Abstract
Extreme flooding can inundate critical nodes in urban metro network, triggering system failures. However, comprehensive indicators for identifying critical nodes and assessing network impacts under disruptions remain underdeveloped. Research also largely relies on hypothetical disruptions rather than flood events with specific intensities and spatial extents. This study replicates the “7·20” extreme rainstorm that occurred in Zhengzhou and applies it to Shanghai through a numerical flood simulation. Using Shanghai's weighted metro network, which includes stations, lines and passenger flow, we employ a complex network method to identify critical nodes and simulate the impact of their failure due to flooding on network functionality. Finally, we discuss the recovery priorities for failed nodes in terms of network functionality. Our results indicate that 54 stations, comprising 14.2 % of the total, may experience water inflow and failure. Stations with high centrality scores, primarily located within the city's inner ring, are particularly prone to flooding. Successive station failures lead to a loss of 1.6 %–76.5 % in the network efficiency (NE) and 0.7 %–82.7 % in the giant connected component (GCC). When the number of failed stations reaches 13, network functionality experiences a dramatic loss of approximately 50 %. Compared to the greedy algorithm (GA), the criticality-based approach is less efficient in restoring network functionality but more computationally feasible. By incorporating flood scenarios and weighted indicators, our approach offers more context-driven and practical insights. Our study provides a methodology for the rapid assessment of critical node exposure, functional impact, and optimal recovery strategies for metro network based on various scenarios.