E-scooter dynamics in a tourism-driven city: Exploring spatiotemporal patterns and nonlinear effects of the built environment in Antalya, Türkiye

Exploring the travel behavior of e-scooter users and the relationship between the built environment and e-scooter usage helps urban planners and policymakers to allocate urban resources efficiently and meet demand. To the best of our knowledge, this study presents one of the first empirical evidence focusing on e-scooter usage in a tourism-driven city. It explores the nonlinear and threshold effects of the built environment on e-scooter usage using a machine learning technique, i.e., Gradient Boosted Regression Trees (GBRT). To this end, it examines extensive 8-month trajectory records of 16,480 e-scooter trips in Antalya, Türkiye. Trips were recorded mainly in the southwest part of Antalya, specifically along the coastline. In summer, e-scooter is primarily used for tourism in coastal and residential areas, while in winter for educational and residential purposes. Besides, e-scooter usage was impacted negatively by temperature and rain. People tend to ride e-scooters mainly in the evening during the summer and at midday during the winter season. The results indicate that the most influential factors on e-scooter usage were recorded as educational and touristic/offshore land use variables, residential area land use, intersection count, and bus accessibility, with respective 42%, 24%, 8.08%, 5%, and 4% contributions. The results of this research can be useful for transportation policies over dockless e-scooter usage in port and touristic cities such as Antalya.