Abstract:The decision tree analysis method is applied to Doppler radar data to identify hail weather. Based on the radar data of 25 cases of hails in the Tianjin area from April to September, 2005 to 2014, we select 6 parameters: probability of hail (POH), probability of severe hail (POSH), vertically integrated liquid water (VIL), maximum reflectivity (DBZM), the height of storms centroid (HT), and the top of storm (TOP). By the decision tree analysis method, we found that there will be hail weather when DBZM is greater that 545 dBz and POSH is greater than 35 Using the decision tree analysis method,we have better results: the percent of doom (POD) is 902% and critical success index (CSI) is 784%, which are higher than the results by the discriminant analysis method.