Abstract: Land use (LU) change can affect the forecasting performance of numerical models by influencing the energy exchange between land and atmosphere. In this study, the WRF model was used to simulate the high temperature process in the Sichuan Basin from August 1 to August 4, 2021, and the impact of LU types change on the high temperature weather forecast ability of the model was discussed. The results show that: (1) There are significant differences between the recent LU types reflected by the GLC_FCS30-2020 data and the default LU types of WRF model, mainly manifested as sharp decreases in croplands and mixed forests, and significant increases in urban and built-up, savannas, woody savannas, and cropland/natural vegetation mosaic. (2) Updating the LU types significantly improves the temperature forecasting ability for high temperature weather, especially the maximum temperature. This is mainly manifested as a decrease in the cold bias and an increase in the warm bias, that is, the ability to forecast daytime, maximum temperature, and the central area of Sichuan Basin is significantly improved, while the forecst ability of nighttime, minimum temperature, and some areas in the eastern and western parts of the Basin is slightly decreased. (3) LU changes increase surface albedo and reduce net radiation, but the changes of surface water and heat transport are the main cause of surface temperature changes. After updating the LU types, the daytime latent heat flux transport decreases, while the sensible heat flux and surface heat flux transport compensatory increase, resulting in an increase in surface temperature. The temperature changes of stations without changing the LU type may be caused by the horizontal movement of surrounding high temperatures. (4) The clearer the sky condition is, the greater the deviation between the temperature simulation and observation is, and the lower the temperature forecasting ability is. In addition, updating LU types can indirectly cause changes in cloud cover.