While the relationship between built environment, mobility choices and travel productions has been well documented, the impact of built environment on traffic performance, particularly congestion, is not extensively investigated. In existing investigations, the mechanisms by which built environment affects traffic performance-whether direct or indirect-remain unclear. Furthermore, traffic performance is mostly quantified by aggregated and rough performance metrics that fail to capture the operational characteristics of traffic. This study examines both the direct and indirect relationships between built environment and traffic congestion using the macroscopic fundamental diagram (MFD) to connect traffic dynamics with built environment properties. The study hypothesizes that built environment has an indirect impact on congestion and analyzes the possible indirect effects utilizing the bootstrap mediation analysis. Two components, namely the “operational capacity” and “travel demand”, are defined as mediating factors. The analysis is conducted based on data collected from 133 Statistical Area Level 2 (SA2) in the Melbourne region, Australia, including both traffic and built environment data collected from 2019. Results show that “road network connectivity” and “road network structure” can intensify congestion, while “public transit accessibility” is critical in congestion reduction. The analysis also reveals that “operational capacity” serves as a mediating factor in this relationship. Notably, this analysis identifies less pronounced mediating effects, hinting that built environment may exert a more direct influence on congestion levels. The proposed analytical method and observations hold practical value for planners and policymakers in developing strategies for planning and infrastructure development. A critical takeaway is the imperative to assess the impact on the consequent operational traffic efficiency to ensure that potential congestion is proactively mitigated through informed and sustainable planning decisions.
Stored only when both papers are in the corpus — 19 citations exist upstream.
Imported from OpenAlex (CC0) — nothing here is generated.