Accurate estimation of snow properties is important for effective water resources management especially in mountainous areas. In this work, we develop a snow data assimilation scheme based on ensemble Kalman filter (EnKF), which can assimilate remotely sensed snow observations into the Common Land Model (CoLM) to produce spatially continuous and temporally consistent snow variables. The snow cover fraction (SCF) product (MOD10C1) from the moderate resolution imaging spectroradiometer (MODIS) aboard the NASA Terra satellite was used to update CoLM snow properties. The assimilation experiment is conducted during 2003-2004, in Xingjiang province, west China. The preliminary results are very promising and show that distributions of snow variables (such as SCF, snow depth, and SWE) are more reasonable and reliable after assimilating MODIS SCF data. The results also indicate that EnKF is an effective and operationally feasible solution for improve snow properties prediction.