://doi.org/10.5194/egusphere-egu25-12583, 2025.
Lumped and semi-distributed hydrological models commonly do not include representation of intercatchment groundwater flow (IGF) fluxes. However, IGF can be a significant component of a catchment’s water balance and have important water resources implications for rivers. In models that have added a water flux to represent IGF, developments have often been made with limited prior justification based on any evidence of perceived subsurface processes. Here, we follow a perceptualisation pathway, underlined by increasing levels of hydrogeological knowledge, to showcase model structure changes to the DECIPHeR hydrological model to incorporate IGF fluxes. The River Kennet, UK (a tributary of the River Thames), was selected as our test catchment. A perceptual model was first developed, utilising available national data on meteorology, hydrogeology and geology, and reviewing water balance calculations and statistics. Four model structural development scenarios were then selected to provide increasing spatial variability in IGF, whereby IGF is modelled with disconnected, semi-routed, or fully routed connectivity between sub-catchments. We demonstrate the decisions that could be taken by a modeller in the incorporation of IGF fluxes to existing models, given an increasing level of hydrogeological perceptualisation. The inclusion of this missing flux in the DECIPHeR model improves calibrations in heavily groundwater dominated sub-catchments. We also show, however, how a lack of prior hydrogeological perceptualisation could lead to a model structure selection that is at odds with the physical reality of the catchment, whereby modelled IGF losses/gains could readily be used as a proxy for other calibration issues e.g. input data errors etc. Discussion is provided on the balance between improved calibration and realism, the importance of transparent and justifiable structural decisions, and the uncertainties associated with this. The perceptual model and the modelling results also highlight the potential of seasonal variations in IGF flux, which necessitates further investigation.
How to cite: Oldham, L., Coxon, G., Howden, N., Jackson, C., Bloomfield, J., and Freer, J.: From perceptualisation to modelling: Improving the representation of spatially variable intercatchment groundwater flow in hydrological models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-3572, https://doi.org/10.5194/egusphere-egu25-3572, 2025.
Groundwater sustains human water use globally, as it provides about a quarter and half of the total water withdrawn for irrigation and domestic purposes, respectively. Intense groundwater pumping has an impact both under- and above-ground, by lowering the water table and reducing streamflow in surrounding rivers. However, groundwater abstraction is often neglected in hydrological models because of the large uncertainties involved. These modelling uncertainties arise from the lack of data to constrain natural processes (including groundwater recharge and discharge, intercatchment groundwater flow) and anthropogenic processes (abstraction rates and their spatiotemporal patterns). Therefore, there is a need to represent groundwater abstraction in hydrological models, to consider its uncertainties, and to determine the appropriate level of complexity in process representation given data availability.
This study examines the uncertainties in groundwater abstraction for streamflow predictions over a sample of catchments in France. To this end, we use a parsimonious lumped hydrological model at the daily time step (GR6J), which represents groundwater storage through an exponential store (Michel et al, 2003). Groundwater abstraction is modelled by taking water from this exponential reservoir. We account for the uncertainties in both the water withdrawal input data and the hydrological model parameters (that describe the natural processes). We adopt annual abstraction data from the French national dataset (BNPE), that we temporally disaggregate using different assumptions. Regarding the model parameters, we select an ensemble of parameter sets that produce simulations that are consistent with the observations (streamflow, groundwater levels). Our results reveal that, beyond streamflow observations, piezometric data help to reduce the uncertainty in the parameters such as the capacity of the exponential store. Overall, our study shows the importance of accounting for groundwater abstraction and its uncertainties for streamflow predictions.
Michel, C., Perrin, C. & Andréassian, V., 2003. The exponential store: a correct formulation for rainfall-runoff modelling. Hydrological Sciences Journal, 48(1): 109-124, https://dx.doi.org/10.1623/hysj.48.1.109.43484
How to cite: Sarrazin, F., de Lavenne, A., Perrin, C., and Andréassian, V.: Accounting for groundwater abstraction and its uncertainties in hydrological models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-6852, https://doi.org/10.5194/egusphere-egu25-6852, 2025.
Runoff during the snowmelt period (hereafter SMR) is an important indicator of water availability and snowmelt floods, and a vital input to the water-food-energy nexus problem in mid-to-high latitude regions. However, despite the abundance of large-scale runoff models and data products, little is known about their SMR performance. Furthermore, diagnosing the key processes that can explain the SMR differences among models remains challenging. To address this issue, this study first utilized three key indicators, i.e., total/maximum discharge in snowmelt periods (Qsum/Qmax) and centroid timing of snowmelt (CTQ), as the first-order indices to assess 15 state-of-the-art models and datasets. Further, an innovative "tree-based model complexity scoring" (TBMCS) method was proposed to score the snow accumulation and snowmelt processes of these models, aiming to quantitatively reveal the relation between model mechanism complexity and their SMR performance. Under long-term mean conditions, we found that the models' simulation of CTQ is better than that of Qsum and Qmax. Overall, the proportion of stations with a ±20% PBias in the simulated Qsum and Qmax is below 30%, while the proportion of stations with a ±5 days difference in the simulated CTQ is below 60%. Most models exhibit larger biases in high-altitude or high-latitude regions, such as the western United States, northern Europe, and the Siberian Plain. Runoff data products are almost always superior to their model counterparts, verifying the role of observation constraints in improving SMR. By using TBMCS, we further found that models with more (less) complex mechanisms often performed better (worse) on CTQ, but this does not apply to Qsum and Qmax. Models focusing more on water balance tend to perform better in simulating Qsum and Qmax. By contrast, models with better energy balance processes do not necessarily yield better water quantity simulations, but can yield better CTQ simulations. This study is the first assessment of the SMR performance of state-of-the-art runoff models and data products. It also innovatively introduces the TBMCS method to challenge the traditional paradigm of "complex models are not necessarily better than simple models", laying the foundation for identifying prioritized areas for future model development.
