云导风资料在遗传算法同化系统中的应用

利用云导风资料和常规观测资料,采用遗传算法同化系统和中尺度MM5模式对2005年7月9～10日长江流域的暴雨过程进行了模拟和分析.结果表明:遗传算法同化系统可以同化常规资料和非常规资料.使用遗传算法同化系统可以有效改善数值预报模式的初始场,在一定程度上提高物理量场的预报,但对降水场的预报效果改善不明显.加入云导风资料后,可进一步改善风场和温度场,使得物理量场和降水场的预报更加接近实况.

APPLICATION OF CLOUD-DERIVED WIND DATA IN GENETIC ALGORITHM ASSIMILATION SYSTEM

By using cloud-derived wind data and routine observational data, a heavy rain over the reaches of Yangtze River from 9 to 10 July 2005 is simulated and analyzed with a genetic algorithm assimilation system and MM5 models. The results indicate that the genetic algorithm assimilation system can assimilate both conventional and unconventional data. It can improve not only the initial field of the numerical forecast model effectively but also enhance the forecasts of other elements to some extent, though not helpful for the improvement of the precipitation forecast obviously. By adding the cloud-derived wind in the experiment, the forecast of the wind and temperature filed are more significantly improved and the prediction of other elements and precipitation is also more similar to the observation.