Snow parameters are important physical quantities of climatology and hydrology research, improving the accuracy of snow parameters is important for climatology, hydrology and disaster prevention and reduction. The western Jilin Province of China has obvious salinization problem. Meanwhile, it belongs to a typical snow-covered area. In this paper, the western Jilin Province is selected as the study area and the main research focuses on analyzing the snow cover conditions. The FY3B-MWRI passive microwave remote sensing data from year 2011 to 2016 are selected as experimental data. Compared with optical remote sensing data, using MWRI data can better obtain snow information, and it is also the preliminary work to retrieve snow depth and snow water equivalent. Furthermore, a new decision tree algorithm for snow cover identification was built to distinguish different snow cover conditions. Compared with the existing three algorithms reported in other literatures, the proposed algorithm improves the identification accuracy of snow cover up to 95.06%. While the accuracy for Singh’s algorithm, Pan’s algorithm and Li’s algorithm were about 80.19%, 78.79% and 90.13%, respectively. This study provides important information to the research of snow cover in saline-alkali land.