How to cite: Lei, X., Lin, H., and Lin, P.: Snowmelt runoff in global hydrological models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-8423, https://doi.org/10.5194/egusphere-egu25-8423, 2025.
Although regional water resources issues need to be solved using hydrological models that can accurately reproduce phenomena, difficulties exist in many regions and countries owing to the lack of quantitative observed data and computers. In this case, global metrological datasets and terrain elevations are available for analying hydrological processes in such ungauged basins. When using a regional model in hydrological analysis, the forced use of global datasets requires usually assimilation and bias correction, most often with high computational cost. Since the accuracy of global datasets has been improving in recent years, a global dataset was dared to be applied in local-scale hydrological analysis without bias correction in this study. The result is compared with that of hydrological analysis using a ground dataset.
The hydrological model consisting of the Diskin–Nazimov infiltration model and the storage–discharge relationships developed for mountainous basins (Fujimura et al., 2011) was used in this study because of its simple structure that uses small datasets that accurately estimate runoff phenomena at the local scale, which can help solve the regional water issues in water resource management or flood control design. Global Satellite Mapping of Precipitation (GSMaP) and Japanese Reanalysis for Three Quarters of a Century (JRA-3Q) are used in this study as global metrological datasets for precipitation and temperature, respectively. Multi-Error-Removed Improved-Terrain DEM (MERIT DEM) provided by Yamazaki et al. (2017) is used as the global digital elevation model. The hydrological analysis is carried out for a period of 21 years at daily time steps for four snowy mountainous basins with areas from 103 to 331 km2 in the Hokkaido region of Japan, using both the global dataset and the gauge-based dataset. Each simulation was assessed using the average daily runoff relative error (ADRE).
The results show that, when using the ground-based dataset, the ADRE range is from 26.6% to 47.2% and the average is 35.5%, and when using the global dataset it is from 44.0% to 76.7% and the average is 60.4%. The use of a global dataset reduces the accuracy of the analysis, but not considerably.
How to cite: Fujimura, K., Iseri, Y., and Yanagawa, A.: Hydrological simulation applying global meteorological datasets and terrain elevations to local-scale snowy basins in Japan, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2769, https://doi.org/10.5194/egusphere-egu25-2769, 2025.
Descriptive historical cartography of river morpho-dynamics is crucial for understanding the impacts of climate change, the evolution of river catchment land use and land cover and the degree of human influence on river systems.
Beyond studying specific river sections, analysing fluvial dynamics at large spatio-temporal scales, such as national or continental water basins (i.e., Atlantic, Mediterranean) over a 20 year-period, presents numerous limitations and challenges. These include issues in hydrological and geomorphological calculation procedures, as well as data availability.
A key metric for understanding river dynamics and their temporal evolution is the stream power, which is essentially the energy exerted by water flow on different parts of the riverbed (i.e., banks and bottom). The calculation inputs, along with water density and recorded discharge, include the slope and wetted channel width. These latest two inputs are essential due to the challenges in accurately extracting riverbed elevation data and measuring wetted channel width along the entire river length. These challenges arise from factors such as vegetation coverage masking the river and the limitations in the spatial resolution of worldwide satellite-borne remote sensing products used for historical studies (e.g., Landsat products). To address these problems, we developed an automated and cloud-based methodology, that follows the next steps:
i) To implement a cloud-computed procedure using Google Earth Engine to process historical data in large areas, such as trends in vegetation indices and frequency of floods;
ii) To treat wetted channel width as a variable subject to random errors and outliers, but model it as a function of more reliably measurable variables (e.g., surrounding vegetation, flood frequency, slope, transverse channelling and discharge return period) to estimate it in non-measurable river sections;
iii) To perform a series of filtering operations on the input data, such as the RANSAC algorithm, to minimize outliers and eliminate topological inconsistencies in height and derived slope of the riverbed;
iv) To compute error propagation in the calculation of stream power, considering the significant sources of error in the river longitudinal slope and wetted channel width variables, as well as the longitudinal variability of stream power, both as indicators of the calculation reliability.
In the final phase, we examine the correlation and potential causality between the stream flow results and socio-environmental aspects of the study areas (e.g., number of cities, population, land use) to broadly understand the patterns that may influence its spatiotemporal variability.
