Adaptive estimation method for economic losses of transportation network under extreme weather conditions

Extreme weather events increasingly threaten the safe operation and economic viability of transportation networks, making rapid and accurate loss assessment critical for emergency response and resilience planning. This paper proposes an adaptive estimation framework that integrates multi-source heterogeneous data—including meteorological observations, traffic flow detectors, social media geotags, and macroeconomic statistics—to enable real-time situational awareness. A quantitative impact model defines disturbance variables for rainstorms, snowstorms, typhoons, and fog, and employs a power-law function to dynamically map weather intensity to traffic loss rates, with online parameter adaptation for changing conditions. To capture systemic ripple effects, a dual-layer complex network representing conventional bus and subway systems is constructed, and a load-redistribution-driven cascade failure algorithm simulates chain reactions under extreme weather. Network service capacity loss is computed using passenger-flow weights and subsequently converted into both direct and indirect economic losses, forming a closed-loop adaptive estimation process. Experimental results demonstrate that the method tracks dynamic network topology changes with a deviation below 4%, achieves relative estimation errors between 7% and 9%—significantly outperforming benchmark approaches—and completes each evaluation in approximately 28–30 seconds, well within the 15-minute real-time decision window. This framework offers a robust, scalable solution for urban transport authorities to assess economic impacts under extreme weather uncertainty.