Articles | Volume 27, issue 1
https://doi.org/10.5194/hess-27-139-2023
© Author(s) 2023. 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-27-139-2023
© Author(s) 2023. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Research article
| 
09 Jan 2023
Research article |
| 09 Jan 2023
| 09 Jan 2023
Continuous streamflow prediction in ungauged basins: long short-term memory neural networks clearly outperform traditional hydrological models
Richard Arsenault
CORRESPONDING AUTHOR
Hydrology, Climate and Climate Change Laboratory, École de technologie supérieure, 1100 Notre-Dame West, Montréal, Québec H3C 1K3, Canada
Jean-Luc Martel
Hydrology, Climate and Climate Change Laboratory, École de technologie supérieure, 1100 Notre-Dame West, Montréal, Québec H3C 1K3, Canada
Frédéric Brunet
Hydrology, Climate and Climate Change Laboratory, École de technologie supérieure, 1100 Notre-Dame West, Montréal, Québec H3C 1K3, Canada
François Brissette
Hydrology, Climate and Climate Change Laboratory, École de technologie supérieure, 1100 Notre-Dame West, Montréal, Québec H3C 1K3, Canada
Juliane Mai
Department of Civil and Environmental Engineering, University of Waterloo, 200 University Ave W, Waterloo, Ontario N2L 3G1, Canada
Related authors
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.
Frédéric Talbot, Jean-Daniel Sylvain, Guillaume Drolet, Annie Poulin, and Richard Arsenault
EGUsphere, https://doi.org/10.5194/egusphere-2024-3353, https://doi.org/10.5194/egusphere-2024-3353, 2024
Short summary
Short summary
This research explores different modelling approaches within a distributed and physically based hydrological model, emphasizing configurations that integrate groundwater recharge and dynamics. The study demonstrates that precise adjustments in model calibration can enhance the accuracy and representation of hydrological variables, which are crucial for effective water management and climate change adaptation strategies.
Frédéric Talbot, Simon Ricard, Jean-Daniel Sylvain, Guillaume Drolet, Annie Poulin, Jean-Luc Martel, and Richard Arsenault
EGUsphere, https://doi.org/10.5194/egusphere-2024-3037, https://doi.org/10.5194/egusphere-2024-3037, 2024
Preprint archived
Short summary
Short summary
This study compares two hydrological modeling approaches for assessing climate change impacts on water systems. We evaluate the conventional method alongside the asynchronous method, which excels in capturing extreme events but faces challenges with event timing. Our findings suggest that a semi-asynchronous approach, blending the strengths of both methods, could offer better results. This research provides key insights for supporting sustainable resource planning in a changing climate.
Jean-Luc Martel, Richard Arsenault, Richard Turcotte, Mariana Castañeda-Gonzalez, François Brissette, William Armstrong, Edouard Mailhot, Jasmine Pelletier-Dumont, Simon Lachance-Cloutier, Gabriel Rondeau-Genesse, and Louis-Philippe Caron
EGUsphere, https://doi.org/10.5194/egusphere-2024-2134, https://doi.org/10.5194/egusphere-2024-2134, 2024
Short summary
Short summary
This study explores six methods to improve the ability of Long Short-Term Memory (LSTM) neural networks to predict peak streamflows, crucial for flood analysis. By enhancing data inputs and model techniques, the research shows LSTM models can match or surpass traditional hydrological models in simulating peak flows. Tested on 88 catchments in Quebec, Canada, these methods offer promising strategies for better flood prediction.
Mehrad Rahimpour Asenjan, Francois Brissette, Richard Arsenault, and Jean-Luc Martel
EGUsphere, https://doi.org/10.5194/egusphere-2024-1183, https://doi.org/10.5194/egusphere-2024-1183, 2024
Preprint archived
Short summary
Short summary
Our study examines how combining climate models impacts future streamflow predictions, crucial for understanding climate change. Comparing six methods across 3,107 North American catchments, we found unequal weighting significantly improves rainfall and temperature projections. However, for streamflow, both equal and unequal weighting perform similarly with bias correction. Our findings underscore the need to carefully select weighting methods and correct biases for accurate climate projections.
