Geographically weighted regression method has been widely used to explore the relationship between surface thermal environment and surface element coverage due to consideration of spatial heterogeneity.However, most of the current researches only focus on a single time node.This paper selected MODIS surface temperature products and land cover data from 6 time nodes in wuhan from 2002 to 2017, and used GWR method to explore the relationship between surface thermal environment and land cover.The experimental results show that because of the urban evolution, the influence of surface land use on urban surface temperature presents obvious spatial and temporal non-stationarity.And by comparing the GWR and GTWR results of multiple time nodes, it shows that the current GTWR method has no obvious improvement on the experimental results.