By using multisource remote sensing data, the drought data of Ebinur Lake wetland in an arid region was extracted and the impacts of drought on the ecological environment and agriculture were analyzed, providing the basis for reasonable management of water resources and the drought warning system. MODIS vegetation and temperature products and Landsat 8 data were used to extract the NDVI and LST in order to inverse temperature and vegetation dryness index (TVDI), and at the same time construct 2-Dimension feature space and analyze the spatiotemporal change of soil moisture. The results showed that the TVDI model has higher precision to effectively inverse regional soil moisture; from 2013 to 2016 in a 4-year span, the Ebinur Lake soil moisture has shown that the wet, normal, and light drought area is decreasing, while the drought and heavy drought area is increasing. In addition, the team analyzed the driving forces in some respects. In recent years, the increase in temperature, population, wind speed, and precipitation showed downward trends, and drought has reduced sharply. It is better to use multisource remote sensing data to efficiently accomplish drought monitoring in a wide area for a long time. The drought in Ebinur Lake wetland intensified during the four years, and the lake water and glacier area are shrinking at a faster rate. The Ebinur Lake regional soil moisture distribution is not improving. In order to protect the regional ecological system, effective local water resources management should be enhanced. In conclusion, carrying out research on the spatial and temporal dynamic changes and the driving forces for drying of Ebinur Lake has important theoretical and practical significances for this arid zone.