Mehrad Rahimpour Asenjan, Francois Brissette, Jean-Luc Martel, and Richard Arsenault
Hydrol. Earth Syst. Sci., 27, 4355–4367, https://doi.org/10.5194/hess-27-4355-2023, https://doi.org/10.5194/hess-27-4355-2023, 2023
Short summary
Short summary
Climate models are central to climate change impact studies. Some models project a future deemed too hot by many. We looked at how including hot models may skew the result of impact studies. Applied to hydrology, this study shows that hot models do not systematically produce hydrological outliers.
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.
Mostafa Tarek, François Brissette, and Richard Arsenault
Hydrol. Earth Syst. Sci., 25, 3331–3350, https://doi.org/10.5194/hess-25-3331-2021, https://doi.org/10.5194/hess-25-3331-2021, 2021
Short summary
Short summary
It is not known how much uncertainty the choice of a reference data set may bring to impact studies. This study compares precipitation and temperature data sets to evaluate the uncertainty contribution to the results of climate change studies. Results show that all data sets provide good streamflow simulations over the reference period. The reference data sets also provided uncertainty that was equal to or larger than that related to general circulation models over most of the catchments.
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.
Maria Staudinger, Anna Herzog, Ralf Loritz, Tobias Houska, Sandra Pool, Diana Spieler, Paul D. Wagner, Juliane Mai, Jens Kiesel, Stephan Thober, Björn Guse, and Uwe Ehret
EGUsphere, https://doi.org/10.5194/egusphere-2025-1076, https://doi.org/10.5194/egusphere-2025-1076, 2025
Short summary
Short summary
Four process-based and four data-driven hydrological models are compared using different training data. We found process-based models to perform better with small data sets but stop learning soon, while data-driven models learn longer. The study highlights the importance of memory in data and the impact of different data sampling methods on model performance. The direct comparison of these models is novel and provides a clear understanding of their performance under various data conditions.
Frédéric Talbot, Jean-Daniel Sylvain, Guillaume Drolet, Annie Poulin, and Richard Arsenault
EGUsphere, https://doi.org/10.5194/egusphere-2024-3353, https://doi.org/10.5194/egusphere-2024-3353, 2024
Short summary
Short summary
This research explores different modelling approaches within a distributed and physically based hydrological model, emphasizing configurations that integrate groundwater recharge and dynamics. The study demonstrates that precise adjustments in model calibration can enhance the accuracy and representation of hydrological variables, which are crucial for effective water management and climate change adaptation strategies.
Frédéric Talbot, Simon Ricard, Jean-Daniel Sylvain, Guillaume Drolet, Annie Poulin, Jean-Luc Martel, and Richard Arsenault
EGUsphere, https://doi.org/10.5194/egusphere-2024-3037, https://doi.org/10.5194/egusphere-2024-3037, 2024
Preprint archived
Short summary
Short summary
This study compares two hydrological modeling approaches for assessing climate change impacts on water systems. We evaluate the conventional method alongside the asynchronous method, which excels in capturing extreme events but faces challenges with event timing. Our findings suggest that a semi-asynchronous approach, blending the strengths of both methods, could offer better results. This research provides key insights for supporting sustainable resource planning in a changing climate.
Jean-Luc Martel, Richard Arsenault, Richard Turcotte, Mariana Castañeda-Gonzalez, François Brissette, William Armstrong, Edouard Mailhot, Jasmine Pelletier-Dumont, Simon Lachance-Cloutier, Gabriel Rondeau-Genesse, and Louis-Philippe Caron
EGUsphere, https://doi.org/10.5194/egusphere-2024-2134, https://doi.org/10.5194/egusphere-2024-2134, 2024
Short summary
Short summary
This study explores six methods to improve the ability of Long Short-Term Memory (LSTM) neural networks to predict peak streamflows, crucial for flood analysis. By enhancing data inputs and model techniques, the research shows LSTM models can match or surpass traditional hydrological models in simulating peak flows. Tested on 88 catchments in Quebec, Canada, these methods offer promising strategies for better flood prediction.
Mehrad Rahimpour Asenjan, Francois Brissette, Richard Arsenault, and Jean-Luc Martel
EGUsphere, https://doi.org/10.5194/egusphere-2024-1183, https://doi.org/10.5194/egusphere-2024-1183, 2024
Preprint archived
Short summary
Short summary
Our study examines how combining climate models impacts future streamflow predictions, crucial for understanding climate change. Comparing six methods across 3,107 North American catchments, we found unequal weighting significantly improves rainfall and temperature projections. However, for streamflow, both equal and unequal weighting perform similarly with bias correction. Our findings underscore the need to carefully select weighting methods and correct biases for accurate climate projections.
