Articles | Volume 25, issue 4
https://doi.org/10.5194/hess-25-2045-2021
© Author(s) 2021. 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-25-2045-2021
© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
19 Apr 2021
Research article |
| 19 Apr 2021
| 19 Apr 2021
Rainfall–runoff prediction at multiple timescales with a single Long Short-Term Memory network
Institute for Machine Learning, Johannes Kepler University Linz, Linz, Austria
David R. Cheriton School of Computer Science, University of Waterloo, Waterloo, Canada
Frederik Kratzert
Institute for Machine Learning, Johannes Kepler University Linz, Linz, Austria
Daniel Klotz
Institute for Machine Learning, Johannes Kepler University Linz, Linz, Austria
Grey Nearing
Google Research, Mountain View, CA, USA
Department of Land, Air and Water Resources, University of California Davis, Davis, CA, USA
Jimmy Lin
David R. Cheriton School of Computer Science, University of Waterloo, Waterloo, Canada
Sepp Hochreiter
Institute for Machine Learning, Johannes Kepler University Linz, Linz, Austria
Related authors
Martin Gauch, Frederik Kratzert, Daniel Klotz, Grey Nearing, Deborah Cohen, and Oren Gilon
EGUsphere, https://doi.org/10.5194/egusphere-2025-1224, https://doi.org/10.5194/egusphere-2025-1224, 2025
Short summary
Short summary
Missing input data are one of the most common challenges when building deep learning hydrological models. We present and analyze different methods that can produce predictions when certain inputs are missing during training or inference. Our proposed strategies provide high accuracy while allowing for more flexible data handling and being robust to outages in operational scenarios.
Eduardo Acuña Espinoza, Frederik Kratzert, Daniel Klotz, Martin Gauch, Manuel Álvarez Chaves, Ralf Loritz, and Uwe Ehret
Hydrol. Earth Syst. Sci., 29, 1749–1758, https://doi.org/10.5194/hess-29-1749-2025, https://doi.org/10.5194/hess-29-1749-2025, 2025
Short summary
Short summary
Long short-term memory (LSTM) networks have demonstrated state-of-the-art performance for rainfall-runoff hydrological modelling. However, most studies focus on predictions at a daily scale, limiting the benefits of sub-daily (e.g. hourly) predictions in applications like flood forecasting. In this study, we introduce a new architecture, multi-frequency LSTM (MF-LSTM), designed to use inputs of various temporal frequencies to produce sub-daily (e.g. hourly) predictions at a moderate computational cost.
Eduardo Acuña Espinoza, Ralf Loritz, Frederik Kratzert, Daniel Klotz, Martin Gauch, Manuel Álvarez Chaves, and Uwe Ehret
Hydrol. Earth Syst. Sci., 29, 1277–1294, https://doi.org/10.5194/hess-29-1277-2025, https://doi.org/10.5194/hess-29-1277-2025, 2025
Short summary
Short summary
Data-driven techniques have shown the potential to outperform process-based models in rainfall–runoff simulations. Hybrid models, combining both approaches, aim to enhance accuracy and maintain interpretability. Expanding the set of test cases to evaluate hybrid models under different conditions, we test their generalization capabilities for extreme hydrological events.
Daniel Klotz, Peter Miersch, Thiago V. M. do Nascimento, Fabrizio Fenicia, Martin Gauch, and Jakob Zscheischler
Earth Syst. Sci. Data Discuss., https://doi.org/10.5194/essd-2024-450, https://doi.org/10.5194/essd-2024-450, 2025
Revised manuscript under review for ESSD
Short summary
Short summary
Data availability is central to hydrological science. It is the basis for advancing our understanding of hydrological processes, building prediction models, and anticipatory water management. We present a data-driven daily runoff reconstruction product for natural streamflow. We name it EARLS: European aggregated reconstruction for large-sample studies. The reconstructions represent daily simulations of natural streamflow across Europe and cover the period from 1953 to 2020.
Frederik Kratzert, Martin Gauch, Daniel Klotz, and Grey Nearing
Hydrol. Earth Syst. Sci., 28, 4187–4201, https://doi.org/10.5194/hess-28-4187-2024, https://doi.org/10.5194/hess-28-4187-2024, 2024
Short summary
Short summary
Recently, a special type of neural-network architecture became increasingly popular in hydrology literature. However, in most applications, this model was applied as a one-to-one replacement for hydrology models without adapting or rethinking the experimental setup. In this opinion paper, we show how this is almost always a bad decision and how using these kinds of models requires the use of large-sample hydrology data sets.
Andreas Auer, Martin Gauch, Frederik Kratzert, Grey Nearing, Sepp Hochreiter, and Daniel Klotz
Hydrol. Earth Syst. Sci., 28, 4099–4126, https://doi.org/10.5194/hess-28-4099-2024, https://doi.org/10.5194/hess-28-4099-2024, 2024
Short summary
Short summary
This work examines the impact of temporal and spatial information on the uncertainty estimation of streamflow forecasts. The study emphasizes the importance of data updates and global information for precise uncertainty estimates. We use conformal prediction to show that recent data enhance the estimates, even if only available infrequently. Local data yield reasonable average estimations but fall short for peak-flow events. The use of global data significantly improves these predictions.
Daniel Klotz, Martin Gauch, Frederik Kratzert, Grey Nearing, and Jakob Zscheischler
Hydrol. Earth Syst. Sci., 28, 3665–3673, https://doi.org/10.5194/hess-28-3665-2024, https://doi.org/10.5194/hess-28-3665-2024, 2024
Short summary
Short summary
The evaluation of model performance is essential for hydrological modeling. Using performance criteria requires a deep understanding of their properties. We focus on a counterintuitive aspect of the Nash–Sutcliffe efficiency (NSE) and show that if we divide the data into multiple parts, the overall performance can be higher than all the evaluations of the subsets. Although this follows from the definition of the NSE, the resulting behavior can have unintended consequences in practice.
Grey S. Nearing, Daniel Klotz, Jonathan M. Frame, Martin Gauch, Oren Gilon, Frederik Kratzert, Alden Keefe Sampson, Guy Shalev, and Sella Nevo
Hydrol. Earth Syst. Sci., 26, 5493–5513, https://doi.org/10.5194/hess-26-5493-2022, https://doi.org/10.5194/hess-26-5493-2022, 2022
Short summary
Short summary
When designing flood forecasting models, it is necessary to use all available data to achieve the most accurate predictions possible. This manuscript explores two basic ways of ingesting near-real-time streamflow data into machine learning streamflow models. The point we want to make is that when working in the context of machine learning (instead of traditional hydrology models that are based on
bio-geophysics), it is not necessary to use complex statistical methods for injecting sparse data.
Juliane Mai, Hongren Shen, Bryan A. Tolson, Étienne Gaborit, Richard Arsenault, James R. Craig, Vincent Fortin, Lauren M. Fry, Martin Gauch, Daniel Klotz, Frederik Kratzert, Nicole O'Brien, Daniel G. Princz, Sinan Rasiya Koya, Tirthankar Roy, Frank Seglenieks, Narayan K. Shrestha, André G. T. Temgoua, Vincent Vionnet, and Jonathan W. Waddell
Hydrol. Earth Syst. Sci., 26, 3537–3572, https://doi.org/10.5194/hess-26-3537-2022, https://doi.org/10.5194/hess-26-3537-2022, 2022
Short summary
Short summary
Model intercomparison studies are carried out to test various models and compare the quality of their outputs over the same domain. In this study, 13 diverse model setups using the same input data are evaluated over the Great Lakes region. Various model outputs – such as streamflow, evaporation, soil moisture, and amount of snow on the ground – are compared using standardized methods and metrics. The basin-wise model outputs and observations are made available through an interactive website.
Jonathan M. Frame, Frederik Kratzert, Daniel Klotz, Martin Gauch, Guy Shalev, Oren Gilon, Logan M. Qualls, Hoshin V. Gupta, and Grey S. Nearing
Hydrol. Earth Syst. Sci., 26, 3377–3392, https://doi.org/10.5194/hess-26-3377-2022, https://doi.org/10.5194/hess-26-3377-2022, 2022
Short summary
Short summary
The most accurate rainfall–runoff predictions are currently based on deep learning. There is a concern among hydrologists that deep learning models may not be reliable in extrapolation or for predicting extreme events. This study tests that hypothesis. The deep learning models remained relatively accurate in predicting extreme events compared with traditional models, even when extreme events were not included in the training set.
Thomas Lees, Steven Reece, Frederik Kratzert, Daniel Klotz, Martin Gauch, Jens De Bruijn, Reetik Kumar Sahu, Peter Greve, Louise Slater, and Simon J. Dadson
Hydrol. Earth Syst. Sci., 26, 3079–3101, https://doi.org/10.5194/hess-26-3079-2022, https://doi.org/10.5194/hess-26-3079-2022, 2022
Short summary
Short summary
Despite the accuracy of deep learning rainfall-runoff models, we are currently uncertain of what these models have learned. In this study we explore the internals of one deep learning architecture and demonstrate that the model learns about intermediate hydrological stores of soil moisture and snow water, despite never having seen data about these processes during training. Therefore, we find evidence that the deep learning approach learns a physically realistic mapping from inputs to outputs.
Daniel Klotz, Frederik Kratzert, Martin Gauch, Alden Keefe Sampson, Johannes Brandstetter, Günter Klambauer, Sepp Hochreiter, and Grey Nearing
Hydrol. Earth Syst. Sci., 26, 1673–1693, https://doi.org/10.5194/hess-26-1673-2022, https://doi.org/10.5194/hess-26-1673-2022, 2022
Short summary
Short summary
This contribution evaluates distributional runoff predictions from deep-learning-based approaches. We propose a benchmarking setup and establish four strong baselines. The results show that accurate, precise, and reliable uncertainty estimation can be achieved with deep learning.
Martin Gauch, Frederik Kratzert, Daniel Klotz, Grey Nearing, Deborah Cohen, and Oren Gilon
EGUsphere, https://doi.org/10.5194/egusphere-2025-1224, https://doi.org/10.5194/egusphere-2025-1224, 2025
Short summary
Short summary
Missing input data are one of the most common challenges when building deep learning hydrological models. We present and analyze different methods that can produce predictions when certain inputs are missing during training or inference. Our proposed strategies provide high accuracy while allowing for more flexible data handling and being robust to outages in operational scenarios.
