The Comparison Experiments Of Cloud Drift Wind Assimilation

These years, the accuracy of weather forecast has been greatly promoted with the development of numerical weather prediction because numerical weather prediction play an important role in the weather forecast.Regular stations distribute sparsely,observation time is limited so it is difficult to improve the accuracy of numerical weather prediction.Along with the air sounding technology development, radar and satellites data and other unconventional information as a supplement to the conventional information applied to numerical prediction models, especially satellite cloud drift wind data become one of the primary means that apply lots of upper-wind data and improve the initial field of mesoscale numerical model effectively.This paper on the basis of using the mesoscale numerical model MM5,reanalysis data of NCEP and the FY-2C cloud drift wind data inversed by National Satellite Meteorological Center. Cloud drift wind data are assimilated into numerical experiments by direct assimilation and adjoint-assimilation, the improvements degree after assimilation in temperature, height and wind field are tested.With a case of heavy rain fall taking place from 00h June 17 to 00h June 18, 2006 in the south of china, the comparison experiments of different schemes are peiformed.After assimilate cloud drift wind data,the forecast on the precipitations was enhanced, in direct assimilation experiment the forecast on the heavy rain was enhanced to acertain degree,in adjoint-assimilation experiment the forecast on the rainstorm with a breakthrough.The output of model is been analyzed by RMS. The results are as follows. The accuracy of forecasting precipitations can be improved after assimilation cloud drift wind data and temperature, height, and wind field are modified to a certain extent. Particularly, adjoint assimilation experiment shows more advantages than direct assimilation experiment in accuracy of precipitation and in temperature, height, wind field.