Articles | Volume 23, issue 12
https://doi.org/10.5194/hess-23-5089-2019
© Author(s) 2019. 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-23-5089-2019
© Author(s) 2019. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
17 Dec 2019
Research article |
| 17 Dec 2019
| 17 Dec 2019
Towards learning universal, regional, and local hydrological behaviors via machine learning applied to large-sample datasets
LIT AI Lab & Institute for Machine Learning, Johannes Kepler University Linz, Linz, Austria
Daniel Klotz
LIT AI Lab & Institute for Machine Learning, Johannes Kepler University Linz, Linz, Austria
Guy Shalev
Google Research, Tel Aviv, Israel
Günter Klambauer
LIT AI Lab & Institute for Machine Learning, Johannes Kepler University Linz, Linz, Austria
Sepp Hochreiter
LIT AI Lab & Institute for Machine Learning, Johannes Kepler University Linz, Linz, Austria
Grey Nearing
Department of Geological Sciences, University of Alabama, Tuscaloosa, AL, USA
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
This preprint is open for discussion and under review for Hydrology and Earth System Sciences (HESS).
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.
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 under review 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.
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.
Martin Gauch, Frederik Kratzert, Daniel Klotz, Grey Nearing, Jimmy Lin, and Sepp Hochreiter
Hydrol. Earth Syst. Sci., 25, 2045–2062, https://doi.org/10.5194/hess-25-2045-2021, https://doi.org/10.5194/hess-25-2045-2021, 2021
Short summary
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.
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.
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
This preprint is open for discussion and under review for Hydrology and Earth System Sciences (HESS).
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
Preprint 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.
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 under review 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.
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.
Martin Gauch, Frederik Kratzert, Daniel Klotz, Grey Nearing, Jimmy Lin, and Sepp Hochreiter
Hydrol. Earth Syst. Sci., 25, 2045–2062, https://doi.org/10.5194/hess-25-2045-2021, https://doi.org/10.5194/hess-25-2045-2021, 2021
Short summary
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.
Jianxiu Qiu, Wade T. Crow, Jianzhi Dong, and Grey S. Nearing
Hydrol. Earth Syst. Sci., 24, 581–594, https://doi.org/10.5194/hess-24-581-2020, https://doi.org/10.5194/hess-24-581-2020, 2020
Short summary
Short summary
Accurately estimating coupling of evapotranspiration (ET) and soil water content (θ) at different depths is key to investigating land–atmosphere interaction. Here we examine whether the model can accurately represent surface θ (θs) versus ET coupling and vertically integrated θ (θv) versus ET coupling. We find that all models agree with observations that θs contains slightly more information with fPET than θv. In addition, an ET scheme is crucial for accurately estimating coupling of θ and ET.
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).
S. V. Kumar, C. D. Peters-Lidard, J. A. Santanello, R. H. Reichle, C. S. Draper, R. D. Koster, G. Nearing, and M. F. Jasinski
Hydrol. Earth Syst. Sci., 19, 4463–4478, https://doi.org/10.5194/hess-19-4463-2015, https://doi.org/10.5194/hess-19-4463-2015, 2015
Related subject area
Subject: Global hydrology | Techniques and Approaches: Modelling approaches
Mapping groundwater-dependent ecosystems using a high-resolution global groundwater model
Can large-scale tree cover change negate climate change impacts on future water availability?
Impact of runoff schemes on global flow discharge: a comprehensive analysis using the Noah-MP and CaMa-Flood models
The benefits and trade-offs of multi-variable calibration of the WaterGAP global hydrological model (WGHM) in the Ganges and Brahmaputra basins
Representation of a two-way coupled irrigation system in the Common Land Model
The effect of climate change on the simulated streamflow of six Canadian rivers based on the CanRCM4 regional climate model
Drivers of global irrigation expansion: the role of discrete global grid choice
Benchmarking historical performance and future projections from a global hydrologic model with a basin-scale model
Changes in mean evapotranspiration dominate groundwater recharge in semi-arid regions
Drought decreases streamflow response to precipitation especially in arid regions
Merging modelled and reported flood impacts in Europe in a combined flood event catalogue for 1950–2020
Future changes in seasonal drought in Australia
Global-scale evaluation of precipitation datasets for hydrological modelling
Toward Routing River Water in Land Surface Models with Recurrent Neural Networks
Influence of irrigation on root zone storage capacity estimation
River flow in the near future: a global perspective in the context of a high-emission climate change scenario
A high-resolution perspective of extreme rainfall and river flow under extreme climate change in Southeast Asia
Unveiling hydrological dynamics in data-scarce regions: experiences from the Ethiopian Rift Valley Lakes Basin
Technical note: Comparing three different methods for allocating river points to coarse-resolution hydrological modelling grid cells
Representing farmer irrigated crop area adaptation in a large-scale hydrological model
Combined impacts of climate and land-use change on future water resources in Africa
Deep learning for quality control of surface physiographic fields using satellite Earth observations
Global dryland aridity changes indicated by atmospheric, hydrological, and vegetation observations at meteorological stations
Root zone soil moisture in over 25 % of global land permanently beyond pre-industrial variability as early as 2050 without climate policy
Assessment of pluri-annual and decadal changes in terrestrial water storage predicted by global hydrological models in comparison with the GRACE satellite gravity mission
Improving the quantification of climate change hazards by hydrological models: a simple ensemble approach for considering the uncertain effect of vegetation response to climate change on potential evapotranspiration
Towards reducing the high cost of parameter sensitivity analysis in hydrologic modeling: a regional parameter sensitivity analysis approach
Point-scale multi-objective calibration of the Community Land Model (version 5.0) using in situ observations of water and energy fluxes and variables
Methodology for constructing a flood-hazard map for a future climate
Diagnosing modeling errors in global terrestrial water storage interannual variability
Hyper-resolution PCR-GLOBWB: opportunities and challenges from refining model spatial resolution to 1 km over the European continent
Poor correlation between large-scale environmental flow violations and freshwater biodiversity: implications for water resource management and the freshwater planetary boundary
Accuracy of five ground heat flux empirical simulation methods in the surface-energy-balance-based remote-sensing evapotranspiration models
Coupling a global glacier model to a global hydrological model prevents underestimation of glacier runoff
Revisiting large-scale interception patterns constrained by a synthesis of global experimental data
Investigating coastal backwater effects and flooding in the coastal zone using a global river transport model on an unstructured mesh
Using a long short-term memory (LSTM) neural network to boost river streamflow forecasts over the western United States
Quantifying overlapping and differing information of global precipitation for GCM forecasts and El Niño–Southern Oscillation
Globally widespread and increasing violations of environmental flow envelopes
Inundation prediction in tropical wetlands from JULES-CaMa-Flood global land surface simulations
Soil moisture estimation in South Asia via assimilation of SMAP retrievals
Toward hyper-resolution global hydrological models including human activities: application to Kyushu island, Japan
Towards hybrid modeling of the global hydrological cycle
The importance of vegetation in understanding terrestrial water storage variations
Large-scale sensitivities of groundwater and surface water to groundwater withdrawal
A hydrography upscaling method for scale-invariant parametrization of distributed hydrological models
A novel method to identify sub-seasonal clustering episodes of extreme precipitation events and their contributions to large accumulation periods
Bright and blind spots of water research in Latin America and the Caribbean
Land surface modeling over the Dry Chaco: the impact of model structures, and soil, vegetation and land cover parameters
Nonstationary weather and water extremes: a review of methods for their detection, attribution, and management
Nicole Gyakowah Otoo, Edwin H. Sutanudjaja, Michelle T. H. van Vliet, Aafke M. Schipper, and Marc F. P. Bierkens
Hydrol. Earth Syst. Sci., 29, 2153–2165, https://doi.org/10.5194/hess-29-2153-2025, https://doi.org/10.5194/hess-29-2153-2025, 2025
Short summary
Short summary
The contribution of groundwater to groundwater-dependent ecosystems (GDEs) is declining as a result of an increase in groundwater abstractions and climate change. This may lead to loss of habitat and biodiversity. This proposed framework enables the mapping and understanding of the temporal and spatial dynamics of GDEs on a large scale. The next step is to assess the global impacts of climate change and water use on GDE extent and health.
