Exploring the relationship between built environment and spatiotemporal heterogeneity of dockless bike-sharing usage: A case study of Shenzhen, China

Dockless bike-sharing (DBS) system provides a convenient and flexible travel mode and is believed to be an essential means to improve travel efficiency and promote sustainable urban mobility. Despite an increasing body of scholarship on this issue, the spatial and temporal disparities of the relationship between built environment and DBS usage are less discussed in existing literature. In particular, few studies have taken the coexistence of complementary and competitive relation between bike-sharing and existing transit systems into account. Using 250 m grid level data in Shenzhen, this study applied multi-scale geographically weighted regression (MGWR) to reveal the local effects of built environment on bike utilization. The results found that population and employment density are generally positively related to bike usage. A significant competitive effect on DBS is observed in areas with dense metro stations, while complementary effect is found in other areas. In addition, bus coverage is found to negatively affect bike usage, meaning that public bus would substitute bike-sharing to travel. Mixed land use can attract bike usage on weekday but not on weekend, while residential as well as commercial areas have the potential to generate more bike demand in urban center and inner suburbs but not in exurban areas. Conversely, industrial areas in suburbs contribute to bike usage. Furthermore, office building plays a critical and positive role in bike usage, especially in suburbs, whereas restaurant shows negative effects during peak times. This research sheds light on the relationships between built environment and travel behavior, and the findings can provide tailor-made policies towards cycling-friendly environment and sustainable transportation.