Direct Assimilation Experiment Research Of Observation Data Of Fengyun-4 Satellite A Based On GSI Assimilation System

Using a more reasonable background error covariance in the assimilation system is crucial to achieve better assimilation results.In this paper,firstly,the NMC method is adopted to construct the B matrix more suitable for the WRF-ARW regional prediction system for the Chinese region.And the characteristic difference between the B matrix and the NCEP prediction system preset by GSI assimilation system is compared and analyzed.The single-point observation test and background error covariance tuning parameter sensitivity test were designed with reference to these feature differences to determine the optimal tuning parameters for constructing B matrix for the Chinese region.The results show that the Sen6 experiment with the optimal tuning parameters and the B matrix constructed for the Chinese region has the best prediction effect.In particular,the prediction performance of V wind component field and relative humidity field has been improved significantly.At the same time,the dilemma of background error covariance parameter adjustment(especially horizontal length scale parameter adjustment)in GSI assimilation system is also introduced.On this basis,the Hyb3 cyclic assimilation experiment with Hybrid assimilation method and B matrix constructed for Chinese region can further improve the prediction effect.To some extent,the location of the simulated rain belt was modified.FY-4A is a new generation of geostationary meteorological satellite in China.It is of great theoretical and practical significance to study the impacts of FY-4A satellite observations on data assimilation and numerical weather prediction.In this paper,based on WRF/GSI framework,the direct assimilation capability of FY-4A AGRI infrared channels radiances under clear sky in GSI assimilation system is constructed.The process of analyzing and assimilating FY-4A AGRI infrared channels radiances by GSI assimilation system is introduced in detail from several important aspects such as pretreatment of radiances,cloud detection,preliminary quality control and data thinning,bias correction and quality control.A comparison of data assimilation and prediction experiments was designed to demonstrate its impacts of cycle assimilation and prediction on a monsoon low pressure heavy precipitation case.The results show that the E12_IMAGERexperiment to directly assimilate four GOES-like AGRI channels and the E14_MIXexperiment to directly assimilate four GOES-like with three AHI-like AGRI channels generated the analysis increment of the positive water vapor mixing ratio with larger central strength,wider horizontal distribution and deeper vertical coverage at key locations on the ocean surface.Meanwhile significantly enhances the prediction performance of U wind component,V wind component and temperature field.In addition,obviously increases the low-layer warm and wet conditions in regional high resolution numerical weather prediction,and significantly improves the range of the precipitation fall area and the rain belt shape.