同化多普勒雷达风资料的两种方法比较  

Comparison of Two Methods of Assimilating Doppler Radar Wind Data

作　　者：杨毅[1] 邱崇践[1] 龚建东[2] 邓莲堂[2] 希爽[1] 

机构地区：[1]兰州大学大气科学学院,甘肃兰州730000 [2]中国气象局国家气象中心,北京100081

出　　处：《高原气象》2007年第3期547-555,共9页Plateau Meteorology

基　　金：科技部科研院所社会公益研究专项国家科技基础条件平台工作重点项目(多部单多普勒雷达风场反演系统研制);青年科学基金项目(集合卡曼滤波同化多普勒雷达资料方法研究;40305015)共同资助

摘　　要：以美国新近研发的天气研究预报模式（WRF）配置的三维变分同化系统WRF 3D-Var为平台,比较了两种不同的同化多普勒雷达径向风资料的方法。一种是WRF 3D-Var系统现有的径向风资料直接同化方法;另一种是首先用两步变分法由多普勒资料反演出水平风,再同化反演风场。针对2003年7月4-5日的一次淮河暴雨过程进行的同化试验结果表明,同化了雷达风资料后得到的水平风场包含了更多的中尺度特征;从降水预报评分和预报的雷达回波来看,两种方法都能够明显改进降水预报,这种正作用能维持6 h左右;相对而言,同化反演的水平风场的效果略优于直接同化雷达径向风的效果。The comparison of two methods of assimilating Doppler radar wind data based on WRF 3D-Var is presented.The first method is afforded by WRF 3D-Var 2.0 in which Doppler radar radial velocity observations are directly assimilated.The second methods are to assimilate the horizontal wind retrieved from Doppler radar data by using a two-step variational methods.Both methods are tested with a heavy rainfall case that occurred in the region of Huaihe River on 5 July 2003.Results show that the analyzed wind contains more mesoscale information.The threat scores with radar wind data are higher than the experiment without radar data assimilation.The predicted results are also verified with reflectivity observations.Both methods show positive impacts on the short-range forecasting for about 6 hours.However,the second methods that assimilated retrieved wind gives better forecast than the first method does.

关 键 词：多普勒雷达 WRF 3D-VAR 同化 

分 类 号：P412.25[天文地球—大气科学及气象学]