In recent years, Regional land use/cover in China change a lot, and climate simulation need to do more research, especially related to impacts on climate simulation by using new accuracy land cover data sets in region climate model. In this paper, With the original land use / land cover data based on (USGS) and high accuracy land use/cover data (LUC90), Regional Integrated Environmental Model System (RIEMS2.0) were used for 10 consecutive years numerical simulation about impacts on climate, and the analysis was focused on the surface temperature and surface fluxes, testing statistical significance. The results showed that: 1)After using LUC90 data simulation, winter cold bias value compared to observation value in most regions in China was reduced, and different seasons temperature bias of northeast area were reduced, passing the test of significance in the summer(P < 0.01). The annual mean temperature in China, North China and South China in the winter of interannual variability simulation was better than in the summer. 2) Effects of land use change on the distribution of the surface fluxes in different regions and different seasons have different performance, mainly through changes in net radiation and latent heat fluxes to the impact of temperature. We found in the winter contribution of net radiation flux change to mean temperature change was large in summer, and in the summer the latent heat flux change contribution to temperature change was more than in winter. Different land use/cover change process in different regions in different seasons resulted to different effect on the temperature.