A Quality Control Procedure for FY-3A MWTS Measurements with Emphasis on Cloud Detection Using VIRR Cloud Fraction and its Applications in NWP System

A quality control procedure for satellite radiance data is proposed and applied to observations from Microwave Temperature Sounder (MWTS) onboard the first satellite in the Chinese polar-orbiting Fengyun-three series (FY-3A). A cloud detection algorithm is incorporated based on cloud fraction provided by a visible/infrared instrument called Visible and InfraRed Radiometer (VIRR) onboard the same satellite platform FY-3A. An analysis of test results conducted in July 2011 indicates that most clouds are identifiable by applying an FY-3A VIRR cloud fraction threshold of 37%. A verification with cloud liquid water path (LWP) data from MetOp-A further demonstrates that FY-3A VIRR cloud detection method is much more effective than using a simple O-B scheme of surface MWTS channel 1. On average, more than 65% global clouds are removed by the FY3-A VIRR-based cloud detection method. Observations removed by the proposed quality control procedure of this study includes the following steps: (i) channels 2-3 if VIRR cloud fraction is greater than 37%; (ii) two outmost field-of-views (FOV) (i.e., FOV 1, 2, 14 and 15); (iii) channels 2-3 if terrain altitude is greater than 500 m; (iv) channel 4 over sea ice; (v) coastal FOVs; and (vi) outliers whose differences with model simulated brightness temperatures deviate from the biweighting mean by more than two biweight standard deviation. About 82.1%, 73.6% and 28.8% observations are removed for MWTS channels 2-4, respectively. After QC, FY-3A MWTS global data compared more homogeneously with NCEP forecast. The global biases and standard deviations are significantly reduced after QC, especially for channel 2. The frequency distribution of the differences between observations and model simulations also becomes more Gaussian-like after QC. The use of FY-3A MWTS radiance data in NWP system has been assessed.