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.
Mehrad Rahimpour Asenjan, Francois Brissette, Jean-Luc Martel, and Richard Arsenault
Hydrol. Earth Syst. Sci., 27, 4355–4367, https://doi.org/10.5194/hess-27-4355-2023, https://doi.org/10.5194/hess-27-4355-2023, 2023
Short summary
Short summary
Climate models are central to climate change impact studies. Some models project a future deemed too hot by many. We looked at how including hot models may skew the result of impact studies. Applied to hydrology, this study shows that hot models do not systematically produce hydrological outliers.
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.
Robert Chlumsky, Juliane Mai, James R. Craig, and Bryan A. Tolson
Hydrol. Earth Syst. Sci. Discuss., https://doi.org/10.5194/hess-2023-69, https://doi.org/10.5194/hess-2023-69, 2023
Revised manuscript not accepted
Short summary
Short summary
A blended model allows multiple hydrologic processes to be represented in a single model, which allows for a model to achieve high performance without the need to modify its structure for different catchments. Here, we improve upon the initial blended version by testing more than 30 blended models in twelve catchments to improve the overall model performance. We validate our proposed, updated blended model version with independent catchments, and make this version available for open use.
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.
Michelle Viswanathan, Tobias K. D. Weber, Sebastian Gayler, Juliane Mai, and Thilo Streck
Biogeosciences, 19, 2187–2209, https://doi.org/10.5194/bg-19-2187-2022, https://doi.org/10.5194/bg-19-2187-2022, 2022
Short summary
Short summary
We analysed the evolution of model parameter uncertainty and prediction error as we updated parameters of a maize phenology model based on yearly observations, by sequentially applying Bayesian calibration. Although parameter uncertainty was reduced, prediction quality deteriorated when calibration and prediction data were from different maize ripening groups or temperature conditions. The study highlights that Bayesian methods should account for model limitations and inherent data structures.
Nicolas Gasset, Vincent Fortin, Milena Dimitrijevic, Marco Carrera, Bernard Bilodeau, Ryan Muncaster, Étienne Gaborit, Guy Roy, Nedka Pentcheva, Maxim Bulat, Xihong Wang, Radenko Pavlovic, Franck Lespinas, Dikra Khedhaouiria, and Juliane Mai
Hydrol. Earth Syst. Sci., 25, 4917–4945, https://doi.org/10.5194/hess-25-4917-2021, https://doi.org/10.5194/hess-25-4917-2021, 2021
Short summary
Short summary
In this paper, we highlight the importance of including land-data assimilation as well as offline precipitation analysis components in a regional reanalysis system. We also document the performance of the first multidecadal 10 km reanalysis performed with the GEM atmospheric model that can be used for seamless land-surface and hydrological modelling in North America. It is of particular interest for transboundary basins, as existing datasets often show discontinuities at the border.
Mostafa Tarek, François Brissette, and Richard Arsenault
Hydrol. Earth Syst. Sci., 25, 3331–3350, https://doi.org/10.5194/hess-25-3331-2021, https://doi.org/10.5194/hess-25-3331-2021, 2021
Short summary
Short summary
It is not known how much uncertainty the choice of a reference data set may bring to impact studies. This study compares precipitation and temperature data sets to evaluate the uncertainty contribution to the results of climate change studies. Results show that all data sets provide good streamflow simulations over the reference period. The reference data sets also provided uncertainty that was equal to or larger than that related to general circulation models over most of the catchments.
Juliane Mai, James R. Craig, and Bryan A. Tolson
Hydrol. Earth Syst. Sci., 24, 5835–5858, https://doi.org/10.5194/hess-24-5835-2020, https://doi.org/10.5194/hess-24-5835-2020, 2020
Cited articles
Abrahart, R. J., Anctil, F., Coulibaly, P., Dawson, C. W., Mount, N. J., See, L. M., Shamseldin, A. Y., Solomatine, D. P., Toth, E., and Wilby, R. L.: Two decades of anarchy? Emerging themes and outstanding challenges for neural network river forecasting, Progress in Physical Geography: Earth and Environment, 36, 480–513, https://doi.org/10.1177/0309133312444943, 2012.
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, 2017.
Ali, A. S. A., Ebrahimi, S., Ashiq, M. M., Alasta, M. S., and Azari, B.: CNN-Bi LSTM neural network for simulating groundwater level, Environ. Eng., 8, 1–7, https://doi.org/10.52547/crpase.8.1.2748, 2022.
