Snow albedo is an important parameter in the study of local or global energy balance and climate change.The snow physical properties such as particle size shape and the impurities are the key reason that affects the snow albedo.The snow bidirectional reflectance distribution function(BRDF)is the premise in the remote sensing retrieve snow parameters.In this paper,the radiative transfer theory(ART)considers the snow particle size,particle shape and the effect of the impurities which was used to retrieve the snow surface albedo.Base on the ART model,a new MODIS snow albedo retrieval algorithm is developed.Then using the new algorithm to get the MODIS 8-day snow albedo data.Taking the Greenland as a study area, the MODIS surface reflectance data(MOD09)and MODIS snow cover data(MOD10L2G)were acquired. Then the GC-Net in-situ measurements data were used to analyze the evaluation of the retrieval snow albedo. Compared with the snow albedo in-situ data,the root mean square error(RMSE)and correlation coefficient for the new MODIS snow albedo over the Greenland is 0.018and 0.83.The result shows that the MODIS snow albedo based on the ART model works well and needs less parameter.