Articles | Volume 29, issue 15
https://doi.org/10.5194/hess-29-3589-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/hess-29-3589-2025
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
07 Aug 2025
Research article |
| 07 Aug 2025
| 07 Aug 2025
A distributed hybrid physics–AI framework for learning corrections of internal hydrological fluxes and enhancing high-resolution regionalized flood modeling
Ngo Nghi Truyen Huynh
CORRESPONDING AUTHOR
INRAE, Aix-Marseille Université, RECOVER, 3275 Route Cézanne, 13182 Aix-en-Provence, France
Pierre-André Garambois
CORRESPONDING AUTHOR
INRAE, Aix-Marseille Université, RECOVER, 3275 Route Cézanne, 13182 Aix-en-Provence, France
Benjamin Renard
INRAE, Aix-Marseille Université, RECOVER, 3275 Route Cézanne, 13182 Aix-en-Provence, France
François Colleoni
INRAE, Aix-Marseille Université, RECOVER, 3275 Route Cézanne, 13182 Aix-en-Provence, France
Jérôme Monnier
INSA, Institut de Mathématiques de Toulouse (IMT), Université de Toulouse, 31400 Toulouse, France
Hélène Roux
Institut de Mécanique des Fluides de Toulouse (IMFT), Université de Toulouse, CNRS, 31400 Toulouse, France
Related authors
Ngo Nghi Truyen Huynh, Pierre-André Garambois, François Colleoni, and Jérôme Monnier
EGUsphere, https://doi.org/10.5194/egusphere-2025-2797, https://doi.org/10.5194/egusphere-2025-2797, 2025
This preprint is open for discussion and under review for Geoscientific Model Development (GMD).
Short summary
Short summary
To better understand hydrological processes and improve flood simulation, combining artificial intelligence (AI) with process-based models is a promising direction. We introduce a hybrid physics-AI approach that seamlessly integrates neural networks into a distributed hydrological model to refine water flow dynamics within an implicit numerical scheme. The hybrid models demonstrate strong performance and interpretable results, leading to reliable streamflow simulations for flood modeling.
François Colleoni, Ngo Nghi Truyen Huynh, Pierre-André Garambois, Maxime Jay-Allemand, Didier Organde, Benjamin Renard, Thomas De Fournas, Apolline El Baz, Julie Demargne, and Pierre Javelle
EGUsphere, https://doi.org/10.5194/egusphere-2025-690, https://doi.org/10.5194/egusphere-2025-690, 2025
Short summary
Short summary
We present smash, an open-source framework for high-resolution hydrological modeling and data assimilation. It combines process-based models with neural networks for regionalization, enabling accurate simulations from catchment to country scale. With an efficient, differentiable solver, smash supports large-scale calibration and parallel computing. Tested on open datasets, it shows strong performance in river flow prediction, making it a valuable tool for research and operational use.
Guillaume Evin, Benoit Hingray, Guillaume Thirel, Agnès Ducharne, Laurent Strohmenger, Lola Corre, Yves Tramblay, Jean-Philippe Vidal, Jérémie Bonneau, François Colleoni, Joël Gailhard, Florence Habets, Frédéric Hendrickx, Louis Héraut, Peng Huang, Matthieu Le Lay, Claire Magand, Paola Marson, Céline Monteil, Simon Munier, Alix Reverdy, Jean-Michel Soubeyroux, Yoann Robin, Jean-Pierre Vergnes, Mathieu Vrac, and Eric Sauquet
EGUsphere, https://doi.org/10.5194/egusphere-2025-2727, https://doi.org/10.5194/egusphere-2025-2727, 2025
This preprint is open for discussion and under review for Hydrology and Earth System Sciences (HESS).
Short summary
Short summary
Explore2 provides hydrological projections for 1,735 French catchments. Using QUALYPSO, this study assesses uncertainties, including internal variability. By the end of the century, low flows are projected to decline in southern France under high emissions, while other indicators remain uncertain. Emission scenarios and regional climate models are key uncertainty sources. Internal variability is often as large as climate-driven changes.
Mohamed Saadi, Louis Guichard, Gabrielle Cognot, Laurent Labbouz, and Hélène Roux
EGUsphere, https://doi.org/10.5194/egusphere-2025-3393, https://doi.org/10.5194/egusphere-2025-3393, 2025
This preprint is open for discussion and under review for Hydrology and Earth System Sciences (HESS).
