Yan et al. (2025) Evaluating the Hydrological Applicability of Satellite Precipitation Products Using a Differentiable, Physics-Based Hydrological Model in the Xiangjiang River Basin, China

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Remote Sensing
• Year: 2025
• Date: 2025-12-31
• Authors: Shixiong Yan, Changbo Jiang, Yuannan Long, Xinkui Wang
• DOI: 10.3390/rs18010137

Research Groups

Not explicitly mentioned in the provided text.

Short Summary

This study systematically evaluates the suitability of multi-source satellite precipitation products for driving a distributed physics-informed deep learning (DPDL) model and a SWAT model in the Xiangjiang River Basin, finding that DPDL outperforms SWAT and that product-specific recalibration significantly improves streamflow simulation accuracy, with overall utility depending on both model architecture and training strategy.

Objective

• To systematically evaluate the suitability and compatibility of multi-source satellite precipitation products within the modeling frameworks of a distributed physics-informed deep learning (DPDL) model and a Soil and Water Assessment Tool (SWAT) model, under different training strategies, for streamflow simulation.

Study Configuration

• Spatial Scale: Xiangjiang River Basin, southern China.
• Temporal Scale: Not explicitly specified in the provided text (implies a period for training and validation).

Methodology and Data

• Models used: Distributed physics-informed deep learning model (DPDL), Soil and Water Assessment Tool (SWAT).
• Data sources: Satellite precipitation products (GSMaP, IMERG-F, CMORPH). Streamflow observations (implied for model evaluation).

Main Results

• In the Xiangjiang River Basin, GSMaP demonstrated the best overall performance with a Critical Success Index of 0.70 and a correlation coefficient of 0.79.
• IMERG-F showed acceptable accuracy with a correlation coefficient of 0.75 but had a relatively high false alarm rate of 0.32.
• CMORPH exhibited the most significant systematic underestimation with a relative bias of -8.48%.
• The DPDL model more effectively captured watershed hydrological dynamics, achieving a validation period correlation coefficient of 0.82 and a Nash–Sutcliffe efficiency of 0.79, outperforming the SWAT model.
• The DPDL model showed a higher relative bias of +16.69% during the validation period, along with greater overestimation fluctuations during dry periods, revealing limitations when training samples are limited.
• The S2 strategy (product-specific recalibration) improved the streamflow simulation accuracy for most precipitation products, with the maximum increase in the Nash–Sutcliffe efficiency coefficient reaching 15.8%.
• The hydrological utility of satellite products is jointly determined by model architecture and training strategy. For the DPDL model, IMERG-F demonstrated the best overall robustness, while GSMaP achieved the highest accuracy under the S2 strategy.

Contributions

• Provides a systematic evaluation of multi-source satellite precipitation products for emerging differentiable, physics-based hydrological models (DPDL).
• Offers theoretical support for optimizing differentiable hydrological modeling.
• Presents new perspectives for evaluating the hydrological utility of satellite precipitation products, highlighting the joint influence of model architecture and training strategy.

Funding

Not explicitly mentioned in the provided text.

Citation

@article{Yan2025Evaluating,
  author = {Yan, Shixiong and Jiang, Changbo and Long, Yuannan and Wang, Xinkui},
  title = {Evaluating the Hydrological Applicability of Satellite Precipitation Products Using a Differentiable, Physics-Based Hydrological Model in the Xiangjiang River Basin, China},
  journal = {Remote Sensing},
  year = {2025},
  doi = {10.3390/rs18010137},
  url = {https://doi.org/10.3390/rs18010137}
}