Using Deep Learning in Ensemble Streamflow Forecasting: Exploring the Predictive Value of Explicit Snowpack Information.
Saved in:
| Title: | Using Deep Learning in Ensemble Streamflow Forecasting: Exploring the Predictive Value of Explicit Snowpack Information. |
|---|---|
| Authors: | Modi, Parthkumar1,2 (AUTHOR) parthkumar.modi@colorado.edu, Jennings, Keith3 (AUTHOR), Kasprzyk, Joseph1 (AUTHOR), Small, Eric4 (AUTHOR), Wobus, Cameron5 (AUTHOR), Livneh, Ben1,2,6 (AUTHOR) |
| Source: | Journal of Advances in Modeling Earth Systems. Mar2025, Vol. 17 Issue 3, p1-22. 22p. |
| Subject Terms: | *Conservation of natural resources, *Water supply, Parameter estimation, Snow accumulation, Statistical services, Deep learning |
| Abstract: | The Ensemble Streamflow Prediction (ESP) framework combines a probabilistic forecast structure with process‐based models for water supply predictions. However, process‐based models require computationally intensive parameter estimation, increasing uncertainties and limiting usability. Motivated by the strong performance of deep learning models, we seek to assess whether the Long Short‐Term Memory (LSTM) model can provide skillful forecasts and replace process‐based models within the ESP framework. Given challenges in implicitly capturing snowpack dynamics within LSTMs for streamflow prediction, we also evaluated the added skill of explicitly incorporating snowpack information to improve hydrologic memory representation. LSTM‐ESPs were evaluated under four different scenarios: one excluding snow and three including snow with varied snowpack representations. The LSTM models were trained using information from 664 GAGES‐II basins during WY1983–2000. During a testing period, WY2001–2010, 80% of basins exhibited Nash‐Sutcliffe Efficiency (NSE) above 0.5 with a median NSE of around 0.70, indicating satisfactory utility in simulating seasonal water supply. LSTM‐ESP forecasts were then tested during WY2011–2020 over 76 western US basins with operational Natural Resources Conservation Services (NRCS) forecasts. A key finding is that in high snow regions, LSTM‐ESP forecasts using simplified ablation assumptions performed worse than those excluding snow, highlighting that snow data do not consistently improve LSTM‐ESP performance. However, LSTM‐ESP forecasts that explicitly incorporated past years' snow accumulation and ablation performed comparably to NRCS forecasts and better than forecasts excluding snow entirely. Overall, integrating deep learning within an ESP framework shows promise and highlights important considerations for including snowpack information in forecasting. Plain Language Summary: The Ensemble Streamflow Prediction (ESP) framework generates probabilistic water supply forecasts using process‐based models. However, process‐based models often face challenges because estimating their parameters is complex and introduces uncertainties. Inspired by emerging deep learning techniques, our study investigates whether the Long Short‐Term Memory (LSTM) model can provide skillful forecasts and potentially replace process‐based models within ESP. We also explore how including explicit information about snow, which significantly influences water flow, could enhance these forecasts. Our findings indicate that in regions with heavy snowfall, using a simpler representation of snowpack led to less accurate forecasts than those excluding snow information, suggesting that snowpack information does not consistently improve forecast performance. However, LSTM‐ESP forecasts incorporating a sophisticated representation of snowpack information performed comparably to current operational forecasts and better than those excluding snowpack information. Integrating deep learning techniques within an ESP could improve water supply forecasts, but careful consideration of how to incorporate snowpack information is crucial for optimal results. Key Points: A new combination of deep learning and Ensemble Streamflow Prediction is evaluatedDeep learning based seasonal streamflow forecasts perform similarly to Natural Resources Conservation Services statistical forecasts across different lead timesForecasts that incorporate comprehensive historical snowpack information perform better than those that exclude snow data [ABSTRACT FROM AUTHOR] |
| Copyright of Journal of Advances in Modeling Earth Systems is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | GreenFILE |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | The Ensemble Streamflow Prediction (ESP) framework combines a probabilistic forecast structure with process‐based models for water supply predictions. However, process‐based models require computationally intensive parameter estimation, increasing uncertainties and limiting usability. Motivated by the strong performance of deep learning models, we seek to assess whether the Long Short‐Term Memory (LSTM) model can provide skillful forecasts and replace process‐based models within the ESP framework. Given challenges in implicitly capturing snowpack dynamics within LSTMs for streamflow prediction, we also evaluated the added skill of explicitly incorporating snowpack information to improve hydrologic memory representation. LSTM‐ESPs were evaluated under four different scenarios: one excluding snow and three including snow with varied snowpack representations. The LSTM models were trained using information from 664 GAGES‐II basins during WY1983–2000. During a testing period, WY2001–2010, 80% of basins exhibited Nash‐Sutcliffe Efficiency (NSE) above 0.5 with a median NSE of around 0.70, indicating satisfactory utility in simulating seasonal water supply. LSTM‐ESP forecasts were then tested during WY2011–2020 over 76 western US basins with operational Natural Resources Conservation Services (NRCS) forecasts. A key finding is that in high snow regions, LSTM‐ESP forecasts using simplified ablation assumptions performed worse than those excluding snow, highlighting that snow data do not consistently improve LSTM‐ESP performance. However, LSTM‐ESP forecasts that explicitly incorporated past years' snow accumulation and ablation performed comparably to NRCS forecasts and better than forecasts excluding snow entirely. Overall, integrating deep learning within an ESP framework shows promise and highlights important considerations for including snowpack information in forecasting. Plain Language Summary: The Ensemble Streamflow Prediction (ESP) framework generates probabilistic water supply forecasts using process‐based models. However, process‐based models often face challenges because estimating their parameters is complex and introduces uncertainties. Inspired by emerging deep learning techniques, our study investigates whether the Long Short‐Term Memory (LSTM) model can provide skillful forecasts and potentially replace process‐based models within ESP. We also explore how including explicit information about snow, which significantly influences water flow, could enhance these forecasts. Our findings indicate that in regions with heavy snowfall, using a simpler representation of snowpack led to less accurate forecasts than those excluding snow information, suggesting that snowpack information does not consistently improve forecast performance. However, LSTM‐ESP forecasts incorporating a sophisticated representation of snowpack information performed comparably to current operational forecasts and better than those excluding snowpack information. Integrating deep learning techniques within an ESP could improve water supply forecasts, but careful consideration of how to incorporate snowpack information is crucial for optimal results. Key Points: A new combination of deep learning and Ensemble Streamflow Prediction is evaluatedDeep learning based seasonal streamflow forecasts perform similarly to Natural Resources Conservation Services statistical forecasts across different lead timesForecasts that incorporate comprehensive historical snowpack information perform better than those that exclude snow data [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 19422466 |
| DOI: | 10.1029/2024MS004582 |
