Abstract: The exploitation of underground coal resources will result in a series of environmental disasters. In order to quantitatively analyze the spatial-temporal evolution characteristics of surface subsidence in Xiaohuigou Coal Mine, it processes Sentinel-1 satellite data from 2017 to 2023 by coherence-based SBAS-InSAR method, obtaining a time series of surface deformation in the mining area that aligns with actual mining conditions. Measurements from the joint surface movement observation station is used to quantitatively analyze the temporal and spatial evolution characteristics of surface subsidence, horizontal movement, and development of ground fissures. Based on the dynamic subsidence results of the mining area, invert the parameters of the Logistic time function model and establish a dynamic prediction model for surface subsidence. The results show that the coherence based SBAS method and the common point superposition method can effectively improve the InSAR monitoring accuracy. The maximum surface subsidence is located at the working face 2204, and the subsidence value reaches 685.52 mm. According to the measured data of the joint surface movement observation station, the Q survey line moves horizontally in the working face 2204, and the M survey line and the N survey line move eastward in the vertical trend. Logistic model is consistent with the spatio-temporal evolution characteristics of surface movement. Based on the model, the measured data are verified and analyzed. The average median error of fitting is about 13.67 mm, and the average relative median error of fitting is about 3.71%. The results of this paper can provide scientific basis for dynamic monitoring and prediction of surface movement in Xiaohuigou Coal Mine.