Abstract: In order to investigate the spatial and temporal distribution characteristics and the variation trends of soil moisture in Zhundong mining area, 11 periods of Landsat remote sensing images from 2001 to 2021 were selected to calculate the normalized vegetation index and surface temperature.Based on normalized difference vegetation index (NDVI) - land surface temperature (LST) characteristic space, the temperature vegetation dryness index(TVDI) model was constructed to invert and classify the soil moisture in Zhundong mining area, and the soil moisture area at all grades was obtained. The grey prediction GM (1, 1) model was used to predict the soil moisture distribution. The results show that from 2001 to 2021, the total soil moisture in Zhundong mining area is low, mainly the grade 4 and the grade 5 soil moisture, and areas of the two accounts for more than 85% of the total area; the soil moisture was higher in the northern and eastern regions, and lower in the central and southwestern regions; there is a little annual variation of soil moisture for the grade 1, the grade 2 and the grade 3 areas, and fluctuates within the range of 0 km² to 2 000 km²; the changes of soil moisture for the grade 4 and the grade 5 areas show opposite trend. The results of the GM(1, 1) prediction show that from 2023 to 2039, the grade 1 area of soil moisture will increases slightly within the range of 0 km² to 1 000 km², the grade 2 area of soil moisture will gradually decrease and tends to 0 km², the grade 3 area of soil moisture will decrease annually, the grade 4 area of soil moisture shows an increasing trend, and the grade 5 area of soil moisture stays stable at about 3 000 km². This indicates an overall decreasing trend for soil moisture in this mining area.