Articles | Volume 28, issue 17
https://doi.org/10.5194/hess-28-4187-2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Special issue:
https://doi.org/10.5194/hess-28-4187-2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
HESS Opinions: Never train a Long Short-Term Memory (LSTM) network on a single basin
Google Research, Vienna, Austria
Martin Gauch
Google Research, Zurich, Switzerland
Daniel Klotz
Helmholtz Centre for Environmental Research – UFZ, Leipzig, Germany
Grey Nearing
Google Research, Mountain View, California, USA
Viewed
Total article views: 5,377 (including HTML, PDF, and XML)
Cumulative views and downloads
(calculated since 18 Jan 2024)
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
4,104 | 1,174 | 99 | 5,377 | 66 | 54 |
- HTML: 4,104
- PDF: 1,174
- XML: 99
- Total: 5,377
- BibTeX: 66
- EndNote: 54
Total article views: 1,865 (including HTML, PDF, and XML)
Cumulative views and downloads
(calculated since 12 Sep 2024)
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
1,611 | 235 | 19 | 1,865 | 18 | 21 |
- HTML: 1,611
- PDF: 235
- XML: 19
- Total: 1,865
- BibTeX: 18
- EndNote: 21
Total article views: 3,512 (including HTML, PDF, and XML)
Cumulative views and downloads
(calculated since 18 Jan 2024)
HTML | XML | Total | BibTeX | EndNote | |
---|---|---|---|---|---|
2,493 | 939 | 80 | 3,512 | 48 | 33 |
- HTML: 2,493
- PDF: 939
- XML: 80
- Total: 3,512
- BibTeX: 48
- EndNote: 33
Viewed (geographical distribution)
Total article views: 5,377 (including HTML, PDF, and XML)
Thereof 5,011 with geography defined
and 366 with unknown origin.
Total article views: 1,865 (including HTML, PDF, and XML)
Thereof 1,802 with geography defined
and 63 with unknown origin.
Total article views: 3,512 (including HTML, PDF, and XML)
Thereof 3,209 with geography defined
and 303 with unknown origin.
Country | # | Views | % |
---|
Country | # | Views | % |
---|
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1
1
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1
1
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1
1
Cited
31 citations as recorded by crossref.
- Are LSTM and conceptual rainfall-runoff models able to cope with limited training datasets under diverse hydrometeorological conditions? F. Boodoo et al. 10.1007/s40808-025-02316-z
- A diversity-centric strategy for the selection of spatio-temporal training data for LSTM-based streamflow forecasting E. Snieder & U. Khan 10.5194/hess-29-785-2025
- Regional scale simulations of daily suspended sediment concentration at gauged and ungauged rivers using deep learning A. Gupta & D. Feng 10.1016/j.jhydrol.2025.133111
- Ensemble learning of catchment-wise optimized LSTMs enhances regional rainfall-runoff modelling − case Study: Basque Country, Spain F. Hosseini et al. 10.1016/j.jhydrol.2024.132269
- A nonstationary stochastic simulator for clustered regional hydroclimatic extremes to characterize compound flood risk A. Nayak et al. 10.1016/j.hydroa.2024.100189
- Local vs regional neural air pollution forecasting models M. Sangiorgio & G. Guariso 10.1016/j.ifacsc.2025.100298
- Analyzing the generalization capabilities of a hybrid hydrological model for extrapolation to extreme events E. Acuña Espinoza et al. 10.5194/hess-29-1277-2025
- A differentiable, physics-based hydrological model and its evaluation for data-limited basins W. Ouyang et al. 10.1016/j.jhydrol.2024.132471
- On the value of a history of hydrology and the establishment of a History of Hydrology Working Group K. Beven et al. 10.1080/02626667.2025.2452357