The proposed cloud-assisted pipeline enables the analysis of large-scale river systems, accounting for their temporal evolution and providing an initial estimate of stream power with an associated confidence index. This method advances previous global studies. It automatically generates essential data for river basin management, assesses the level of human impact on river systems, and facilitates comparisons across different hydrographic regions.
ACKNOWLEDGMENTS: This work is funded by the European Research Council (ERC) through the Horizon Europe 2021 Starting Grant program under REA grant agreement number 101039181 - SEDAHEAD.
How to cite: Vallejo, M. and Juez, C.: A remote sensing cloud-assisted pipeline to estimate river stream power at catchment scale, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-2278, https://doi.org/10.5194/egusphere-egu25-2278, 2025.
Accurate representation of hydrological processes across varying spatio-temporal scales is essential for effective flood risk assessment and water resource management. Traditional distributed hydrological models often struggle to integrate rainfall-runoff processes with floodplain inundation dynamics, limiting their ability to assess flood risk at the river basin scale. The Rainfall-Runoff-Inundation (RRI) model addresses this challenge by providing a unified framework that couples both rainfall-runoff and flood inundation in a two-dimensional context.
The original version, RRI v.1, demonstrated the potential for integrated hydrological modeling but had limitations, particularly in underestimating peak discharges in small to medium-sized basins. This issue arose from the model’s structural design, where rainfall passed through multiple slope grid cells before reaching the river channel, delaying flow accumulation, especially in steep or small catchments. Additionally, the model's uniform grid treatment for both rainfall-runoff and inundation increased computational demands for high-resolution inundation simulations. The simplified floodplain representation also failed to differentiate between left and right bank floodplains, limiting its application in complex flood inundation.
To address these issues, RRI v.2 introduces a dual-grid approach and improved hydrologic process-representations with the following key advancements.
- Dual-grid Framework: Coarse grids (e.g., 150 m) for rainfall-runoff computations and finer grids (e.g., 30 m) for floodplain inundation enhance spatial resolution where necessary while optimizing computational efficiency.
- Improved Slope Representation: River channels are included in all grid cells, and slope length is directly incorporated into runoff computations, providing more accurate flow routing and addressing peak discharge underestimation in small basins.
- Enhanced Floodplain Dynamics: Differentiating left and right bank interactions improves floodplain process representation, enhancing model reliability in complex inundation settings.
Additionally, RRI v.2 integrates observed soil characteristics into the runoff model, improving the representation of infiltration and subsurface flow processes referring to our recent work (Sugawara and Sayama, Journal of Hydrology, 2024). Using high-resolution terrain data (e.g., 10 m DEM) and reflecting localized hydrological conditions, the model captures small-scale basin dynamics with greater accuracy.
Preliminary applications to the September 2024 Noto Peninsula heavy rainfall event demonstrate the ability of RRI v.2 to simulate observed flood patterns, peak discharges, and inundation extents. The dual-grid approach not only increases computational efficiency but also ensures scalability for more complex rainfall-runoff and inundation processes. This new development provides a versatile tool for real-time flood forecasting and risk assessment under climate change.
How to cite: Sayama, T., Tanaka, T., and Sugawara, Y.: Advancing Hydrological Process Representation with the Dual-grid Rainfall-Runoff-Inundation Model, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-13699, https://doi.org/10.5194/egusphere-egu25-13699, 2025.
Hydrological models are influenced by multiple choices, which can significantly affect all phases of their application. These choices impact the calibration process by influencing the estimation of optimized parameters, the validation phase, and the model's overall performance in forecasting applications. Among these sources, input data, such as meteorological variables, play a pivotal role. While accurate collection and validation of such data are essential, they are often insufficient. For example, in the case of semi-distributed hydrological models applied to a basin divided into multiple hydrological response units (HRUs), most of them typically lack adequate instrumentation. Consequently, it becomes necessary to estimate or simulate meteorological inputs, such as precipitation and air temperature through appropriate geostatistical modeling. Beyond the choice of estimation method, a critical consideration is what constitutes a representative value for an HRU: whether it originates from a single point (e.g. the HRU centroid) or is an appropriate statistic from a grid of points. This decision has substantial implications for the model's computational time, performance, and reliability and can introduce uncertainties in the final modeled product.
This study investigates how different configurations of input data may affect model performance in the upper part of the Noce River, located in the Trento province of Italy. The analysis was conducted using the GEOframe framework, its kriging method and semi-distributed model. Four configurations were analyzed moving from the most simplified and computationally convenient (one representative point over the subbasin) towards the most complex (average of gridded values over the subbasin). The effects of the different scenarios are evaluated over several hydrological processes (river discharges, soil moisture, snow evolution), quantifying the trade-offs between computational efficiency and the accuracy of input data representation. The work offers insights into how different configurations can influence the reliability of hydrological forecasts and the uncertainties in the final results.
How to cite: Andreis, D., Formetta, G., and Rigon, R.: Impact of Different Geospatial Meteorological Input Configurations on a Semi-Distributed hydrological model output, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-10093, https://doi.org/10.5194/egusphere-egu25-10093, 2025.
