Assimilation Experiment of Sounding Data on the Rainfall in Southwest China

Using sounding data, T213 forecast field and the NCEP reanalyzed data, the impact on different sounding data preprocessing system to assimilation results and southwestern numerical weather forecast are researched in GRAPES (Global and Regional Assimilation and Prediction Enhanced System) 3D-VAR assimilation system.The main conclusions are the following:i) Different receiving-data method and truncation time affect the number of conventional upper-air stations.The different sounding data preprocessing system leads to the different number of usable data item, and the upper-air stations and vertical levels provided by National Meteorological Center are more than that provided by Chengdu Region Center.ii) The more sounding data are used, the more adjustment extents are obvious.ii) Assimilating sounding data can improve the quality of background field and the improvement effect is in direct proportion to the number of data used.iv) The results of rain prediction show that the more sounding data are assimilated, the more storm rainfall center value is close to the rainfall observation.v) The upper-air data preprocessing system from National Meteorological Center is superior to that of from Chengdu Region Center according to the TS score during July, 2005 in southwest China.