Alvarez-Garreton, C., Mendoza, P. A., Boisier, J. P., Addor, N., Galleguillos, M., Zambrano-Bigiarini, M., Lara, A., Puelma, C., Cortes, G., Garreaud, R., McPhee, J., and Ayala, A.: The CAMELS-CL dataset: catchment attributes and meteorology for large sample studies – Chile dataset, Hydrol. Earth Syst. Sci., 22, 5817–5846, https://doi.org/10.5194/hess-22-5817-2018, 2018.
Anctil, F. and Rat, A.: Evaluation of neural network streamflow forecasting on 47 watersheds, J. Hydrol. Eng., 10, 85–88, https://doi.org/10.1061/(ASCE)1084-0699(2005)10:1(85), 2005.
Arsenault, R. and Brissette, F. P.: Continuous streamflow prediction in ungauged basins: The effects of equifinality and parameter set selection on uncertainty in regionalization approaches, Water Resour. Res., 50, 6135–6153, https://doi.org/10.1002/2013WR014898, 2014.
Arsenault, R. and Brissette, F.: Multi-model averaging for continuous streamflow prediction in ungauged basins, Hydrolog. Sci. J., 61, 2443–2454, https://doi.org/10.1080/02626667.2015.1117088, 2016.
Arsenault, R., Brissette, F., and Martel, J.-L.: The hazards of split-sample validation in hydrological model calibration, J. Hydrol., 566, 346–362, https://doi.org/10.1016/j.jhydrol.2018.09.027, 2018.
Arsenault, R., Brissette, F., Martel, J.-L., Troin, M., Lévesque, G., Davidson-Chaput, J., Castañeda Gonzalez, M., Ameli, A., and Poulin, A.: A comprehensive, multisource database for hydrometeorological modeling of 14 425 North American watersheds, Scientific Data, 7, 243, https://doi.org/10.1038/s41597-020-00583-2, 2020a.
Arsenault, R., Brissette, F., Martel, J. L., Troin, M., Lévesque, G., Davidson-Chaput, J., Castañeda Gonzalez, M., Ameli, A., and Poulin, A.: HYSETS – A 14425 watershed Hydrometeorological Sandbox over North America, Open Science Framework [data set], https://doi.org/10.17605/OSF.IO/RPC3W, 2020b.
Arsenault, R., Brissette, F., and Martel, J. L.: LSTM regionalization datasets and codes, Open Science Framework [code], https://doi.org/10.17605/OSF.IO/3S2PQ, 2022.
Ayzel, G. and Heistermann, M.: The effect of calibration data length on the performance of a conceptual hydrological model versus LSTM and GRU: A case study for six basins from the CAMELS dataset, Computers and Geosciences, 149, 104708, https://doi.org/10.1016/j.cageo.2021.104708, 2021.
Ayzel, G., Kurochkina, L., Kazakov, E., and Zhuravlev, S.: Streamflow prediction in ungauged basins: benchmarking the efficiency of deep learning, E3S Web Conf., 163, 01001, https://doi.org/10.1051/e3sconf/202016301001, 2020.
Ayzel, G., Kurochkina, L., Abramov, D., and Zhuravlev, S.: Development of a Regional Gridded Runoff Dataset Using Long Short-Term Memory (LSTM) Networks, Hydrology, 8, 6, https://doi.org/10.3390/hydrology8010006, 2021.
Bengio, Y., Simard, P., and Frasconi, P.: Learning long-term dependencies with gradient descent is difficult, IEEE T. Neural Networ., 5, 157–166, https://doi.org/10.1109/72.279181, 1994.
Blöschl, G., Bierkens, M. F. P., Chambel, A., et al.: Twenty-three unsolved problems in hydrology (UPH) – a community perspective, Hydrolog. Sci. J., 64, 1141–1158, https://doi.org/10.1080/02626667.2019.1620507, 2019.
Castiglioni, S., Castellarin, A., Montanari, A., Skøien, J. O., Laaha, G., and Blöschl, G.: Smooth regional estimation of low-flow indices: physiographical space based interpolation and top-kriging, Hydrol. Earth Syst. Sci., 15, 715–727, https://doi.org/10.5194/hess-15-715-2011, 2011.