The Budyko framework is a strong contender to be a useful tool for studying the impacts of climate change and human activities on the water balance components. This framework proves to be more effective and suitable for catchments with negligible storage changes, which is assumed to be true when taking long-term means. Since the storage change might not be negligible even at decadal scales, the efficacy of the framework at shorter timescales has been debated. Hence, the framework suffers from temporal scaling issues, which have been a central focus in hydrological research. Previous studies have replaced the precipitation term with effective precipitation (Precipitation – Storage change) to tackle this problem at a finer temporal scale. Here, we assess the efficacy of using effective precipitation for various climate change and human intervention scenarios. A closed-loop environment was developed using synthetic data as a business-as-usual scenario and various change scenarios were introduced by adding realistic linear trends in meteorological and hydrological datasets. We found that the effective precipitation strategy works for natural storage variations but fails when human-induced storage changes such as groundwater abstraction is considered. Therefore, we propose an improved Budyko framework that has storage change index as the third axis added to the traditional Budyko framework. We demonstrate that this novel framework acts as a robust way of partitioning precipitation at finer temporal scales with accuracy better than the existing approaches.
How to cite: Shaw, B. and Dutt Vishwakarma, B.: A new way to incorporate storage change term improves the efficacy of the Budyko framework as compared to using effective precipitation, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-15213, https://doi.org/10.5194/egusphere-egu25-15213, 2025.
Coastal regions are dynamic environments where surface and subsurface hydrological processes interact in complex ways, significantly affecting water resources and ecosystem sustainability. In regions affected by land subsidence, the surface water/groundwater interaction is further complicated by possible changes in groundwater flow paths, surface water dynamics, and so on. These changes may exacerbate flood risks and saline water inundations. In addition, human interventions, such as engineered infrastructure, is likely to intensify these challenges. Despite the progress in understanding these processes, developing models to capture the coupled dynamics of surface and subsurface flows still needs further investigations for quantitative discussion. For this purpose, we use numerical code HydroGeoSphere to develop coupled dynamics of surface and subsurface flow model in the Kujukuri Plain, the eastern coastal region of Japan. Groundwater levels and river levels show a similar trend during non-irrigation periods, primarily influenced by rainfall patterns. This similarity is likely to indicate a hydrological connection between the two systems, highlighting the dominant role of precipitation in these periods. River levels are additionally affected by tidal fluctuations, which are not observed in measured groundwater levels. This can be understood by the amplitude decay of high frequency signal in subsurface environment. The simulation results show a reasonably good agreement with observed river levels and groundwater levels. The sensitive parameters identified in the analysis are hydraulic conductivity, surface roughness coefficient, and recharge which mainly affect model results. These findings highlight the need for careful calibration of these parameters to ensure model reliability for the model results to be applied to effective water resource management strategies.
How to cite: Tsai, C. S., Liu, J., Ito, Y., and Tokunaga, T.: Understanding surface and subsurface flow dynamics in a coastal subsided region, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-14704, https://doi.org/10.5194/egusphere-egu25-14704, 2025.
Land degradation is a major concern in the Mediterranean region. In southwest Morocco, the Argan agroforestry system, part of the UNESCO Biosphere Reserve network, is the primary income source for rural communities. However, it faces growing threats from increased drought, soil erosion, and overgrazing by goats and camels. To face these challenges, combining modeling tools and new water-saving technologies is a promising approach for promoting sustainable argan production. In this study, we used the HYDRUS-2D model to assess the effectiveness of subsurface water retention technology (SWRT) in improving the survival of argan seedlings transplanted for reforestation on coarse-textured soils. A total of 460 argan tree seedlings were transplanted in the Essaouira Living Lab of the PRIMA SALAM-MED Project. Biodegradable SWRT membranes were applied to 50% of the seedlings, while the remaining 50% were left without SWRT. Ground-based data on soil properties, irrigation, climate, and soil moisture were collected from the study site and used to set up and run the model. The adopted methodology involved multisite calibration of soil water content at three depths (10 cm, 20 cm, and 40 cm) to estimate water losses across 10 sites, comprising five sites with SWRT and five sites without SWRT. The results obtained showed that the HYDRUS-2D model correctly simulated the observed soil water content in nearly all sites with and without SWRT. Furthermore, the highest reduction rates in simulated water losses were observed in soil profiles with SWRT compared to those without SWRT which exhibited higher loss rates. Overall, our findings highlight that SWRT is an effective solution for enhancing water-use efficiency and improving root zone water storage, promoting argan tree growth in the Essaouira region, particularly in soils with high infiltration capacity and permeability. Implementing SWRT can also contribute to sustainable land management practices and support local communities by fostering resilient agroforestry systems and securing their primary income source.
Keywords: Mediterranean region; Forest degradation; Argan tree; Land management; SWRT; HYDRUS-2D.
Acknowledgement & funding
This research was carried out within the SALAM-MED project funded under the Partnership for Research and Innovation in the Mediterranean Area (PRIMA) programme supported by the European Union. Grant Agreement number: [2123] [SALAM-MED] [Call 2021 Section 1 Water RIA].
The content of this abstract reflects the views only of the authors, and the PRIMA Foundation is not responsible for any use that may be made of the information it contains.