Short summary
Short summary
LSTM networks are excellent deep-learning tools to reproduce stream temperature observations, but their performances over the range of extreme (summer) stream temperature values have been overlooked. We close this gap by looking at strategies to improve the LSTM performances over the highest 10 % values of stream temperature observations. We found that the best strategy is to train the LSTM models at several locations with input variables that include static catchment and reach attributes.
Ngo Nghi Truyen Huynh, Pierre-André Garambois, François Colleoni, and Jérôme Monnier
EGUsphere, https://doi.org/10.5194/egusphere-2025-2797, https://doi.org/10.5194/egusphere-2025-2797, 2025
This preprint is open for discussion and under review for Geoscientific Model Development (GMD).
Short summary
Short summary
To better understand hydrological processes and improve flood simulation, combining artificial intelligence (AI) with process-based models is a promising direction. We introduce a hybrid physics-AI approach that seamlessly integrates neural networks into a distributed hydrological model to refine water flow dynamics within an implicit numerical scheme. The hybrid models demonstrate strong performance and interpretable results, leading to reliable streamflow simulations for flood modeling.
Juliette Godet, Pierre Nicolle, Nabil Hocini, Eric Gaume, Philippe Davy, Frederic Pons, Pierre Javelle, Pierre-André Garambois, Dimitri Lague, and Olivier Payrastre
Earth Syst. Sci. Data, 17, 2963–2983, https://doi.org/10.5194/essd-17-2963-2025, https://doi.org/10.5194/essd-17-2963-2025, 2025
Short summary
Short summary
This paper describes a dataset that includes input, output, and validation data for the simulation of flash flood hazards and three specific flash flood events in the French Mediterranean region. This dataset is particularly valuable as flood mapping methods often lack sufficient benchmark data. Additionally, we demonstrate how the hydraulic method we used, named Floodos, produces highly satisfactory results.
François Colleoni, Ngo Nghi Truyen Huynh, Pierre-André Garambois, Maxime Jay-Allemand, Didier Organde, Benjamin Renard, Thomas De Fournas, Apolline El Baz, Julie Demargne, and Pierre Javelle
EGUsphere, https://doi.org/10.5194/egusphere-2025-690, https://doi.org/10.5194/egusphere-2025-690, 2025
Short summary
Short summary
We present smash, an open-source framework for high-resolution hydrological modeling and data assimilation. It combines process-based models with neural networks for regionalization, enabling accurate simulations from catchment to country scale. With an efficient, differentiable solver, smash supports large-scale calibration and parallel computing. Tested on open datasets, it shows strong performance in river flow prediction, making it a valuable tool for research and operational use.
Mathieu Lucas, Michel Lang, Benjamin Renard, and Jérôme Le Coz
Hydrol. Earth Syst. Sci., 28, 5031–5047, https://doi.org/10.5194/hess-28-5031-2024, https://doi.org/10.5194/hess-28-5031-2024, 2024
Short summary
Short summary
The proposed flood frequency model accounts for uncertainty in the perception threshold S and the starting date of the historical period. Using a 500-year-long case study, inclusion of historical floods reduces the uncertainty in flood quantiles, even when only the number of exceedances of S is known. Ignoring threshold uncertainty leads to underestimated flood quantile uncertainty. This underlines the value of using a comprehensive framework for uncertainty estimation.
Maxime Jay-Allemand, Julie Demargne, Pierre-André Garambois, Pierre Javelle, Igor Gejadze, François Colleoni, Didier Organde, Patrick Arnaud, and Catherine Fouchier
Proc. IAHS, 385, 281–290, https://doi.org/10.5194/piahs-385-281-2024, https://doi.org/10.5194/piahs-385-281-2024, 2024
Short summary
Short summary
This work targets the improvement of a hydrologic model used for flash flood warnings. A gridded model is used to spatially describe the hydrological processes. We develop a method to estimate the best model setup based on scarce river flow observations. It uses a complex algorithm combined with geographical descriptors to generate gridded parameters that better capture catchment characteristics. Results are promising, improving the discharge estimations where no observations are available.
