Exploring the performance and interpretability of hybrid hydrologic model coupling physical mechanisms and deep learning DOI

Miao He,

S. S. Jiang, Liliang Ren

et al.

Journal of Hydrology, Journal Year: 2024, Volume and Issue: 649, P. 132440 - 132440

Published: Dec. 3, 2024

Language: Английский

Advancing streamflow prediction in data-scarce regions through vegetation-constrained distributed hybrid ecohydrological models DOI
L. Zhong, Huimin Lei, Zhiyuan Li

et al.

Journal of Hydrology, Journal Year: 2024, Volume and Issue: 645, P. 132165 - 132165

Published: Oct. 19, 2024

Language: Английский

Citations

4
Enhancing representation of data-scarce reservoir-regulated river basins using a hybrid DL-process based approach DOI
Liangkun Deng, Xiang Zhang, Louise Slater

et al.

Journal of Hydrology, Journal Year: 2025, Volume and Issue: 655, P. 132895 - 132895

Published: Feb. 20, 2025

Language: Английский

Citations

0
On the future of hydroecological models of everywhere DOI
Keith Beven

Environmental Modelling & Software, Journal Year: 2025, Volume and Issue: unknown, P. 106431 - 106431

Published: March 1, 2025

Language: Английский

Citations

0
High‐Resolution National‐Scale Water Modeling Is Enhanced by Multiscale Differentiable Physics‐Informed Machine Learning DOI Creative Commons
Yalan Song, Tadd Bindas, Chaopeng Shen

et al.

Water Resources Research, Journal Year: 2025, Volume and Issue: 61(4)

Published: April 1, 2025

Abstract The National Water Model (NWM) is a key tool for flood forecasting, planning, and water management. Key challenges facing the NWM include calibration parameter regionalization when confronted with big data. We present two novel versions of high‐resolution (∼37 km 2 ) differentiable models (a type hybrid model): one implicit, unit‐hydrograph‐style routing another explicit Muskingum‐Cunge in river network. former predicts streamflow at basin outlets whereas latter presents discretized product that seamlessly covers rivers conterminous United States (CONUS). Both use neural networks to provide multiscale parameterization process‐based equations structural backbone, which were trained simultaneously (“end‐to‐end”) on 2,807 basins across CONUS evaluated 4,997 basins. show great potential elevate future performance extensively calibrated as well ungauged sites: median daily Nash‐Sutcliffe efficiency all improved around 0.68 from 0.48 NWM3.0. As they resolve spatial heterogeneity, both greatly simulations western also Prairie Pothole Region, long‐standing modeling challenge. version further >10,000 . Overall, our results how neural‐network‐based parameterizations can improve providing operational predictions while maintaining interpretability multivariate outputs. system supports Basic Interface (BMI), allows seamless integration next‐generation NWM. CONUS‐scale hydrologic data set evaluation use.

Language: Английский

Citations

0
Exploring the performance and interpretability of hybrid hydrologic model coupling physical mechanisms and deep learning DOI

Miao He,

S. S. Jiang, Liliang Ren

et al.

Journal of Hydrology, Journal Year: 2024, Volume and Issue: 649, P. 132440 - 132440

Published: Dec. 3, 2024

Language: Английский

Citations

1