Abstract: Z_DR column, the quasi-vertical continuous column of enhanced differential reflectivity (i.e., Z_DR≥1 dB) observed by dual-polarization radar, can extend well above the environmental 0℃ level. Z_DR column can provide information about the location and intensity of convective storm updraft, which makes it a useful tool for analyzing the evolution of convective storms. This paper introduces an automatic Z_DR column identification algorithm, which is designed to provide diagnostic information pertinent to convective storm warning. Based on the 3D structure characteristics of convective storms, the algorithm for Z_DR column identification is designed and its morphological parameters are calculated. The application of Z_DR column morphological parameters in quantitative analysis of convective storms is explored by using Xiamen dual polarization radar and automatic weather station data. The study yields the following results. (1) Statistically significant differences exist between severe and non-severe storms in terms of the Z_DR column morphological parameters, indicating that these products can provide references for forecasters to distinguish the two types of convective storms. Once the Z_DR column depth reaches 1500 m, at least 60% of the volume scans are associated with severe storms. Similarly, once the thresholds for Z_DR column volume, centroid height and maximum Z_DR value reach 20 m³, 500 m and 3 dB, respectively, at least 70%, 70% and 50% of the volume scans are associated with severe storms. (2) The evolution of Z_DR columns is an appropriate index that can reflect the development of convective storms, and the peak values of its morphological parameters precede the occurrence of severe convective weather. During the continuous severe convective weather process, the re-development of the Z_DR column is earlier than that of the convective storm. (3) Z_DR columns can predict the merging and splitting process of convective storms. The process of storm merging (splitting) is accompanied by the Z_DR column merging (splitting). Z_DR column merging (splitting) occurs earlier than that of the convective storm in 57% (69%) of the processes. (4) There is a correlation betweenthe position of the Z_DR column and subsequent propagation direction of the convective storm, which can provide a reference for improving the prediction of the movement path of convective storm.