To better characterize the microphysical differences among different types of rainstorms, this study conducts a comparative analysis on two rainstorm events in Hunan in June 2025, using data from precipitation spectrometers and Doppler weather radars, as well as ERA5 reanalysis data. The two events include a short-duration heavy precipitation dominated rainstorm from 7 to 9 (the “6·7” event) and a persistent precipitation dominated rainstorm from 21 to 23 (the “6·21” event). The results are shown below. (1) The synoptic situations of the two events are relatively similar, both influenced by a 500-hPa trough and a vortex shear line in the middle-lower level. However, compared with the “6·7” event, the synoptic systems in the “6·21” event are more stable with slower movement, leading to overlapped heavy rainfall areas, prolonged duration and large accumulated precipitation. (2) The “6·7” event exhibits vigorous convective development, with low radar echo centroid and high precipitation efficiency. In contrast, the “6·21” event has lower cloud tops and strong echo centers are located below 0 ℃, indicating predominantly stratiform cloud precipitation. (3) In both events, raindrops with diameters less than 1 mm contribute most to the number concentration, while raindrops with diameters of 1-3 mm contribute most to the amount of precipitation. The D_m and lg N_w of the two events are compared. The lg N_w and D_m values of the short-duration heavy precipitation event are 3.7-3.9 mm^-1·m^-3 and 2.0-2.3 mm, respectively; the persistent precipitation event, however, displays higher lg N_w 4.0-4.2 mm^-1·m^-3) and lower D_m(1.6-1.7 mm). (4) The analysis of the radar reflectivity-rainfall rate (Z-R) relationship reveals that the convective precipitation in the “6·7” event is characterized by a larger coefficient and a smaller exponent, while that in the “6·21” event has a smaller coefficient and a larger exponent. This study reveals the differences in the cloud physical mechanisms between the two types of rainstorms and can provide references for radar-based quantitative precipitation estimation of rainstorms in Hunan.