Spatial-temporal patterns and influencing factors of the Building Green View Index: A new approach for quantifying 3D urban greenery visibility

Abstract

The vertical expansion of urbanization has increased the morphological heterogeneity of the urban landscape, affecting the physical and emotional wellbeing of urban dwellers by obstructing the view of greenery. In this study, multisource spatial data was used to calculate the building green view index (BGVI). Baidu Street View (BSV) images were collected for comparison with the corresponding BGVI results. A random forest model was used to analyze the contributions and marginal effects of multiple influencing factors on BGVI. The results indicated that approximately 76.10% of the sampled sites had a higher BGVI than the street view green view index, indicating buildings' superiority of visible greenery in height. The western edge of the research region frequently had the highest BGVI. Meanwhile, the hotspot regions were primarily located in the west, which was more consistent with the distribution of high value zones. The green area within the maximum visible distance, the maximum visible distance, and average height were the most influential factors of BGVI according to marginal effects analysis with the highest IncNodePurity. As a quantitative measure of urban dwellers’ visual accessibility to green space, the BGVI will contribute to urban green planning and the development of landscape architecture.