Land surface temperature (LST) is a key indicator of the urban heat island effect and is affected by multiple factors. However, existing research mainly focuses on the contributions of urban landscape and meteorology, and the impact of changes in atmospheric environment has not been fully considered. Based on multisource data and a random forest model, this study quantified the independent and interactive effects of aerosols, meteorological conditions, and urban features on LST in Beijing. The results revealed that the effects of the meteorological factors and aerosol optical depth (AOD) on LST were significantly greater than those of the urban landscape index. The response of LST to multiple factors is nonlinear, and the interactions of precipitation with wind speed and vegetation have the strongest cooling effects on LST. The aerosol impact shifts seasonally, with its direct radiative effect dominating in spring and inducing a cooling of up to about 2.0 °C. Notably, the land use type plays a background role in determining the LST, and the average LST decreases by approximately 1.5 °C for every 50% increase in tree coverage. As the building height increases by 10%, the summer LST increases by approximately 2 °C. In addition, the interactions of precipitation with wind speed and vegetation were identified as having the strongest cooling effects on LST. By elucidating the nonlinear interactions among aerosol, meteorological, and urban features, this work moves beyond isolated factor analysis and offers mechanism cognition for urban planning strategies.
Zhang et al. (2025) studied this question.