How to cite: bouizrou, I., Castelli, G., Villani, L., Fassih, B., Nait Douch, A., Ait-El-Mokhtar, M., Wahbi, S., and Bresci, E.: Modeling soil-water dynamics for sustainable Argan reforestation using Subsurface Water Retention Technology (SWRT) and HYDRUS-2D in southwest Morocco, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-9767, https://doi.org/10.5194/egusphere-egu25-9767, 2025.
Flow conditions in rivers and open channels are often incorporated into hydrogeological models with a constant horizontal grid resolution, typically without considering real river channel widths. Such grid mismatches can lead to an underestimation of flow velocity where river widths are narrower than the model’s grid size. Furthermore, the exchange between rivers and the subsurface is often too large, leading to erroneously high vertical exchange rates. To address these challenges, this study approximates subscale river channel flow using the kinematic wave equation for overland flow calculation of the ParFlow integrated hydrological model, enhanced by a scaled roughness coefficient as proposed by Schalge et al. (2019). The scaling exploits a relationship between grid cell size and river width, derived from a simplified modification of the Manning-Strickler equation. Additionally, subsurface-river exchange rates, including exfiltration and infiltration rates, are adjusted along riverbeds based on grid resolution. These adjustments correct the exchange rates even when the grid size is relatively coarse. The proposed scaling approach was implemented and validated using the ParFlow integrated hydrological model with its integrated land surface model CLM. The model setup for the test runs features a spatial resolution of 611m over Germany and surrounding regions. The reliability of the results was assessed using an innovative application of the First Order Reliability Method (FORM) that shows significantly improved streamflow predictions. A cross-validation with observations from gauging stations confirms these improvements, underscoring the effectiveness of the proposed river-width parameterization.
How to cite: Soltani, S., belleflamme, A., Hammoudeh, S., and Kollet, S.: Enhancing Streamflow Predictions with a River Parameterization in an Integrated Hydrological Model, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-11292, https://doi.org/10.5194/egusphere-egu25-11292, 2025.
Rainfall-runoff models are often calibrated by defining feasible parameter ranges and constraining them with streamflow data, and occasionally other hydrological variables. Traditionally, global prior ranges have served as a baseline, containing a wide range of parameter values suitable for various catchment types. However, there might be more information available to reduce a priori parameter uncertainty in a structured way. This study addresses this gap by defining plausible prior parameter ranges based on the distribution of identifiable parameters and their relationship to catchment characteristics. Using a version of the conceptual Probability Distributed Moisture (PDM) model, the study focuses on a large sample of catchments in the United States, covering diverse climatic, land cover, geological, and landscape types. Thus, investigating the effects of these physical and climatic properties on parameter prior ranges. The combination of automatic grouping and catchment attributes resulted in significant reductions in parameter space, with high average predictive accuracies for traditional efficiency measures. Surprisingly, we find distinct and spatially coherent regions within the US where specific prior parameter ranges maintain high levels of performance. More than 75% of the catchments show NSE values above 0.6 and KGE values above 0.7. Our results suggest that regionalizing prior parameter ranges can significantly reduce parameter uncertainty. These findings have significant implications for the prediction of hydrological responses in ungauged catchments.
How to cite: Jarajapu, D. C., Francke, T., and Wagener, T.: Regionalizing prior parameter ranges for rainfall-runoff models, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-5976, https://doi.org/10.5194/egusphere-egu25-5976, 2025.
Actual evapotranspiration (AET) estimates are crucial for hydrological processes, especially in Predictions in Ungauged Basins (PUB). However, traditional hydrological models often struggle to provide reliable ET values, which limits their performance in regions with sparse observational data. To address this, we integrate evapotranspiration constraints derived from the PML-V2 model, calibrated using flux tower data, into the improved HBV hydrological model (HBV-PML). This approach allows HBV to replicate ET estimates from PML-V2 model, ensuring the two models fully coupled. We applied the HBV-PML model to 4641 small catchments distributed globally, using a basin attribute similarity method to transfer parameters from 10 similar basins for each target basin. Ensemble results from 10 parameter sets showed that the runoff simulations from improved HBV-PML were slightly lower than those from the original HBV model, but both models achieved a median Nash-Sutcliffe Efficiency (NSE) greater than 0.6 for monthly runoff and greater than 0.35 for daily runoff. This is remarkable for HBV-PML since its ET output was identical to that of PML-V2. Our work demonstrates the potential of incorporating more accurate ET values into hydrological models, achieving reliable runoff simulations while dramatically improving the accuracy of evapotranspiration estimates.
How to cite: Wei, H., Zhang, Y., and Liu, C.: Runoff Simulation in Predictions in Ungauged Basins (PUB) Using the Improved HBV-PML Model with Evapotranspiration Constraints, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-5505, https://doi.org/10.5194/egusphere-egu25-5505, 2025.
The spatial discretization of hydrological sub units (HSU) is an inevitable and effective way to achieve refined distributed simulation. It can not only strengthen the distributed characteristics of the model, but also help simulate the runoff process on a small-scale watershed. However, the spatial variability of the runoff characteristics of HSU divided on a smaller spatial scale increases, and the hydrological process mechanism shows different characteristics under the refined spatio-temporal scale, which then forms the spatio-temporal scale effect. Therefore, it is of great significance to study the theory of spatial discretization of basin hydrological simulation at different spatio-temporal scales, construct a division method for runoff simulation HSUs, quantitatively describe and calculate the corresponding runoff characteristics, develop corresponding runoff simulation methods, and construct a refined distributed basin hydrological model.
