Abstract：Based on dual polarization radar data,a new precipitation type identification and prediction algorithm is proposed.This algorithm utilizes polarization parameters such as horizontal reflectance factor,differential reflectance,cross-correlation coefficient,and differential propagation phase shift obtained from dual polarization radar inversion,combined with fuzzy logic and support vector machine methods,to achieve precise identification of precipitation types.A precipitation type prediction model based on long short-term memory network has been constructed,which can predict the evolution of precipitation types 30 minutes in advance.The algorithm has been validated in actual precipitation cases,with a recognition accuracy of over 90% and a prediction accuracy of over 85%,demonstrating good application prospects.The research results have a positive impact on improving the quantitative precipitation estimation and forecasting capabilities.
余代辉, 伦绪勇, 兰世怀, 李圳坤, 蒙延雷. 基于双偏振雷达数据的降水类型识别与预测算法研究[J]. 气象水文海洋仪器, 2025, 42(1): 11-14.
Yu Daihui, Lun Xuyong, Lan Shihuai, Li Zhenkun, Meng Yanlei. Research on precipitation type identification and prediction algorithm based on dual polarization radar data. Meteorological Hydrological and Marine Instrument, 2025, 42(1): 11-14.
2025, Vol. 42 
