Abstract: The raindrop size distribution characteristics for extreme precipitation in central and southern Hebei from 29 Jul to 1 Aug 2023, were analyzed by using the raindrop spectrometer observation data. Results show that when the precipitation intensity R value is less than 50 mm/h, the raindrop diameter increases significantly with an increase in the value of R. However, as the R value increases, the raindrop diameter reaches a plateau, stabilizing at 2 mm without further increase. There are more data points where precipitation occurred in the oceanic convection characteristic area, whereas there are fewer data points where precipitation occurred in the continent characteristic area; however, there are still many data points in the transition interval between the two areas. Using the K-means clustering algorithm, precipitation samples were divided into five categories according to their droplet spectral characteristics, revealing obvious differences across various intensities. When the precipitation intensity is higher than 50 mm/h, the droplet spectral properties for precipitation exhibit large values. Because of the sustained stability of large particle diameters and high number concentrations, a sudden change in the raindrop spectrum and the continuous flow of water vapor transported by typhoons can produce heavy precipitation over a short period. Using the droplet spectral data to fit the radar precipitation estimation relationship, the authors obtain fitting formulas for radar reflectivity Z, differential reflectivity K_DP, and precipitation intensity R, Z=224R^1.44 and R=45.8K_DP^0.77. Results obtained using R–K_DP for quantifying the precipitation estimation relationship are better than those yielded using the Z–R relationship, especially over 40 mm/h precipitation, and the estimation error for the former is significantly better than that for the latter.