Impacts of 2D/3D building morphology on vegetation greening trends in Hong Kong: An urban-rural contrast perspective

Building morphology profoundly impacts the microclimate, potentially affecting vegetation greening. However, the effects of 2D/3D building morphology on vegetation greening, especially the urban-rural disparities, remains understudied. In this study, we examined the effects of building morphology on vegetation greening in urban and rural areas in Hong Kong by employing a machine learning model. Vegetation greening trends were derived using the Enhanced Vegetation Index (EVI) through the Theil-Sen median method and the Mann-Kendall (MK) test. Results indicated a prevalent greening from 2010 to 2020, with a slope of 0.0024, and more significant in rural. Statistically significant but low correlation existed between building morphology and vegetation greening. Their relationship exhibited notable urban-rural differences and non-monotonic nonlinearity, with 3D indexes showing a stronger impact than 2D indexes. Specifically, sky view factor (SVF) dominated in urban areas, contributing 23.60 %, while landscape shape index (LSI) was the key contributor in rural, accounting for 27.30 %. SVF, and mean building height (MBH) transitioned from negative to positive effects, whereas landscape patch index (LPI) and edge density (ED) shifted from positive to negative effects, each with distinct "turning points" in urban and rural. LSI’s impact showed a negative-positive-negative shift in urban and a negative-positive shift in rural. Building volume density (BVD) presented a positive to negative shift in urban and negative to positive shift in rural. The identified complicated relationship deepens our understanding of the drivers of vegetation greening in the built environment, informing the optimal building morphology threshold for efficient greening effect toward sustainable development.