Attributing Urban Traffic Performance Loss to Spatiotemporal Storm Patterns: A Full-Spectrum Analysis across a 22-Year Rainfall Record

Urban transportation resilience is increasingly threatened by the complex spatiotemporal dynamics of rainfall, which trigger cascading disruptions through pluvial flooding-induced network perturbations. However, the resulting impact patterns of rainfalls on transportation network performance remain ill understood, underscoring the need for systematic assessment across the full spectrum of rainfall conditions. Thus, this work integrates high-resolution pluvial flood modeling with microscopic traffic simulation to investigate traffic performance degradation in the Beijing Municipal Administrative Center across 22-year high-resolution rainfall scenarios. SHapley Additive exPlanations (SHAP) are utilized to attribute variations in network performance to specific spatiotemporal rainfall characteristics, identifying the dominant drivers of traffic congestion. Building on these mechanistic insights from the full-spectrum series, we systematically reveal the critical thresholds that trigger undesirable transitions from stability to failure. These thresholds serve as a vital scientific reference for the development of impact-based early warning systems, facilitating proactive disaster mitigation and enhancing urban resilience.