In this paper, We studies the influence of the spatial scale of HSU on the runoff simulation of hydrological models, and proposes an optimization method based on the division of HSU scales, the matching of input data time scales, the quantitative calculation of model parameters, and the evaluation of basin applicability. With the refinement of the HSU scales, the spatio-temporal resolution of precipitation and evaporation input data should be improved accordingly to ensure the precise matching of rainfall evaporation process with hydrological response. Specifically, the calculation units were first divided into different scales, gradually refined from large scale to small scale, and the time scale changes of precipitation and evaporation data were simulated, using time steps of 3 hours, 2 hours and 1 hour respectively to improve the resolution of the hydrological response of the basin. Secondly, the input parameters of the Xin'anjiang (XAJ) model were optimized based on the quantitative calculation of the basin's underlying surface characteristics (such as area, morphological factors, river network density, slope, etc.). By combining quantitative calculations and empirical derivations at different scales, the effects of confluence parameters at different scales were analyzed. Finally, this paper verifies the rationality of the basin division, especially by evaluating the closure of the calculation unit based on DEM and bedrock depth data to ensure that each calculation unit has the hydrological mechanism characteristics required for hydrological model. In order to optimize the spatial scale of the calculation unit, a multi-objective optimization algorithm (Pareto frontier optimization) was used, and the rationality of the basin selection was verified through empirical research, thus ensuring the the model’s reliability. The results show that with the refinement of the calculation unit scale, the temporal and spatial scales of the precipitation evaporation input data and the hydrological response are better matched, but if the unit scale is too small, it may not meet the requirements of basin closure and hydrological mechanism. Therefore, the selection of the calculation unit scale should comprehensively consider the basin characteristics, the temporal and spatial scales of the data and the model mechanism. The reasonable calculation unit scale should usually not be less than 100 km² to ensure the accuracy of the model and the reliability of the mechanism.
How to cite: Qian, R., Zhang, X., Fang, Y., Sheng, K., and Cao, Y.: Scale Effects of Distributed Hydrological Simulation: Forcing, Structure and Mechanism, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-7756, https://doi.org/10.5194/egusphere-egu25-7756, 2025.
This work presents two country-wide datasets for Mexico: (1) Mexico's High Resolution Climate Database (MexHiResClimDB) and (2) Mexico's watershed database (MexWatDB). The MexHiResClimDB is a long-term (1951-2020) climate database at a spatial resolution of approximately 600 metres (that comprises daily minimum, average and maximum temperature as well as precipitation), while the MexWatDB is a new watershed classification for Mexico based on the Pfafstetter classification system. This new classification system is proposed because the currently used watershed classification in Mexico was developed in the 1950s (and is based on letters without a logical sequence, i.e. no upstream or downstream criteria was used for its coding). The MexWatDB consists of three nested watershed levels: the first division (L1) comprises 359 watersheds, while L2 has 1980 watersheds and L3 a total of 6262 hierarchically ordered watersheds (whereas the currently used scheme has a total of 976). While the MexHiresClimDB comprises daily, monthly and yearly rasters of Tmin, Tavg, Tmax and Precip (with their corresponding normals for the 1951-1980, 1961-1990, 1971-2000, 1981-2010 and 1991-2020 periods), the MexWatDB includes the aforementioned variables with the previously mentioned temporal aggregation on a watershed basis for each division level (i.e, L1, L2 and L3). An adequate watershed classification and a high resolution climate database is needed in Mexico, because daily precipitation can vary from 0 to more than 300 mm per day (or more than fivefold on a monthly basis for adjacent watersheds). These two new databases will be helpful to develop hydrological models from regional to local scales and to quantify the spatial variability of climate change in Mexico.
How to cite: Carrera-Hernandez, J. J.: The MexHiResClimDB and the MexWatDB: Two new country-wide databases in Mexico to improve hydrological modelling from regional to local scales., EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-4925, https://doi.org/10.5194/egusphere-egu25-4925, 2025.
Irrigation can significantly influence the hydrological system of agriculturally productive catchments thus understanding the dynamics of the water balance is indispensable for sustainable (agricultural) water management. Accurate estimation of catchment water balance, using distributed hydrological modelling and earth observation data, requires thorough consideration of the associated uncertainties from different sources such as uncertainties in the model representation of system processes and the errors and uncertainties embedded in the remotely sensed data.
In this study, we estimate the spatial distribution and temporal dynamics of the water balance of the Hindon River Basin, a sub-basin of the Ganga River Basin in India, where intense irrigation has driven overexploitation of water resources, highly influencing the hydrological regime. For this estimation the open-source distributed hydrological model wflow_sbm is used and evaluated with open global datasets, demonstrating the applicability of this method in data scarce regions. We emphasize on representing relevant hydrological features of the Hindon Basin that are often neglected in hydrological model applications: irrigation, domestic and industrial water use, and (infiltrating) river-aquifer interaction. Our analysis includes the quantification and assessment of the uncertainties associated with the global datasets used and model representation of important processes. First, prior uncertainties of the input variables were analyzed by comparing the errors and correlations of the products from different sources. Second, the impact of different schematizations of irrigation application and subsurface flow (including river infiltration to aquifer) on the modelled water balance is assessed. Finally, model output is evaluated by comparing the estimates and uncertainties of modelled water balance components with estimates from various global datasets.