Guillaume Evin, Matthieu Le Lay, Catherine Fouchier, David Penot, Francois Colleoni, Alexandre Mas, Pierre-André Garambois, and Olivier Laurantin
Hydrol. Earth Syst. Sci., 28, 261–281, https://doi.org/10.5194/hess-28-261-2024, https://doi.org/10.5194/hess-28-261-2024, 2024
Short summary
Short summary
Hydrological modelling of mountainous catchments is challenging for many reasons, the main one being the temporal and spatial representation of precipitation forcings. This study presents an evaluation of the hydrological modelling of 55 small mountainous catchments of the northern French Alps, focusing on the influence of the type of precipitation reanalyses used as inputs. These evaluations emphasize the added value of radar measurements, in particular for the reproduction of flood events.
Rabab Yassine, Ludovic Cassan, Hélène Roux, Olivier Frysou, and François Pérès
Earth Surf. Dynam., 11, 1199–1221, https://doi.org/10.5194/esurf-11-1199-2023, https://doi.org/10.5194/esurf-11-1199-2023, 2023
Short summary
Short summary
Predicting river morphology evolution is very complicated, especially for mountain rivers with complex morphologies such as the Lac des Gaves reach in France. A 2D hydromorphological model was developed to reproduce the channel's evolution and provide reliable volumetric predictions while revealing the challenge of choosing adapted sediment transport and friction laws. Our model can provide decision-makers with reliable predictions to design suitable restoration measures for this reach.
Reyhaneh Hashemi, Pierre Brigode, Pierre-André Garambois, and Pierre Javelle
Hydrol. Earth Syst. Sci., 26, 5793–5816, https://doi.org/10.5194/hess-26-5793-2022, https://doi.org/10.5194/hess-26-5793-2022, 2022
Short summary
Short summary
Hydrologists have long dreamed of a tool that could adequately predict runoff in catchments. Data-driven long short-term memory (LSTM) models appear very promising to the hydrology community in this respect. Here, we have sought to benefit from traditional practices in hydrology to improve the effectiveness of LSTM models. We discovered that one LSTM parameter has a hydrologic interpretation and that there is a need to increase the data and to tune two parameters, thereby improving predictions.
Léo Pujol, Pierre-André Garambois, and Jérôme Monnier
Geosci. Model Dev., 15, 6085–6113, https://doi.org/10.5194/gmd-15-6085-2022, https://doi.org/10.5194/gmd-15-6085-2022, 2022
Short summary
Short summary
This contribution presents a new numerical model for representing hydraulic–hydrological quantities at the basin scale. It allows modeling large areas at a low computational cost, with fine zooms where needed. It allows the integration of local and satellite measurements, via data assimilation methods, to improve the model's match to observations. Using this capability, good matches to in situ observations are obtained on a model of the complex Adour river network with fine zooms on floodplains.
François Colleoni, Pierre-André Garambois, Pierre Javelle, Maxime Jay-Allemand, and Patrick Arnaud
EGUsphere, https://doi.org/10.5194/egusphere-2022-506, https://doi.org/10.5194/egusphere-2022-506, 2022
Preprint archived
Short summary
Short summary
This contribution presents the first evaluation of Variational Data Assimilation successfully applied over a large sample to the spatially distributed calibration of a newly taylored grid-based parsimonious model structure and corresponding adjoint. High performances are obtained in spatio-temporal validation and at flood time scales, especially for mediterranenan and oceanic catchments. Regional sensitivity analysis revealed the importance of the non conservative and production components.
Jérôme Le Coz, Guy D. Moukandi N'kaya, Jean-Pierre Bricquet, Alain Laraque, and Benjamin Renard
Proc. IAHS, 384, 25–29, https://doi.org/10.5194/piahs-384-25-2021, https://doi.org/10.5194/piahs-384-25-2021, 2021
Abubakar Haruna, Pierre-Andre Garambois, Helene Roux, Pierre Javelle, and Maxime Jay-Allemand
Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2021-414, https://doi.org/10.5194/hess-2021-414, 2021
Manuscript not accepted for further review
Short summary
Short summary
We compared three hydrological models in a flash flood modelling framework. We first identified the sensitive parameters of each model, then compared their performances in terms of outlet discharge and soil moisture simulation. We found out that resulting from the differences in their complexities/process representation, performance depends on the aspect/measure used. The study then highlights and proposed some future investigations/modifications to improve the models.
