多普勒雷达资料的直接同化对降雨预报的影响

Impacts of the Direct Assimilation of Doppler Radar Data on the Rainfall Forecast

利用中尺度数值模式WRFV3.1及其三维变分同化系统WRFVAR3.1,对2008年8月10日奥运期间北京地区发生的一次强对流天气过程进行了多普勒雷达资料变分同化试验,并对同化结果和降水预报进行了对比分析。结果表明:(1)雷达资料的同化能有效改善初始场中风场、水汽、温度场的空间分布、量值大小,使其包含更多中尺度信息;(2)雷达反射率同化对大气的湿度场和温度场影响较大,而雷达径向速度同化则主要调整大气的风场;(3)对短时0～6h内的降水预报,反射率的同化对降水预报改善明显,径向速度的同化对降水预报影响相对较小,而两者的联合同化降水预报效果最好;(4)雷达资料的同化对较长时间6～12h的降水预报仍有较好效果。

Using the Model WRFV3.1 and its 3DVAR system WRFVAR3.l Doppler radar data variational assimilation experiments are performed on a heavy convective rainfall process occurred at Beijing on 10 August 2008 when the Olympic Games was held. A contrastive analysis are put into practice on assimilation products and rainfall forecast. The analyzed ICs （initial conditions） show that the direct assimilations of radar data can effectively adjust the distribution and intensity of wind field, water vapor and temperature field of the background, and more mesoscale information can be included in the analysis. Better rainfall forecast is generated with the reflectivity assimilation than that with the radial velocity assimilation, and the best one is derived with the assimilation of both observations during the first six hours. Moreover, the predicted reflectivity is corresponded well with the reflectivity observations in distribution and intensity after radar data assimilation during the second six hours.