Eduardo Acuña Espinoza, Frederik Kratzert, Daniel Klotz, Martin Gauch, Manuel Álvarez Chaves, Ralf Loritz, and Uwe Ehret
Hydrol. Earth Syst. Sci., 29, 1749–1758, https://doi.org/10.5194/hess-29-1749-2025, https://doi.org/10.5194/hess-29-1749-2025, 2025
Short summary
Short summary
Long short-term memory (LSTM) networks have demonstrated state-of-the-art performance for rainfall-runoff hydrological modelling. However, most studies focus on predictions at a daily scale, limiting the benefits of sub-daily (e.g. hourly) predictions in applications like flood forecasting. In this study, we introduce a new architecture, multi-frequency LSTM (MF-LSTM), designed to use inputs of various temporal frequencies to produce sub-daily (e.g. hourly) predictions at a moderate computational cost.
Eduardo Acuña Espinoza, Ralf Loritz, Frederik Kratzert, Daniel Klotz, Martin Gauch, Manuel Álvarez Chaves, and Uwe Ehret
Hydrol. Earth Syst. Sci., 29, 1277–1294, https://doi.org/10.5194/hess-29-1277-2025, https://doi.org/10.5194/hess-29-1277-2025, 2025
Short summary
Short summary
Data-driven techniques have shown the potential to outperform process-based models in rainfall–runoff simulations. Hybrid models, combining both approaches, aim to enhance accuracy and maintain interpretability. Expanding the set of test cases to evaluate hybrid models under different conditions, we test their generalization capabilities for extreme hydrological events.
Sanika Baste, Daniel Klotz, Eduardo Acuña Espinoza, Andras Bardossy, and Ralf Loritz
EGUsphere, https://doi.org/10.5194/egusphere-2025-425, https://doi.org/10.5194/egusphere-2025-425, 2025
Short summary
Short summary
This study evaluates the extrapolation performance of Long Short-Term Memory (LSTM) networks in rainfall-runoff modeling, specifically under extreme conditions. The findings reveal that the LSTM cannot predict discharge values beyond a theoretical limit, which is well below the extremity of its training data. This behavior results from the LSTM's gating structures rather than saturation of cell states alone.
Daniel Klotz, Peter Miersch, Thiago V. M. do Nascimento, Fabrizio Fenicia, Martin Gauch, and Jakob Zscheischler
Earth Syst. Sci. Data Discuss., https://doi.org/10.5194/essd-2024-450, https://doi.org/10.5194/essd-2024-450, 2025
Revised manuscript under review for ESSD
Short summary
Short summary
Data availability is central to hydrological science. It is the basis for advancing our understanding of hydrological processes, building prediction models, and anticipatory water management. We present a data-driven daily runoff reconstruction product for natural streamflow. We name it EARLS: European aggregated reconstruction for large-sample studies. The reconstructions represent daily simulations of natural streamflow across Europe and cover the period from 1953 to 2020.
Gab Abramowitz, Anna Ukkola, Sanaa Hobeichi, Jon Cranko Page, Mathew Lipson, Martin G. De Kauwe, Samuel Green, Claire Brenner, Jonathan Frame, Grey Nearing, Martyn Clark, Martin Best, Peter Anthoni, Gabriele Arduini, Souhail Boussetta, Silvia Caldararu, Kyeungwoo Cho, Matthias Cuntz, David Fairbairn, Craig R. Ferguson, Hyungjun Kim, Yeonjoo Kim, Jürgen Knauer, David Lawrence, Xiangzhong Luo, Sergey Malyshev, Tomoko Nitta, Jerome Ogee, Keith Oleson, Catherine Ottlé, Phillipe Peylin, Patricia de Rosnay, Heather Rumbold, Bob Su, Nicolas Vuichard, Anthony P. Walker, Xiaoni Wang-Faivre, Yunfei Wang, and Yijian Zeng
Biogeosciences, 21, 5517–5538, https://doi.org/10.5194/bg-21-5517-2024, https://doi.org/10.5194/bg-21-5517-2024, 2024
Short summary
Short summary
This paper evaluates land models – computer-based models that simulate ecosystem dynamics; land carbon, water, and energy cycles; and the role of land in the climate system. It uses machine learning and AI approaches to show that, despite the complexity of land models, they do not perform nearly as well as they could given the amount of information they are provided with about the prediction problem.
Raoul A. Collenteur, Ezra Haaf, Mark Bakker, Tanja Liesch, Andreas Wunsch, Jenny Soonthornrangsan, Jeremy White, Nick Martin, Rui Hugman, Ed de Sousa, Didier Vanden Berghe, Xinyang Fan, Tim J. Peterson, Jānis Bikše, Antoine Di Ciacca, Xinyue Wang, Yang Zheng, Maximilian Nölscher, Julian Koch, Raphael Schneider, Nikolas Benavides Höglund, Sivarama Krishna Reddy Chidepudi, Abel Henriot, Nicolas Massei, Abderrahim Jardani, Max Gustav Rudolph, Amir Rouhani, J. Jaime Gómez-Hernández, Seifeddine Jomaa, Anna Pölz, Tim Franken, Morteza Behbooei, Jimmy Lin, and Rojin Meysami
Hydrol. Earth Syst. Sci., 28, 5193–5208, https://doi.org/10.5194/hess-28-5193-2024, https://doi.org/10.5194/hess-28-5193-2024, 2024
Short summary
Short summary
We show the results of the 2022 Groundwater Time Series Modelling Challenge; 15 teams applied data-driven models to simulate hydraulic heads, and three model groups were identified: lumped, machine learning, and deep learning. For all wells, reasonable performance was obtained by at least one team from each group. There was not one team that performed best for all wells. In conclusion, the challenge was a successful initiative to compare different models and learn from each other.
Claudia Färber, Henning Plessow, Simon Mischel, Frederik Kratzert, Nans Addor, Guy Shalev, and Ulrich Looser
Earth Syst. Sci. Data Discuss., https://doi.org/10.5194/essd-2024-427, https://doi.org/10.5194/essd-2024-427, 2024
Revised manuscript accepted for ESSD
Short summary
Short summary
Large-sample datasets are essential in hydrological science to support modelling studies and advance process understanding. Caravan is a community initiative to create a large-sample hydrology dataset of meteorological forcing data, catchment attributes, and discharge data for catchments around the world. This dataset is a subset of hydrological discharge data and station-based watersheds from the Global Runoff Data Centre (GRDC), which are covered by an open data policy.
Frederik Kratzert, Martin Gauch, Daniel Klotz, and Grey Nearing
Hydrol. Earth Syst. Sci., 28, 4187–4201, https://doi.org/10.5194/hess-28-4187-2024, https://doi.org/10.5194/hess-28-4187-2024, 2024
Short summary
Short summary
Recently, a special type of neural-network architecture became increasingly popular in hydrology literature. However, in most applications, this model was applied as a one-to-one replacement for hydrology models without adapting or rethinking the experimental setup. In this opinion paper, we show how this is almost always a bad decision and how using these kinds of models requires the use of large-sample hydrology data sets.
Andreas Auer, Martin Gauch, Frederik Kratzert, Grey Nearing, Sepp Hochreiter, and Daniel Klotz
Hydrol. Earth Syst. Sci., 28, 4099–4126, https://doi.org/10.5194/hess-28-4099-2024, https://doi.org/10.5194/hess-28-4099-2024, 2024
Short summary
Short summary
This work examines the impact of temporal and spatial information on the uncertainty estimation of streamflow forecasts. The study emphasizes the importance of data updates and global information for precise uncertainty estimates. We use conformal prediction to show that recent data enhance the estimates, even if only available infrequently. Local data yield reasonable average estimations but fall short for peak-flow events. The use of global data significantly improves these predictions.
Daniel Klotz, Martin Gauch, Frederik Kratzert, Grey Nearing, and Jakob Zscheischler
Hydrol. Earth Syst. Sci., 28, 3665–3673, https://doi.org/10.5194/hess-28-3665-2024, https://doi.org/10.5194/hess-28-3665-2024, 2024
Short summary
Short summary
The evaluation of model performance is essential for hydrological modeling. Using performance criteria requires a deep understanding of their properties. We focus on a counterintuitive aspect of the Nash–Sutcliffe efficiency (NSE) and show that if we divide the data into multiple parts, the overall performance can be higher than all the evaluations of the subsets. Although this follows from the definition of the NSE, the resulting behavior can have unintended consequences in practice.
Qiutong Yu, Bryan A. Tolson, Hongren Shen, Ming Han, Juliane Mai, and Jimmy Lin
Hydrol. Earth Syst. Sci., 28, 2107–2122, https://doi.org/10.5194/hess-28-2107-2024, https://doi.org/10.5194/hess-28-2107-2024, 2024
Short summary
Short summary
It is challenging to incorporate input variables' spatial distribution information when implementing long short-term memory (LSTM) models for streamflow prediction. This work presents a novel hybrid modelling approach to predict streamflow while accounting for spatial variability. We evaluated the performance against lumped LSTM predictions in 224 basins across the Great Lakes region in North America. This approach shows promise for predicting streamflow in large, ungauged basin.
Louise J. Slater, Louise Arnal, Marie-Amélie Boucher, Annie Y.-Y. Chang, Simon Moulds, Conor Murphy, Grey Nearing, Guy Shalev, Chaopeng Shen, Linda Speight, Gabriele Villarini, Robert L. Wilby, Andrew Wood, and Massimiliano Zappa
Hydrol. Earth Syst. Sci., 27, 1865–1889, https://doi.org/10.5194/hess-27-1865-2023, https://doi.org/10.5194/hess-27-1865-2023, 2023
Short summary
Short summary
Hybrid forecasting systems combine data-driven methods with physics-based weather and climate models to improve the accuracy of predictions for meteorological and hydroclimatic events such as rainfall, temperature, streamflow, floods, droughts, tropical cyclones, or atmospheric rivers. We review recent developments in hybrid forecasting and outline key challenges and opportunities in the field.