Judith Eeckman, Hélène Roux, Audrey Douinot, Bertrand Bonan, and Clément Albergel
Hydrol. Earth Syst. Sci., 25, 1425–1446, https://doi.org/10.5194/hess-25-1425-2021, https://doi.org/10.5194/hess-25-1425-2021, 2021
Short summary
Short summary
The risk of flash flood is of growing importance for populations, particularly in the Mediterranean area in the context of a changing climate. The representation of soil processes in models is a key factor for flash flood simulation. The importance of the various methods for soil moisture estimation are highlighted in this work. Local measurements from the field as well as data derived from satellite imagery can be used to assess the performance of model outputs.
Cited articles
Agency, E. E.: CORINE Land Cover 2012 (raster 100 m), European Environment Agency [data set], https://doi.org/10.2909/a84ae124-c5c5-4577-8e10-511bfe55cc0d, 2019. a, b
Arnaud, P., Aubert, Y., Organde, D., Cantet, P., Fouchier, C., and Folton, N.: Estimation de l'aléa hydrométéorologique par une méthode par simulation : la méthode SHYREG : présentation – performances – bases de données, Houille Blanche, 100, 20–26, https://doi.org/10.1051/lhb/2014012, 2014. a
Artigue, G., Johannet, A., Borrell, V., and Pistre, S.: Flash flood forecasting in poorly gauged basins using neural networks: case study of the Gardon de Mialet basin (southern France), Nat. Hazards Earth Syst. Sci., 12, 3307–3324, https://doi.org/10.5194/nhess-12-3307-2012, 2012. a
Astagneau, P. C., Bourgin, F., Andréassian, V., and Perrin, C.: When does a parsimonious model fail to simulate floods? Learning from the seasonality of model bias, Hydrolog. Sci. J., 66, 1288–1305, https://doi.org/10.1080/02626667.2021.1923720, 2021. a
Astagneau, P. C., Bourgin, F., Andréassian, V., and Perrin, C.: Catchment response to intense rainfall: evaluating modelling hypotheses, Hydrol. Process., 36, e14676, https://doi.org/10.1002/hyp.14676, 2022. a
Beven, K.: Towards a new paradigm in hydrology, in: Water for the Future: Hydrology in Perspective, IAHS Publication, 164, 393–403, 1987. a
Brigode, P., Génot, B., Lobligeois, F., and Delaigue, O.: Summary sheets of watershed-scale hydroclimatic observed data for France, Recherche Data Gouv [data set], https://doi.org/10.15454/UV01P1, 2020. a
Chen, C., Jiang, J., Liao, Z., Zhou, Y., Wang, H., and Pei, Q.: A short-term flood prediction based on spatial deep learning network: a case study for Xi County, China, J. Hydrol., 607, 127535, https://doi.org/10.1016/j.jhydrol.2022.127535, 2022. a
Chen, R. T. Q., Rubanova, Y., Bettencourt, J., and Duvenaud, D.: Neural Ordinary Differential Equations, arXiv [preprint], https://doi.org/10.48550/arXiv.1806.07366, 2019. a
Colleoni, F., Huynh, N. N. T., Garambois, P.-A., Jay-Allemand, M., Organde, D., Renard, B., De Fournas, T., El Baz, A., Demargne, J., and Javelle, P.: SMASH v1.0: A Differentiable and Regionalizable High-Resolution Hydrological Modeling and Data Assimilation Framework, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2025-690, 2025. a
Delaigue, O., Guimarães, G. M., Brigode, P., Génot, B., Perrin, C., Soubeyroux, J.-M., Janet, B., Addor, N., and Andréassian, V.: CAMELS-FR dataset: a large-sample hydroclimatic dataset for France to explore hydrological diversity and support model benchmarking, Earth Syst. Sci. Data, 17, 1461–1479, https://doi.org/10.5194/essd-17-1461-2025, 2025. a
Dooge, J. C. I.: Looking for hydrologic laws, Water Resour. Res., 22, 46S–58S, https://doi.org/10.1029/WR022i09Sp0046S, 1986. a