Freek Engel, Anne J. Hoek van Dijke, Caspar T. J. Roebroek, and Imme Benedict
Hydrol. Earth Syst. Sci., 29, 1895–1918, https://doi.org/10.5194/hess-29-1895-2025, https://doi.org/10.5194/hess-29-1895-2025, 2025
Short summary
Short summary
A warming climate alters the freshwater availability over land, and, due to related tree cover change and potential forestation, this availability can be further enhanced or negated. We find that large-scale change in tree cover may counteract climate-driven changes on a global scale, whereas, regionally, the climate and tree cover impacts can differ extensively. Current ecosystem restoration projects should account for the effects of (re-)forestation on (non-)local water availability.
Mohamed Hamitouche, Giorgia Fosser, Alessandro Anav, Cenlin He, and Tzu-Shun Lin
Hydrol. Earth Syst. Sci., 29, 1221–1240, https://doi.org/10.5194/hess-29-1221-2025, https://doi.org/10.5194/hess-29-1221-2025, 2025
Short summary
Short summary
This study evaluates how different methods of simulating runoff impact river flow predictions globally. By comparing seven approaches within the Noah-Multi-parameterisation (Noah-MP) land surface model, we found significant differences in accuracy, with some methods underestimating or overestimating runoff. The results are crucial for improving water resource management and flood prediction. Our work highlights the need for precise modelling to better prepare for climate-related challenges.
Howlader Mohammad Mehedi Hasan, Petra Döll, Seyed-Mohammad Hosseini-Moghari, Fabrice Papa, and Andreas Güntner
Hydrol. Earth Syst. Sci., 29, 567–596, https://doi.org/10.5194/hess-29-567-2025, https://doi.org/10.5194/hess-29-567-2025, 2025
Short summary
Short summary
We calibrate a global hydrological model using multiple observations to analyse the benefits and trade-offs of multi-variable calibration. We found such an approach to be very important for understanding the real-world system. However, some observations are very essential to the system, in particular, streamflow. We also showed uncertainties in the calibration results, which are often useful for making informed decisions. We emphasize considering observation uncertainty in model calibration.
Shulei Zhang, Hongbin Liang, Fang Li, Xingjie Lu, and Yongjiu Dai
EGUsphere, https://doi.org/10.5194/egusphere-2024-4093, https://doi.org/10.5194/egusphere-2024-4093, 2025
Short summary
Short summary
This study enhances irrigation modeling in the Common Land Model by capturing the full irrigation process, detailing water supplies from various sources, and enabling bidirectional coupling between water demand and supply. The proposed model accurately simulates irrigation water withdrawals, energy fluxes, river flow, and crop yields. It offers insights into irrigation-related climate impacts and water scarcity, contributing to sustainable water management and improved Earth system modeling.
Vivek K. Arora, Aranildo Lima, and Rajesh Shrestha
Hydrol. Earth Syst. Sci., 29, 291–312, https://doi.org/10.5194/hess-29-291-2025, https://doi.org/10.5194/hess-29-291-2025, 2025
Short summary
Short summary
This study presents a Canada-wide assessment of climate change impacts on the hydro-climatology of the region's major river basins. We find that precipitation, runoff, and temperature are all expected to increase over Canada in the future. The northerly Mackenzie and Yukon rivers are relatively less affected by climate change compared to the southerly Fraser and Columbia rivers, which are located in the milder northwestern Pacific region.
Sophie Wagner, Fabian Stenzel, Tobias Krueger, and Jana de Wiljes
Hydrol. Earth Syst. Sci., 28, 5049–5068, https://doi.org/10.5194/hess-28-5049-2024, https://doi.org/10.5194/hess-28-5049-2024, 2024
Short summary
Short summary
Statistical models that explain global irrigation rely on location-referenced data. Traditionally, a system based on longitude and latitude lines is chosen. However, this introduces bias to the analysis due to the Earth's curvature. We propose using a system based on hexagonal grid cells that allows for distortion-free representation of the data. We show that this increases the model's accuracy by 28 % and identify biophysical and socioeconomic drivers of historical global irrigation expansion.
Rajesh R. Shrestha, Alex J. Cannon, Sydney Hoffman, Marie Whibley, and Aranildo Lima
EGUsphere, https://doi.org/10.5194/egusphere-2024-3170, https://doi.org/10.5194/egusphere-2024-3170, 2024
Short summary
Short summary
We evaluate the historical performance and future projections from a global hydrologic model, the Community Water Model, against a watershed hydrologic model, the Variable Infiltration Capacity for the Liard River basin in Canada. Results from the two models are generally consistent at annual and monthly time scales, suggesting that a calibrated global hydrologic model can provide robust projections. We explain the differences in projections in terms of model uncertainties.
Tuvia Turkeltaub and Golan Bel
Hydrol. Earth Syst. Sci., 28, 4263–4274, https://doi.org/10.5194/hess-28-4263-2024, https://doi.org/10.5194/hess-28-4263-2024, 2024
Short summary
Short summary
Future climate projections suggest that climate change will impact groundwater recharge, with its exact effects being uncertain due to incomplete understanding of rainfall, evapotranspiration, and recharge relations. We studied the effects of changes in the average, spread, and frequency of extreme events of rainfall and evapotranspiration on groundwater recharge. We found that increasing or decreasing the potential evaporation has the most dominant effect on groundwater recharge.
Alessia Matanó, Raed Hamed, Manuela I. Brunner, Marlies H. Barendrecht, and Anne F. Van Loon
EGUsphere, https://doi.org/10.5194/egusphere-2024-2715, https://doi.org/10.5194/egusphere-2024-2715, 2024
Short summary
Short summary
Persistent droughts change how rivers respond to rainfall. Our study of over 5,000 catchments worldwide found that hydrological and soil moisture droughts decrease river flow response to rain, especially in arid regions, while vegetation decline slightly increases it. Snow-covered areas are more resilient due to stored water buffering changes. Droughts can also cause long-lasting changes, with short, intense droughts reducing river response to rainfall and prolonged droughts increasing it.
