An improved structure–function–risk (SFR) gravity model for identifying critical road segments in expressway networks during typhoon disaster

To address the limitations of traditional topology-only critical-section identification under typhoon-disaster scenarios, this study proposes an improved Structure–Function–Risk (SFR) gravity model. Existing studies are often limited to static topological analysis and have difficulty capturing the spatial heterogeneity of typhoon exposure and the multidimensional criticality of road sections. To address this limitation, this study reconstructs the road-section mass by coupling structural attributes, traffic-related functional attributes, and cumulative typhoon exposure derived from historical typhoon tracks and wind-field buffers from 2014 to 2023. Taking the Fujian expressway network as an empirical case, a representative gravity influence radius is selected through sensitivity analysis, and critical-section identification and network-robustness assessments based on sequential road-section removal are conducted. The results show that the highly ranked road sections identified by the SFR model are relatively concentrated in areas with high cumulative typhoon exposure and substantial traffic loads, indicating that the incorporation of exposure and functional information changes the rankings obtained from topology-only indicators. In addition, the topology-updated dynamic road-section removal strategy produces a nonlinear divergence in network performance compared with the static fixed-ranking strategy, revealing a clear performance scissors effect. This study shifts critical-section evaluation from topology-only importance toward disaster-oriented multidimensional criticality and provides insights for disaster-prevention resource allocation and climate-resilience improvement in coastal expressway networks.