Three Dimension Variational Study On Microwave Atmospheric Observations Of Meteorological Satellite

Typhoon usually starts in the tropical oceans,and its landing will cause huge life and property damage to the coastal areas,so how to improve the accuracy of its forecast has become the study's focus and difficulty of the oceanographers and meteorologists.Numerical weather forecast(NWP)plays an important role in weather forecast,and an important factor restricting the accuracy of forecast is the initial value,currently the way to solve this problem is data assimilation,which can correct the short-term forecast with the observation constantly to produce a more accurate initial value.At present,one widely used data assimilation method is the three-dimensional variational method,and the satellite observations have the characteristics of high spatial-temporal resolution and wide coverage,which can compensate the conventional observation in the ocean and plateau,so it has a high prospect in typhoon forecast.The mesoscale numerical weather prediction mode(WRF mode)and its three-dimensional variational assimilation system(WRF-3DVAR)with the new background error covariance matrix(B matrix)are used to assimilate directly the microwave emissivity data of AMSUA,MHS,ATMS and so on.Then the assimilation experiments for the typhoon “ Kai-tak” in 2012 and “Noul” in 2015 are carried out,which will assimilate the satellite microwave radiation data of 3h before and after the assimilation time into the background field to get the analysis value,which will be the initial value for a numerical weather forecast of 3 days,to compare and analyze the influence of different observations on the assimilation result and the prediction of typhoon track.The results show that the assimilation of satellite observations can improve the background field effectively,and then improve the prediction accuracy of typhoon track.The assimilation effect of AMSUA and ATMS data is better,and the assimilation effect of MHS data is less obvious;The typhoon scene is different,satellite observations are different,assimilation effect is also different,the assimilation effect of AMSUA + MHS + ATMS data compared with ATMS data and AMSUA + MHS data,is poor in the Noul scene,but is better in the Kai Tak scene.