Dominik Paprotny, Belinda Rhein, Michalis I. Vousdoukas, Paweł Terefenko, Francesco Dottori, Simon Treu, Jakub Śledziowski, Luc Feyen, and Heidi Kreibich
Hydrol. Earth Syst. Sci., 28, 3983–4010, https://doi.org/10.5194/hess-28-3983-2024, https://doi.org/10.5194/hess-28-3983-2024, 2024
Short summary
Short summary
Long-term trends in flood losses are regulated by multiple factors, including climate variation, population and economic growth, land-use transitions, reservoir construction, and flood risk reduction measures. Here, we reconstruct the factual circumstances in which almost 15 000 potential riverine, coastal and compound floods in Europe occurred between 1950 and 2020. About 10 % of those events are reported to have caused significant socioeconomic impacts.
Anna M. Ukkola, Steven Thomas, Elisabeth Vogel, Ulrike Bende-Michl, Steven Siems, Vjekoslav Matic, and Wendy Sharples
EGUsphere, https://doi.org/10.31223/X56110, https://doi.org/10.31223/X56110, 2024
Short summary
Short summary
Future drought changes in Australia –the driest inhabited continent on Earth– have remained stubbornly uncertain. We assess future drought changes in Australia using projections from climate and hydrological models. We show an increasing probability of drought over highly-populated and agricultural regions of Australia in coming decades, suggesting potential impacts on agricultural activities, ecosystems and urban water supply.
Solomon H. Gebrechorkos, Julian Leyland, Simon J. Dadson, Sagy Cohen, Louise Slater, Michel Wortmann, Philip J. Ashworth, Georgina L. Bennett, Richard Boothroyd, Hannah Cloke, Pauline Delorme, Helen Griffith, Richard Hardy, Laurence Hawker, Stuart McLelland, Jeffrey Neal, Andrew Nicholas, Andrew J. Tatem, Ellie Vahidi, Yinxue Liu, Justin Sheffield, Daniel R. Parsons, and Stephen E. Darby
Hydrol. Earth Syst. Sci., 28, 3099–3118, https://doi.org/10.5194/hess-28-3099-2024, https://doi.org/10.5194/hess-28-3099-2024, 2024
Short summary
Short summary
This study evaluated six high-resolution global precipitation datasets for hydrological modelling. MSWEP and ERA5 showed better performance, but spatial variability was high. The findings highlight the importance of careful dataset selection for river discharge modelling due to the lack of a universally superior dataset. Further improvements in global precipitation data products are needed.
Mauricio Lima, Katherine Deck, Oliver R. A. Dunbar, and Tapio Schneider
EGUsphere, https://doi.org/10.48550/arXiv.2404.14212, https://doi.org/10.48550/arXiv.2404.14212, 2024
Short summary
Short summary
Machine learning is playing an increasingly important role in hydrological modeling. In this paper, we introduce an adaptation of existing machine learning models forecasting streamflow in river basins, redesigning them with the goal of integrating them into climate models. We demonstrate the effectiveness of our adapted model by showing that it outperforms a physics-based river model. These results motivate further studies of the use of machine learning based river models inside climate models.
Fransje van Oorschot, Ruud J. van der Ent, Andrea Alessandri, and Markus Hrachowitz
Hydrol. Earth Syst. Sci., 28, 2313–2328, https://doi.org/10.5194/hess-28-2313-2024, https://doi.org/10.5194/hess-28-2313-2024, 2024
Short summary
Short summary
Vegetation plays a crucial role in regulating the water cycle by transporting water from the subsurface to the atmosphere via roots; this transport depends on the extent of the root system. In this study, we quantified the effect of irrigation on roots at a global scale. Our results emphasize the importance of accounting for irrigation in estimating the vegetation root extent, which is essential to adequately represent the water cycle in hydrological and climate models.
Omar V. Müller, Patrick C. McGuire, Pier Luigi Vidale, and Ed Hawkins
Hydrol. Earth Syst. Sci., 28, 2179–2201, https://doi.org/10.5194/hess-28-2179-2024, https://doi.org/10.5194/hess-28-2179-2024, 2024
Short summary
Short summary
This work evaluates how rivers are projected to change in the near future compared to the recent past in the context of a warming world. We show that important rivers of the world will notably change their flows, mainly during peaks, exceeding the variations that rivers used to exhibit. Such large changes may produce more frequent floods, alter hydropower generation, and potentially affect the ocean's circulation.
Mugni Hadi Hariadi, Gerard van der Schrier, Gert-Jan Steeneveld, Samuel J. Sutanto, Edwin Sutanudjaja, Dian Nur Ratri, Ardhasena Sopaheluwakan, and Albert Klein Tank
Hydrol. Earth Syst. Sci., 28, 1935–1956, https://doi.org/10.5194/hess-28-1935-2024, https://doi.org/10.5194/hess-28-1935-2024, 2024
Short summary
Short summary
We utilize the high-resolution CMIP6 for extreme rainfall and streamflow projection over Southeast Asia. This region will experience an increase in both dry and wet extremes in the near future. We found a more extreme low flow and high flow, along with an increasing probability of low-flow and high-flow events. We reveal that the changes in low-flow events and their probabilities are not only influenced by extremely dry climates but also by the catchment characteristics.
Ayenew D. Ayalew, Paul D. Wagner, Dejene Sahlu, and Nicola Fohrer
Hydrol. Earth Syst. Sci., 28, 1853–1872, https://doi.org/10.5194/hess-28-1853-2024, https://doi.org/10.5194/hess-28-1853-2024, 2024
Short summary
Short summary
The study presents a pioneering comprehensive integrated approach to unravel hydrological complexities in data-scarce regions. By integrating diverse data sources and advanced analytics, we offer a holistic understanding of water systems, unveiling hidden patterns and driving factors. This innovative method holds immense promise for informed decision-making and sustainable water resource management, addressing a critical need in hydrological science.
Juliette Godet, Eric Gaume, Pierre Javelle, Pierre Nicolle, and Olivier Payrastre
Hydrol. Earth Syst. Sci., 28, 1403–1413, https://doi.org/10.5194/hess-28-1403-2024, https://doi.org/10.5194/hess-28-1403-2024, 2024
Short summary
Short summary
This work was performed in order to precisely address a point that is often neglected by hydrologists: the allocation of points located on a river network to grid cells, which is often a mandatory step for hydrological modelling.
Jim Yoon, Nathalie Voisin, Christian Klassert, Travis Thurber, and Wenwei Xu
Hydrol. Earth Syst. Sci., 28, 899–916, https://doi.org/10.5194/hess-28-899-2024, https://doi.org/10.5194/hess-28-899-2024, 2024
Short summary
Short summary
Global and regional models used to evaluate water shortages typically neglect the possibility that irrigated crop areas may change in response to future hydrological conditions, such as the fallowing of crops in response to drought. Here, we enhance a model used for water shortage analysis with farmer agents that dynamically adapt their irrigated crop areas based on simulated hydrological conditions. Results indicate that such cropping adaptation can strongly alter simulated water shortages.