Grey S. Nearing, Daniel Klotz, Jonathan M. Frame, Martin Gauch, Oren Gilon, Frederik Kratzert, Alden Keefe Sampson, Guy Shalev, and Sella Nevo
Hydrol. Earth Syst. Sci., 26, 5493–5513, https://doi.org/10.5194/hess-26-5493-2022, https://doi.org/10.5194/hess-26-5493-2022, 2022
Short summary
Short summary
When designing flood forecasting models, it is necessary to use all available data to achieve the most accurate predictions possible. This manuscript explores two basic ways of ingesting near-real-time streamflow data into machine learning streamflow models. The point we want to make is that when working in the context of machine learning (instead of traditional hydrology models that are based on
bio-geophysics), it is not necessary to use complex statistical methods for injecting sparse data.
Sella Nevo, Efrat Morin, Adi Gerzi Rosenthal, Asher Metzger, Chen Barshai, Dana Weitzner, Dafi Voloshin, Frederik Kratzert, Gal Elidan, Gideon Dror, Gregory Begelman, Grey Nearing, Guy Shalev, Hila Noga, Ira Shavitt, Liora Yuklea, Moriah Royz, Niv Giladi, Nofar Peled Levi, Ofir Reich, Oren Gilon, Ronnie Maor, Shahar Timnat, Tal Shechter, Vladimir Anisimov, Yotam Gigi, Yuval Levin, Zach Moshe, Zvika Ben-Haim, Avinatan Hassidim, and Yossi Matias
Hydrol. Earth Syst. Sci., 26, 4013–4032, https://doi.org/10.5194/hess-26-4013-2022, https://doi.org/10.5194/hess-26-4013-2022, 2022
Short summary
Short summary
Early flood warnings are one of the most effective tools to save lives and goods. Machine learning (ML) models can improve flood prediction accuracy but their use in operational frameworks is limited. The paper presents a flood warning system, operational in India and Bangladesh, that uses ML models for forecasting river stage and flood inundation maps and discusses the models' performances. In 2021, more than 100 million flood alerts were sent to people near rivers over an area of 470 000 km2.
Juliane Mai, Hongren Shen, Bryan A. Tolson, Étienne Gaborit, Richard Arsenault, James R. Craig, Vincent Fortin, Lauren M. Fry, Martin Gauch, Daniel Klotz, Frederik Kratzert, Nicole O'Brien, Daniel G. Princz, Sinan Rasiya Koya, Tirthankar Roy, Frank Seglenieks, Narayan K. Shrestha, André G. T. Temgoua, Vincent Vionnet, and Jonathan W. Waddell
Hydrol. Earth Syst. Sci., 26, 3537–3572, https://doi.org/10.5194/hess-26-3537-2022, https://doi.org/10.5194/hess-26-3537-2022, 2022
Short summary
Short summary
Model intercomparison studies are carried out to test various models and compare the quality of their outputs over the same domain. In this study, 13 diverse model setups using the same input data are evaluated over the Great Lakes region. Various model outputs – such as streamflow, evaporation, soil moisture, and amount of snow on the ground – are compared using standardized methods and metrics. The basin-wise model outputs and observations are made available through an interactive website.
Jonathan M. Frame, Frederik Kratzert, Daniel Klotz, Martin Gauch, Guy Shalev, Oren Gilon, Logan M. Qualls, Hoshin V. Gupta, and Grey S. Nearing
Hydrol. Earth Syst. Sci., 26, 3377–3392, https://doi.org/10.5194/hess-26-3377-2022, https://doi.org/10.5194/hess-26-3377-2022, 2022
Short summary
Short summary
The most accurate rainfall–runoff predictions are currently based on deep learning. There is a concern among hydrologists that deep learning models may not be reliable in extrapolation or for predicting extreme events. This study tests that hypothesis. The deep learning models remained relatively accurate in predicting extreme events compared with traditional models, even when extreme events were not included in the training set.
Thomas Lees, Steven Reece, Frederik Kratzert, Daniel Klotz, Martin Gauch, Jens De Bruijn, Reetik Kumar Sahu, Peter Greve, Louise Slater, and Simon J. Dadson
Hydrol. Earth Syst. Sci., 26, 3079–3101, https://doi.org/10.5194/hess-26-3079-2022, https://doi.org/10.5194/hess-26-3079-2022, 2022
Short summary
Short summary
Despite the accuracy of deep learning rainfall-runoff models, we are currently uncertain of what these models have learned. In this study we explore the internals of one deep learning architecture and demonstrate that the model learns about intermediate hydrological stores of soil moisture and snow water, despite never having seen data about these processes during training. Therefore, we find evidence that the deep learning approach learns a physically realistic mapping from inputs to outputs.
Daniel Klotz, Frederik Kratzert, Martin Gauch, Alden Keefe Sampson, Johannes Brandstetter, Günter Klambauer, Sepp Hochreiter, and Grey Nearing
Hydrol. Earth Syst. Sci., 26, 1673–1693, https://doi.org/10.5194/hess-26-1673-2022, https://doi.org/10.5194/hess-26-1673-2022, 2022
Short summary
Short summary
This contribution evaluates distributional runoff predictions from deep-learning-based approaches. We propose a benchmarking setup and establish four strong baselines. The results show that accurate, precise, and reliable uncertainty estimation can be achieved with deep learning.
Frederik Kratzert, Daniel Klotz, Sepp Hochreiter, and Grey S. Nearing
Hydrol. Earth Syst. Sci., 25, 2685–2703, https://doi.org/10.5194/hess-25-2685-2021, https://doi.org/10.5194/hess-25-2685-2021, 2021
Short summary
Short summary
We investigate how deep learning models use different meteorological data sets in the task of (regional) rainfall–runoff modeling. We show that performance can be significantly improved when using different data products as input and further show how the model learns to combine those meteorological input differently across time and space. The results are carefully benchmarked against classical approaches, showing the supremacy of the presented approach.
Frederik Kratzert, Daniel Klotz, Guy Shalev, Günter Klambauer, Sepp Hochreiter, and Grey Nearing
Hydrol. Earth Syst. Sci., 23, 5089–5110, https://doi.org/10.5194/hess-23-5089-2019, https://doi.org/10.5194/hess-23-5089-2019, 2019
Short summary
Short summary
A new approach for regional rainfall–runoff modeling using long short-term memory (LSTM)-based models is presented and benchmarked against a range of well-known hydrological models. The approach significantly outperforms regionally calibrated hydrological models but also basin-wise calibrated models. Furthermore, we propose an adaption of the LSTM that allows us to extract the learned catchment understanding of the model and show that it matches our hydrology expert knowledge.
Frederik Kratzert, Daniel Klotz, Claire Brenner, Karsten Schulz, and Mathew Herrnegger
Hydrol. Earth Syst. Sci., 22, 6005–6022, https://doi.org/10.5194/hess-22-6005-2018, https://doi.org/10.5194/hess-22-6005-2018, 2018
Short summary
Short summary
In this paper, we propose a novel data-driven approach for
rainfall–runoff modelling, using the long short-term memory (LSTM) network, a special type of recurrent neural network. We show in three different experiments that this network is able to learn to predict the discharge purely from meteorological input parameters (such as precipitation or temperature) as accurately as (or better than) the well-established Sacramento Soil Moisture Accounting model, coupled with the Snow-17 snow model.
Karsten Schulz, Reinhard Burgholzer, Daniel Klotz, Johannes Wesemann, and Mathew Herrnegger
Hydrol. Earth Syst. Sci., 22, 2607–2613, https://doi.org/10.5194/hess-22-2607-2018, https://doi.org/10.5194/hess-22-2607-2018, 2018
Short summary
Short summary
The unit hydrograph has been one of the most widely employed modelling techniques to predict rainfall-runoff behaviour of hydrological catchments. We developed a lecture theatre experiment including some student involvement to illustrate the principles behind this modelling technique. The experiment only uses very simple and cheap material including a set of plastic balls (representing rainfall), magnetic stripes (tacking the balls to the white board) and sieves (for ball/water gauging).
Related subject area
Subject: Catchment hydrology | Techniques and Approaches: Modelling approaches
Hydrological regime index for non-perennial rivers
Assessing the adequacy of traditional hydrological models for climate change impact studies: a case for long short-term memory (LSTM) neural networks
Assessing the value of high-resolution data and parameter transferability across temporal scales in hydrological modeling: a case study in northern China
Technical note: How many models do we need to simulate hydrologic processes across large geographical domains?
CONCN: a high-resolution, integrated surface water–groundwater ParFlow modeling platform of continental China
Evaluating the effects of topography and land use change on hydrological signatures: a comparative study of two adjacent watersheds
Technical note: What does the Standardized Streamflow Index actually reflect? Insights and implications for hydrological drought analysis
Long short-term memory networks for enhancing real-time flood forecasts: a case study for an underperforming hydrologic model
Assessing the value of high-resolution rainfall and streamflow data for hydrological modeling: an analysis based on 63 catchments in southeast China
Catchments do not strictly follow Budyko curves over multiple decades, but deviations are minor and predictable
Scale dependency in modeling nivo-glacial hydrological systems: the case of the Arolla basin, Switzerland
Extended-range forecasting of stream water temperature with deep-learning models
Technical note: An approach for handling multiple temporal frequencies with different input dimensions using a single LSTM cell
Projections of streamflow intermittence under climate change in European drying river networks
Economic valuation of subsurface water contributions to watershed ecosystem services using a fully integrated groundwater–surface-water model
Analyzing the generalization capabilities of a hybrid hydrological model for extrapolation to extreme events
CH-RUN: a deep-learning-based spatially contiguous runoff reconstruction for Switzerland
Runoff component quantification and future streamflow projection in a large mountainous basin based on a multidata-constrained cryospheric–hydrological model
Multi-variable process-based calibration of a behavioural hydrological model
Exploring the potential processes controlling changes in precipitation–runoff relationships in non-stationary environments
A diversity-centric strategy for the selection of spatio-temporal training data for LSTM-based streamflow forecasting
Simulating the Tone River eastward diversion project in Japan carried out 4 centuries ago
Lack of robustness of hydrological models: a large-sample diagnosis and an attempt to identify hydrological and climatic drivers
Achieving water budget closure through physical hydrological process modelling: insights from a large-sample study
Heavy-tailed flood peak distributions: what is the effect of the spatial variability of rainfall and runoff generation?
