P7 Calculation of local wind speed near surface by using 4DVAR

Yoneda, Jiro, Ryohji Ohba, Kazuki Okabayashi, Mitsubishi Heavy Industries, Ltd., Japan, Hirohiko Ishikawa, Kyoto University, Japan, and Kenichi Tatsumi, Tokyo University of Agriculture and Technology, Japan

In Japan there are about 1300 surface observation stations of JMA (Japan Meteorological Agency) called as AMeDAS (Automated Meteorological Data Acquistion System). Most of AMeDAS stations measure 10 minites averaged wind speed every hour. The AMeDAS data are available from JMA near-real time. Therefore in this study, we tried to incorporate these surface wind data in addition to the NCEP reanalysis data for the better reproduction of local wind field near the surface using 4DVAR technique. As the result, both initial condition and prediction accuracy were improved at several observation sites. However the wind field was unrealistically twisted at some other areas. These may be caused since the input data are to much affected by local topography which are not reflected in the model topography. Therefore it is important to select the represented data for application of 4DVAR using AMeDAS.