Currently, most of the researches based on dualpolarization weather radar focus on the way to identify the various phases of particles and detect the vertical structure of the clouds, and the researches on the identification and characteristics of Z_DR column are few. In order to strengthen the application of dual-polarization radar in Hainan, the identification algorithm of the Z_DR column based on the dual-polarization radar (2020－2021) is proposed according to the characteristics of convective cloud systems on the Hainan Island. Based on the method, combined with multi-sources data, two convective weather processes in Hainan are analyzed and studied to verify the Z_DR column. The results show that the algorithm can identify the Z_DR column on the Hainan Island; the algorithm could obtain the spatial distribution of the Z_DR column in real-time to analyze the convective weather processes, weather modification and short-term disaster warning. The Z_DR column is usually deep in the early and middle stages of the convective clouds, and the maximum can reach 3 to 5 km above the 0 ℃ layer; the Z_DR column with deeper spatial distribution is more likely to appear in the direction of the convective cloud movement. The position of the leading edge has a corresponding relationship with the area with large radar reflectivity (≥50 dBz), which continues to deepen with the evolution of the clouds. The average Z_DR column depth variation trend has a weak correlation with the radar echo top height and the vertical accumulated liquid water content. The changes of the radar echo top height and the vertical accumulated liquid water content usually lag the Z_DR column depth (12 to 15 min).