Reconstruct the Meso-scale Information of Typhoon with BDA Method Combined with AMSU-A Data Assimilation Method

Typhoon is one of the most frequent disasters affecting human being. The numerical prediction accuracy of typhoon is closely related to the quality of initial fields. Bogus constructing technique, BDA (Bogus Data Assimilation) scheme, and directly assimilation of multifold observed satellite data are all the very important factors in improving typhoon initial fields. In this paper, taking advantage of Advanced Research WRF model, RTTOV model and 3DVAR technique (3-dimensional variational data assimilation), by the research of Bogus construction method and implantation method, the directly 3DVAR of multifold observed data (include: Bogus data, NOAA-16 ATOVS HIRS data, NOAA-16 ATOVS AMSU-A data, and NOAA-16 ATOVS AMSU-B data etc.) are studied, and a CAMOD (Cooperative Assimilation of Multifold Observed Data) method is proposed.This method can reconstruct the structure of initial typhoon, improve the quality of initial fields, and finally improve track simulation and intensity simulation of typhoon. In this paper, aiming at typhoon "Kalmaegi" (No.0807), initializing from different times separately, many data assimilation experiments and numerical simulation experiments are performed, and the reason why data assimilation can improve typhoon numerical simulation are discussed respectively from wind fields, temperature fields, pressure fields, and geopotential height fields. The numerical results indicate: (1) The assimilation of satellite data has positive effect on typhoon numerical simulation. But for different satellite data, there is different effect on improving typhoon track simulation. In general, the improvement effect of AMSU-A data is obviously better than AMSU-B data and HIRS data. Satellite data assimilation can quite affects the steering flow of typhoon. When the subtropical high enhances and extends westward, causing by assimilation of satellite data, the steering flow will be beneficial to the moving westward of typhoon,and vice versa. (2)Bogus data are very important for both typhoon track simulation and typhoon intensity simulation. As long as Bogus data are joined, the improving effect on typhoon track forecast and typhoon intensity forecast will be obvious; (3) When multifold observed data are assimilated cooperatively, the improving effect of typhoon numerical simulation will be the best; (4)By analyzing the optimal initial fields which assimilating multifold observed data, it can be found that after assimilation typhoon sea level wind fields and sea level pressure fields are both improved obviously, warm core structure at high level is enhanced, geopotential height of typhoon center is decreased and the central axis falls to northwestward. These are the reasons which improve typhoon track simulation and intensity simulation. The numerical results in this paper should be tested in more typhoon cases. And the CAMOD of multi-satellite data and multi-orbit data are also the emphases in further research work.

This work is jointly supported by the International Scientific and Technological Cooperation Projects (2010DFB33880), the Commonwealth Trade (Meteorology) Program (No. GYHY200906009 & No. GYHY201106004), the National Natural Science Foundation of China (No. 40805046 & No. 10871099) and the National High Technology Research and Development Program 863 (No. 2010AA012304).