Douinot, A., Roux, H., Garambois, P.-A., and Dartus, D.: Using a multi-hypothesis framework to improve the understanding of flow dynamics during flash floods, Hydrol. Earth Syst. Sci., 22, 5317–5340, https://doi.org/10.5194/hess-22-5317-2018, 2018. a
Feng, D., Liu, J., Lawson, K., and Shen, C.: Differentiable, learnable, regionalized process-based models with multiphysical outputs can approach state-of-the-art hydrologic prediction accuracy, Water Resour. Res., 58, e2022WR032404, https://doi.org/10.1029/2022WR032404, 2022. a, b
Ficchì, A., Perrin, C., and Andréassian, V.: Hydrological modelling at multiple sub-daily time steps: model improvement via flux-matching, J. Hydrol., 575, 1308–1327, https://doi.org/10.1016/j.jhydrol.2019.05.084, 2019. a
Finke, P., Hartwich, R., Dudal, R., Ibanez, J., Jamagne, M., King, D., Montanarella, L., and Yassoglou, N.: Geo-referenced soil database for Europe. Manual of procedures, version 1.0, European Communities [data set], https://esdac.jrc.ec.europa.eu/ESDB_Archive/ESBN/Backup_old/docs/1998-rep5/250k-manual-v1.pdf (last access: 1 August 2025), 1998. a
Fleischmann, A. S., Paiva, R. C. D., Collischonn, W., Siqueira, V. A., Paris, A., Moreira, D. M., Papa, F., Bitar, A. A., Parrens, M., Aires, F., and Garambois, P. A.: Trade-offs between 1-D and 2-D regional river hydrodynamic models, Water Resour. Res., 56, e2019WR026812, https://doi.org/10.1029/2019WR026812, 2020. a
Garambois, P.-A., Roux, H., Larnier, K., Labat, D., and Dartus, D.: Parameter regionalization for a process-oriented distributed model dedicated to flash floods, J. Hydrol., 525, 383–399, 2015. a
Garambois, P.-A., Calmant, S., Roux, H., Paris, A., Monnier, J., Finaud-Guyot, P., Samine Montazem, A., and Santos da Silva, J.: Hydraulic visibility: using satellite altimetry to parameterize a hydraulic model of an ungauged reach of a braided river, Hydrol. Process., 31, 756–767, https://doi.org/10.1002/hyp.11033, 2017. a
Hascoet, L. and Pascual, V.: The Tapenade automatic differentiation tool: principles, model, and specification, ACM T. Math Software, 39, 1–43, 2013. a
Hashemi, R., Brigode, P., Garambois, P.-A., and Javelle, P.: How can we benefit from regime information to make more effective use of long short-term memory (LSTM) runoff models?, Hydrol. Earth Syst. Sci., 26, 5793–5816, https://doi.org/10.5194/hess-26-5793-2022, 2022. a, b
He, Q., Barajas-Solano, D., Tartakovsky, G., and Tartakovsky, A. M.: Physics-informed neural networks for multiphysics data assimilation with application to subsurface transport, Adv. Water Resour., 141, 103610, https://doi.org/10.1016/j.advwatres.2020.103610, 2020. a
Hochreiter, S. and Schmidhuber, J.: Long short-term memory, Neural Comput., 9, 1735–1780, https://doi.org/10.1162/neco.1997.9.8.1735, 1997. a
Höge, M., Scheidegger, A., Baity-Jesi, M., Albert, C., and Fenicia, F.: Improving hydrologic models for predictions and process understanding using neural ODEs, Hydrol. Earth Syst. Sci., 26, 5085–5102, https://doi.org/10.5194/hess-26-5085-2022, 2022. a, b
Huynh, N. N. T.: Medest Region and France 235-Catchments Data, Zenodo [data set], https://doi.org/10.5281/zenodo.13826145, 2024. a
Huynh, N. N. T. and Colleoni, F.: SMASH: Version 1.1-dev, Zenodo [code], https://doi.org/10.5281/zenodo.13696078, 2024. a
Huynh, N. N. T., Garambois, P.-A., Colleoni, F., and Javelle, P.: Signatures-and-sensitivity-based multi-criteria variational calibration for distributed hydrological modeling applied to Mediterranean floods, J. Hydrol., 625, 129992, https://doi.org/10.1016/j.jhydrol.2023.129992, 2023. a
Huynh, N. N. T., Garambois, P.-A., Colleoni, F., Renard, B., Monnier, J., and Roux, H.: Multiscale learnable physical modeling and data assimilation framework: application to high-resolution regionalized hydrological simulation of flash flood, Authorea [preprint], 1–26, https://doi.org/10.22541/au.170709054.44271526/v2, 2024a. a, b