- HESS Opinions: Never train a Long Short-Term Memory (LSTM) network on a single basin F. Kratzert et al. 10.5194/hess-28-4187-2024
- Detection and attribution of eco-hydrological alteration based on deep learning-driven gap-filled runoff in a large-scale catchment Z. Dong et al. 10.1016/j.ejrh.2025.102228
- HydroEcoLSTM: A Python package with graphical user interface for hydro-ecological modeling with long short-term memory neural network T. Nguyen et al. 10.1016/j.ecoinf.2025.102994
- Empowering Regional Rainfall-Runoff Modeling Through Encoder–Decoder Based on Convolutional Neural Networks W. Jiang et al. 10.3390/w17030339
- Fine-tuning long short-term memory models for seamless transition in hydrological modelling: From pre-training to post-application X. Chen et al. 10.1016/j.envsoft.2025.106350
- Technical note: An approach for handling multiple temporal frequencies with different input dimensions using a single LSTM cell E. Acuña Espinoza et al. 10.5194/hess-29-1749-2025
- Defense and Security Mechanisms in the Internet of Things: A Review S. Szymoniak et al. 10.3390/app15020499
- CH-RUN: a deep-learning-based spatially contiguous runoff reconstruction for Switzerland B. Kraft et al. 10.5194/hess-29-1061-2025
- Learning extreme vegetation response to climate drivers with recurrent neural networks F. Martinuzzi et al. 10.5194/npg-31-535-2024
- Machine Learning-Based Process Optimization in Biopolymer Manufacturing: A Review I. Malashin et al. 10.3390/polym16233368
- Assessing groundwater level modelling using a 1-D convolutional neural network (CNN): linking model performances to geospatial and time series features M. Gomez et al. 10.5194/hess-28-4407-2024
- A short history of philosophies of hydrological model evaluation and hypothesis testing K. Beven 10.1002/wat2.1761
- Improved streamflow simulations in hydrologically diverse basins using physically-informed deep learning models B. Magotra et al. 10.1080/02626667.2025.2458545
- Hyperparameter optimization of regional hydrological LSTMs by random search: A case study from Basque Country, Spain F. Hosseini et al. 10.1016/j.jhydrol.2024.132003
- To bucket or not to bucket? Analyzing the performance and interpretability of hybrid hydrological models with dynamic parameterization E. Acuña Espinoza et al. 10.5194/hess-28-2705-2024
- Improving streamflow predictions across CONUS by integrating advanced machine learning models and diverse data K. Tayal et al. 10.1088/1748-9326/ad6fb7
- Large-sample hydrology – a few camels or a whole caravan? F. Clerc-Schwarzenbach et al. 10.5194/hess-28-4219-2024
- Temporal Fusion Transformers for streamflow Prediction: Value of combining attention with recurrence S. Rasiya Koya & T. Roy 10.1016/j.jhydrol.2024.131301
- Enhancing long short-term memory (LSTM)-based streamflow prediction with a spatially distributed approach Q. Yu et al. 10.5194/hess-28-2107-2024
- Investigating the Model Hypothesis Space: Benchmarking Automatic Model Structure Identification With a Large Model Ensemble D. Spieler & N. Schütze 10.1029/2023WR036199
- A data-centric perspective on the information needed for hydrological uncertainty predictions A. Auer et al. 10.5194/hess-28-4099-2024
- Stream Nitrate Dynamics Driven Primarily by Discharge and Watershed Physical and Soil Characteristics at Intensively Monitored Sites: Insights From Deep Learning G. Gorski et al. 10.1029/2023WR036591
20 citations as recorded by crossref.