Celray James Chawanda, Albert Nkwasa, Wim Thiery, and Ann van Griensven
Hydrol. Earth Syst. Sci., 28, 117–138, https://doi.org/10.5194/hess-28-117-2024, https://doi.org/10.5194/hess-28-117-2024, 2024
Short summary
Short summary
Africa's water resources are being negatively impacted by climate change and land-use change. The SWAT+ hydrological model was used to simulate the hydrological cycle in Africa, and results show likely decreases in river flows in the Zambezi and Congo rivers and highest flows in the Niger River basins due to climate change. Land cover change had the biggest impact in the Congo River basin, emphasizing the importance of including land-use change in studies.
Tom Kimpson, Margarita Choulga, Matthew Chantry, Gianpaolo Balsamo, Souhail Boussetta, Peter Dueben, and Tim Palmer
Hydrol. Earth Syst. Sci., 27, 4661–4685, https://doi.org/10.5194/hess-27-4661-2023, https://doi.org/10.5194/hess-27-4661-2023, 2023
Short summary
Short summary
Lakes play an important role when we try to explain and predict the weather. More accurate and up-to-date description of lakes all around the world for numerical models is a continuous task. However, it is difficult to assess the impact of updated lake description within a weather prediction system. In this work, we develop a method to quickly and automatically define how, where, and when updated lake description affects weather prediction.
Haiyang Shi, Geping Luo, Olaf Hellwich, Xiufeng He, Alishir Kurban, Philippe De Maeyer, and Tim Van de Voorde
Hydrol. Earth Syst. Sci., 27, 4551–4562, https://doi.org/10.5194/hess-27-4551-2023, https://doi.org/10.5194/hess-27-4551-2023, 2023
Short summary
Short summary
Using evidence from meteorological stations, this study assessed the climatic, hydrological, and ecological aridity changes in global drylands and their associated mechanisms. A decoupling between atmospheric, hydrological, and vegetation aridity was found. This highlights the added value of using station-scale data to assess dryland change as a complement to results based on coarse-resolution reanalysis data and land surface models.
En Ning Lai, Lan Wang-Erlandsson, Vili Virkki, Miina Porkka, and Ruud J. van der Ent
Hydrol. Earth Syst. Sci., 27, 3999–4018, https://doi.org/10.5194/hess-27-3999-2023, https://doi.org/10.5194/hess-27-3999-2023, 2023
Short summary
Short summary
This research scrutinized predicted changes in root zone soil moisture dynamics across different climate scenarios and different climate regions globally between 2021 and 2100. The Mediterranean and most of South America stood out as regions that will likely experience permanently drier conditions, with greater severity observed in the no-climate-policy scenarios. These findings underscore the impact that possible future climates can have on green water resources.
Julia Pfeffer, Anny Cazenave, Alejandro Blazquez, Bertrand Decharme, Simon Munier, and Anne Barnoud
Hydrol. Earth Syst. Sci., 27, 3743–3768, https://doi.org/10.5194/hess-27-3743-2023, https://doi.org/10.5194/hess-27-3743-2023, 2023
Short summary
Short summary
The GRACE (Gravity Recovery And Climate Experiment) satellite mission enabled the quantification of water mass redistributions from 2002 to 2017. The analysis of GRACE satellite data shows here that slow changes in terrestrial water storage occurring over a few years to a decade are severely underestimated by global hydrological models. Several sources of errors may explain such biases, likely including the inaccurate representation of groundwater storage changes.
Thedini Asali Peiris and Petra Döll
Hydrol. Earth Syst. Sci., 27, 3663–3686, https://doi.org/10.5194/hess-27-3663-2023, https://doi.org/10.5194/hess-27-3663-2023, 2023
Short summary
Short summary
Hydrological models often overlook vegetation's response to CO2 and climate, impairing their ability to forecast impacts on evapotranspiration and water resources. To address this, we suggest involving two model variants: (1) the standard method and (2) a modified approach (proposed here) based on the Priestley–Taylor equation (PT-MA). While not universally applicable, a dual approach helps consider uncertainties related to vegetation responses to climate change, enhancing model representation.
Samah Larabi, Juliane Mai, Markus Schnorbus, Bryan A. Tolson, and Francis Zwiers
Hydrol. Earth Syst. Sci., 27, 3241–3263, https://doi.org/10.5194/hess-27-3241-2023, https://doi.org/10.5194/hess-27-3241-2023, 2023
Short summary
Short summary
The computational cost of sensitivity analysis (SA) becomes prohibitive for large hydrologic modeling domains. Here, using a large-scale Variable Infiltration Capacity (VIC) deployment, we show that watershed classification helps identify the spatial pattern of parameter sensitivity within the domain at a reduced cost. Findings reveal the opportunity to leverage climate and land cover attributes to reduce the cost of SA and facilitate more rapid deployment of large-scale land surface models.
Tanja Denager, Torben O. Sonnenborg, Majken C. Looms, Heye Bogena, and Karsten H. Jensen
Hydrol. Earth Syst. Sci., 27, 2827–2845, https://doi.org/10.5194/hess-27-2827-2023, https://doi.org/10.5194/hess-27-2827-2023, 2023
Short summary
Short summary
This study contributes to improvements in the model characterization of water and energy fluxes. The results show that multi-objective autocalibration in combination with mathematical regularization is a powerful tool to improve land surface models. Using the direct measurement of turbulent fluxes as the target variable, parameter optimization matches simulations and observations of latent heat, whereas sensible heat is clearly biased.
Yuki Kimura, Yukiko Hirabayashi, Yuki Kita, Xudong Zhou, and Dai Yamazaki
Hydrol. Earth Syst. Sci., 27, 1627–1644, https://doi.org/10.5194/hess-27-1627-2023, https://doi.org/10.5194/hess-27-1627-2023, 2023
Short summary
Short summary
Since both the frequency and magnitude of flood will increase by climate change, information on spatial distributions of potential inundation depths (i.e., flood-hazard map) is required. We developed a method for constructing realistic future flood-hazard maps which addresses issues due to biases in climate models. A larger population is estimated to face risk in the future flood-hazard map, suggesting that only focusing on flood-frequency change could cause underestimation of future risk.
Hoontaek Lee, Martin Jung, Nuno Carvalhais, Tina Trautmann, Basil Kraft, Markus Reichstein, Matthias Forkel, and Sujan Koirala
Hydrol. Earth Syst. Sci., 27, 1531–1563, https://doi.org/10.5194/hess-27-1531-2023, https://doi.org/10.5194/hess-27-1531-2023, 2023
Short summary
Short summary
We spatially attribute the variance in global terrestrial water storage (TWS) interannual variability (IAV) and its modeling error with two data-driven hydrological models. We find error hotspot regions that show a disproportionately large significance in the global mismatch and the association of the error regions with a smaller-scale lateral convergence of water. Our findings imply that TWS IAV modeling can be efficiently improved by focusing on model representations for the error hotspots.
