Through idealized experiments, this study is to test the performance of multi-grid three-dimensional variant ional (3D-Var) on three-dimensional (3D) Doppler radar radial velocity data assimilation, and to what degree the 3D Doppler radar radial velocity can improve the conventional (in situ) wind observation analysis. A two-scale idealized wind field is constructed, and then random-distributed conventional wind data and 3D Doppler radar radial velocity data are generated based on this idealized wind field. By assimilating these data and comparing the analyses with the true idealized circulation field, the multi-grid 3D-Var performance in 3D Doppler radar radial velocity analysis is evaluated. The effects of weak constraint and strong constraint on the multi-grid 3D-Var analyses are also presented. Results show that the 3D Doppler radar radial velocity data do provide additional useful information especially in the sparse distributed conventional observation situation, and the multi-grid 3D-Var with strong constraint can make better analyses on not only horizontal but also vertical velocity.
展开▼
