Spatiotemporal diagnosis towards mismatch between traffic crash risks and safety governance in rapidly urbanizing...

The velocity of urban functional expansion often outpaces the evolution of safety governance, creating a critical spatiotemporal mismatch that undermines traffic safety. However, existing studies typically rely on static snapshots, failing to capture the dynamic migration of these traffic accident risks or decode the evolving underlying mechanisms behind the governance deficit. To investigate the spatiotemporal evolution of this mismatch and analyze its underlying behavioral and structural causes, this research proposed a novel diagnostic framework integrating Geographically Weighted Regression with Random Forest interpretation. Using a decade of crash data (2013-2022) from a rapidly urbanizing Chinese city, we reconstructed the spatiotemporal trajectory of high-risk mismatch zones, where observed crash frequencies systematically exceed the levels predicted by static built environment factors. The empirical results indicate that this safety-governance mismatch manifests through three critical dimensions: spatiotemporal migration, mechanism heterogeneity, and spatial variation in crash patterns. First, spatially, the center of gravity of safety risks has shifted systematically from the mature city center to the expanding suburban fringe, empirically verifying the lagging governance hypothesis. Second, temporally, the dominant risk factors have fundamentally shifted. Behavioral factors and conflict patterns consistently dominate the risk hierarchy over the decade, while infrastructure factors remain marginal. This confirms that the governance deficit is a management deficit rather than a physical infrastructure gap. Finally, through mechanism deciphering, a differential diagnosis isolates the unique crash patterns of these high-risk zones. Unlike the congestion-induced passive errors in the core, these fringe zones are characterized by Improper Operation and severe Vehicle-Non-Motor conflicts, reflecting a critical gap in managing mixed traffic flows. The successful application of this diagnostic framework demonstrates its efficacy in identifying the blind spots of traditional static management. These findings challenge the traditional "build-first, manage-later" paradigm. Instead, a dynamic, precision-based governance framework is advocated to synchronize safety management with the spatiotemporal evolution of urban risks.