Jannis M. Hoch, Edwin H. Sutanudjaja, Niko Wanders, Rens L. P. H. van Beek, and Marc F. P. Bierkens
Hydrol. Earth Syst. Sci., 27, 1383–1401, https://doi.org/10.5194/hess-27-1383-2023, https://doi.org/10.5194/hess-27-1383-2023, 2023
Short summary
Short summary
To facilitate locally relevant simulations over large areas, global hydrological models (GHMs) have moved towards ever finer spatial resolutions. After a decade-long quest for hyper-resolution (i.e. equal to or smaller than 1 km), the presented work is a first application of a GHM at 1 km resolution over Europe. This not only shows that hyper-resolution can be achieved but also allows for a thorough evaluation of model results at unprecedented detail and the formulation of future research.
Chinchu Mohan, Tom Gleeson, James S. Famiglietti, Vili Virkki, Matti Kummu, Miina Porkka, Lan Wang-Erlandsson, Xander Huggins, Dieter Gerten, and Sonja C. Jähnig
Hydrol. Earth Syst. Sci., 26, 6247–6262, https://doi.org/10.5194/hess-26-6247-2022, https://doi.org/10.5194/hess-26-6247-2022, 2022
Short summary
Short summary
The relationship between environmental flow violations and freshwater biodiversity at a large scale is not well explored. This study intended to carry out an exploratory evaluation of this relationship at a large scale. While our results suggest that streamflow and EF may not be the only determinants of freshwater biodiversity at large scales, they do not preclude the existence of relationships at smaller scales or with more holistic EF methods or with other biodiversity data or metrics.
Zhaofei Liu
Hydrol. Earth Syst. Sci., 26, 6207–6226, https://doi.org/10.5194/hess-26-6207-2022, https://doi.org/10.5194/hess-26-6207-2022, 2022
Short summary
Short summary
Ground heat flux (G) accounts for a significant fraction of the surface energy balance (SEB), but there is insufficient research on these models compared with other flux. The accuracy of G simulation methods in the SEB-based remote sensing evapotranspiration models is evaluated. Results show that the accuracy of each method varied significantly at different sites and at half-hour intervals. Further improvement of G simulations is recommended for the remote sensing evapotranspiration modelers.
Pau Wiersma, Jerom Aerts, Harry Zekollari, Markus Hrachowitz, Niels Drost, Matthias Huss, Edwin H. Sutanudjaja, and Rolf Hut
Hydrol. Earth Syst. Sci., 26, 5971–5986, https://doi.org/10.5194/hess-26-5971-2022, https://doi.org/10.5194/hess-26-5971-2022, 2022
Short summary
Short summary
We test whether coupling a global glacier model (GloGEM) with a global hydrological model (PCR-GLOBWB 2) leads to a more realistic glacier representation and to improved basin runoff simulations across 25 large-scale basins. The coupling does lead to improved glacier representation, mainly by accounting for glacier flow and net glacier mass loss, and to improved basin runoff simulations, mostly in strongly glacier-influenced basins, which is where the coupling has the most impact.
Feng Zhong, Shanhu Jiang, Albert I. J. M. van Dijk, Liliang Ren, Jaap Schellekens, and Diego G. Miralles
Hydrol. Earth Syst. Sci., 26, 5647–5667, https://doi.org/10.5194/hess-26-5647-2022, https://doi.org/10.5194/hess-26-5647-2022, 2022
Short summary
Short summary
A synthesis of rainfall interception data from past field campaigns is performed, including 166 forests and 17 agricultural plots distributed worldwide. These site data are used to constrain and validate an interception model that considers sub-grid heterogeneity and vegetation dynamics. A global, 40-year (1980–2019) interception dataset is generated at a daily temporal and 0.1° spatial resolution. This dataset will serve as a benchmark for future investigations of the global hydrological cycle.
Dongyu Feng, Zeli Tan, Darren Engwirda, Chang Liao, Donghui Xu, Gautam Bisht, Tian Zhou, Hong-Yi Li, and L. Ruby Leung
Hydrol. Earth Syst. Sci., 26, 5473–5491, https://doi.org/10.5194/hess-26-5473-2022, https://doi.org/10.5194/hess-26-5473-2022, 2022
Short summary
Short summary
Sea level rise, storm surge and river discharge can cause coastal backwater effects in downstream sections of rivers, creating critical flood risks. This study simulates the backwater effects using a large-scale river model on a coastal-refined computational mesh. By decomposing the backwater drivers, we revealed their relative importance and long-term variations. Our analysis highlights the increasing strength of backwater effects due to sea level rise and more frequent storm surge.
Kieran M. R. Hunt, Gwyneth R. Matthews, Florian Pappenberger, and Christel Prudhomme
Hydrol. Earth Syst. Sci., 26, 5449–5472, https://doi.org/10.5194/hess-26-5449-2022, https://doi.org/10.5194/hess-26-5449-2022, 2022
Short summary
Short summary
In this study, we use three models to forecast river streamflow operationally for 13 months (September 2020 to October 2021) at 10 gauges in the western US. The first model is a state-of-the-art physics-based streamflow model (GloFAS). The second applies a bias-correction technique to GloFAS. The third is a type of neural network (an LSTM). We find that all three are capable of producing skilful forecasts but that the LSTM performs the best, with skilful 5 d forecasts at nine stations.
Tongtiegang Zhao, Haoling Chen, Yu Tian, Denghua Yan, Weixin Xu, Huayang Cai, Jiabiao Wang, and Xiaohong Chen
Hydrol. Earth Syst. Sci., 26, 4233–4249, https://doi.org/10.5194/hess-26-4233-2022, https://doi.org/10.5194/hess-26-4233-2022, 2022
Short summary
Short summary
This paper develops a novel set operations of coefficients of determination (SOCD) method to explicitly quantify the overlapping and differing information for GCM forecasts and ENSO teleconnection. Specifically, the intersection operation of the coefficient of determination derives the overlapping information for GCM forecasts and the Niño3.4 index, and then the difference operation determines the differing information in GCM forecasts (Niño3.4 index) from the Niño3.4 index (GCM forecasts).
Vili Virkki, Elina Alanärä, Miina Porkka, Lauri Ahopelto, Tom Gleeson, Chinchu Mohan, Lan Wang-Erlandsson, Martina Flörke, Dieter Gerten, Simon N. Gosling, Naota Hanasaki, Hannes Müller Schmied, Niko Wanders, and Matti Kummu
Hydrol. Earth Syst. Sci., 26, 3315–3336, https://doi.org/10.5194/hess-26-3315-2022, https://doi.org/10.5194/hess-26-3315-2022, 2022
Short summary
Short summary
Direct and indirect human actions have altered streamflow across the world since pre-industrial times. Here, we apply a method of environmental flow envelopes (EFEs) that develops the existing global environmental flow assessments by methodological advances and better consideration of uncertainty. By assessing the violations of the EFE, we comprehensively quantify the frequency, severity, and trends of flow alteration during the past decades, illustrating anthropogenic effects on streamflow.