A Distributed Hybrid Physics-AI Framework for Learning Corrections of Internal Hydrological Fluxes and Enhancing High-Resolution Regionalized Flood Modeling
State updating of the Xin'anjiang model: joint assimilating streamflow and multi-source soil moisture data via the asynchronous ensemble Kalman filter with enhanced error models
Improving the hydrological consistency of a process-based solute-transport model by simultaneous calibration of streamflow and stream concentrations
Understanding the relationship between streamflow forecast skill and value across the western US
Leveraging soil diversity to mitigate hydrological extremes with nature-based solutions in productive catchments
Leveraging a time-series event separation method to disentangle time-varying hydrologic controls on streamflow – application to wildfire-affected catchments
The significance of the leaf area index for evapotranspiration estimation in SWAT-T for characteristic land cover types of West Africa
Improved representation of soil moisture processes through incorporation of cosmic-ray neutron count measurements in a large-scale hydrologic model
Spatio-temporal patterns and trends of streamflow in water-scarce Mediterranean basins
A large-sample modelling approach towards integrating streamflow and evaporation data for the Spanish catchments
Seasonal variation in land cover estimates reveals sensitivities and opportunities for environmental models
Soil moisture and precipitation intensity control the transit time distribution of quick flow in a flashy headwater catchment
Estimating response times, flow velocities, and roughness coefficients of Canadian Prairie basins
Learning landscape features from streamflow with autoencoders
Constraining pesticide degradation in conceptual distributed catchment models with compound-specific isotope analysis (CSIA)
On the use of streamflow transformations for hydrological model calibration
Unveiling the Impact of Potential Evapotranspiration Method Selection on Trends in Hydrological Cycle Components Across Europe
Simulation-based inference for parameter estimation of complex watershed simulators
Comparative Hydrological Modeling of Snow-Cover and Frozen Ground Impacts Under Topographically Complex Conditions
Multi-scale soil moisture data and process-based modeling reveal the importance of lateral groundwater flow in a subarctic catchment
Catchment response to climatic variability: implications for root zone storage and streamflow predictions
Hybrid hydrological modeling for large alpine basins: a semi-distributed approach
Controls on spatial and temporal variability of soil moisture across a heterogeneous boreal forest landscape
Where can rewetting of forested peatland reduce extreme flows?
Karst aquifer discharge response to rainfall interpreted as anomalous transport
Pablo Fernando Dornes and Rocío Noelia Comas
Hydrol. Earth Syst. Sci., 29, 2901–2923, https://doi.org/10.5194/hess-29-2901-2025, https://doi.org/10.5194/hess-29-2901-2025, 2025
Short summary
Short summary
The Desaguadero–Salado–Chadiluevú–Curacó (DSCC) River is a semiarid river which is heavily dammed at its tributaries which collect the snowmelt runoff. This runoff feeds mostly gravitational irrigation systems of very low efficiency. As a result, the DSCC River does not have natural runoff. The proposed hydrological regime index (HRI) is able to discriminate and quantify regime alterations under permanent and non-permanent flow conditions and with low- and high-impoundment conditions.
Jean-Luc Martel, François Brissette, Richard Arsenault, Richard Turcotte, Mariana Castañeda-Gonzalez, William Armstrong, Edouard Mailhot, Jasmine Pelletier-Dumont, Gabriel Rondeau-Genesse, and Louis-Philippe Caron
Hydrol. Earth Syst. Sci., 29, 2811–2836, https://doi.org/10.5194/hess-29-2811-2025, https://doi.org/10.5194/hess-29-2811-2025, 2025
Short summary
Short summary
This study compares long short-term memory (LSTM) neural networks with traditional hydrological models to predict future streamflow under climate change. Using data from 148 catchments, it finds that LSTM models, which learn from extensive data sequences, perform differently and often better than traditional hydrological models. The continental LSTM model, which includes data from diverse climate zones, is particularly effective for understanding climate impacts on water resources.
Mahmut Tudaji, Yi Nan, and Fuqiang Tian
Hydrol. Earth Syst. Sci., 29, 2633–2654, https://doi.org/10.5194/hess-29-2633-2025, https://doi.org/10.5194/hess-29-2633-2025, 2025
Short summary
Short summary
We assessed the value of high-resolution data and parameter transferability across temporal scales based on seven catchments in northern China. We found that higher-resolution data do not always improve model performance, questioning the need for such data. Model parameters are transferable across different data resolutions but not across computational time steps. It is recommended to utilize a smaller computational time step when building hydrological models even without high-resolution data.
Wouter J. M. Knoben, Ashwin Raman, Gaby J. Gründemann, Mukesh Kumar, Alain Pietroniro, Chaopeng Shen, Yalan Song, Cyril Thébault, Katie van Werkhoven, Andrew W. Wood, and Martyn P. Clark
Hydrol. Earth Syst. Sci., 29, 2361–2375, https://doi.org/10.5194/hess-29-2361-2025, https://doi.org/10.5194/hess-29-2361-2025, 2025
Short summary
Short summary
Hydrologic models are needed to provide simulations of water availability, floods, and droughts. The accuracy of these simulations is often quantified with so-called performance scores. A common thought is that different models are more or less applicable to different landscapes, depending on how the model works. We show that performance scores are not helpful in distinguishing between different models and thus cannot easily be used to select an appropriate model for a specific place.
Chen Yang, Zitong Jia, Wenjie Xu, Zhongwang Wei, Xiaolang Zhang, Yiguang Zou, Jeffrey McDonnell, Laura Condon, Yongjiu Dai, and Reed Maxwell
Hydrol. Earth Syst. Sci., 29, 2201–2218, https://doi.org/10.5194/hess-29-2201-2025, https://doi.org/10.5194/hess-29-2201-2025, 2025
Short summary
Short summary
We developed the first high-resolution, integrated surface water–groundwater hydrologic model of the entirety of continental China using ParFlow. The model shows good performance in terms of streamflow and water table depth when compared to global data products and observations. It is essential for water resources management and decision-making in China within a consistent framework in the changing world. It also has significant implications for similar modeling in other places in the world.
Haifan Liu, Haochen Yan, and Mingfu Guan
Hydrol. Earth Syst. Sci., 29, 2109–2132, https://doi.org/10.5194/hess-29-2109-2025, https://doi.org/10.5194/hess-29-2109-2025, 2025
Short summary
Short summary
Land changes and landscape features critically impact water systems. Studying two watersheds in China’s Greater Bay Area, we found slope strongly influences water processes in mountainous areas. However, this relationship is weak in the lower regions of steeper watersheds. Urbanization leads to an increase in annual surface runoff, while flatter watersheds exhibit a buffering capacity against this effect. However, this buffering capacity diminishes with increasing annual rainfall intensity.
Fabián Lema, Pablo A. Mendoza, Nicolás A. Vásquez, Naoki Mizukami, Mauricio Zambrano-Bigiarini, and Ximena Vargas
Hydrol. Earth Syst. Sci., 29, 1981–2002, https://doi.org/10.5194/hess-29-1981-2025, https://doi.org/10.5194/hess-29-1981-2025, 2025
Short summary
Short summary
Hydrological droughts affect ecosystems and socioeconomic activities worldwide. Despite the fact that they are commonly described with the Standardized Streamflow Index (SSI), there is limited understanding of what they truly reflect in terms of water cycle processes. Here, we used state-of-the-art hydrological models in Andean basins to examine drivers of SSI fluctuations. The results highlight the importance of careful selection of indices and timescales for accurate drought characterization and monitoring.
Sebastian Gegenleithner, Manuel Pirker, Clemens Dorfmann, Roman Kern, and Josef Schneider
Hydrol. Earth Syst. Sci., 29, 1939–1962, https://doi.org/10.5194/hess-29-1939-2025, https://doi.org/10.5194/hess-29-1939-2025, 2025
Short summary
Short summary
Accurate early-warning systems are crucial for reducing the damage caused by flooding events. In this study, we explored the potential of long short-term memory networks for enhancing the forecast accuracy of hydrologic models employed in operational flood forecasting. The presented approach elevated the investigated hydrologic model’s forecast accuracy for further ahead predictions and at flood event runoff.
Mahmut Tudaji, Yi Nan, and Fuqiang Tian
Hydrol. Earth Syst. Sci., 29, 1919–1937, https://doi.org/10.5194/hess-29-1919-2025, https://doi.org/10.5194/hess-29-1919-2025, 2025
Short summary
Short summary
Common intuition holds that higher input data resolution leads to better results. To assess the benefits of high-resolution data, we conduct simulation experiments using data with various temporal resolutions across multiple catchments and find that higher-resolution data do not always improve model performance, challenging the necessity of pursuing such data. In catchments with small areas or significant flow variability, high-resolution data is more valuable.
Muhammad Ibrahim, Miriam Coenders-Gerrits, Ruud van der Ent, and Markus Hrachowitz
Hydrol. Earth Syst. Sci., 29, 1703–1723, https://doi.org/10.5194/hess-29-1703-2025, https://doi.org/10.5194/hess-29-1703-2025, 2025
Short summary
Short summary
The quantification of precipitation into evaporation and runoff is vital for water resources management. The Budyko framework, based on aridity and evaporative indices of a catchment, can be an ideal tool for that. However, recent research highlights deviations of catchments from the expected evaporative index, casting doubt on its reliability. This study quantifies deviations of 2387 catchments, finding them minor and predictable. Integrating these into predictions upholds the framework's efficacy.
