Based on the C-band radar echo intensity and radial wind data of Guangzhou Baiyun Airport, the precipitation amount estimation and the rain belt moving prediction are carried out for the heavy precipitation process in South China during the first flood season. In the strong precipitation background of abundant rainfall areas, the C-band radar has better ability to detect cloud rain refinement structures, although the echo intensity is attenuated. The radar data quality control processing can preserve the original characteristics of the radar data, effectively filter out clutter and noise, and alleviate the problem that the radar data has many clutter at low elevation angle scanning surface and the altitude difference along radial direction. The data processing provides a technique support for precipitation estimation. The optimized radar Z-I precipitation estimation method is only based on information of each scanning procedure, so its calculation is simple and fast. However, the cloud type difference on the same scanning surface is sensitive to the parameter values of the optimization method, when the same parameters are used in different clouds that will affect the estimation effect of precipitation. By separating areas of the radar cloud echoes on the same scanning surface and selecting rainfall stations, the optimization parameters can be determined according to different cloud echoes, which can improve the optimal radar precipitation estimation. According to the short characteristics of radar scanning time, the radar radial wind convergence line and the strong echo center position are identified, and the extrapolation method is used to predict the rain belt orientation, position and moving speed. The verification effect is good.