Abstract:Combining the ATOVS emissivity data and GTS routine observation data, a simulation analysis for the Meiyu front rainstorm process from Jul. 4—5, 2013 was conducted in this paper through the control test, 3DVAR assimilation test and ETKF-3DVAR blending assimilation test of any inassimilable observation data. The results showed that the blending assimilation scheme can obviously improve the accuracy of rainstorm forecast, and effectively improve the initial field structures of wind field, temperature field and relative humidity field and other meteorological elements, while these meteorological elements play an important role in the accuracy of rainstorm forecast. Besides, the blending assimilation scheme has an evident effect on improving the middle and low atmosphere. In the process of numerical integration, the advantages of the blending assimilation scheme gradually diminish over time.