Abstract: There are sometimes great differences in the raindrop spectrum characteristics between short-term heavy rain (SHR) and persistent heavy rain (PHR) due to the different of cold cloud and warm cloud processes. Analyzing the difference of raindrop spectrum characteristics between SHR and PHR is helpful to understand the microphysical characteristics of the different types of heavy precipitation, which can play an important role in improving the estimation accuracy of radar quantitative precipitation. Taking an event with SHR and PHR processes caused by the Southwest China vortex in June 2018 in Hubei as an example, based on the observations from the automatic weather stations, CINRADA/SA Doppler weather radar data, raindrop spectrum from DSG5 precipitation phenomenon instruments, and the ERA5 reanalysis data, we comparatively analyzed the raindrop spectrum characteristics between SHR and PHR, and discussed their difference in the radar reflectivity (Z)-rainfall (R) relationship (Z=aR^b) by fitting analysis. Results are as follows. (1) In the SHR process of this event, the average number concentration (N_D) in the each particle size (D) is high, and the particle size is large for the convective cloud precipitation, which is related to the active ice phase process in the internal and the coalescence and collision-breakup process of raindrops in the warm cloud layer. While in the PHR process of this event, the ND of small particle size (D < 2mm) is high but the N_D of medium and large particle size is low for the stratiform cloud precipitation. (2) Distribution of the normalized Gamma spectral intercept parameter (lgN_w) and the mass weighted average diameter (D_m) shows that the SHR process has a wider raindrop spectrum, larger D_m and smaller lgN_w, and the slope of separation line of convective-stratiform cloud precipitation is smaller. (3) The Z-R fitting formulas for the convective cloud precipitation in the SHR and PHR processes are Z=183.33 R^1.56 and Z=169.74 R^1.49, respectively. When the rain intensity (R) is greater than 60 mm·h^-1, compared with the typical Z-R fitting formula (Z=300R^1.40) for Doppler weather radar, the Z-R fitting formula (Z=428.38R^1.37) for the convective cloud precipitation in the SHR process have a similar b value but a larger a value. If given Z, the R estimated by the Z-R fitting formula of SHR is weaker than that of PHR, which is caused by the existence of a small number of large particles.