Anne-Laure Argentin, Pascal Horton, Bettina Schaefli, Jamal Shokory, Felix Pitscheider, Leona Repnik, Mattia Gianini, Simone Bizzi, Stuart N. Lane, and Francesco Comiti
Hydrol. Earth Syst. Sci., 29, 1725–1748, https://doi.org/10.5194/hess-29-1725-2025, https://doi.org/10.5194/hess-29-1725-2025, 2025
Short summary
Short summary
In this article, we show that by taking the optimal parameters calibrated with a semi-lumped model for the discharge at a catchment's outlet, we can accurately simulate runoff at various points within the study area, including three nested and three neighboring catchments. In addition, we demonstrate that employing more intricate melt models, which better represent physical processes, enhances the transfer of parameters in the simulation, until we observe overparameterization.
Ryan S. Padrón, Massimiliano Zappa, Luzi Bernhard, and Konrad Bogner
Hydrol. Earth Syst. Sci., 29, 1685–1702, https://doi.org/10.5194/hess-29-1685-2025, https://doi.org/10.5194/hess-29-1685-2025, 2025
Short summary
Short summary
We generate operational forecasts of daily maximum stream water temperature for 32 consecutive days at 54 stations in Switzerland with our best-performing data-driven model. The average forecast error is 0.38 °C for 1 d ahead and increases to 0.90 °C for 32 d ahead given the uncertainty in the meteorological variables influencing water temperature. Here we compare the skill of several models, how well they can forecast at new and ungauged stations, and the importance of different model inputs.
Eduardo Acuña Espinoza, Frederik Kratzert, Daniel Klotz, Martin Gauch, Manuel Álvarez Chaves, Ralf Loritz, and Uwe Ehret
Hydrol. Earth Syst. Sci., 29, 1749–1758, https://doi.org/10.5194/hess-29-1749-2025, https://doi.org/10.5194/hess-29-1749-2025, 2025
Short summary
Short summary
Long short-term memory (LSTM) networks have demonstrated state-of-the-art performance for rainfall-runoff hydrological modelling. However, most studies focus on predictions at a daily scale, limiting the benefits of sub-daily (e.g. hourly) predictions in applications like flood forecasting. In this study, we introduce a new architecture, multi-frequency LSTM (MF-LSTM), designed to use inputs of various temporal frequencies to produce sub-daily (e.g. hourly) predictions at a moderate computational cost.
Louise Mimeau, Annika Künne, Alexandre Devers, Flora Branger, Sven Kralisch, Claire Lauvernet, Jean-Philippe Vidal, Núria Bonada, Zoltán Csabai, Heikki Mykrä, Petr Pařil, Luka Polović, and Thibault Datry
Hydrol. Earth Syst. Sci., 29, 1615–1636, https://doi.org/10.5194/hess-29-1615-2025, https://doi.org/10.5194/hess-29-1615-2025, 2025
Short summary
Short summary
Our study projects how climate change will affect the drying of river segments and stream networks in Europe, using advanced modelling techniques to assess changes in six river networks across diverse ecoregions. We found that drying events will become more frequent and intense and will start earlier or last longer, potentially turning some river sections from perennial to intermittent. The results are valuable for river ecologists for evaluating the ecological health of river ecosystem.
Tariq Aziz, Steven K. Frey, David R. Lapen, Susan Preston, Hazen A. J. Russell, Omar Khader, Andre R. Erler, and Edward A. Sudicky
Hydrol. Earth Syst. Sci., 29, 1549–1568, https://doi.org/10.5194/hess-29-1549-2025, https://doi.org/10.5194/hess-29-1549-2025, 2025
Short summary
Short summary
This study determines the value of subsurface water for ecosystem services' supply in an agricultural watershed in Ontario, Canada. Using a fully integrated water model and an economic valuation approach, the research highlights subsurface water's critical role in maintaining watershed ecosystem services. The study informs on the sustainable use of subsurface water and introduces a new method for managing watershed ecosystem services.
Eduardo Acuña Espinoza, Ralf Loritz, Frederik Kratzert, Daniel Klotz, Martin Gauch, Manuel Álvarez Chaves, and Uwe Ehret
Hydrol. Earth Syst. Sci., 29, 1277–1294, https://doi.org/10.5194/hess-29-1277-2025, https://doi.org/10.5194/hess-29-1277-2025, 2025
Short summary
Short summary
Data-driven techniques have shown the potential to outperform process-based models in rainfall–runoff simulations. Hybrid models, combining both approaches, aim to enhance accuracy and maintain interpretability. Expanding the set of test cases to evaluate hybrid models under different conditions, we test their generalization capabilities for extreme hydrological events.
Basil Kraft, Michael Schirmer, William H. Aeberhard, Massimiliano Zappa, Sonia I. Seneviratne, and Lukas Gudmundsson
Hydrol. Earth Syst. Sci., 29, 1061–1082, https://doi.org/10.5194/hess-29-1061-2025, https://doi.org/10.5194/hess-29-1061-2025, 2025
Short summary
Short summary
This study reconstructs daily runoff in Switzerland (1962–2023) using a deep-learning model, providing a spatially contiguous dataset on a medium-sized catchment grid. The model outperforms traditional hydrological methods, revealing shifts in Swiss water resources, including more frequent dry years and declining summer runoff. The reconstruction is publicly available.
Mengjiao Zhang, Yi Nan, and Fuqiang Tian
Hydrol. Earth Syst. Sci., 29, 1033–1060, https://doi.org/10.5194/hess-29-1033-2025, https://doi.org/10.5194/hess-29-1033-2025, 2025
Short summary
Short summary
Owing to differences in the existing published results, we conducted a detailed analysis of the runoff components and future trends in the Yarlung Tsangpo River basin and found that the contributions of snowmelt and glacier melt runoff to streamflow (both ~5 %) are limited and much lower than previous results. The streamflow in this area will continuously increase in the future, but the overestimated contribution of glacier melt could lead to an underestimation of this increasing trend.
Moritz Maximilian Heuer, Hadysa Mohajerani, and Markus Christian Casper
EGUsphere, https://doi.org/10.5194/egusphere-2025-636, https://doi.org/10.5194/egusphere-2025-636, 2025
Short summary
Short summary
This study presents a calibration approach for water balance models. The different calibration steps aim at calibrating different hydrological processes: evapotranspiration, the runoff partitioning into surface runoff, interflow and groundwater recharge, as well as the groundwater behaviour. This allows for selection of a model parameterisation that correctly predicts the discharge at catchment outlet and simultaneously correctly depicts the underlying hydrological processes.
Tian Lan, Tongfang Li, Hongbo Zhang, Jiefeng Wu, Yongqin David Chen, and Chong-Yu Xu
Hydrol. Earth Syst. Sci., 29, 903–924, https://doi.org/10.5194/hess-29-903-2025, https://doi.org/10.5194/hess-29-903-2025, 2025
Short summary
Short summary
This study develops an integrated framework based on the novel Driving index for changes in Precipitation–Runoff Relationships (DPRR) to explore the controlling changes in precipitation–runoff relationships in non-stationary environments. According to the quantitative results of the candidate driving factors, the possible process explanations for changes in the precipitation–runoff relationships are deduced. The main contribution offers a comprehensive understanding of hydrological processes.
Everett Snieder and Usman T. Khan
Hydrol. Earth Syst. Sci., 29, 785–798, https://doi.org/10.5194/hess-29-785-2025, https://doi.org/10.5194/hess-29-785-2025, 2025
Short summary
Short summary
Improving the accuracy of flood forecasts is paramount to minimising flood damage. Machine learning (ML) models are increasingly being applied for flood forecasting. Such models are typically trained on large historic hydrometeorological datasets. In this work, we evaluate methods for selecting training datasets that maximise the spatio-temporal diversity of the represented hydrological processes. Empirical results showcase the importance of hydrological diversity in training ML models.
Joško Trošelj and Naota Hanasaki
Hydrol. Earth Syst. Sci., 29, 753–766, https://doi.org/10.5194/hess-29-753-2025, https://doi.org/10.5194/hess-29-753-2025, 2025
Short summary
Short summary
This study presents the first distributed hydrological simulation which confirms claims raised by historians that the eastward diversion project of the Tone River in Japan was conducted 4 centuries ago to increase low flows and subsequent travelling possibilities surrounding the capital, Edo (Tokyo), using inland navigation. We showed that great steps forward can be made for improving quality of life with small human engineering waterworks and small interventions in the regime of natural flows.
Léonard Santos, Vazken Andréassian, Torben O. Sonnenborg, Göran Lindström, Alban de Lavenne, Charles Perrin, Lila Collet, and Guillaume Thirel
Hydrol. Earth Syst. Sci., 29, 683–700, https://doi.org/10.5194/hess-29-683-2025, https://doi.org/10.5194/hess-29-683-2025, 2025
Short summary
Short summary
This work investigates how hydrological models are transferred to a period in which climate conditions are different to the ones of the period in which they were set up. The robustness assessment test built to detect dependencies between model error and climatic drivers was applied to three hydrological models in 352 catchments in Denmark, France and Sweden. Potential issues are seen in a significant number of catchments for the models, even though the catchments differ for each model.
Xudong Zheng, Dengfeng Liu, Shengzhi Huang, Hao Wang, and Xianmeng Meng
Hydrol. Earth Syst. Sci., 29, 627–653, https://doi.org/10.5194/hess-29-627-2025, https://doi.org/10.5194/hess-29-627-2025, 2025
Short summary
Short summary
Water budget non-closure is a widespread phenomenon among multisource datasets which undermines the robustness of hydrological inferences. This study proposes a Multisource Dataset Correction Framework grounded in Physical Hydrological Process Modelling to enhance water budget closure, termed PHPM-MDCF. We examined the efficiency and robustness of the framework using the CAMELS dataset and achieved an average reduction of 49 % in total water budget residuals across 475 CONUS basins.
Elena Macdonald, Bruno Merz, Viet Dung Nguyen, and Sergiy Vorogushyn
Hydrol. Earth Syst. Sci., 29, 447–463, https://doi.org/10.5194/hess-29-447-2025, https://doi.org/10.5194/hess-29-447-2025, 2025
Short summary
Short summary
Flood peak distributions indicate how likely the occurrence of an extreme flood is at a certain river. If the distribution has a so-called heavy tail, extreme floods are more likely than might be anticipated. We find heavier tails in small catchments compared to large catchments, and spatially variable rainfall leads to a lower occurrence probability of extreme floods. Spatially variable runoff does not show effects. The results can improve estimations of probabilities of extreme floods.
