Abstract:

In 2017,China launched FY-3D,a next-generation meteorological satellite,into polar orbit.One of its core sensors is the MediumResolution Spectral Imager II (MERSI-II) that is capable of high-precision quantitative inversion of atmospheric,terrestrial,and oceanic parameters,including metrics of cloud,aerosol,water vapor,land surface characteristics,and ocean color.We established the true snow cover extent of the middle Tianshan Mountains,Xinjiang,China with visual interpretation of highresolution Landsat-8 images.We then used FY-3D/MERSI-II data,captured during cloudless conditions in July and August 2018,to calculate the normalized snow cover index.We compared the accuracy of FY-3D/MERSI-II data calculated by confusion matrix against MODIS daily snow product data.With an average overall snow cover extraction accuracy of 0.855,the FY-3D/MERSI-II imagery outperformed MODIS that had an accuracy of 0.820.The average Kappa coefficient of FY-3D/MERSI-II (0.659) was also higher than that of MODIS (0.558).Overall,the accuracy of the snow cover area extracted from FY-3D/MERSI-II data was higher than from MODIS,achieving a closer match to the high-resolution Landsat-8 images.

Key words: FY-3D/MERSI-II')">

FY-3D/MERSI-II, MOD10A1, NDSI, accuracy contrast analysis