A rapid identification and warning method for severe weather via Doppler radar based on an improved TITAN algorithm

Abstract

An adaptive fast-recognition warning algorithm for severe weather using ground-based Doppler radar and Tropical Rainfall Measuring Mission (TRMM) sensor data is proposed for different seasons, regions and topography conditions in the East and South China Seas. An improved regional segmentation method and the Thunderstorm Identification, Tracking, Analysis and Nowcasting (TITAN) algorithm were applied to identify three-dimensional strong storm cells and their physical features. Multiple logistic linear regression was used to establish a probabilistic warning model for strong convective weather, such as hail and lightning. Doppler radar identification and warning experiments were carried out for a strong squall line near the East China Sea and a supercell storm near the South China Sea. The experimental results showed that the proposed method has higher identification accuracy based on the azimuth and distance than the traditional algorithm. Moreover, the false negative and false report rates of the proposed method are low, which helps to quickly identify and warn against severe weather and protect lives and property.