Ngo Nghi Truyen Huynh, Pierre-André Garambois, Benjamin Renard, François Colleoni, Jérôme Monnier, and Hélène Roux
EGUsphere, https://doi.org/10.5194/egusphere-2024-3665, https://doi.org/10.5194/egusphere-2024-3665, 2025
Short summary
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 development of learnable and interpretable process-based models.
Junfu Gong, Xingwen Liu, Cheng Yao, Zhijia Li, Albrecht H. Weerts, Qiaoling Li, Satish Bastola, Yingchun Huang, and Junzeng Xu
Hydrol. Earth Syst. Sci., 29, 335–360, https://doi.org/10.5194/hess-29-335-2025, https://doi.org/10.5194/hess-29-335-2025, 2025
Short summary
Short summary
Our study introduces a new method to improve flood forecasting by combining soil moisture and streamflow data using an advanced data assimilation technique. By integrating field and reanalysis soil moisture data and assimilating this with streamflow measurements, we aim to enhance the accuracy of flood predictions. This approach reduces the accumulation of past errors in the initial conditions at the start of the forecast, helping to better prepare for and respond to floods.
Jordy Salmon-Monviola, Ophélie Fovet, and Markus Hrachowitz
Hydrol. Earth Syst. Sci., 29, 127–158, https://doi.org/10.5194/hess-29-127-2025, https://doi.org/10.5194/hess-29-127-2025, 2025
Short summary
Short summary
To increase the predictive power of hydrological models, it is necessary to improve their consistency, i.e. their physical realism, which is measured by the ability of the model to reproduce observed system dynamics. Using a model to represent the dynamics of water and nitrate and dissolved organic carbon concentrations in an agricultural catchment, we showed that using solute-concentration data for calibration is useful to improve the hydrological consistency of the model.
Parthkumar A. Modi, Jared C. Carbone, Keith S. Jennings, Hannah Kamen, Joseph R. Kasprzyk, Bill Szafranski, Cameron W. Wobus, and Ben Livneh
EGUsphere, https://doi.org/10.5194/egusphere-2024-4046, https://doi.org/10.5194/egusphere-2024-4046, 2025
Short summary
Short summary
This study shows that in unmanaged snow-dominated basins, high forecast accuracy doesn’t always lead to high economic value, especially during extreme conditions like droughts. It highlights how irregular errors in modern forecasting systems weaken the connection between accuracy and value. These findings call for forecast evaluations to focus not only on accuracy but also on economic impacts, providing valuable guidance for better water resource management under uncertainty.
Benjamin Guillaume, Adrien Michez, and Aurore Degré
EGUsphere, https://doi.org/10.5194/egusphere-2024-3978, https://doi.org/10.5194/egusphere-2024-3978, 2025
Short summary
Short summary
Nature-based solutions (NbS) can mitigate floods and agricultural droughts by enhancing soil health and restoring hydrological cycles. This study highlights that leveraging soil diversity is key to optimizing NbS performance.
Haley A. Canham, Belize Lane, Colin B. Phillips, and Brendan P. Murphy
Hydrol. Earth Syst. Sci., 29, 27–43, https://doi.org/10.5194/hess-29-27-2025, https://doi.org/10.5194/hess-29-27-2025, 2025
Short summary
Short summary
The influence of watershed disturbances has proved challenging to disentangle from natural streamflow variability. This study evaluates the influence of time-varying hydrologic controls on rainfall–runoff in undisturbed and wildfire-disturbed watersheds using a novel time-series event separation method. Across watersheds, water year type and season influenced rainfall–runoff patterns. Accounting for these controls enabled clearer isolation of wildfire effects.
Fabian Merk, Timo Schaffhauser, Faizan Anwar, Ye Tuo, Jean-Martial Cohard, and Markus Disse
Hydrol. Earth Syst. Sci., 28, 5511–5539, https://doi.org/10.5194/hess-28-5511-2024, https://doi.org/10.5194/hess-28-5511-2024, 2024
Short summary
Short summary
Evapotranspiration (ET) is computed from the vegetation (plant transpiration) and soil (soil evaporation). In western Africa, plant transpiration correlates with vegetation growth. Vegetation is often represented using the leaf area index (LAI). In this study, we evaluate the importance of the LAI for ET calculation. We take a close look at this interaction and highlight its relevance. Our work contributes to the understanding of terrestrial water cycle processes .
Eshrat Fatima, Rohini Kumar, Sabine Attinger, Maren Kaluza, Oldrich Rakovec, Corinna Rebmann, Rafael Rosolem, Sascha E. Oswald, Luis Samaniego, Steffen Zacharias, and Martin Schrön
Hydrol. Earth Syst. Sci., 28, 5419–5441, https://doi.org/10.5194/hess-28-5419-2024, https://doi.org/10.5194/hess-28-5419-2024, 2024
Short summary
Short summary
This study establishes a framework to incorporate cosmic-ray neutron measurements into the mesoscale Hydrological Model (mHM). We evaluate different approaches to estimate neutron counts within the mHM using the Desilets equation, with uniformly and non-uniformly weighted average soil moisture, and the physically based code COSMIC. The data improved not only soil moisture simulations but also the parameterisation of evapotranspiration in the model.
Laia Estrada, Xavier Garcia, Joan Saló-Grau, Rafael Marcé, Antoni Munné, and Vicenç Acuña
Hydrol. Earth Syst. Sci., 28, 5353–5373, https://doi.org/10.5194/hess-28-5353-2024, https://doi.org/10.5194/hess-28-5353-2024, 2024
Short summary
Short summary
Hydrological modelling is a powerful tool to support decision-making. We assessed spatio-temporal patterns and trends of streamflow for 2001–2022 with a hydrological model, integrating stakeholder expert knowledge on management operations. The results provide insight into how climate change and anthropogenic pressures affect water resources availability in regions vulnerable to water scarcity, thus raising the need for sustainable management practices and integrated hydrological modelling.
Patricio Yeste, Matilde García-Valdecasas Ojeda, Sonia R. Gámiz-Fortis, Yolanda Castro-Díez, Axel Bronstert, and María Jesús Esteban-Parra
Hydrol. Earth Syst. Sci., 28, 5331–5352, https://doi.org/10.5194/hess-28-5331-2024, https://doi.org/10.5194/hess-28-5331-2024, 2024
Short summary
Short summary
Integrating streamflow and evaporation data can help improve the physical realism of hydrologic models. We investigate the capabilities of the Variable Infiltration Capacity (VIC) to reproduce both hydrologic variables for 189 headwater located in Spain. Results from sensitivity analyses indicate that adding two vegetation parameters is enough to improve the representation of evaporation and that the performance of VIC exceeded that of the largest modelling effort currently available in Spain.
Daniel T. Myers, David Jones, Diana Oviedo-Vargas, John Paul Schmit, Darren L. Ficklin, and Xuesong Zhang
Hydrol. Earth Syst. Sci., 28, 5295–5310, https://doi.org/10.5194/hess-28-5295-2024, https://doi.org/10.5194/hess-28-5295-2024, 2024
Short summary
Short summary
We studied how streamflow and water quality models respond to land cover data collected by satellites during the growing season versus the non-growing season. The land cover data showed more trees during the growing season and more built areas during the non-growing season. We next found that the use of non-growing season data resulted in a higher modeled nutrient export to streams. Knowledge of these sensitivities would be particularly important when models inform water resource management.
Hatice Türk, Christine Stumpp, Markus Hrachowitz, Karsten Schulz, Peter Strauss, Günter Blöschl, and Michael Stockinger
Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2024-359, https://doi.org/10.5194/hess-2024-359, 2024
Revised manuscript accepted for HESS
Short summary
Short summary
Using advances in transit time estimation and tracer data, we tested if fast-flow transit times are controlled solely by soil moisture or are also controlled by precipitation intensity. We used soil moisture-dependent and precipitation intensity-conditional transfer functions. We showed that significant portion of event water bypasses the soil matrix through fast flow paths (overland flow, tile drains, preferential flow paths) in dry soil conditions for both low and high-intensity precipitation.
Kevin R. Shook, Paul H. Whitfield, Christopher Spence, and John W. Pomeroy
Hydrol. Earth Syst. Sci., 28, 5173–5192, https://doi.org/10.5194/hess-28-5173-2024, https://doi.org/10.5194/hess-28-5173-2024, 2024
Short summary
Short summary
Recent studies suggest that the velocities of water running off landscapes in the Canadian Prairies may be much smaller than generally assumed. Analyses of historical flows for 23 basins in central Alberta show that many of the rivers responded more slowly and that the flows are much slower than would be estimated from equations developed elsewhere. The effects of slow flow velocities on the development of hydrological models of the region are discussed, as are the possible causes.
Alberto Bassi, Marvin Höge, Antonietta Mira, Fabrizio Fenicia, and Carlo Albert
Hydrol. Earth Syst. Sci., 28, 4971–4988, https://doi.org/10.5194/hess-28-4971-2024, https://doi.org/10.5194/hess-28-4971-2024, 2024
Short summary
Short summary
The goal is to remove the impact of meteorological drivers in order to uncover the unique landscape fingerprints of a catchment from streamflow data. Our results reveal an optimal two-feature summary for most catchments, with a third feature associated with aridity and intermittent flow that is needed for challenging cases. Baseflow index, aridity, and soil or vegetation attributes strongly correlate with learnt features, indicating their importance for streamflow prediction.
Sylvain Payraudeau, Pablo Alvarez-Zaldivar, Paul van Dijk, and Gwenaël Imfeld
EGUsphere, https://doi.org/10.5194/egusphere-2024-2840, https://doi.org/10.5194/egusphere-2024-2840, 2024
Short summary
Short summary
Our study focuses on the rising concern of pesticides damaging aquatic ecosystems, which puts drinking water, the environment, and human health at risk. We provided more accurate estimates of how pesticides break down and spread in small water systems, helping to improve pesticide management practices. By using unique chemical markers in our analysis, we enhanced the accuracy of our predictions, offering important insights for better protection of water sources and natural ecosystems.
