新一代天气雷达资料在2003年江淮流域暴雨模拟中的初步应用:模拟降水和风场的对比

Application of the China New Generation Weather Radar Data to the Torrential Rain Simulation over the Yangtze River-Huaihe River Basin in 2003: Contrast of Precipitation and Wind

利用时间尺度密集的长沙、常德、合肥、南昌、南京、武汉和宜昌共7个站的新一代天气雷达（CINRADSA雷达）观测反演资料、TBB资料、常规的探空和地面观测资料、NCEP分析资料，与中尺度MM5模式相结合，以NCEP格点资料作为大尺度背静场，加入12h间隔的探空、3h间隔的常规地面观测资料及1h间隔的雷达反演风廓线资料，进行全程四维同化模拟，考察中尺度数值模式MM5对2003年梅雨期间发生在湖南、安徽和江苏的暴雨过程（7月8～9日）的模拟能力。分析表明：除了模拟降水与实况接近以外（雨量和雨区），MM5模式输出的风廓线和从雷达观测反演得到的风廓线结果有很好的相似性，加入雷达反演风廓线资料后对模拟结果有一定改进，为进一步利用模式输出结果研究造成2003年江淮流域暴雨洪涝的中尺度对流系统的结构和机理提供了可能。

China new generation weather radar （CINRAD-SA radar） takes on volume scan model （VCP31 OR VCP21）, which can observe weather systems successively and obtain four-dimensional data. Therefore, the wind profile retrieved from the radar data can be got continuously, namely, the space distribution of the wind and its temporal variation can be known. It makes up the insufficiency of the general observation on temporal and spatial distribution. Once these wind data were merged into the meso-scale model, it would provide more information of the atmosphere and improve the simulation results. Utilizing the retrieving observation data of CINRAD-SA radar at Changsha, Changde, Hefei, Nanchang, Nanjing, Wuhan, and Yichang, the black body temperature （TBB） data, the general sounding and surface observational data, the National Center for Environmental Prediction/National Center for Atmospheric Research （NCEP/NCAR） 1°× 1° grid data, linked with the meso-scale model MMS, taking the NCEP/NCAR data as the first guess field, a-dopting the wind-profile data retrieved from the above seven radars with 1-h interval, the sounding data with 12-h interval, the surface observational data with 3 h interval into nudging four-dimensional data assimilation process of MMS, the torrential rain occurred at Hunan, Anhui and Jiangsu provinces during 8- 9 July 2003 was simulated. Two simulation schemes were done in this study. One included the general observed data from meteorological station, but no retrieved wind data of the seven radars, which was called control scheme, the other was the same as the control scheme but adopted the wind profile retrieved from the radar observation with 1-h interval, which was called test scheme. Both the two schemes simulated weather pattern correctly, and the distributions of total precipitation during simulated time （from 1200 UTC 8 July to 1200 UTC 9 July 2003） took on southwest - northeast direction, which is similar to the 24 h observed rainfall distribution. There were several rainfall centers located at this belt, which cannot be found in the observed data. The precipitation of the test scheme was little more than that of control scheme in coarse domain. The temporal variation of the wind direction from the two schemes at the radar station were almost like to the retrieved, but wind speed of test scheme at lower levels was less stronger than that of control scheme sometimes, which reflected the effect of the radar data on the model simulation. The results from the simulation manifested that besides the rainfall value and rain region are similar to the fact, the wind profiles from the simulation are also similar to that from the retrieved radar data, which provided the possibility to study the meso- scale convective system caused the torrential rain over the Yangtze River- Huaihe River basin in 2003.