- Are LSTM and conceptual rainfall-runoff models able to cope with limited training datasets under diverse hydrometeorological conditions? F. Boodoo et al. 10.1007/s40808-025-02316-z
- A diversity-centric strategy for the selection of spatio-temporal training data for LSTM-based streamflow forecasting E. Snieder & U. Khan 10.5194/hess-29-785-2025
- Regional scale simulations of daily suspended sediment concentration at gauged and ungauged rivers using deep learning A. Gupta & D. Feng 10.1016/j.jhydrol.2025.133111
- Ensemble learning of catchment-wise optimized LSTMs enhances regional rainfall-runoff modelling − case Study: Basque Country, Spain F. Hosseini et al. 10.1016/j.jhydrol.2024.132269
- A nonstationary stochastic simulator for clustered regional hydroclimatic extremes to characterize compound flood risk A. Nayak et al. 10.1016/j.hydroa.2024.100189
- Local vs regional neural air pollution forecasting models M. Sangiorgio & G. Guariso 10.1016/j.ifacsc.2025.100298
- Analyzing the generalization capabilities of a hybrid hydrological model for extrapolation to extreme events E. Acuña Espinoza et al. 10.5194/hess-29-1277-2025
- A differentiable, physics-based hydrological model and its evaluation for data-limited basins W. Ouyang et al. 10.1016/j.jhydrol.2024.132471
- On the value of a history of hydrology and the establishment of a History of Hydrology Working Group K. Beven et al. 10.1080/02626667.2025.2452357
- HESS Opinions: Never train a Long Short-Term Memory (LSTM) network on a single basin F. Kratzert et al. 10.5194/hess-28-4187-2024
- Detection and attribution of eco-hydrological alteration based on deep learning-driven gap-filled runoff in a large-scale catchment Z. Dong et al. 10.1016/j.ejrh.2025.102228
- HydroEcoLSTM: A Python package with graphical user interface for hydro-ecological modeling with long short-term memory neural network T. Nguyen et al. 10.1016/j.ecoinf.2025.102994
- Empowering Regional Rainfall-Runoff Modeling Through Encoder–Decoder Based on Convolutional Neural Networks W. Jiang et al. 10.3390/w17030339
- Fine-tuning long short-term memory models for seamless transition in hydrological modelling: From pre-training to post-application X. Chen et al. 10.1016/j.envsoft.2025.106350
- Technical note: An approach for handling multiple temporal frequencies with different input dimensions using a single LSTM cell E. Acuña Espinoza et al. 10.5194/hess-29-1749-2025
- Defense and Security Mechanisms in the Internet of Things: A Review S. Szymoniak et al. 10.3390/app15020499
- CH-RUN: a deep-learning-based spatially contiguous runoff reconstruction for Switzerland B. Kraft et al. 10.5194/hess-29-1061-2025
- Learning extreme vegetation response to climate drivers with recurrent neural networks F. Martinuzzi et al. 10.5194/npg-31-535-2024
- Machine Learning-Based Process Optimization in Biopolymer Manufacturing: A Review I. Malashin et al. 10.3390/polym16233368
- Assessing groundwater level modelling using a 1-D convolutional neural network (CNN): linking model performances to geospatial and time series features M. Gomez et al. 10.5194/hess-28-4407-2024
11 citations as recorded by crossref.
- A short history of philosophies of hydrological model evaluation and hypothesis testing K. Beven 10.1002/wat2.1761
- Improved streamflow simulations in hydrologically diverse basins using physically-informed deep learning models B. Magotra et al. 10.1080/02626667.2025.2458545
- Hyperparameter optimization of regional hydrological LSTMs by random search: A case study from Basque Country, Spain F. Hosseini et al. 10.1016/j.jhydrol.2024.132003
- To bucket or not to bucket? Analyzing the performance and interpretability of hybrid hydrological models with dynamic parameterization E. Acuña Espinoza et al. 10.5194/hess-28-2705-2024
- Improving streamflow predictions across CONUS by integrating advanced machine learning models and diverse data K. Tayal et al. 10.1088/1748-9326/ad6fb7
- Large-sample hydrology – a few camels or a whole caravan? F. Clerc-Schwarzenbach et al. 10.5194/hess-28-4219-2024
- Temporal Fusion Transformers for streamflow Prediction: Value of combining attention with recurrence S. Rasiya Koya & T. Roy 10.1016/j.jhydrol.2024.131301
- Enhancing long short-term memory (LSTM)-based streamflow prediction with a spatially distributed approach Q. Yu et al. 10.5194/hess-28-2107-2024
- Investigating the Model Hypothesis Space: Benchmarking Automatic Model Structure Identification With a Large Model Ensemble D. Spieler & N. Schütze 10.1029/2023WR036199
- A data-centric perspective on the information needed for hydrological uncertainty predictions A. Auer et al. 10.5194/hess-28-4099-2024
- Stream Nitrate Dynamics Driven Primarily by Discharge and Watershed Physical and Soil Characteristics at Intensively Monitored Sites: Insights From Deep Learning G. Gorski et al. 10.1029/2023WR036591
Latest update: 23 Apr 2025
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.
Recently, a special type of neural-network architecture became increasingly popular in hydrology...
Special issue