Shen et al. (2025) An effective gauge-satellite fusion approach for daily precipitation bias correction based on multi-dimensional precipitation feature space (BCFS)

Identification

• Journal: Atmospheric Research
• Year: 2025
• Date: 2025-11-21
• Authors: Gaoyun Shen, Li Zeng, Lin Fu, Lei Wang, Junann Xiong, Wei Wang
• DOI: 10.1016/j.atmosres.2025.108624

Research Groups

• Southwest Petroleum University, School of Civil Engineering and Geomatics, China
• Xizang Autonomous Region Key Laboratory of Satellite Remote Sensing and Application, China
• Wuhan University, State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, China

Short Summary

This paper proposes a novel daily precipitation bias-correction approach, BCFS, which integrates spatial patterns with temporal attributes into a unified precipitation feature space for gauge-satellite fusion. The method significantly improves the accuracy of daily precipitation estimates, especially in gauge-sparse regions, by effectively reducing biases across various precipitation intensities.

Objective

• To develop a novel daily precipitation bias-correction approach (BCFS) that addresses limitations of existing methods by integrating spatiotemporal variability into a unified precipitation feature space for gauge-satellite fusion.

Study Configuration

• Spatial Scale: Han River basin, Jinsha River basin (China)
• Temporal Scale: 1998–2020

Methodology and Data

• Models used: BCFS (proposed method), WHU-SGCC (comparison method)
• Data sources: Gauge observations, CHIRPS (Climate Hazards Group Infrared Precipitation with Station data)

Main Results

• BCFS reduced relative bias by 74.46 %, 85.49 %, 99.98 %, 98.69 %, 35.78 %, and 27.78 % across different precipitation intensities compared to uncorrected CHIRPS data.
• BCFS outperformed the WHU-SGCC method, improving accuracy for below-average precipitation and reducing biases at higher intensities.
• Validation in the Jinsha River basin confirmed regional applicability, showing improved extreme precipitation indices and revealing increasing trends in spring extreme events in the basin’s upper reaches.

Contributions

• Proposes a novel BCFS method that constructs a unified precipitation feature space by integrating spatial patterns with temporal attributes, overcoming limitations of existing gauge-satellite fusion methods.
• Demonstrates substantial improvements in daily precipitation estimates, particularly in gauge-sparse regions, by effectively capturing spatiotemporal characteristics and significantly reducing biases across various precipitation intensities.
• Provides robust support for hydrometeorological monitoring and extreme-event forecasting through enhanced high-resolution precipitation data accuracy.

Funding

[Not mentioned in the provided text.]

Citation

@article{Shen2025effective,
  author = {Shen, Gaoyun and Zeng, Li and Fu, Lin and Wang, Lei and Xiong, Junann and Wang, Wei},
  title = {An effective gauge-satellite fusion approach for daily precipitation bias correction based on multi-dimensional precipitation feature space (BCFS)},
  journal = {Atmospheric Research},
  year = {2025},
  doi = {10.1016/j.atmosres.2025.108624},
  url = {https://doi.org/10.1016/j.atmosres.2025.108624}
}