Guillaume Thirel, Léonard Santos, Olivier Delaigue, and Charles Perrin
Hydrol. Earth Syst. Sci., 28, 4837–4860, https://doi.org/10.5194/hess-28-4837-2024, https://doi.org/10.5194/hess-28-4837-2024, 2024
Short summary
Short summary
We discuss how mathematical transformations impact calibrated hydrological model simulations. We assess how 11 transformations behave over the complete range of streamflows. Extreme transformations lead to models that are specialized for extreme streamflows but show poor performance outside the range of targeted streamflows and are less robust. We show that no a priori assumption about transformations can be taken as warranted.
Vishal Thakur, Yannis Markonis, Rohini Kumar, Johanna Ruth Thomson, Mijael Rodrigo Vargas Godoy, Martin Hanel, and Oldrich Rakovec
Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2024-341, https://doi.org/10.5194/hess-2024-341, 2024
Revised manuscript accepted for HESS
Short summary
Short summary
Understanding the changes in water movement in earth is crucial for everyone. To quantify this water movement there are several techniques. We examined how different methods of estimating evaporation impact predictions of various types of water movement across Europe. We found that, while these methods generally agree on whether changes are increasing or decreasing, they differ in magnitude. This means selecting the right evaporation method is crucial for accurate predictions of water movement.
Robert Hull, Elena Leonarduzzi, Luis De La Fuente, Hoang Viet Tran, Andrew Bennett, Peter Melchior, Reed M. Maxwell, and Laura E. Condon
Hydrol. Earth Syst. Sci., 28, 4685–4713, https://doi.org/10.5194/hess-28-4685-2024, https://doi.org/10.5194/hess-28-4685-2024, 2024
Short summary
Short summary
Large-scale hydrologic simulators are a needed tool to explore complex watershed processes and how they may evolve with a changing climate. However, calibrating them can be difficult because they are costly to run and have many unknown parameters. We implement a state-of-the-art approach to model calibration using neural networks with a set of experiments based on streamflow in the upper Colorado River basin.
Nan Wu, Ke Zhang, Amir Naghibi, Hossein Hashemi, Zhongrui Ning, Qinuo Zhang, Xuejun Yi, Haijun Wang, Wei Liu, Wei Gao, and Jerker Jarsjö
Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2024-324, https://doi.org/10.5194/hess-2024-324, 2024
Revised manuscript accepted for HESS
Short summary
Short summary
The hydrology of cold regions in the human population is poorly understood due to complex motion and limited data, hindering streamflow analysis. Using existing models, we compared runoff from an extended model with snowmelt and frozen ground, validating its reliability and integration. This study focuses on the effects of snowmelt and frozen ground on runoff, affecting precipitation type, surface-groundwater partitioning, and evapotranspiration.
Jari-Pekka Nousu, Kersti Leppä, Hannu Marttila, Pertti Ala-aho, Giulia Mazzotti, Terhikki Manninen, Mika Korkiakoski, Mika Aurela, Annalea Lohila, and Samuli Launiainen
Hydrol. Earth Syst. Sci., 28, 4643–4666, https://doi.org/10.5194/hess-28-4643-2024, https://doi.org/10.5194/hess-28-4643-2024, 2024
Short summary
Short summary
We used hydrological models, field measurements, and satellite-based data to study the soil moisture dynamics in a subarctic catchment. The role of groundwater was studied with different ways to model the groundwater dynamics and via comparisons to the observational data. The choice of groundwater model was shown to have a strong impact, and representation of lateral flow was important to capture wet soil conditions. Our results provide insights for ecohydrological studies in boreal regions.
Nienke Tempel, Laurène Bouaziz, Riccardo Taormina, Ellis van Noppen, Jasper Stam, Eric Sprokkereef, and Markus Hrachowitz
Hydrol. Earth Syst. Sci., 28, 4577–4597, https://doi.org/10.5194/hess-28-4577-2024, https://doi.org/10.5194/hess-28-4577-2024, 2024
Short summary
Short summary
This study explores the impact of climatic variability on root zone water storage capacities and, thus, on hydrological predictions. Analysing data from 286 areas in Europe and the US, we found that, despite some variations in root zone storage capacity due to changing climatic conditions over multiple decades, these changes are generally minor and have a limited effect on water storage and river flow predictions.
Bu Li, Ting Sun, Fuqiang Tian, Mahmut Tudaji, Li Qin, and Guangheng Ni
Hydrol. Earth Syst. Sci., 28, 4521–4538, https://doi.org/10.5194/hess-28-4521-2024, https://doi.org/10.5194/hess-28-4521-2024, 2024
Short summary
Short summary
This paper developed hybrid semi-distributed hydrological models by employing a process-based model as the backbone and utilizing deep learning to parameterize and replace internal modules. The main contribution is to provide a high-performance tool enriched with explicit hydrological knowledge for hydrological prediction and to improve understanding about the hydrological sensitivities to climate change in large alpine basins.
Francesco Zignol, William Lidberg, Caroline Greiser, Johannes Larson, Raúl Hoffrén, and Anneli M. Ågren
EGUsphere, https://doi.org/10.5194/egusphere-2024-2909, https://doi.org/10.5194/egusphere-2024-2909, 2024
Short summary
Short summary
We investigated the factors influencing soil moisture variations across a boreal forest catchment in northern Sweden, where data is usually scarce. We found that soil moisture is shaped by topographical features, vegetation and soil characteristics, and weather conditions. The insights presented in this study will help improve models that predict soil moisture over space and time, which is crucial for forest management and nature conservation in the face of climate change and biodiversity loss.
Maria Elenius, Charlotta Pers, Sara Schützer, and Berit Arheimer
Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2024-271, https://doi.org/10.5194/hess-2024-271, 2024
Revised manuscript accepted for HESS
Short summary
Short summary
Simulations of peatland rewetting in Sweden under various conditions of climate, local hydrology and rewetting practices showed insubstantial changes in landscape flow extremes due to mixing with runoff from various landcover. The impact on local hydrological extremes are governed by groundwater levels prior to rewetting and reduced tree cover, hence wetland allocation and management practices are crucial if the purpose is to reduce flow extremes in peatland streams.
Dan Elhanati, Nadine Goeppert, and Brian Berkowitz
Hydrol. Earth Syst. Sci., 28, 4239–4249, https://doi.org/10.5194/hess-28-4239-2024, https://doi.org/10.5194/hess-28-4239-2024, 2024
Short summary
Short summary
A continuous time random walk framework was developed to allow modeling of a karst aquifer discharge response to measured rainfall. The application of the numerical model yielded robust fits between modeled and measured discharge values, especially for the distinctive long tails found during recession times. The findings shed light on the interplay of slow and fast flow in the karst system and establish the application of the model for simulating flow and transport in such systems.
Cited articles
Addor, N., Newman, A. J., Mizukami, N., and Clark, M. P.: The CAMELS data set: catchment attributes and meteorology for large-sample studies, Hydrol. Earth Syst. Sci., 21, 5293–5313, https://doi.org/10.5194/hess-21-5293-2017, 2017a. a
Addor, N., Newman, A., Mizukami, M., and Clark, M. P.: Catchment attributes for large-sample studies [data set], Boulder, CO, UCAR/NCAR, https://doi.org/10.5065/D6G73C3Q (last access: 14 April 2021), 2017. a
Addor, N., Nearing, G., Prieto, C., Newman, A. J., Le Vine, N., and Clark,
M. P.: A Ranking of Hydrological Signatures Based on Their Predictability in
Space, Water Resour. Res., 54, 8792–8812, https://doi.org/10.1029/2018WR022606,
2018. a, b
Araya, I. A., Valle, C., and Allende, H.: A Multi-Scale Model based on the
Long Short-Term Memory for day ahead hourly wind speed forecasting, Pattern
Recognition Letters, 136, 333–340, https://doi.org/10.1016/j.patrec.2019.10.011, 2019. a
Bengio, Y., Simard, P., and Frasconi, P.: Learning long-term dependencies with
gradient descent is difficult, IEEE Transactions on Neural Networks, 5,
157–166, https://doi.org/10.1109/72.279181, 1994. a
Chung, J., Ahn, S., and Bengio, Y.: Hierarchical Multiscale Recurrent Neural
Networks, arXiv preprint, arXiv:1609.01704, 2016. a
Clausen, B. and Biggs, B. J. F.: Flow variables for ecological studies in
temperate streams: groupings based on covariance, J. Hydrol., 237,
184–197, https://doi.org/10.1016/S0022-1694(00)00306-1, 2000. a, b
Court, A.: Measures of streamflow timing, J. Geophys. Res.