Choi, J., Lee, J., and Kim, S.: Utilization of the Long Short-Term Memory network for predicting streamflow in ungauged basins in Korea, Ecol. Eng., 182, 106699, https://doi.org/10.1016/j.ecoleng.2022.106699, 2022.
Coulibaly, P., Anctil, F., and Bobée, B.: Daily reservoir inflow forecasting using artificial neural networks with stopped training approach, J. Hydrol., 230, 244–257, https://doi.org/10.1016/S0022-1694(00)00214-6, 2000.
Coxon, G., Addor, N., Bloomfield, J. P., Freer, J., Fry, M., Hannaford, J., Howden, N. J. K., Lane, R., Lewis, M., Robinson, E. L., Wagener, T., and Woods, R.: CAMELS-GB: hydrometeorological time series and landscape attributes for 671 catchments in Great Britain, Earth Syst. Sci. Data, 12, 2459–2483, https://doi.org/10.5194/essd-12-2459-2020, 2020.
Dawson, C. W. and Wilby, R. L.: Hydrological modelling using artificial neural networks, Progress in Physical Geography: Earth and Environment, 25, 80–108, https://doi.org/10.1177/030913330102500104, 2001.
Delaigue, O., Brigode, P., Andréassian, V., Perrin, C., Etchevers, P., Soubeyroux, J.-M., Janet, B., and Nans, A.: CAMELS-FR: A large sample hydroclimatic dataset for France to explore hydrological diversity and support model benchmarking, in: IAHS-2022 Scientific Assembly, May 2022, Montpellier, France, hal-03687235, 2022.
Duan, S. and Ullrich, P.: A comprehensive investigation of machine learning models for estimating daily snow water equivalent over the Western U. S., Earth and Space Science Open Archive, 14, 1–14, https://doi.org/10.1002/essoar.10509011.1, 2021.
Feng, D., Lawson, K., and Shen, C.: Prediction in ungauged regions with sparse flow duration curves and input-selection ensemble modeling, arXiv [preprint], https://doi.org/10.48550/arXiv.2011.13380, 2020.
Fortin, V.: Le modèle météo-apport HSAMI: historique, théorie et application, Institut de Recherche d'Hydro-Québec, 68 pp., 2000.
Frame, J. M., Kratzert, F., Klotz, D., Gauch, M., Shalev, G., Gilon, O., Qualls, L. M., Gupta, H. V., and Nearing, G. S.: Deep learning rainfall–runoff predictions of extreme events, Hydrol. Earth Syst. Sci., 26, 3377–3392, https://doi.org/10.5194/hess-26-3377-2022, 2022.
Fry, L. M., Gronewold, A. D., Fortin, V., Buan, S., Clites, A. H., Luukkonen, C., Holtschlag, D., Diamond, L., Hunter, T., Seglenieks, F., Durnford, D., Dimitrijevic, M., Subich, C., Klyszejko, E., Kea, K., and Restrepo, P.: The Great Lakes Runoff Intercomparison Project Phase 1: Lake Michigan (GRIP-M), J. Hydrol., 519, 3448–3465, https://doi.org/10.1016/j.jhydrol.2014.07.021, 2014.
Gauch, M., Kratzert, F., Klotz, D., Nearing, G., Lin, J., and Hochreiter, S.: Rainfall–runoff prediction at multiple timescales with a single Long Short-Term Memory network, Hydrol. Earth Syst. Sci., 25, 2045–2062, https://doi.org/10.5194/hess-25-2045-2021, 2021a.
Gauch, M., Mai, J., and Lin, J.: The proper care and feeding of CAMELS: How limited training data affects streamflow prediction, Environ. Modell. Softw., 135, 104926, https://doi.org/10.1016/j.envsoft.2020.104926, 2021b.
Guo, Y., Liu, Y., Oerlemans, A., Lao, S., Wu, S., and Lew, M. S.: Deep learning for visual understanding: A review, Neurocomputing, 187, 27–48, https://doi.org/10.1016/j.neucom.2015.09.116, 2016.
Guo, Y., Zhang, Y., Zhang, L., and Wang, Z.: Regionalization of hydrological modeling for predicting streamflow in ungauged catchments: A comprehensive review, WIREs Water, 8, e1487, https://doi.org/10.1002/wat2.1487, 2021.
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.
Hansen, N., Müller, S. D., and Koumoutsakos, P.: Reducing the time complexity of the derandomized evolution strategy with Covariance Matrix Adaptation (CMA-ES), Evol. Comput., 11, 1–18, https://doi.org/10.1162/106365603321828970, 2003.
