中文摘要：

针对一次华南暴雨过程，采用WRF区域中尺度模式进行了控制试验和同化试验。利用WRF-3DVAR同化系统同化了常规探空和地面观测资料，分析了两种资料对初值场的影响，以及对降水和各物理量预报效果的影响。结果表明：同化能改进初始场，并可改进暴雨落区和强度预报；同化可提高WRF模式对风场、温度场、高度场以及水汽场的预报能力，但有一定的时效性；同时同化探空和地面资料，比仅同化探空资料对大气低层物理量的预报能力要提高较多。

英文摘要：

In this paper, the control and assimilation experiments of WRF regional mesoscale model prediction are verified in the South China during a course of heavy rain. Based on the WRF-3DVAR system, the sounding data and surface observations are assimilated into WRF model. The effect of the two kinds data on the initialization and the prediction of rain and physical variables are analyzed. Results show that the assimilation of the either data can well improve the initialization, and has positive impact on prediction of the area and intensity of heavy rain. The assimilation can make great improvement on the forecast of physical variables, such as wind, temperature, height, vapor with the the time limitation. For forecasting physical variables in the lower levels, the assimilation both of sounding data and surface observations will be better than the assimilation of the sounding data.