Characteristics of land surface temperature clusters: Case study of the central urban area of Guangzhou

Abstract Land surface temperature (LST) clusters within cities are typical regions reflecting the spatial heterogeneity of the urban thermal environment. To explore the spatial heterogeneity of the urban thermal environment in the central urban area of Guangzhou, we used the localized contour tree method to identify high- and low-LST clusters. In addition, to understand the characteristics of LST clusters (characteristics of the thermal environment and landscape distribution within the LST cluster), we obtained thermal field value (TFV) data based on the rules of equal-area division and determined the urban functional grid (UFG) based on multi-source data. We found that the characteristics of the high- and low-LST clusters significantly differed. Compared with that of the low-LST clusters, the spatial distribution of the high-LST clusters is denser. The number and total area of low-LST clusters are smaller than that of high-LST clusters, although the average area of low-LST clusters is larger. The spatial heterogeneity of the 10 types of UFGs within the high- and low-LST clusters is significant. The developed and ecological UFGs classified from 10 types of UFGs affect the formation of LST clusters. In addition, there are main UFGs that affect the mean TFV of the LST clusters. This study provides a new approach for the identification of LST clusters and a new perspective for exploring the spatial heterogeneity of urban thermal environments.