jaue2023053: Exploring Nonlinear Impact of Streetscape Greenery on Pedestrian Volumes

Abstract

Streetscape greenery plays an indispensable role in contemporary urban planning and design, widely regarded as a potent instrument for enhancing the comfort and allure of urban pedestrian spaces. However, past research predominantly focused on the analysis of individual pedestrian behavior, such as walking duration and frequency, while neglecting the interplay between collective pedestrian activities imbued with vitality and sociability and their relationship with greenery. Particularly, the responsiveness of pedestrian volumes to streetscape greenery remained understudied. Therefore, this study employed Baidu Street View images to assess streetscape greenery and utilized a machine learning model (random forest model) to explore the non-linear relationship and threshold effects between streetscape greenery and pedestrian volumes. Our findings reveal that: (1) the evaluation of streetscape greenery through street view images exhibits an inverted "U"-shaped impact on pedestrian volumes. (2) A conspicuous threshold effect emerges when the streetscape greenery reaches 0.33. Our research underscores the need for urban planners and policymakers to consider the visibility of streetscape greenery from the collective pedestrian perspective, optimizing the cost-effective enhancement of residents' exposure to green spaces.