中文摘要：

通过对近两年收集的降水过程的雷达回波资料进行统计分析，得到本地区主要降水过程类型及其回波演变特征；利用反射率因子的基数据和风廓线产品数据等资料，运用交叉相关法计算区域速度、读取风廓线得到环境风速度，自动判别降水回波类型、识别回波整体及单体移动速度；应用本地实时的dBz-I关系及线性外推法计算未来1～2h站点降水量及面雨量，实现降水临近预报自动化；应用自动站雨量资料及雷达联合自动站资料所作的雨强实时分析场对1h站点雨量预报及区域雨量预报进行初步检验，1～24．9mm和≥25mm以上站点预报的平均相对误差分别为58％、33％；区域雨量预报的雨区范围及中心分布与实况具有较好的一致性；2h降水预报在回波较稳定时，仍有较好的预报效果。

英文摘要：

Based on the statistical analysis of the radar echo and precipitation data in the last two years, how to identify the major types of precipitation and their corresponding retrieved echo features are investigated. By means of reflectivities and wind profile products, rainfall echo styles, integer and cell echo moving speed can be identified automatically according to surrounding wind from VWP and area wind calculated by cross-correlation method. Rainfall nowcasting of station and surface rainfall during 1- 2 h is automation using local real-time dBz-I relation and extrapolation method. One h station and surface rainfall forecast are preliminarily examined according to rainfall data from automatic weather station and an h real-time rainfall field calculated by optimum interpolation, the average relative errors of station rainfall from 1 mm to 24.9mm and above 25mm are 58% and 33% respectively. In area rainfall nowcasting, both of the rainfall area and centre are very consistent between forecasting value and the real value, the 2 h forecasting is satisfactory when rainfall echoes are relative stable.