Abstract: In order to understand the surface temperature and vegetation coverage changes and trends in the Shuixigou fire area of the Junggar coalfield, Landsat image data in the summer of from 1998 to 2018 are selected, and retrieved by a single-window algorithm and pixel dichotomy model to get the surface temperature and vegetation coverage of the Shuixigou fire area in the southern Junggar coalfield, and combined the artificial threshold method and survey data to extract temperature anomaly area and first-level vegetation coverage area.The GM(1, 1) is used to predict and analyze its change characteristics.The results show that the maximum temperature, minimum temperature, average temperature, and temperature threshold changes in the study area have obvious correlations on the time scale.West in temperature anomaly area extended to the southeast, the center to the northwest, and keeps extending to the east-west, but the area of the east decreased significantly.The area of first-level vegetation coverage is greatly affected by the temperature anomaly area, and the overall fluctuation changes, showing an increasing trend from 1998 to 2018.The results of the GM(1, 1) show that during the forecast period, the area of temperature anomaly area and the area of primary vegetation coverage are in a positive relationship.As the area of temperature anomaly area increases, the area of primary vegetation coverage shows an increasing trend.