Research On Quality Control Of FY-3A Microwave Humidity Sounder Material

Quality control of FY-3A Microwave Humidity Sounder （MWHS） measurement is the main purpose of this study. We choose the global data in L1C format from January17th to23th in2010as research object. In this study, bi-weight algorithm is used to do quality control for the value O-B in order to eliminate the outliers which affected by surface emissivity and cloud. Value O is the actual observed radiation brightness temperature of FY-3A MWHS. Value B represents simulated radiation brightness temperature of FY-3A MWHS, which simulated by rapid radiation transfer mode RTTOV₉₃（model background field is6hour forecast field in NCEP GFS global forecast system）. It shows that, standard deviation of the value O-B of channel3、4and5has obviously reduced2-3K after bi-weight quality control method. In addition, the change of average deviation of value O-B with the change of time is also very stable after bi-weight quality control. It is thus clear that this QC method can be effective to remove the outliers and can be used in quality control for Satellites.Now, most assimilation system assimilate the satellite observation in clear sky directly home and abroad, so cloud detection test is done for FY-3A MWHS observation only through threshold method, which considers it’s cloud when observation incremental （O-B） of channel2is less than or equal to a certain threshold. Then compare the cloud detection results with the cloud product of VIRR matched by MWHS measurements in L1C format. After several tests, the cloud detection results are reasonable when the threshold value equal to-1k.It shows that observation brightness temperature value is maximum at subpoint by analyzing distribution of global average observation brightness temperature O and simulated brightness temperature B （of ch3, ch4and ch5） with the changes of scanning angle for FY-3A microwave humidity sounder from January17th to23th in2010, presenting decrease trend with the increase of the scanning angle. As a result, the bias correction for scanning angle for MWHS is needed. It’s done by computing bias correction coefficients at each scan position and each latitude of MWHS measurements form70°S to70°N, where5°is set for a latitude belt. The experiment shows that the bias correction coefficients can be provided for scan bias correction in data assimilation system.