Abstract: | The Weather Research and Forecasting (WRF) model and its three-dimensional variational data assimilation system are used in this study to assimilate the INSAT-3D, a recently launched Indian geostationary meteorological satellite derived from atmospheric motion vectors (AMVs) over the South Asian region during peak Indian summer monsoon month (i.e., July 2014). A total of four experiments were performed daily with and without assimilation of INSAT-3D-derived AMVs and the other AMVs available through Global Telecommunication System (GTS) for the entire month of July 2014. Before assimilating these newly derived INSAT-3D AMVs in the numerical model, a preliminary evaluation of these AMVs is performed with National Centers for Environmental Prediction (NCEP) final model analyses. The preliminary validation results show that root-mean-square vector difference (RMSVD) for INSAT-3D AMVs is ~3.95, 6.66, and 5.65 ms?1 at low, mid, and high levels, respectively, and slightly more RMSVDs are noticed in GTS AMVs (~4.0, 8.01, and 6.43 ms?1 at low, mid, and high levels, respectively). The assimilation of AMVs has improved the WRF model of produced wind speed, temperature, and moisture analyses as well as subsequent model forecasts over the Indian Ocean, Arabian Sea, Australia, and South Africa. Slightly more improvements are noticed in the experiment where only the INSAT-3D AMVs are assimilated compared to the experiment where only GTS AMVs are assimilated. The results also show improvement in rainfall predictions over the Indian region after AMV assimilation. Overall, the assimilation of INSAT-3D AMVs improved the WRF model short-range predictions over the South Asian region as compared to control experiments. |