Abstract: Based on Landsat 5 thematic mapper (TM) data, the land surface temperature (LST) of urban area in Nanyang were retrieved by using a mono-window algorithm method. The spatial differentiation of LST and its relationship with land surface parameters were analyzed. The results indicated that the high LST patches are mainly concentrated in warehouse area and public service area. The effect of the high LST in these areas on urban environment is negative. The water body by blocking river is the larger low LST patch, and it is positive to adjust the urban thermal environment and blocking river measurement of small-medium city. The spatial autocorrelation of LST has obvious differences in the different directions, and its range is large in east-west and northeast-southwest directions, which is related to the distribution of water body. There is a significant positive correlation between LST and road density index. Accordingly, it suggests that more water bodies and green spaces should be elaborately inserted into the warehouse lands to discretize large-area urban heat island effect for improving the urban environment.

Key words: Land surface temperature (LST), Spatial differentiation, Landsat, Road density