Human-Centric Risk Assessment of Metro Systems Under Extreme Rainfall

The process of urbanization and climate change are stressing metro systems, particularly posing risks to passenger safety and mobility. As a complex and interconnected network, the operational status of different components within a metro system is highly interdependent. When a station or tunnel segment is compromised by extreme rainfall (e.g., due to flooding or structural failure), the resulting disruption can propagate throughout the network, impacting passenger flow and travel efficiency. From a human-centric perspective, this study develops a novel risk assessment indicator based on average passenger travel time, integrating the Dynamic Inoperability Input-Output Model (DIIM) with a topological metric of network efficiency. The proposed model dynamically evaluates the degradation of the system performance by analyzing the initial inoperability of affected components and its propagation throughout the network. Taking the Shanghai Metro System as a case study, model parameters are calibrated using historical data, including publicly available ridership records and meteorological data. The analysis reveals the temporal evolution of system performance and identifies key risk propagation pathways under extreme rainfall conditions.

Keywords: Metro System, Extreme Rainfall, Human-Centric Risk Assessment, Dynamic Inoperability Input-Output Model (DIIM), Network Efficiency.