(1896–1977), 67, 4335–4339, https://doi.org/10.1029/JZ067i011p04335, 1962. a
Frame, J., Nearing, G., Kratzert, F., and Rahman, M.: Post processing the
U.S. National Water Model with a Long Short-Term Memory network, EarthArXiv,
https://doi.org/10.31223/osf.io/4xhac, 2020. a
Gauch, M.: Code for “Rainfall-Runoff Prediction at Multiple Timescales with a Single Long Short-Term Memory Network”, Zenodo [code], https://doi.org/10.5281/zenodo.4687991 (last access: 14 April 2021), 2021. a
Gauch, M. and Lin, J.: A Data Scientist's Guide to Streamflow Prediction, arXiv
preprint, arXiv:2006.12975, 2020. a
Gauch, M., Kratzert, F., Klotz, D., Nearing, G., Lin, J., and Hochreiter, S.: Models and Predictions for “Rainfall-Runoff Prediction at Multiple Timescales with a Single Long Short-Term Memory Network” [data set], Zenodo, https://doi.org/10.5281/zenodo.4095485, 2020a. a
Gauch, M., Kratzert, F., Klotz, D., Nearing, G., Lin, J., and Hochreiter, S.: Data for “Rainfall-Runoff Prediction at Multiple Timescales with a Single Long Short-Term Memory Network” [data set], Zenodo, https://doi.org/10.5281/zenodo.4072701, 2020b. a
Gers, F. A., Schmidhuber, J., and Cummins, F.: Learning to forget: continual
prediction with LSTM, IET Conference Proceedings, pp. 850–855, 1999. a
Gochis, D. J., Barlage, M., Cabell, R., Casali, M., Dugger, A., FitzGerald, K.,
McAllister, M., McCreight, J., RafieeiNasab, A., Read, L., Sampson, K.,
Yates, D., and Zhang, Y.: The WRF-Hydro® modeling system
technical description, available at:
https://ral.ucar.edu/sites/default/files/public/projects/Technical%20Description%20%26amp%3B%20User%20Guides/wrfhydrov511technicaldescription.pdf (last access: 14 April 2021),
2020. a
Graves, A., Fernández, S., and Schmidhuber, J.: Multi-dimensional Recurrent
Neural Networks, in: Artificial Neural Networks – ICANN 2007, edited by:
de Sá, J. M., Alexandre, L. A., Duch, W., and Mandic, D., pp. 549–558,
Springer Berlin Heidelberg, Berlin, Heidelberg, 2007. a
Greff, K., Srivastava, R. K., Koutník, J., Steunebrink, B. R., and
Schmidhuber, J.: LSTM: A Search Space Odyssey, IEEE Transactions on Neural
Networks and Learning Systems, 28, 2222–2232,
https://doi.org/10.1109/TNNLS.2016.2582924, 2017. a
Gupta, H. V., Kling, H., Yilmaz, K. K., and Martinez, G. F.: Decomposition of
the mean squared error and NSE performance criteria: implications for
improving hydrological modelling, J. Hydrol., 377, 80–91,
https://doi.org/10.1016/j.jhydrol.2009.08.003, 2009. a, b, c
He, K., Zhang, X., Ren, S., and Sun, J.: Deep Residual Learning for Image
Recognition, in: Proceedings of the IEEE Conference on Computer Vision and
Pattern Recognition (CVPR), June 2016, Las Vegas, Nevada, 770–778, 2016. a
Hochreiter, S. and Schmidhuber, J.: Long Short-Term Memory, Neural
Computation, 9, 1735–1780, https://doi.org/10.1162/neco.1997.9.8.1735, 1997. a, b
Hoedt, P.-J., Kratzert, F., Klotz, D., Halmich, C., Holzleitner, M., Nearing,
G., Hochreiter, S., and Klambauer, G.: MC-LSTM: Mass-Conserving LSTM, available at: https://arxiv.org/abs/2101.05186,
2021. a
Jozefowicz, R., Zaremba, W., and Sutskever, I.: An Empirical Exploration of
Recurrent Network Architectures, in: Proceedings of the 32nd International
Conference on Machine Learning, edited by: Bach, F. and Blei, D., vol. 37 of
Proceedings of Machine Learning Research, pp. 2342–2350, PMLR,
Lille, France, 2015. a, b
Klotz, D., Kratzert, F., Gauch, M., Keefe Sampson, A., Brandstetter, J., Klambauer, G., Hochreiter, S., and Nearing, G.: Uncertainty Estimation with Deep Learning for Rainfall–Runoff Modelling, Hydrol. Earth Syst. Sci. Discuss. [preprint], https://doi.org/10.5194/hess-2021-154, in review, 2021. a
Koutník, J., Greff, K., Gomez, F., and Schmidhuber, J.: A Clockwork RNN,
arXiv preprint, arXiv:1402.3511, 2014. 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
Kratzert, F., Gauch, M., and Klotz, D.: NeuralHydrology Python Library, Zenodo [code], https://doi.org/10.5281/zenodo.4688003 (last access: 14 April 2021), 2020. a
Kratzert, F.: CAMELS Extended Maurer Forcing Data, HydroShare [data set], https://doi.org/10.4211/hs.17c896843cf940339c3c3496d0c1c077 (last access: 14 April 2021), 2019. a
Kratzert, F., Klotz, D., Shalev, G., Klambauer, G., Hochreiter, S., and Nearing, G.: Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets, Hydrol. Earth Syst. Sci., 23, 5089–5110, https://doi.org/10.5194/hess-23-5089-2019, 2019. a, b, c, d, e, f, g
Ladson, T. R., Brown, R., Neal, B., and Nathan, R.: A Standard Approach to
Baseflow Separation Using The Lyne and Hollick Filter, Australasian
J. Water Res., 17, 25–34, available at: https://www.tandfonline.com/doi/ref/10.7158/13241583.2013.11465417 (last access: 14 April 2021),
2013. a
Lechner, M. and Hasani, R.: Learning Long-Term Dependencies in
Irregularly-Sampled Time Series, arXiv preprint, arXiv:2006.04418, 2020. a
Mozer, M.: Induction of Multiscale Temporal Structure, in: Advances in Neural
Information Processing Systems 4, edited by: Moody, J. E., Hanson, S. J., and
Lippmann, R., pp. 275–282, Morgan Kaufmann, 1991. a
Nash, J. E. and Sutcliffe, J. V.: River flow forecasting through conceptual
models part I – A discussion of principles, J. Hydrol., 10,
282–290, https://doi.org/10.1016/0022-1694(70)90255-6, 1970. a
Neil, D., Pfeiffer, M., and Liu, S.-C.: Phased LSTM: Accelerating Recurrent
Network Training for Long or Event-based Sequences, in: Advances in Neural
Information Processing Systems 29, edited by: Lee, D. D., Sugiyama, M.,
Luxburg, U. V., Guyon, I., and Garnett, R., pp. 3882–3890, Curran
Associates, Inc., 2016. a
Newman, A., Sampson, K., Clark, M. P., Bock, A., Viger, R., and Blodgett, D.: A
large-sample watershed-scale hydrometeorological dataset for the contiguous
USA, UCAR/NCAR [data set], https://doi.org/10.5065/d6mw2f4d, 2014. a, b
Newman, A., Sampson, K., Clark, M. P., Bock, A., Viger, R., and Blodgett, D.: CAMELS: Catchment Attributes and Meteorology for Large-sample Studies [data set], Boulder, CO, UCAR/NCAR, https://ral.ucar.edu/sites/default/files/public/product-tool/camels-catchment-attributes-and-meteorology-for-large-sample-studies-dataset-downloads/basin_timeseries_v1p2_metForcing_obsFlow.zip (last access: 14 April 2021), 2014. a
Newman, A., Mizukami, N., Clark, M. P., Wood, A. W., Nijssen, B., and Nearing,
G.: Benchmarking of a Physically Based Hydrologic Model, J.
Hydrometeorol., 18, 2215–2225, https://doi.org/10.1175/JHM-D-16-0284.1, 2017. a
Olah, C.: Understanding LSTM Networks, colah's blog, available at:
https://colah.github.io/posts/2015-08-Understanding-LSTMs/ (last access: 14 April 2021), 2015. a
Olden, J. D. and Poff, N. L.: Redundancy and the choice of hydrologic indices
for characterizing streamflow regimes, River Res. Appl., 19,
101–121, https://doi.org/10.1002/rra.700, 2003. a, b
Salas, F. R., Somos-Valenzuela, M. A., Dugger, A., Maidment, D. R., Gochis,
D. J., David, C. H., Yu, W., Ding, D., Clark, E. P., and Noman, N.: Towards
Real-Time Continental Scale Streamflow Simulation in Continuous and Discrete
Space, J. Am. Water Resour. Assoc., 54, 7–27,
https://doi.org/10.1111/1752-1688.12586, 2018. a, b
Sankarasubramanian, A., Vogel, R. M., and Limbrunner, J. F.: Climate elasticity
of streamflow in the United States, Water Resour. Res., 37,
1771–1781, https://doi.org/10.1029/2000WR900330, 2001. a
Sawicz, K., Wagener, T., Sivapalan, M., Troch, P. A., and Carrillo, G.: Catchment classification: empirical analysis of hydrologic similarity based on catchment function in the eastern USA, Hydrol. Earth Syst. Sci., 15, 2895–2911, https://doi.org/10.5194/hess-15-2895-2011, 2011. a, b
Schmidhuber, J.: Neural Sequence Chunkers, Tech. rep. FKI 148 91, Technische Universität München, Institut für Informatik, 1991. a
United States Geological Survey: USGS Instantaneous Values Web Service, available at:
https://waterservices.usgs.gov/rest/IV-Service.html (last access:
15 October 2020), 2021. a
Westerberg, I. K. and McMillan, H. K.: Uncertainty in hydrological signatures, Hydrol. Earth Syst. Sci., 19, 3951–3968, https://doi.org/10.5194/hess-19-3951-2015, 2015. a, b, c, d
Xia, Y., Mitchell, K., Ek, M., Sheffield, J., Cosgrove, B., Wood, E., Luo, L.,
Alonge, C., Wei, H., Meng, J., Livneh, B., Lettenmaier, D., Koren, V., Duan,
Q., Mo, K., Fan, Y., and Mocko, D.: Continental-scale water and energy flux
analysis and validation for the North American Land Data Assimilation
System project phase 2 (NLDAS-2): 1. Intercomparison and application of
model products, J. Geophys. Res.-Atmos., 117, D03109,
https://doi.org/10.1029/2011JD016048, 2012. a, b
Yilmaz, K. K., Gupta, H. V., and Wagener, T.: A process-based diagnostic
approach to model evaluation: Application to the NWS distributed hydrologic
model, Water Resour. Res., 44, W09417, https://doi.org/10.1029/2007WR006716, 2008. a, b, c
Zamir, A. R., Sax, A., Cheerla, N., Suri, R., Cao, Z., Malik, J., and Guibas,
L. J.: Robust Learning Through Cross-Task Consistency, in: The IEEE/CVF
Conference on Computer Vision and Pattern Recognition (CVPR), June 2020 (online), 11197–11206, 2020. a
Short summary
We present multi-timescale Short-Term Memory (MTS-LSTM), a machine learning approach that predicts discharge at multiple timescales within one model. MTS-LSTM is significantly more accurate than the US National Water Model and computationally more efficient than an individual LSTM model per timescale. Further, MTS-LSTM can process different input variables at different timescales, which is important as the lead time of meteorological forecasts often depends on their temporal resolution.
We present multi-timescale Short-Term Memory (MTS-LSTM), a machine learning approach that...
