The Yulong Snow Mountain is characterized by marine glaciers, which is sensitive to earth's climate change and has important ecological research value. To study the variation of surface temperature in the Yulong Snow Mountain area, a quantitative analysis method based on a three-dimensional (3D) surface temperature field is proposed in this paper. The digital surface model is constructed using high-resolution stereo mapping satellite images to provide accurate 3D terrain reference. Multi-temporal thermal infrared satellite image data are used to invert the surface temperature field model to provide a basis for the change of surface temperature. The digital surface model is combined with the multi-temporal surface temperature field model to analyze the spatio-temporal variation of surface temperature in detail. Based on Landsat satellite images acquired by same season from 1987 to 2018, a digital surface model is constructed by inversion of surface temperature and 3D images of the ZY3 satellite. The fourth order temperature interval was used to analyze the surface temperature variation in the Yulong Snow Mountain. The low temperature area and the sub-low temperature area are shrinking with a speed of 2.096 km²/a and 2.662 km²/a, respectively. The medium temperature area and the warm temperature area are expanding with a speed of 2.902 km²/a and 1.703 km²/a within the experimental area in recent years. The multi-temporal surface temperature changes in the Yulong Snow Mountain area are analyzed. The results show that the climate is warming around the Yulong Snow Mountain area, and could provide reference for global ecological environment change.