Measuring the Topological Robustness of Transportation Networks to Disaster-Induced Failures: A Percolation Approach

Abstract

This paper presents a framework that integrates network theory methods into infrastructure network assessment in order to investigate transportation network robustness from a topological perspective. The objectives of this paper are threefold: (1) develop a framework that can effectively measure network performance under different stress levels (hazard scales); (2) determine the constraints of infrastructure networks’ spatially embedded nature in network robustness assessment; and (3) characterize the percolation transition in infrastructure networks and systematically evaluate cities’ robustness, and in doing so provide an evidence-driven evaluation framework for urban resilience planning. In this research, 13 city and 3 state transportation networks were investigated through both the proposed simulation and an analytical framework. The results show that (1) certain theoretical methods, such as generating functions, do not apply to transportation network analysis due to their unique spatially embedded features; (2) different city and state transportation networks’ degree distribution and percolation dynamics display similar patterns; (3) critical percolation threshold identifies the robustness feature in extreme cases and provides an early warning for urban traffic disruption, and the robustness index accurately measures the network robustness from a holistic point of view; and (4) reducing network spatial complexity by decreasing the node assortativity will make the empirical simulation comply with the theoretical results, which unveils the source of spatial essence inherent in infrastructure networks. Through the validation of different transportation networks, the proposed robustness assessment framework is verified to be able to extend the robustness analysis to a general network.