利用多普勒天气雷达资料对一次暴雨过程的同化模拟

A Rainstorm Simulation with Assimilated Doppler Radar Observations

以CAPS(Center for Analysis and Prediction of Storm)研发的ARPS模式(The Advanced RegionalPrediction System V5.2.4)为基础,结合我国多普勒雷达资料,模拟2001年7月13日安徽省的一次暴雨过程,采用3DVAR(3-dimensional variational data assimilation)同化方法,做多时次同化雷达资料试验,前一时次模拟的结果作为下一时次的初始场,不断调整。结果表明,加入雷达资料后的风场、湿度场等都有明显调整,可以明显提高3 h降水模拟效果;同化的雷达时次越多,对上述各要素场和降水的模拟与实际观测的对应效果越好。

The rainstorm in Anhui Province on July 13th,2001 is simulated by numerical model ARPS （Advanced regional prediction system） and its 3D-VAR （3-dimensional variational data assimilation） system developed by CAPS （Center for analysis and prediction of storm） and Doppler radar data in China. Assimilation experiments are carried on by adding Doppler radar data at 10 and 30 minutes intervals, respectively, with the former results as the followed one＇s initial fields adjusted continuously. It is found that the winds and water vapor fields are obviously improved after inducing Doppler radar data, especially the 3-h precipitation simulations. And more times radar data induced, better the forecast simulations.