Our analysis reveals the impact of different model choices, highlighting the necessity of proper model representation of hydrological processes and uncertainty assessment to achieve a more reliable estimation of catchment water balance.
How to cite: Mendoza, R., van Verseveld, W., Weerts, A., Sperna Weiland, F., and Seijger, C.: Quantifying water balance dynamics and associated uncertainties in an irrigated catchment using open-source gridded datasets and hydrological modelling with improved process representation: case of Hindon River Basin, India, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-18743, https://doi.org/10.5194/egusphere-egu25-18743, 2025.
Variable Infiltration Capacity (VIC) model is a large-scale, semi-distributed hydrologic model that operates at grid cell level. An effective application of VIC model requires proper model calibration. The total set of parameters from the two models (VIC and routing model) can be quite large but out of these parameters four soil parameters namely, exponent of soil moisture capacity curve (binfil), maximum velocity of base flow that occurs from the lowest soil layer (Dm), the fraction of Dm where non-linear baseflow occurs (Ds), and fraction of maximum soil moisture where non-linear baseflow occurs (Ws) were considered as these parameters cannot be estimated based on the available soil information. The present study attempts to model the hydrology of two synthetic and one real catchment, Ashti Catchment (sub catchment of Godavari River basin in India) using the VIC two soil layer model. The model was set in water balance mode at 0.25×0.25° spatial resolution with a daily time step. The calibration was carried out using Genetic Algorithm (GA) for six objective functions, Root mean square of error (RMSE), Maximum absolute relative error (MAXARE), Mean absolute error (MAE), Maximum absolute error (MAX_AE), Nash Sutcliffe Efficiency (NSE), and Kling-Gupta Efficiency (KGE).
Out of six objective functions, MAXARE in simulated discharge was found to be the best objective function for VIC when compared with other objective functions based on how close the generated parameter values were to the true values and the amount of computational time taken. Throughout the calibration, the GA parameters were kept as - solution per population: 32, num of parent mating: 2, mutation type: random, mutation probability: 0.09, crossover type: uniform, and crossover probability: 0.7. The overall result of calibration for both synthetic and real catchments for twenty years of discharge data (1971-1990) indicated that the simulated discharge was more sensitive to parameter binfil as compared to Dm, Ds and Ws. Individually, parameters Ds and Dm showed an insensitivity to the GA parameters up to three hundred generations. Further, out of seventeen set of initial values of binfil, Ds, Dm, and Ws, five sets provide different final parameter values after calibration but same calibration results in terms of MAXARE = 2.037%, NSE = 1.0, Coefficient of correlation = 0.999 Coefficient of determination = 0.998 and RMSE = 4.812 cumec. This indicated the presence of equifinality—where multiple parameters set produced similar model outputs. This study offers a foundation for further refining calibration approaches to address equifinality and improve model robustness.
Keywords: Hydrological modeling, VIC, Genetic algorithm, Equifinality, optimization
How to cite: Dwivedi, B., Guha, S., and Tripathi, S.: Equifinality in calibration of the Variable Infiltration Capacity (VIC) Model Parameters , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-6402, https://doi.org/10.5194/egusphere-egu25-6402, 2025.
The definition of the objective function for hydrological model calibration and goodness-of-fit metrics for validation are crucial in characterising model performance and driving model development. Ideally, model developers and users should identify the most suitable metric to characterize model performance based on their specific use cases and objectives. The benefits of this are evident; for example, the optimal objective function to calibrate a hydrological model only used for flood forecasting should be different from the optimal one to be used for a model focusing only on low flows. However, in practice, most hydrological models are used for multiple applications and a standard generalist function is adopted for their calibration. The two most widely used generalist functions are the Nash-Sutcliffe Efficiency (NSE) and the Kling-Gupta Efficiency (KGE), in its standard and modified versions. While the NSE is a simple normalization of the mean square error (MSE), the KGE overcomes some of the NSE limitations based on the decomposition of the MSE into three components, i.e., the error in the mean simulated streamflow, the relative variability, and the linear correlation between simulations and observations. Still, KGE presents some limitations, including a large sensitivity to outliers and an assumption of linearity and normality in the error distribution, which are impactful especially for the characterization of performances over low flows. Some alternative calibration functions have been proposed in the literature to overcome these limitations, but no calibration function overcomes the traditional two options (NSE and KGE) in improving the simulation along the whole range of the flow duration curve. Here we present the results of an extensive comparison of hydrological models calibrated and validated with multiple functions, including new variants and combinations of KGE and information-theory metrics, that can be suitable to characterize the performance over high, low, and regime flows. Two hydrological models (GR4J and Open Source LISFLOOD) were calibrated with several alternative calibration functions over more than 200 catchments in Europe with a varied range of hydroclimatic conditions. Both models were evaluated using multiple metrics, including use-case specific hydrological signatures focusing on flood characteristics, average regime and low flows. Based on our analysis, a new function is proposed which combines the three KGE components with an additional component based on the Jensen-Shannon Divergence. The performance of the two models calibrated with the new function is shown to outperform the standard KGE and NSE over low flows with minimal change in performance over regimes and high flows. This study shows that more effort should be devoted to the choice of the optimal calibration function for hydrological model applications when aiming to improve specific aspects of model performance.