Toby R. Marthews, Simon J. Dadson, Douglas B. Clark, Eleanor M. Blyth, Garry D. Hayman, Dai Yamazaki, Olivia R. E. Becher, Alberto Martínez-de la Torre, Catherine Prigent, and Carlos Jiménez
Hydrol. Earth Syst. Sci., 26, 3151–3175, https://doi.org/10.5194/hess-26-3151-2022, https://doi.org/10.5194/hess-26-3151-2022, 2022
Short summary
Short summary
Reliable data on global inundated areas remain uncertain. By matching a leading global data product on inundation extents (GIEMS) against predictions from a global hydrodynamic model (CaMa-Flood), we found small but consistent and non-random biases in well-known tropical wetlands (Sudd, Pantanal, Amazon and Congo). These result from known limitations in the data and the models used, which shows us how to improve our ability to make critical predictions of inundation events in the future.
Jawairia A. Ahmad, Barton A. Forman, and Sujay V. Kumar
Hydrol. Earth Syst. Sci., 26, 2221–2243, https://doi.org/10.5194/hess-26-2221-2022, https://doi.org/10.5194/hess-26-2221-2022, 2022
Short summary
Short summary
Assimilation of remotely sensed data into a land surface model to improve the spatiotemporal estimation of soil moisture across South Asia exhibits potential. Satellite retrieval assimilation corrects biases that are generated due to an unmodeled hydrologic phenomenon, i.e., irrigation. The improvements in fine-scale, modeled soil moisture estimates by assimilating coarse-scale retrievals indicates the utility of the described methodology for data-scarce regions.
Naota Hanasaki, Hikari Matsuda, Masashi Fujiwara, Yukiko Hirabayashi, Shinta Seto, Shinjiro Kanae, and Taikan Oki
Hydrol. Earth Syst. Sci., 26, 1953–1975, https://doi.org/10.5194/hess-26-1953-2022, https://doi.org/10.5194/hess-26-1953-2022, 2022
Short summary
Short summary
Global hydrological models (GHMs) are usually applied with a spatial resolution of about 50 km, but this time we applied the H08 model, one of the most advanced GHMs, with a high resolution of 2 km to Kyushu island, Japan. Since the model was not accurate as it was, we incorporated local information and improved the model, which revealed detailed water stress in subregions that were not visible with the previous resolution.
Basil Kraft, Martin Jung, Marco Körner, Sujan Koirala, and Markus Reichstein
Hydrol. Earth Syst. Sci., 26, 1579–1614, https://doi.org/10.5194/hess-26-1579-2022, https://doi.org/10.5194/hess-26-1579-2022, 2022
Short summary
Short summary
We present a physics-aware machine learning model of the global hydrological cycle. As the model uses neural networks under the hood, the simulations of the water cycle are learned from data, and yet they are informed and constrained by physical knowledge. The simulated patterns lie within the range of existing hydrological models and are plausible. The hybrid modeling approach has the potential to tackle key environmental questions from a novel perspective.
Tina Trautmann, Sujan Koirala, Nuno Carvalhais, Andreas Güntner, and Martin Jung
Hydrol. Earth Syst. Sci., 26, 1089–1109, https://doi.org/10.5194/hess-26-1089-2022, https://doi.org/10.5194/hess-26-1089-2022, 2022
Short summary
Short summary
We assess the effect of how vegetation is defined in a global hydrological model on the composition of total water storage (TWS). We compare two experiments, one with globally uniform and one with vegetation parameters that vary in space and time. While both experiments are constrained against observational data, we found a drastic change in the partitioning of TWS, highlighting the important role of the interaction between groundwater–soil moisture–vegetation in understanding TWS variations.
Marc F. P. Bierkens, Edwin H. Sutanudjaja, and Niko Wanders
Hydrol. Earth Syst. Sci., 25, 5859–5878, https://doi.org/10.5194/hess-25-5859-2021, https://doi.org/10.5194/hess-25-5859-2021, 2021
Short summary
Short summary
We introduce a simple analytical framework that allows us to estimate to what extent large-scale groundwater withdrawal affects groundwater levels and streamflow. It also calculates which part of the groundwater withdrawal comes out of groundwater storage and which part from a reduction in streamflow. Global depletion rates obtained with the framework are compared with estimates from satellites, from global- and continental-scale groundwater models, and from in situ datasets.
Dirk Eilander, Willem van Verseveld, Dai Yamazaki, Albrecht Weerts, Hessel C. Winsemius, and Philip J. Ward
Hydrol. Earth Syst. Sci., 25, 5287–5313, https://doi.org/10.5194/hess-25-5287-2021, https://doi.org/10.5194/hess-25-5287-2021, 2021
Short summary
Short summary
Digital elevation models and derived flow directions are crucial to distributed hydrological modeling. As the spatial resolution of models is typically coarser than these data, we need methods to upscale flow direction data while preserving the river structure. We propose the Iterative Hydrography Upscaling (IHU) method and show it outperforms other often-applied methods. We publish the multi-resolution MERIT Hydro IHU hydrography dataset and the algorithm as part of the pyflwdir Python package.
Jérôme Kopp, Pauline Rivoire, S. Mubashshir Ali, Yannick Barton, and Olivia Martius
Hydrol. Earth Syst. Sci., 25, 5153–5174, https://doi.org/10.5194/hess-25-5153-2021, https://doi.org/10.5194/hess-25-5153-2021, 2021
Short summary
Short summary
Episodes of extreme rainfall events happening in close temporal succession can lead to floods with dramatic impacts. We developed a novel method to individually identify those episodes and deduced the regions where they occur frequently and where their impact is substantial. Those regions are the east and northeast of the Asian continent, central Canada and the south of California, Afghanistan, Pakistan, the southwest of the Iberian Peninsula, and north of Argentina and south of Bolivia.
Alyssa J. DeVincentis, Hervé Guillon, Romina Díaz Gómez, Noelle K. Patterson, Francine van den Brandeler, Arthur Koehl, J. Pablo Ortiz-Partida, Laura E. Garza-Díaz, Jennifer Gamez-Rodríguez, Erfan Goharian, and Samuel Sandoval Solis
Hydrol. Earth Syst. Sci., 25, 4631–4650, https://doi.org/10.5194/hess-25-4631-2021, https://doi.org/10.5194/hess-25-4631-2021, 2021
Short summary
Short summary
Latin America and the Caribbean face many water-related stresses which are expected to worsen with climate change. To assess the vulnerability, we reviewed over 20 000 multilingual research articles using machine learning and an understanding of the regional landscape. Results reveal that the region’s inherent vulnerability is compounded by research blind spots in niche topics (reservoirs and risk assessment) and subregions (Caribbean nations), as well as by its reliance on one country (Brazil).
