A cross-city exploratory analysis of the robustness of bus transit networks using open-source data

A robust bus transit network is of fundamental importance for sustainable development by alleviating urban problems. This paper aims to explore the robustness of 57 bus transit networks from the aspect of transferability. Bus transit networks are constructed using open-source data from the same data source for ensuring a consistent comparison, and the network robustness is analyzed using giant component (GC) to represent the maximum scale of transferability and network efficiency (NE) that characterizes the overall efficiency of transferability. (1) The results reveal a universal heavy-tailed distribution of network betweenness irrespective of cities and indicate target attack is more destructive to network robustness than random attack. (2) Different cities have different degrees of robustness, where most large cities tend to be more vulnerable than small cities and NE is more likely to be affected by target attack than GC. (3) The impact of target attack may become weaker than random attack after removing a certain percentage of nodes, which varies in different cities. (4) Thereafter, we present clusters of cities according to similarities of their network robustness. Thus, our comparative results can benefit transit planners and policymakers by enhancing the robustness of bus transit networks. (C) 2021 Published by Elsevier B.V.

Automated summary

A strong bus transit network is crucial for sustainable development by addressing urban issues. This paper examines the robustness of 57 bus transit networks in terms of transferability. Results show a universal heavy-tailed distribution of network betweenness, with target attacks being more detrimental to network robustness than random attacks. Different cities exhibit varying degrees of robustness, with larger cities being more vulnerable than smaller ones. NE is more likely to be affected by target attacks than GC, and the impact of target attacks weakens compared to random attacks after removing a certain percentage of nodes. Additionally, clusters of cities are identified based on similarities in network robustness for the benefit of transit planners and policymakers.