Abstract：To address the limitations of traditional numerical weather prediction models in short-term and nowcasting precipitation forecasting,this study investigates high-resolution radar data processing and feature extraction techniques.The proposed improved bio-echo extraction network,MFF-PSPnet,demonstrates superior adaptability and performance in small-sample scenarios through structural innovations.Results indicate that this fusion technique effectively improves forecast accuracy,and its application in operational meteorology can strengthen disaster early warning capabilities and decision support,holding significant implications for the future development of precipitation forecasting technologies.
范婷婷, 马千惠, 李晓杰. 高分辨力气象雷达回波特征提取与降水类型识别模型研究[J]. 气象水文海洋仪器, 2026, 43(5): 13-16.
Fan Tingting, Ma Qianhui, Li Xiaojie. Research on high resolution weather radar echo feature extraction and precipitation type recognition model. Meteorological Hydrological and Marine Instrument, 2026, 43(5): 13-16.
2026, Vol. 43 