Michiel Maertens, Gabriëlle J. M. De Lannoy, Sebastian Apers, Sujay V. Kumar, and Sarith P. P. Mahanama
Hydrol. Earth Syst. Sci., 25, 4099–4125, https://doi.org/10.5194/hess-25-4099-2021, https://doi.org/10.5194/hess-25-4099-2021, 2021
Short summary
Short summary
In this study, we simulated the water balance over the South American Dry Chaco and assessed the impact of land cover changes thereon using three different land surface models. Our simulations indicated that different models result in a different partitioning of the total water budget, but all showed an increase in soil moisture and percolation over the deforested areas. We also found that, relative to independent data, no specific land surface model is significantly better than another.
Louise J. Slater, Bailey Anderson, Marcus Buechel, Simon Dadson, Shasha Han, Shaun Harrigan, Timo Kelder, Katie Kowal, Thomas Lees, Tom Matthews, Conor Murphy, and Robert L. Wilby
Hydrol. Earth Syst. Sci., 25, 3897–3935, https://doi.org/10.5194/hess-25-3897-2021, https://doi.org/10.5194/hess-25-3897-2021, 2021
Short summary
Short summary
Weather and water extremes have devastating effects each year. One of the principal challenges for society is understanding how extremes are likely to evolve under the influence of changes in climate, land cover, and other human impacts. This paper provides a review of the methods and challenges associated with the detection, attribution, management, and projection of nonstationary weather and water extremes.
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, b
Addor, N., Newman, A. J., Mizukami, N., and Clark, M. P.: Catchment attributes for large-sample studies, UCAR/NCAR, Boulder, CO, USA, https://doi.org/10.5065/D6G73C3Q, 2017b. a, b, c, d
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 Resources Res., 54, 8792–8812, https://doi.org/10.1029/2018WR022606, 2018. a, b
Anderson, E. A.: National Weather Service river forecast system: Snow
accumulation and ablation model, NOAA Tech. Memo. NWS HYDRO-17, 87 pp., 1973. a
Beck, H. E., van Dijk, A. I. J. M., de Roo, A., Miralles, D. G., McVicar, T. R., Schellekens, J., and Bruijnzeel, L. A.: Global-scale regionalization of hydrologic model parameters, Water Resour. Res., 52, 3599–3622, https://doi.org/10.1002/2015WR018247, 2016. a
Beven, K. and Freer, J.: Equifinality, data assimilation, and uncertainty estimation in mechanistic modelling of complex environmental systems using the GLUE methodology, J. Hydrol., 249, 11–29, 2001. a
Blöschl, G. and Sivapalan, M.: Scale issues in hydrological modelling: a
review, Hydrol. Process., 9, 251–290, 1995. a
Blöschl, G., Sivapalan, M., Savenije, H., Wagener, T., and Viglione, A.: Runoff prediction in ungauged basins: synthesis across processes, places and scales, Cambridge University Press, Cambridge, 2013. a
Burnash, R. J. C.: The NWS river forecast system–catchment modeling, in:
Computer models of watershed hydrology, edited by: Singh, V. P., Water Resources Publications, Littleton, CO, 311–366, 1995. a
Burnash, R. J., Ferral, R. L., and McGuire, R. A.: A generalized streamflow simulation system, conceptual modeling for digital computers, Joint Federal and State River Forecast Center, U.S. National Weather Service, and California Departmentof Water Resources Tech. Rep., 204 pp., 1973. a
Clark, M. P., Slater, A. G., Rupp, D. E., Woods, R. A., Vrugt, J. A., Gupta, H. V., Wagener, T., and Hay, L. E.: Framework for Understanding Structural Errors (FUSE): A modular framework to diagnose differences between hydrological models, Water Resour. Res., 44, W00B02, https://doi.org/10.1029/2007WR006735, 2008. a
Cohen, J.: Statistical power analysis for the behavioral sciences, 2nd Edn., Erlbaum, Hillsdale, NJ, 1988. a
Gupta, H. V., Sorooshian, S., and Yapo, P. O.: Toward improved calibration of hydrologic models: Multiple and noncommensurable measures of information, Water Resour. Res., 34, 751–763, 1998. a
Gupta, H. V., Wagener, T., and Liu, Y.: Reconciling theory with observations: elements of a diagnostic approach to model evaluation, Hydrol. Process., 22, 3802–3813, 2008. a
Gupta, H. V., Perrin, C., Blöschl, G., Montanari, A., Kumar, R., Clark, M., and Andréassian, V.: Large-sample hydrology: a need to balance depth with breadth, Hydrol. Earth Syst. Sci., 18, 463–477, https://doi.org/10.5194/hess-18-463-2014, 2014. a
Henn, B., Clark, M. P., Kavetski, D., and Lundquist, J. D.: Estimating mountain basin-mean precipitation fromstreamflow using Bayesian inference, Water Resour. Res., 51, 8012–8033, 2008. a
Herman, J. D., Kollat, J. B., Reed, P. M., and Wagener, T.: Technical Note: Method of Morris effectively reduces the computational demands of global sensitivity analysis for distributed watershed models, Hydrol. Earth Syst. Sci., 17, 2893–2903, https://doi.org/10.5194/hess-17-2893-2013, 2013. a
Hochreiter, S.: Untersuchungen zu dynamischen neuronalen Netzen, Diploma,
Technische Universität München, Germany, 1991. a
Hochreiter, S. and Schmidhuber, J.: Long short-term memory, Neural Comput., 9, 1735–1780, 1997. a
Hrachowitz, M., Savenije, H., Blöschl, G., McDonnell, J., Sivapalan, M., Pomeroy, J., Arheimer, B., Blume, T., Clark, M., Ehret, U., Fenicia, F., Freer, J. E., Gelfan, A., Gupta, H. V., Hughes, D. A., Hut, R. W., Montanari, A., Pande, S., Tetzlaff, D., Troch, P. A., Uhlenbrook, S., Wagener, T., Winsemius, H. C., Woods, R. A., Zehe, E., and Cudennec, C.: A decade of Predictions in Ungauged Basins (PUB) – a review, Hydrolog. Sci. J., 58, 1198–1255, 2013. a
Hunter, J. D.: Matplotlib: A 2D graphics environment, Comput. Sci. Eng., 9, 90–95, 2007. a
Jakeman, A. J. and Hornberger, G. M.: How much complexity is warranted in a rainfall-runoff model?, Water Resour. Res., 29, 2637–2649, 1993. a
Kirchner, J. W.: Getting the right answers for the right reasons: Linking measurements, analyses, and models to advance the science of hydrology, Water Resour. Res., 42, W03S04, https://doi.org/10.1029/2005WR004362, 2006. a
Kratzert, F.: Benchmark models, HydroShare, https://doi.org/10.4211/hs.474ecc37e7db45baa425cdb4fc1b61e1, 2019a. a
