Understanding seasonal contributions of urban morphology to thermal environment based on boosted regression tree approach

Urban morphology significantly affects the urban thermal environment. Seasonal impacts of two-dimensional (2D) and three-dimensional (3D) urban morphology on land surface temperature (LST) remain uncertain, and the impacts exist scale effects. Thus, taking Beijing as the study area, boosted regression tree (BRT) model was used to investigate the seasonal contributions of urban morphology to the thermal environment. Building density (BD), building height (BH), floor area ratio (FAR), sky view factor (SVF), and frontal area index (FAI), were used to comprehensively characterize urban morphology, and 13 scales ranging from 30 m to 600 m were used to investigate scale effects. The results showed that there are obvious spatial differences in LST and urban morphology indicators in the study area. 270 m was determined as the optimal scale for modeling in the study area. BH and BD are the domain indicators, which together contribute more than 75% of the variance of LST among four seasons, while the relative influences of SVF, FAR, and FAI are relatively low. Relationships between urban morphology indicators and LST are nonlinear among four seasons. The findings provide a scientific un-derstanding for urban planners on mitigating the UHI effects through optimizing buildings.

Automated summary

Urban morphology significantly affects the urban thermal environment, and there are scale effects in the seasonal impacts of 2D and 3D urban morphology. Using the boosted regression tree model, the study investigates the contributions of urban morphology to the thermal environment in Beijing. The results show that building density and building height have the greatest influence, while sky view factor, floor area ratio, and frontal area index have relatively low impacts. These findings have scientific implications for urban planners in mitigating the urban heat island effects through optimizing buildings.