How to cite: Grimaldi, S., Bavera, D., Ficchi, A., Moschini, F., Toreti, A., Pistocchi, A., and Salamon, P.: Towards an enhanced objective function for hydrological model calibration to improve the performance along the whole flow duration curve, EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-16423, https://doi.org/10.5194/egusphere-egu25-16423, 2025.
The hydrological response of small basins remains complex and challenging to quantify in an accurate way, particularly during extreme events such as floods as well as in the context of sustainable water resources management (Sellami et al., 2016). The application of hydrological models at basin scale offers a promising solution to this challenge by providing valuable tools for water resources management, enabling the analysis of past and current basin conditions as well as the evaluation of the implications of management decisions and imposed changes. In this study, the distributed hydrological model DREAM (Manfreda et al., 2005; Perrini et al., 2024), which incorporates Dunnian and Hortonian mechanisms, was applied to simulate flood events in the Fiumarella di Corleto basin (32.5 km²) and its sub-basin (0.65 km²) in the Italian region of Basilicata. Simulations were conducted for flood events occurring over a 20-year period (2002–2022). These simulations were based on a detailed hydrological and geomorphological characterization of the study area, integrating a hydro-meteorological dataset and initial soil moisture conditions derived from monitoring instruments and a geographical database (Dal Sasso et al., 2023) . Significant flood events were selected for model parameter optimization, due to their representativeness, allowing for the verification of the model’s performance, ensuring its ability to accurately reproduce hydrological behavior as well as for belonging to a dataset characterized by complete hydrological information. The results show that the hydrological model, with the Hortonian runoff mechanism, outperforms in capturing the basin’s immediate response to rainfall events. Preliminary results revealed a satisfactory match between simulated and observed data, as evidenced by the Nash-Sutcliffe efficiency coefficient ranging from 0.52 to 0.73 and the Kling-Gupta efficiency coefficient between 0.56 and 0.75. While errors in simulated and observed peak outflows varied, ranging from acceptable (2–3%) to more significant (up to 20%), the overall performance metrics indicate reliable alignment. These findings underscore the model’s capability to accurately reproduce flood processes, confirming its reliability for simulating extreme hydrological events and supporting its application in watershed management and flood risk mitigation.
DISCLAIMERS
The present research has been carried out within the RETURN Extended Partnership and received funding from the European Union Next-Generation EU (National Recovery and Resilience Plan - NRRP, Mission 4, Component 2, Investment 1.3 - D.D. 1243 2/8/2022, PE0000005).
This abstract is part of the project NODES which has received fundining from the MUR-M4C2 1.5 of PNRR funded by the European Union - NextGenerationEU (Grant agreement no. ECS00000036).
How to cite: Sileo, B., Dal Sasso, S. F., Onorati, B., Margiotta, M. R., Gioia, A., Iacobellis, V., and Fiorentino, M.: Advancing hydrological modeling in small watersheds: the fiumarella case study (southern Italy), EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-19368, https://doi.org/10.5194/egusphere-egu25-19368, 2025.
The medicane Daniel occurred in early September of 2023 and hit several countries of the Mediterranean area with catastrophic consequences and thousands of fatalities. Specifically in Greece, the event can be divided in two phases: a) the flash floods observed mainly in the catchments near Volos city and b) the fluvial flooding of Pineios river after two days due to the levee failures in several points, which were stressed due to the increasing water stage of the river. In this work, we focus only on the first part of the event, simulating the flood impact in the catchment which drains through Volos city and has an area of 33.5 km2. We used the physics-based 2D integrated flood simulator SERGHEI in a High-Performance Computing environment, since the simulation is characterized by increased computational burden. For the rainfall, we used two sources of meteorological data: a) a satellite-based ERA5 reanalysis of the European Centre for Medium-Range Weather Forecasts (ECMWF); b) a synthetic rainfall based on the statistical processing of the data recorded at the meteorological stations of the Hellenic Integrated Marine Inland water Observing, Forecasting and offshore Technology System (HIMIOFoTS), coupled with the Intensity-Duration-Frequency (IDF) curves of Greece. In order to perform a plausible check regarding the modelling efficiency, crowd-sourced information is collected from social media and compared with the SERGHEI outcome.
How to cite: Vagenas, S., Gkouma, A., Caviedes-Voullième, D., and Bellos, V.: Simulating the impact of medicane “Daniel” in an urban catchment using a 2D integrated flood simulator , EGU General Assembly 2025, Vienna, Austria, 27 Apr–2 May 2025, EGU25-562, https://doi.org/10.5194/egusphere-egu25-562, 2025.