Kratzert, F.: CAMELS extended Maurer forcings, HydroShare, https://doi.org/10.4211/hs.17c896843cf940339c3c3496d0c1c077, 2019b. a
Kratzert, F.: kratzert/ealstm_regional_modeling: Code to reproduce paper experiments/results, Zenodo, https://doi.org/10.5281/zenodo.3530884, 2019c. a
Kratzert, F.: Pre-trained models, HydroShare,
https://doi.org/10.4211/hs.83ea5312635e44dc824eeb99eda12f06, 2019d. a
Kratzert, F., Herrnegger, M., Klotz, D., Hochreiter, S., and Klambauer, G.: Do internals of neural networks make sense in the context of hydrology?, in: AGU Fall Meeting Abstracts, 2018AGUFM.H13B..06K, 2018a. 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, 2018b. a, b
Kratzert, F., Klotz, D., Herrnegger, M., and Hochreiter, S.: A glimpse into the Unobserved: Runoff simulation for ungauged catchments with LSTMs, in: Workshop on Modeling and Decision-Making in the Spatiotemporal Domain, 32nd Conference on Neural Information Processing Systems (NeurIPS 2018), Montreal, Canada, 3–8 December 2018c. a
Kratzert, F., Herrnegger, M., Klotz, D., Hochreiter, S., and Klambauer, G.: NeuralHydrology-Interpreting LSTMs in Hydrology, arXiv preprint arXiv:1903.07903, 2019. a
Kumar, R., Samaniego, L., and Attinger, S.: Implications of distributed hydrologic model parameterization on water fluxes at multiple scales and locations, Water Resour. Res., 49, 360–379, 2013. a
Liang, X., Lettenmaier, D. P., Wood, E. F., and Burges, S. J.: A simple hydrologically based model of land surface water and energy fluxes for general circulation models, J. Geophys. Res., 99, 14415–14428, 1994. a
McKinney, W.: Data Structures for Statistical Computing in Python, Proceedings of the 9th Python in Science Conference, Austin, Texas, 28 June–2 July 2010, 1697900, 51–56, 2010. a
Mizukami, N., Rakovec, O., Newman, A. J., Clark, M. P., Wood, A. W., Gupta, H. V., and Kumar, R.: On the choice of calibration metrics for “high-flow” estimation using hydrologic models, Hydrol. Earth Syst. Sci., 23, 2601–2614, https://doi.org/10.5194/hess-23-2601-2019, 2019. a
Morris, M. D.: Factorial sampling plans for preliminary computational experiments, Technometrics, 33, 161–174, 1991. a
Naef, F.: Can we model the rainfall-runoff process today?/Peut-on actuellement mettre en modèle le processus pluie-écoulement?, Hydrol. Sci. B., 26, 281–289, 1981. 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, 1970. a
Newman, A., Sampson, K., Clark, M., Bock, A., Viger, R., and Blodgett, D.: A large-sample watershed-scale hydrometeorological dataset for the contiguous USA, UCAR/NCAR, Boulder, CO, USA, https://doi.org/10.5065/D6MW2F4D, 2014. a
Newman, A. J., Clark, M. P., Sampson, K., Wood, A., Hay, L. E., Bock, A., Viger, R. J., Blodgett, D., Brekke, L., Arnold, J. R., Hopson, T., and Duan, Q.: Development of a large-sample watershed-scale hydrometeorological data set for the contiguous USA: data set characteristics and assessment of regional variability in hydrologic model performance, Hydrol. Earth Syst. Sci., 19, 209–223, https://doi.org/10.5194/hess-19-209-2015, 2015. a
Paszke, A., Gross, S., Chintala, S., Chanan, G., Yang, E., DeVito, Z., Lin, Z., Desmaison, A., Antiga, L., and Lerer, A.: Automatic differentiation in PyTorch, in: NIPS 2017 Autodiff Workshop: The Future of Gradient-based Machine Learning Software and Techniques, Long Beach, CA, US, 9 December 2017. a
Perrin, C., Michel, C., and Andréassian, V.: Does a large number of parameters enhance model performance? Comparative assessment of common catchment model structures on 429 catchments, J. Hydrol., 242, 275–301, 2001. a
Peters-Lidard, C. D., Clark, M., Samaniego, L., Verhoest, N. E. C., van Emmerik, T., Uijlenhoet, R., Achieng, K., Franz, T. E., and Woods, R.: Scaling, similarity, and the fourth paradigm for hydrology, Hydrol. Earth Syst. Sci., 21, 3701–3713, https://doi.org/10.5194/hess-21-3701-2017, 2017. a
Saltelli, A., Tarantola, S., Campolongo, F., and Ratto, M.: Sensitivity analysis in practice: a guide to assessing scientific models, Wiley Online Library, 94–100, 2004. a
Samaniego, L., Kumar, R., and Attinger, S.: Multiscale parameter regionalization of a grid-based hydrologic model at the mesoscale, Water Resour. Res., 46, W05523, https://doi.org/10.1029/2008WR007327, 2010. a, b
Seibert, J.: Regionalisation of parameters for a conceptual rainfall–runoff model, Agr. Forest Meteorol., 98–99, 279–293, 1999. a
Seibert, J. and Vis, M. J. P.: Teaching hydrological modeling with a user-friendly catchment-runoff-model software package, Hydrol. Earth Syst. Sci., 16, 3315–3325, https://doi.org/10.5194/hess-16-3315-2012, 2012. a
Sivapalan, M., Takeuchi, K., Franks, S. W., Gupta, V. K., Karambiri, H., Lakshmi, V., Liang, X., McDonnell, J. J., Mendiondo, E. M., O'Connell, P. E., Oki, T., Pomeroy, J. W., Schertzer, D., Uhlenbrook, S., and Zehe, E.: IAHS Decade on Predictions in Ungauged Basins (PUB), 2003–2012: Shaping an exciting future for the hydrological sciences, Hydrolog. Sci. J., 48, 857–880, 2003. a
Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., and Salakhutdinov, R.: Dropout: a simple way to prevent neural networks from overfitting, J. Mach. Learn. Res., 15, 1929–1958, 2014. a
Van Der Walt, S., Colbert, S. C., and Varoquaux, G.: The NumPy array: A structure for efficient numerical computation, Comput. Sci. Eng., 13, 22–30, 2011. a
van Rossum, G.: Python tutorial, Technical Report CS-R9526, Centrum voor Wiskunde en Informatica (CWI), Amsterdam, 1995. a
Wang, A. and Solomatine, D. P.: Practical Experience of Sensitivity Analysis: Comparing Six Methods, on Three Hydrological Models, with Three Performance Criteria, Water, 11, 1062, https://doi.org/10.3390/w11051062, 2019. a
Wilcoxon, F.: Individual comparisons by ranking methods, Biometrics Bull., 1, 80–83, 1945. a
Wood, A. W., Maurer, E. P., Kumar, A., and Lettenmaier, D. P.: Long-range experimental hydrologic forecasting for the eastern United States, J. Geophys. Res., 107, 4429, https://doi.org/10.1029/2001JD000659, 2002. a
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.
A new approach for regional rainfall–runoff modeling using long short-term memory (LSTM)-based...