Huynh, N. N. T., Garambois, P.-A., Colleoni, F., Renard, B., Roux, H., Demargne, J., Jay-Allemand, M., and Javelle, P.: Learning regionalization using accurate spatial cost gradients within a differentiable high-resolution hydrological model: application to the French Mediterranean region, Water Resour. Res., 60, e2024WR037544, https://doi.org/10.1029/2024WR037544, 2024b. a, b, c, d, e, f, g
Jiang, S., Zheng, Y., and Solomatine, D.: Improving AI system awareness of geoscience knowledge: symbiotic integration of physical approaches and deep learning, Geophys. Res. Lett., 47, e2020GL088229, https://doi.org/10.1029/2020GL088229, 2020. a
Kingma, D. P. and Ba, J.: Adam: A method for stochastic optimization, arXiv [preprint], https://doi.org/10.48550/arXiv.1412.6980 2014. a
Konapala, G., Kao, S.-C., Painter, S. L., and Lu, D.: Machine learning assisted hybrid models can improve streamflow simulation in diverse catchments across the conterminous US, Environ. Res. Lett., 15, 104022, https://doi.org/10.1088/1748-9326/aba927, 2020. a
Kratzert, F., Klotz, D., Brenner, C., Schulz, K., and Herrnegger, M.: Rainfall–runoff modelling using Long Short-Term Memory (LSTM) networks, Hydrol. Earth Syst. Sci., 22, 6005–6022, https://doi.org/10.5194/hess-22-6005-2018, 2018. a, b
Kratzert, F., Nearing, G., Addor, N., Erickson, T., Gauch, M., Gilon, O., Gudmundsson, L., Hassidim, A., Klotz, D., Nevo, S., Shalev, G., and Matias, Y.: Caravan – a global community dataset for large-sample hydrology, Scientific Data, 10, 61, https://doi.org/10.1038/s41597-023-01975-w, 2023. a
Kumanlioglu, A. A. and Fistikoglu, O.: Performance enhancement of a conceptual hydrological model by integrating artificial intelligence, J. Hydrol. Eng., 24, 04019047, https://doi.org/10.1061/(ASCE)HE.1943-5584.0001850, 2019. a, b
Larnier, K., Garambois, P.-A., Emery, C., Pujol, L., Monnier, J., Gal, L., Paris, A., Yesou, H., Ledauphin, T., and Calmant, S.: Estimating Channel Parameters and Discharge at River Network Scale Using Hydrological-Hydraulic Models, SWOT and Multi-Satellite Data, Water Resour. Res., 61, e2024WR038455, https://doi.org/10.1029/2024WR038455, 2025. a, b, c
LeCun, Y., Bengio, Y., and Hinton, G.: Deep learning, Nature, 521, 436–444, https://doi.org/10.1038/nature14539, 2015. a
Li, B., Sun, T., Tian, F., Tudaji, M., Qin, L., and Ni, G.: Hybrid hydrological modeling for large alpine basins: a semi-distributed approach, Hydrol. Earth Syst. Sci., 28, 4521–4538, https://doi.org/10.5194/hess-28-4521-2024, 2024. a
Liu, Y. and Gupta, H. V.: Uncertainty in hydrologic modeling: toward an integrated data assimilation framework, Water Resour. Res., 43, W07401, https://doi.org/10.1029/2006WR005756, 2007. a
Maier, H. R. and Dandy, G. C.: Neural networks for the prediction and forecasting of water resources variables: a review of modelling issues and applications, Environ. Modell. Softw., 15, 101–124, https://doi.org/10.1016/S1364-8152(99)00007-9, 2000. a
Malou, T., Garambois, P.-A., Paris, A., Monnier, J., and Larnier, K.: Generation and analysis of stage-fall-discharge laws from coupled hydrological-hydraulic river network model integrating sparse multi-satellite data, J. Hydrol., 603, 126993, https://doi.org/10.1016/j.jhydrol.2021.126993, 2021. a
Meyer Oliveira, A., Fleischmann, A. S., and Paiva, R. C. D.: On the contribution of remote sensing-based calibration to model hydrological and hydraulic processes in tropical regions, J. Hydrol., 597, 126184, https://doi.org/10.1016/j.jhydrol.2021.126184, 2021. a
