Abstract:Hail weather is characterized by sudden onset, pronounced locality, and significant destructive power, posing multiple adverse impacts on human production and daily life. Accurate and timely identification of hail weather holds critical importance for disaster early warning and prevention. Although Doppler weather radar observations play a vital role in hail identification, the limited spatiotemporal coverage of single-source radar data and the insufficient timeliness and accuracy of traditional identification methods remain challenging. To address the limitations of single data sources, this study proposes a hail identification method based on satellite and radar data fusion. Leveraging the spatiotemporal complementarity between satellite and radar data, the method combines threshold characteristics of satellite and radar observations before and after hailfall to achieve efficient multi-source data fusion and identification through deep learning algorithms. Experimental results demonstrate that the proposed method effectively integrates satellite and radar data, with the YOLOv7 model achieving a recognition accuracy of 90.83%. It successfully identifies hail-affected regions, providing crucial references for hail weather early warning. Notably, in areas where radar data are susceptible to terrain occlusion, the method significantly mitigates inaccurate hail region identification caused by poor data quality, demonstrating high practical value.