FY-3Dis a new generation of polar orbiting meteorological satellites in China.The Medium Resolution Spectral Imager(MERSI-Ⅱ)is one of the core sensors it carries.It is of great significance for global numerical weather prediction,atmospheric quantitative detection,and climate change monitoring.The snow area ratio product is one of many land surface products and is the main input parameter for hydrological models and regional climate models.based on MERSI-Ⅱdata,this paper develops an algorithm for extracting the proportion of snow cover area.The core of the algorithm is mixed pixel decomposition.The Spatial Spectral Endmember Extraction(SSEE)algorithm automatically extracts the endmembers,and the Fully Constrained Least Squares(FCLS)solves the linear mixed model.The unmixed results were superimposed on the cloud mask to obtain FY-3D/MERSI-Ⅱsnow area ratio data(FY-FSC).The FY-FSC was verified by using the Landsat 8snow area ratio data(L-FSC)as a reference value,and the FY-FSC and MODIS snow area ratio data(M-FSC)were compared.The results show that the overall root mean square error (RMSE)of FY-FSC is 0.17,the correlation coefficient(R)is 0.54,the Absolute Mean Error(AME)is 0.10,the overall R of M-FSC is 0.41,RMSE is 0.26,and AME is 0.29.Using the accuracy evaluation factor K of the snow area extraction to compare the accuracy of the total snow area obtained by FY-FSC and M-FSC.The results show that the average K values of FY-FSC and M-FSC data are 88.51%and 86.78%, respectively,and the accuracy of FY-FSC is higher than that of M-FSC.FY-FSC will be included as a test parameter in the FY-3D/MERSI-Ⅱsnow cover business product,which can fill the blank of the domestic satellite operational inversion sub-pixel snow parameters.