Zhao et al. (2025) Correction–fusion of NWP precipitation conditioned by rainfall stations and multivariate environmental information

Identification

• Journal: Advances in Space Research
• Year: 2025
• Date: 2025-11-11
• Authors: Qingzhi Zhao, Peng Geng, Zufeng Li, Yibin Yao, Yatong Li, Xiaohua Fu, Qiong Wu
• DOI: 10.1016/j.asr.2025.11.024

Research Groups

• College of Geomatics, Xi’an University of Science and Technology, Xi’an, China
• Powerchina Northwest Engineering Corporation Limited, Xi’an, China
• School of Geodesy and Geomatics, Wuhan University, Wuhan, China

Short Summary

This study proposes a novel correction-fusion method for numerical weather prediction (NWP) precipitation, integrating station observations and multivariate environmental information to generate high-resolution, high-precision precipitation products. The merged precipitation significantly improves accuracy, precipitation detection, and hydrological utility compared to raw NWP model output.

Objective

• To develop a correction–fusion method for NWP precipitation that considers multiple environmental information and combines Weather Research and Forecasting (WRF) model grid precipitation and station observation precipitation to obtain high-resolution, high-precision precipitation products.

Study Configuration

• Spatial Scale: Regional (implied high-resolution output).
• Temporal Scale: Sub-daily to daily (evidenced by 12-hour accumulated rain analysis and precipitation occurrence time).

Methodology and Data

• Models used: Weather Research and Forecasting (WRF) model, WRF-Hydro model, MGWR (Multiscale Geographically Weighted Regression, implied by keywords).
• Data sources: Numerical Weather Prediction (NWP) model grid precipitation (specifically WRF model), rainfall station observation precipitation, multivariate environmental information.

Main Results

• The accuracy of the merged precipitation data at four rainfall stations is superior to that of WRF precipitation.
• The average improvement rates for the correlation coefficient and root-mean-square error (RMSE) are 32.7 % and 25.9 %, respectively.
• The merged precipitation product demonstrates a strong ability to capture the occurrence time of precipitation, with an average probability of detection (POD) of 0.99 at four rainfall stations.
• The merged precipitation product shows good improvement for 12-hour accumulated light rain (32.5 %), moderate rain (39.6 %), heavy rain (36.7 %), and torrential rain (0.5 %).
• The WRF-Hydro model driven by the merged precipitation effectively depicts rainfall–runoff process curves, leading to improved accuracy in streamflow simulation.

Contributions

• Introduces a novel correction-fusion method that effectively integrates NWP model output, station observations, and environmental information to enhance precipitation product accuracy.
• Provides a high-resolution, high-precision precipitation product that significantly outperforms raw NWP model data in terms of correlation, RMSE, and precipitation event detection.
• Demonstrates the practical hydrological utility of the merged precipitation product by improving streamflow simulation accuracy when used as input for the WRF-Hydro model.

Funding

• Not specified in the provided text.

Citation

@article{Zhao2025Correctionfusion,
  author = {Zhao, Qingzhi and Geng, Peng and Li, Zufeng and Yao, Yibin and Li, Yatong and Fu, Xiaohua and Wu, Qiong},
  title = {Correction–fusion of NWP precipitation conditioned by rainfall stations and multivariate environmental information},
  journal = {Advances in Space Research},
  year = {2025},
  doi = {10.1016/j.asr.2025.11.024},
  url = {https://doi.org/10.1016/j.asr.2025.11.024}
}