集合Kalman滤波同化理论及其数值试验

The Theory of Ensemble Kalman Filter Assimilation and Its Numerical Experiment

集合Kalman滤波用于数值试验有着坚实的理论基础。本文介绍了集合Kalman滤波理论及其技术实现,在此基础上搭建了集合Kalman滤波同化系统,用MM5模式同化了实测探空资料并作了48 h的预报试验,并将预报结果与实测值及4D VAR同化的结果作了比较。试验结果表明:集合Kalman滤波同化探空资料可以改进MM5模式的预报效果,且集合Kalman滤波同化后模式的预报效果明显优于4D VAR同化后模式的预报效果。

There is sufficient theory applying the ensemble Kalman filter （ENKF） to numerical forecast data assimilation. We build the ENKF data assimilation system model based on the theory of ENKF and its technology. The experiment of assimilating observed air sounding data is done by using non-hydrostatic equilibrium mesoscale model （MM5）. The model which integrated 48 hours is assumed perfect in this experiment. By comparing the observation data with the results of forecast using ENKF and 4D VAR data assimilation, it demonstrate that the data using ENKF assimilation can improve the effect of MM5 model. The difference of model＇s forecast effect are also discussed after using ENKF and 4D VAR assimilating observed air sounding data. The results show that the forecast effect using ENKF is superior to that using 4D VAR data assimilation. It is concretely represented in the forecast of the 12 h rainfall distributions, the zonal wind fields at different levels and the forecast error of the wind and temperature fields.