Abstract: Human activity is an important factor affecting regional thermal environment change,and it is of great significance to explore the response relationship between human activity intensity (HAI) and land surface temperature (LST) for regional sustainable development.Based on MOD11A1 ground temperature data and combined with multi-source data such as land use (LU),population density (PD),night light (NTL),grassland use intensity (GUI) and existing biomass (EB),this paper proposed a method suitable for measuring the HAI in Three-River Source Region,and analyzed the spatio-temporal variation characteristics of HAI and LST in Three-River Source Region from 2000 to 2020.Comprehensive use of spatial autocorrelation and spatial autoregressive model to explore the spatial relationship between them.The results show that:(1) According to the regional ecological characteristics of Three-River Source Region,the HAI index model constructed in this paper can effectively identify the spatial distribution of HAI in Three-River Source Region,and can better distinguish the spatial differences of HAI in Three-River Source Region;(2) The average value of HAI in Three-River Source Region in the past 20 years was 0.285,and the overall intensity was low.The spatial distribution characteristics of HAI and LST were both high in the east and low in the west,and the two were significantly positively correlated.(3) LU,PD,GUI and EB in HAI index significantly affect the change of LST in Three-River Source Region,while there is a phenomenon of "warming lag" between NTL and LST,and its response to LST is not obvious.In general,in order to slow down the further rise of LST in this region and alleviate the challenges faced by the regional ecological security system,it is necessary to limit the spread of high human activity areas.