Spatiotemporal analysis of multimodal traffic flow patterns and their influencing factors: A case study of New York

Different travel modes reflect diverse demands and exhibit distinct mobility behavior patterns. Investigating the flow patterns of various transportation modes and their relationships with built environment factors is essential in urban transportation systems. However, few studies have integrated the mobility patterns of different travel modes to capture the heterogeneous effects of built environment factors. This study conducts a comprehensive spatiotemporal analysis of multimodal transportation behaviors in New York City, exploring the differences in mobility patterns and the spatiotemporal heterogeneity of environmental influences across bike-sharing, taxis, and carpool services. We examine the temporal distributions and spatial autocorrelations of different traffic flows and employ multi-scale geographically weighted regression (MGWR) to analyze the spatiotemporal differences in travel behaviors. Furthermore, we investigate the spatial patterns of how built environment factors and transportation resources influence the flow of different travel modes across various urban areas. The results indicate clear behavioral differences in the flow patterns of different transportation modes over time and space, revealing significant spatiotemporal heterogeneity in multimodal traffic flows. Specifically, bike-sharing is more sensitive in residential and land-use complex areas, dominating short-distance commuting during weekday peak hours. Taxis are concentrated in commercial areas with stable local spatial autocorrelation of flow. Carpool services, on the other hand, increase in the evening peak periods and during long-distance travel demands. Our spatial analysis reveals the complex interactions among urban mobility patterns, environment factors, and human behaviors, highlighting significant indirect effects and interdependencies. These findings guide planners in addressing multimodal dynamics, environmental heterogeneity, and diverse user needs.