Abstract: The national coal mining strategic center has moved to the western China and the clustering, high intensity mining makes the surface ecological environment of the western mining areas present a strong disturbance situation, which makes the fragile ecological environment gradually degenerate, and the most intuitive embodiment is the impact on vegetation. The ecological restoration benefits and impact on the ecological environment of the traditional mining area vegetation restoration and reconstruction methods still need to be discussed, which may lead to the contradiction between the high investment of vegetation restoration funds and the low recovery efficiency. The vegetation restoration in mining area should strictly follow the development law of local natural ecosystem, avoid the disturbance to the ecosystem after mining, scientifically configure, optimize the layout and guide the restoration of the self healing ability of the ecosystem.Therefore, it is an important basis and premise for exploring the effective method of vegetation guided restoration in semi arid mining areas to clarify the temporal changes of vegetation coverage in the disturbed areas of coal mining. Based on the Landsat NDVI time series data, taking Shendongcentral mining area as the study area, the temporal changes of vegetation coverage in mining area and non mining area were analyzed. According to the regression slope, correlation coefficient and standard deviation of vegetation coverage in different time orders, the change trend and interannual fluctuation degree of vegetation coverage in the study area were evaluated, and the driving factors influencing the change of vegetation coverage were analyzed. The results show that from 2000 to 2016, the vegetation NDVI time series of Shendongcentral mining area demonstrated a phenological cycle change. Due to the large area of vegetation reconstruction in the mining area, the INDV of the central mining area increased significantly after 2005. The influence of coal mining subsidence on vegetation coverage was not all negative. The area with the decreaseof vegetation coverage accounts for 27.65% of the total area of the central mining area.Soil water content is the most important factor of soil physical and chemical properties, followed by organic matter content, total nitrogen content and bulk density. The groundwater depth of 3.2 m and 8.0 m are two important thresholds affecting the vegetation INDV in Shendong mining area. Rainfall also has a very important impact on the change of vegetation coverage in the central mining area. The vegetation guided restoration in semi arid mining area firstly needs to identify and control the permanent cracks on the surface to create the precondition of soil moisture recovery in subsidence area. Secondly, it is necessary to strengthen the coupling relationship between plant growth and soil moisture and groundwater table depth in subsidence area, and analyze the soil water content and groundwater depth threshold of plant growth under drought stress and plant death, so as to provide the basis for the restoration of soil water content and groundwater table depth. For plant configuration and coverage control, high spatial resolution remote sensing images can be used for a fine classification of surface vegetation in the study area. Combined with the temporal variation trend and fluctuation degree of vegetation INDV, and considering the site conditions of vegetation growth, the vegetation composition in the positive fluctuation area of vegetation INDV is taken as a reference basis. In addition, the local plant species with poor soil resistance, strong drought resistance and developed root system should be selected as far as possible. Finally, it also should strengthen the detection on the distribution of groundwater resources in the mining area, strengthen the recycling of the water in the goaf, and give priority to the exploitation of the areas without or less groundwater in a planned way, so as to reduce or delay the damage to groundwater from the source.