Seasonal urban surface thermal environment analysis based on local climate zones: A case study of Chongqing.

The rapid urbanization has exacerbated the heat island effect, impacting city development and residents' health. This study, using Local Climate Zones (LCZ) as a framework, connects spatial structure, resource allocation, and thermal environment research. It investigates the spatiotemporal heterogeneity of the surface thermal environment and its driving forces, crucial for mitigating heat issues. Utilizing various data sources like remote sensing images, road network data, land use data, high-resolution street view data, and building data, the research employs the random forest algorithm to map LCZs in Chongqing's central urban area. Through mathematical statistics, equi-sector analysis, ring-layer analysis, and Pearson correlation analysis, the study examines seasonal variations in spatiotemporal heterogeneity and driving mechanisms of the surface thermal environment. Key findings include: (1) The central urban area of Chongqing is dominated by open-building and vegetation-type LCZs, with building-type LCZs showing a "clustered" distribution, while natural-type LCZs are mainly found in the suburbs with ribbon and block distribution in the urban area. (2) The surface thermal environment in the study area correlates strongly with surface cover and exhibits significant high temperature effects in summer. (3) The surface thermal conditions vary significantly among different LCZs and exhibit seasonal patterns, natural-type LCZs generally have lower temperatures compared to building-type LCZs.(4) The surface thermal characteristics within the same category of LCZs in different locations display distinct differences and seasonal variations. (5) The internal temperatures of LCZs are significantly linked to four surface attributes, each displaying seasonal fluctuations. Greenness, height, and wetness are inversely related to the surface thermal conditions, while brightness shows a positive correlation. Both seasonal variations and LCZ types differences have a noticeable influence on their respective driving mechanisms to some degree.