He, Y., Bárdossy, A., and Zehe, E.: A review of regionalisation for continuous streamflow simulation, Hydrol. Earth Syst. Sci., 15, 3539–3553, https://doi.org/10.5194/hess-15-3539-2011, 2011.
Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A., Muñoz-Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schepers, D., Simmons, A., Soci, C., Abdalla, S., Abellan, X., Balsamo, G., Bechtold, P., Biavati, G., Bidlot, J., Bonavita, M., De Chiara, G., Dahlgren, P., Dee, D., Diamantakis, M., Dragani, R., Flemming, J., Forbes, R., Fuentes, M., Geer, A., Haimberger, L., Healy, S., Hogan, R. J., Hólm, E., Janisková, M., Keeley, S., Laloyaux, P., Lopez, P., Lupu, C., Radnoti, G., de Rosnay, P., Rozum, I., Vamborg, F., Villaume, S., and Thépaut, J.-N.: The ERA5 global reanalysis, Q. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803, 2020.
Hochreiter, S. and Schmidhuber, J.: Long Short-Term Memory, Neural Comput., 9, 1735–1780, https://doi.org/10.1162/neco.1997.9.8.1735, 1997.
Hoedt, P.-J., Kratzert, F., Klotz, D., Halmich, C., Holzleitner, M., Nearing, G. S., Hochreiter, S., and Klambauer, G.: MC-LSTM: Mass-Conserving LSTM, Proceedings of the 38th International Conference on Machine Learning, PMLR, 139, 4275–4286, 2021.
Hrachowitz, M., Savenije, H. H. G., Blöschl, G., McDonnell, J. J., Sivapalan, M., Pomeroy, J. W., Arheimer, B., Blume, T., Clark, M. P., 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, https://doi.org/10.1080/02626667.2013.803183, 2013.
Jin, H., Song, Q., and Hu, X.: Auto-Keras: An efficient neural architecture search system, Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 4–8 August 2019, Anchorage, AK, USA, https://doi.org/10.48550/arXiv.1806.10282, 2019.
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.
Kratzert, F., Klotz, D., Herrnegger, M., Sampson, A. K., Hochreiter, S., and Nearing, G. S.: Toward improved predictions in ungauged basins: Exploiting the power of machine learning, Water Resour. Res., 55, 11344–11354, https://doi.org/10.1029/2019WR026065, 2019a.
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, 2019b.
Kratzert, F., Gauch, M., Nearing, G., and Klotz, D.: NeuralHydrology – A Python library for Deep Learning, J. Open Source Softw., 7, 4050, https://doi.org/10.21105/joss.04050, 2022.
LeCun, Y., Bengio, Y., and Hinton, G.: Deep learning, Nature, 521, 436–444, https://doi.org/10.1038/nature14539, 2015.
Lees, T., Reece, S., Kratzert, F., Klotz, D., Gauch, M., De Bruijn, J., Kumar Sahu, R., Greve, P., Slater, L., and Dadson, S. J.: Hydrological concept formation inside long short-term memory (LSTM) networks, Hydrol. Earth Syst. Sci., 26, 3079–3101, https://doi.org/10.5194/hess-26-3079-2022, 2022.
Li, Q., Zhu, Y., Shangguan, W., Wang, X., Li, L., and Yu, F.: An attention-aware LSTM model for soil moisture and soil temperature prediction, Geoderma, 409, 115651, https://doi.org/10.1016/j.geoderma.2021.115651, 2022.
Li, X., Khandelwal, A., Jia, X., Cutler, K., Ghosh, R., Renganathan, A., Xu, S., Tayal, K., Nieber, J., Duffy, C., and Steinbach, M.: Regionalization in a global hydrologic deep learning model: from physical descriptors to random vectors, Water Resour. Res., 58, e2021WR031794, https://doi.org/10.1029/2021WR031794, 2022.
Mai, J., Shen, H., Tolson, B. A., Gaborit, É., Arsenault, R., Craig, J. R., Fortin, V., Fry, L. M., Gauch, M., Klotz, D., Kratzert, F., O'Brien, N., Princz, D. G., Rasiya Koya, S., Roy, T., Seglenieks, F., Shrestha, N. K., Temgoua, A. G. T., Vionnet, V., and Waddell, J. W.: The Great Lakes Runoff Intercomparison Project Phase 4: the Great Lakes (GRIP-GL), Hydrol. Earth Syst. Sci., 26, 3537–3572, https://doi.org/10.5194/hess-26-3537-2022, 2022.
