Influences of buildings on urban heat island based on 3D landscape metrics: an investigation of China's 30 megacities at micro grid-cell scale and macro city scale

Context The building landscape greatly affects the urban heat island (UHI), especially in three-dimensional (3D) space, by changing the energy flow between the land surface, the building surface and the lower atmosphere. Objectives This study quantitatively analyzed the relationship between the 3D spatial pattern of buildings and UHI in China's 30 provincial capitals/municipalities and discussed them at grid-cell scale and city scale, respectively. Methods In consideration of the spatial heterogeneity of the urban environment, Geographically Weighted Regression Model (GWR) was selected to identify the effects of 3D building landscape pattern on summer UHI among 30 megacities of China at both micro grid-cell scale and macro city scale. Nine landscape metrics that used to describe the 3D structure of buildings and the UHI that calculated by hot-spot analysis were collected as input variables. Results The floor area ratio (FAR), the average height (AH), and the space congestion degree (SCD) are the most influential factors affecting UHI. AH and SCD are negatively correlated with UHI, while FAR is the opposite. However, these relationships are not static, and they will change when interfered with other factors. The relationship between FAR and UHI becomes negative in the case of relatively low FAR value. In areas with low building coverage ratio, AH is positively correlated with UHI. Conclusions The results of this study revealed the complicated association between the 3D building spatial pattern and UHI at micro and macro urban contexts, which was significant for decision-makers to formulate policies based on local conditions.