The rapid development of urbanization spurred the production of more large cities and the expansion of urban area,which resulted in the deteriorated urban heat island.Satellite remote sensing imagery was used to estimate the spatial information of land surface temperature and vegetation index as well as land surface fluxes over the urban area in Xi'an city.In order to understand the inner mechanism of urban heat island,not only the relationship between land surface temperature and NDVI,but also the spatial structure of land surface fluxes also has been analysed in the built-in and suburb region.A practical evapotranspiration estimation algorithm and the mono-window algorithm were utilized to calculate land surface fluxes and temperature from Landsat ETM+ image respectively in this paper.The spatial characteristic of evapotranspiration factor was obtained with interpolation method according to the feature space of land surface temperature and NDVI,then the land surface fluxes and actual evapotranspiration could be acquired by combining with energy balance equation and Priestly-Taylor formula.The results showed that there were large differences of thermal environment between built-in and suburb area of Xi'an city,and land surface temperature not only was related with NDVI negatively,but also related with land surface latent heat fluxes and actual evapotranspiration negatively,meanwhile it also was related with land surface sensible heat fluxes and soil heat fluxes positively.Further detailed analysis indicated that more obvious urban heat island appeared over Lianhu district where more factories distributed,on the contrary,urban heat island phenomena became unapparent over Baqiao district where more vegetation covered.So it could be concluded that urban heat island could be monitored using Landsat ETM+ image in terms of land surface temperature,NDVI,land surface fluxes and actual evapotranspiration perfectly,which could provide better bases for reducing effects of urban heat island on human being's life.