Seasonal land surface temperature variations and their influencing factors in urban residential and urban village homestead blocks on Xiamen Island, China

Abstract

Coastal cities face unique thermal challenges due to compact spatial layouts, limited land, and monsoon-sensitive climates, yet mechanisms driving land surface temperature (LST) remain underexplored. This study investigates seasonal LST variations on Xiamen Island, China, and their associations with natural and anthropogenic factors. Stepwise regression using 18 variables showed that spatial location, morphology, and demographics significantly influence LST. In summer, adjusted R² reached 0.548 for urban residential blocks (URBs) and 0.852 for urban-village homestead blocks (UVHBs), declining in winter. In URBs, latitude and population density positively correlated with LST, while building density and floor area ratio also affected temperature. In UVHBs, landscape heterogeneity (Shannon diversity index) negatively correlated with LST. This study’s novelty lies in comparing urban and urban-village residential types in a coastal context, highlighting how spatial structure and human activity shape seasonal thermal patterns. The results emphasize careful interpretation of correlations and inform climate-adaptive urban planning.