Applications of the Three-cornered Hat Method to the Error Variance Estimations of FY-4A Atmospheric Temperature Profiles

This study applied the three-cornered hat method to the error variance estimations of temperature profiles retrieved from the Geosynchronous Interferometric Infrared Sounder (GIIRS), the first hyperspectral infrared sounder onboard Fengyun-4A (FY-4A) geostationary satellite. An accurate observational error is a crucial input parameter for FY-4A temperature profile assimilation. Using multiple data sets from 2020, this study examined the error variances of FY-4A temperature profiles in clear-sky and cloudy-sky respectively. Because of the effects of clouds on GIIRS, the error variances in cloudy-sky are much larger than in clear-sky. Provided that error variances less than 4K(2) are significant for data assimilation, the FY-4A temperature profiles in clear-sky from 900 to 150 hPa are meaningful to be assimilated. Results reveal that the latitudinal dependency of error variances is strong at 100 hPa. Encouraged by the vertical distributions of the temperature profile differences between FY-4A and other data sets, the error variance of total potential energy is also examined in this study. Although the detailed structures of temperature profiles are removed, the vertical integration obtains a more minor percentage error. Further studies on data assimilation are worth discussing if total potential energy is more meaningful than the FY-4A temperature profiles.