Nearing, G. S., Kratzert, F., Sampson, A. K., Pelissier, C. S., Klotz, D., Frame, J. M., Prieto, C., and Gupta, H. V.: What role does hydrological science play in the age of machine learning?, Water Resour. Res., 57, e2020WR028091, https://doi.org/10.1029/2020WR028091, 2021. a
Odry, J.: Prédétermination des débits de crues extrêmes en sites non jaugés : régionalisation de la méthode par simulation SHYREG, PhD thesis, Géosciences de l’environnement, Hydrologie Aix-Marseille 2017, http://www.theses.fr/2017AIXM0424 (last access: 1 August 2025), 2017. a
Organde, D., Arnaud, P., Fine, J.-A., Fouchier, C., Folton, N., and Lavabre, J.: Régionalisation d'une méthode de prédétermination de crue sur l'ensemble du territoire français: la méthode SHYREG, Revue des Sciences de l'Eau, 26, 65–78, 2013. a
Oudin, L., Hervieu, F., Michel, C., Perrin, C., Andréassian, V., Anctil, F., and Loumagne, C.: Which potential evapotranspiration input for a lumped rainfall–runoff model?: Part 2 Towards a simple and efficient potential evapotranspiration model for rainfall–runoff modelling, J. Hydrol., 303, 290–306, 2005. a
Oudin, L., Andréassian, V., Perrin, C., Michel, C., and Le Moine, N.: Spatial proximity, physical similarity, regression and ungaged catchments: a comparison of regionalization approaches based on 913 French catchments, Water Resour. Res., 44, W03413, https://doi.org/10.1029/2007WR006240, 2008. a
Papamarkou, T., Hinkle, J., Young, M. T., and Womble, D.: Challenges in Markov chain Monte Carlo for Bayesian neural networks, Stat. Sci., 37, 425–442, https://doi.org/10.1214/21-STS840, 2022. a
Poncelet, C.: Du bassin au paramètre : jusqu'où peut-on régionaliser un modèle hydrologique conceptuel ?, PhD thesis, Hydrologie Paris 6 2016, http://www.theses.fr/2016PA066550 (last access: 1 August 2025), 2016. a
Pujol, L., Garambois, P.-A., Finaud-Guyot, P., Monnier, J., Larnier, K., Mosé, R., Biancamaria, S., Yesou, H., Moreira, D., Paris, A., and Calmant, S.: Estimation of multiple inflows and effective channel by assimilation of multi-satellite hydraulic signatures: The ungauged anabranching Negro river, J. Hydrol., 591, 125331, https://doi.org/10.1016/j.jhydrol.2020.125331, 2020. a
Pujol, L., Garambois, P.-A., and Monnier, J.: Multi-dimensional hydrological-hydraulic model with variational data assimilation for river networks and floodplains, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2022-10, 2022. a, b, c
Quintana-Seguí, P., Le Moigne, P., Durand, Y., Martin, E., Habets, F., Baillon, M., Canellas, C., Franchisteguy, L., and Morel, S.: Analysis of near-surface atmospheric variables: validation of the SAFRAN analysis over France, J. Appl. Meteorol. Clim., 47, 92, https://doi.org/10.1175/2007JAMC1636.1, 2008. a
Rackauckas, C., Ma, Y., Martensen, J., Warner, C., Zubov, K., Supekar, R., Skinner, D., Ramadhan, A., and Edelman, A.: Universal Differential Equations for Scientific Machine Learning, arXiv [preprint], https://doi.org/10.48550/arXiv.2001.04385, 2021. a
Raissi, M., Perdikaris, P., and Karniadakis, G.: Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations, J. Comput. Phys., 378, 686–707, https://doi.org/10.1016/j.jcp.2018.10.045, 2019. a
Roux, H., Labat, D., Garambois, P.-A., Maubourguet, M.-M., Chorda, J., and Dartus, D.: A physically-based parsimonious hydrological model for flash floods in Mediterranean catchments, Nat. Hazards Earth Syst. Sci., 11, 2567–2582, https://doi.org/10.5194/nhess-11-2567-2011, 2011. a
