Multidimensional characteristics of urban green space and its impact in mitigating urban heat Island effects: a case study of Guangzhou

Urban Heat Island (UHI) effect is a significant concern in cities, increasing environmental and living conditions challenges. Urban green spaces are essential in mitigating UHI effects by regulating temperature. However, most studies do not adequately consider the multidimensional characteristics of green spaces in their cooling effects on the UHI. To address this gap, this study systematically investigates the cooling effects of green space by examining global and local mitigation mechanisms, with Guangzhou as a case study. Land surface temperature (LST) data were retrieved using the Mono-Window algorithm to analyze spatial variations in UHI intensity. Green space was further evaluated in terms of distribution, quality, and morphology, providing a robust evaluation. A Random Forest (RF) regression model, combined with SHapley Additive exPlanations (SHAP), was used to quantify the global contributions of UGS indicators to cooling effects. A Multiscale Geographically Weighted Regression (MGWR) model was employed to examine spatial heterogeneity. Results show an increase in high-temperature zones in Guangzhou from 2019 to 2024, particularly in Panyu and Nansha, indicating intensified UHI effects. Northern forested areas exhibited strong cooling effects, with Cooling Effect Index (CEI) values above 0.45. The distance to green space (DGS) was found to be the primary factor influencing cooling, accounting for 75.3% of the variance in CEI in 2024. The MGWR model confirmed that DGS has a negative correlation with LST, indicating that closer proximity to green spaces is associated with lower temperatures. This study provides evidence for optimizing green space planning and thermal environment management in Guangzhou and other megacities.