Search

• Recognized the heterogeneity in the mechanisms influencing LST across housing price areas. • UGS possesses marginal utility and NDVI of 0.5 is the most efficient cooling strategy. • The green spaces' cooling effect is pronounced in areas with medium to low housing prices. • Reducing the building density in affordable housing areas will notably mitigate LST. With rapid urbanization, the urban heat island (UHI) effect has become a major problem worldwide. This study focuses on the relationship between Urban Green Space (UGS) and Land Surface Temperature (LST), providing insights into how natural environment, urban construction, and human activities influence this relationship by analyzing housing areas of different tiers. Utilizing the LightGBM machine learning approach, this study takes Chengdu City in Southwest China as a research area and finds that green space has a more significant cooling effect on LST in the middle-market housing and affordable housing areas. Furthermore, within the NDVI range of 0.37 to 0.48, the LST decreases by 0.63℃ for every 0.1 increase. Additionally, the study reveals that urban construction characteristics, such as building density and building height, exert a more significant influence on LST in middle-market housing and affordable housing areas, with a contribution rate of 23.61 % observed in middle-market housing areas and the lowest contribution rate of 17.84 % observed in high-end housing areas. Overall, our research emphasizes the importance of addressing urban thermal disparities, particularly in diverse housing areas, by prioritizing green space augmentation and optimizing built environments, thereby offering actionable insights for equitable and sustainable urban planning. [ABSTRACT FROM AUTHOR]