Roy, A., Kasiviswanathan, K. S., Patidar, S., Adeloye, A. J., Soundharajan, B.-S., and Ojha, C. S. P.: A novel physics-aware machine learning-based dynamic error correction model for improving streamflow forecast accuracy, Water Resour. Res., 59, e2022WR033318, https://doi.org/10.1029/2022WR033318, 2023. a
Santos, L., Thirel, G., and Perrin, C.: Continuous state-space representation of a bucket-type rainfall-runoff model: a case study with the GR4 model using state-space GR4 (version 1.0), Geosci. Model Dev., 11, 1591–1605, https://doi.org/10.5194/gmd-11-1591-2018, 2018. a, b, c, d
Shen, C. and Lawson, K.: Applications of Deep Learning in Hydrology, chap. 19, John Wiley and Sons, Ltd, https://doi.org/10.1002/9781119646181.ch19, 283–297, 2021. a, b
Shi, X., Chen, Z., Wang, H., Yeung, D.-Y., Wong, W.-k., and Woo, W.-c.: Convolutional LSTM Network: a machine learning approach for precipitation nowcasting, in: Proceedings of the 28th International Conference on Neural Information Processing Systems – Volume 1, NIPS'15, Montreal, Canada, 7–12 December 2015, 802–810, https://dl.acm.org/doi/10.5555/2969239.2969329 (last access: 1 August 2025), 2015. a
Tran, V. N., Ivanov, V. Y., Xu, D., and Kim, J.: Closing in on hydrologic predictive accuracy: combining the strengths of high-fidelity and physics-agnostic models, Geophys. Res. Lett., 50, e2023GL104464, https://doi.org/10.1029/2023GL104464, 2023. a
Tripathy, K. P. and Mishra, A. K.: Deep learning in hydrology and water resources disciplines: concepts, methods, applications, and research directions, J. Hydrol., 628, 130458, https://doi.org/10.1016/j.jhydrol.2023.130458, 2024. a
Vergara, H., Kirstetter, P.-E., Gourley, J. J., Flamig, Z. L., Hong, Y., Arthur, A., and Kolar, R.: Estimating a-priori kinematic wave model parameters based on regionalization for flash flood forecasting in the Conterminous United States, J. Hydrol., 541, 421–433, 2016. a
Vidal, J.-P., Martin, E., Franchistéguy, L., Baillon, M., and Soubeyroux, J.-M.: A 50-year high-resolution atmospheric reanalysis over France with the Safran system, Int. J. Climatol., 30, 1627–1644, https://doi.org/10.1002/joc.2003, 2010. a
Wang, C., Jiang, S., Zheng, Y., Han, F., Kumar, R., Rakovec, O., and Li, S.: Distributed hydrological modeling with physics-encoded deep learning: a general framework and its application in the Amazon, Water Resour. Res., 60, e2023WR036170, https://doi.org/10.1029/2023WR036170, 2024. a, b
Xu, T., Longyang, Q., Tyson, C., Zeng, R., and Neilson, B. T.: Hybrid physically based and deep learning modeling of a snow dominated, mountainous, karst watershed, Water Resour. Res., 58, e2021WR030993, https://doi.org/10.1029/2021WR030993, 2022. a
Yin, Y., Guen, V. L., Dona, J., de Bézenac, E., Ayed, I., Thome, N., and Gallinari, P.: Augmenting physical models with deep networks for complex dynamics forecasting, J. Stat. Mech.-Theory E., 2021, 124012, https://doi.org/10.1088/1742-5468/ac3ae5, 2021. a
Zhu, C., Byrd, R. H., Lu, P., and Nocedal, J.: Algorithm 778: L-BFGS-B: Fortran subroutines for large-scale bound-constrained optimization, ACM Trans. Math. Softw., 23, 550–560, 1997. a
Short summary
Understanding and modeling flash-flood-prone areas remains challenging due to limited data and scale-relevant hydrological theory. While machine learning shows promise, its integration with process-based models is difficult. We present an approach incorporating machine learning into a high-resolution hydrological model to correct internal fluxes and transfer parameters between watersheds. Results show improved accuracy, advancing the development of learnable and interpretable process-based models.
Understanding and modeling flash-flood-prone areas remains challenging due to limited data and...