Martel, J.-L., Demeester, K., Brissette, F., Poulin, A., and Arsenault, R.: HMETS – A simple and efficient hydrology model for teaching hydrological modelling, flow forecasting and climate change impacts, Int. J. Eng. Educ., 33, 1307–1316, 2017.
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.
Nearing, G. S., Kratzert, F., Sampson, A. K., Pelissier, C. S., Klotz, D., Frame, J. M., Prieto, C., and Gupta, H. V.: What role does hydrological science play in the age of machine learning?, Water Resour. Res., 57, e2020WR028091, https://doi.org/10.1029/2020WR028091, 2021.
Nogueira Filho, F. J. M., Souza Filho, F. D. A., Porto, V. C., Vieira Rocha, R., Sousa Estácio, Á. B., and Martins, E. S. P. R.: Deep Learning for Streamflow Regionalization for Ungauged Basins: Application of Long-Short-Term-Memory Cells in Semiarid Regions, Water, 14, 1318, 2022.
Nourani, V., Khodkar, K., and Gebremichael, M.: Uncertainty assessment of LSTM based groundwater level predictions, Hydrolog. Sci. J., 67, 773–790, https://doi.org/10.1080/02626667.2022.2046755, 2022.
Oudin, L., Hervieu, F., Michel, C., Perrin, C., Andréassian, V., Anctil, F., and Loumagne, C.: Which potential evapotranspiration input for a lumped rainfall–runoff model?: Part 2 – Towards a simple and efficient potential evapotranspiration model for rainfall–runoff modelling, J. Hydrol., 303, 290–306, https://doi.org/10.1016/j.jhydrol.2004.08.026, 2005.
Oudin, L., Andréassian, V., Perrin, C., Michel, C., and Le Moine, N.: Spatial proximity, physical similarity, regression and ungaged catchments: A comparison of regionalization approaches based on 913 French catchments, Water Resour. Res., 44, W03413, https://doi.org/10.1029/2007WR006240, 2008.
Oudin, L., Kay, A., Andréassian, V., and Perrin, C.: Are seemingly physically similar catchments truly hydrologically similar?, Water Resour. Res., 46, W11558, https://doi.org/10.1029/2009WR008887, 2010.
Oyebode, O. and Stretch, D.: Neural network modeling of hydrological systems: A review of implementation techniques, Nat. Resour. Model., 32, e12189, https://doi.org/10.1111/nrm.12189, 2019.
Parajka, J., Merz, R., and Blöschl, G.: A comparison of regionalisation methods for catchment model parameters, Hydrol. Earth Syst. Sci., 9, 157–171, https://doi.org/10.5194/hess-9-157-2005, 2005.
Parajka, J., Viglione, A., Rogger, M., Salinas, J. L., Sivapalan, M., and Blöschl, G.: Comparative assessment of predictions in ungauged basins – Part 1: Runoff-hydrograph studies, Hydrol. Earth Syst. Sci., 17, 1783–1795, https://doi.org/10.5194/hess-17-1783-2013, 2013.
Perrin, C., Michel, C., and Andréassian, V.: Improvement of a parsimonious model for streamflow simulation, J. Hydrol., 279, 275–289, https://doi.org/10.1016/S0022-1694(03)00225-7, 2003.
Perrin, C., Oudin, L., Andreassian, V., Rojas-Serna, C., Michel, C., and Mathevet, T.: Impact of limited streamflow data on the efficiency and the parameters of rainfall – runoff models, Hydrolog. Sci. J., 52, 131–151, https://doi.org/10.1623/hysj.52.1.131, 2007.
Poissant, D., Arsenault, R., and Brissette, F.: Impact of parameter set dimensionality and calibration procedures on streamflow prediction at ungauged catchments, Journal of Hydrology: Regional Studies, 12, 220–237, https://doi.org/10.1016/j.ejrh.2017.05.005, 2017.
Razavi, T. and Coulibaly, P.: Streamflow prediction in ungauged basins: Review of regionalization methods, J. Hydrol. Eng., 18, 958–975, https://doi.org/10.1061/(ASCE)HE.1943-5584.0000690, 2013.
