多时刻台风Bogus资料同化对台风“天鹅”的影响

IMPACTS OF CYCLE-BOGUS DATA ASSIMILATION ON TYPHOON "GONI"(0907)

基于GRAPES(Global/Regional Assimilation and Predition System)三维变分同化系统,设计在6小时内分别应用三小时一次和逐小时的台风定位和定强信息对台风进行多次BDA(Bogus Data Assimilation)同化,在多次BDA中只同化近似程度高的涡旋风分量而近似程度较差的辐散风分量则通过多次调整和协调逐步向涡旋风适应。多次BDA能逐步调整台风位置并且在同化过程中随着台风位置更接近实况,使得台风周围流场更趋合理。采用多次BDA方案对"天鹅"(0907)台风进行数值试验,结果表明多次同化方案能逐步改进台风中心附近的环流场信息,与一次BDA相比,得到的初始场结构更协调,对台风的路径和降水预报也有一定改善。

Based on GRAPES（Global/Regional Assimilation and Prediction System） 3D-VAR system,a scheme called Cycle-BDA（bogus data assimilation） is designed to assimilate bogus data with information of location and intensity of typhoon every three hours and every hour respectively within six hours.The tangential winds are only assimilated because of the high accuracy and the radial winds are adjusted gradually to be consistent with the tangential winds during the Cycles,while the location of typhoon center is corrected step by step as well.A numerical simulation of typhoon ＂Goni＂ shows that the assimilation scheme can correct the typhoon＇s initial position and improve the circulation of the initial field,and leads to improving the track and rainfall forecast against the once-BDA scheme.