Robinson, N., Regetz, J., and Guralnick, R. P.: EarthEnv-DEM90: A nearly-global, void-free, multi-scale smoothed, 90 m digital elevation model from fused ASTER and SRTM data, ISPRS J. Photogramm., 87, 57–67, https://doi.org/10.1016/j.isprsjprs.2013.11.002, 2014.
Salinas, J. L., Laaha, G., Rogger, M., Parajka, J., Viglione, A., Sivapalan, M., and Blöschl, G.: Comparative assessment of predictions in ungauged basins – Part 2: Flood and low flow studies, Hydrol. Earth Syst. Sci., 17, 2637–2652, https://doi.org/10.5194/hess-17-2637-2013, 2013.
Shen, H., Tolson, B. A., and Mai, J.: Time to update the split-sample approach in hydrological model calibration, Water Resour. Res., 58, e2021WR031523, https://doi.org/10.1029/2021WR031523, 2022.
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, https://doi.org/10.1623/hysj.48.6.857.51421, 2003.
Skøien, J. O. and Blöschl, G.: Spatiotemporal topological kriging of runoff time series, Water Resour. Res., 43, W09419, https://doi.org/10.1029/2006WR005760, 2007.
Tarek, M., Brissette, F. P., and Arsenault, R.: Evaluation of the ERA5 reanalysis as a potential reference dataset for hydrological modelling over North America, Hydrol. Earth Syst. Sci., 24, 2527–2544, https://doi.org/10.5194/hess-24-2527-2020, 2020a.
Tarek, M., Brissette, F. P., and Arsenault, R.: Large-scale analysis of global gridded precipitation and temperature datasets for climate change impact studies, J. Hydrometeorol., 21, 2623–2640, https://doi.org/10.1175/jhm-d-20-0100.1, 2020b.
Valéry, A., Andréassian, V., and Perrin, C.: “As simple as possible but not simpler”: What is useful in a temperature-based snow-accounting routine? Part 2 – Sensitivity analysis of the Cemaneige snow accounting routine on 380 catchments, J. Hydrol., 517, 1176–1187, https://doi.org/10.1016/j.jhydrol.2014.04.058, 2014.
Viglione, A., Parajka, J., Rogger, M., Salinas, J. L., Laaha, G., Sivapalan, M., and Blöschl, G.: Comparative assessment of predictions in ungauged basins – Part 3: Runoff signatures in Austria, Hydrol. Earth Syst. Sci., 17, 2263–2279, https://doi.org/10.5194/hess-17-2263-2013, 2013.
Wagener, T. and Wheater, H. S.: Parameter estimation and regionalization for continuous rainfall-runoff models including uncertainty, J. Hydrol., 320, 132–154, https://doi.org/10.1016/j.jhydrol.2005.07.015, 2006.
Wagener, T., Wheater, H., and Gupta, H. V.: Rainfall-runoff modelling in gauged and ungauged catchments, World Scientific, ISBN 1-86094-466-3, 2004.
Yadav, M., Wagener, T., and Gupta, H.: Regionalization of constraints on expected watershed response behavior for improved predictions in ungauged basins, Adv. Water Resour., 30, 1756–1774, https://doi.org/10.1016/j.advwatres.2007.01.005, 2007.
Zhang, Y. and Chiew, F. H. S.: Relative merits of different methods for runoff predictions in ungauged catchments, Water Resour. Res., 45, W07412, https://doi.org/10.1029/2008WR007504, 2009.
Zhang, Y., Chiew, F. H. S., Li, M., and Post, D.: Predicting runoff signatures using regression and hydrological modeling approaches, Water Resour. Res., 54, 7859–7878, https://doi.org/10.1029/2018WR023325, 2018.
Zhang, Y., Ragettli, S., Molnar, P., Fink, O., and Peleg, N.: Generalization of an Encoder-Decoder LSTM model for flood prediction in ungauged catchments, J. Hydrol., 614, 128577, https://doi.org/10.1016/j.jhydrol.2022.128577, 2022.
Short summary
Predicting flow in rivers where no observation records are available is a daunting task. For decades, hydrological models were set up on these gauges, and their parameters were estimated based on the hydrological response of similar or nearby catchments where records exist. New developments in machine learning have now made it possible to estimate flows at ungauged locations more precisely than with hydrological models. This study confirms the performance superiority of machine learning models.
Predicting flow in rivers where no observation records are available is a daunting task. For...
