Cyclic subway networks are less risky in metropolises

¹ Guangdong HUST Industrial Technology Research Institute, Guangdong Province Key Lab of Digital Manufacturing Equipment - Dongguan, 523000, China
² Key Laboratory of Image Processing and Intelligent Control, School of Automation, State Key Laboratory of Digital Manufacturing Equipment and Technology, Huazhong University of Science and Techonology, Wuhan, 430074, China
³ Department of Electronic Engineering, City University of Hong Kong - Kowloon, Hong Kong SAR, China

^(a) zht@mail.hust.edu.cn
^(b) chenduxin@hust.edu.cn

Received: 9 February 2018
Accepted: 27 March 2018

Abstract

Subways are crucial in modern transportation systems of metropolises. To quantitatively evaluate the potential risks of subway networks suffered from natural disasters or deliberate attacks, real data from seven Chinese subway systems are collected and their population distributions and anti-risk capabilities are analyzed. Counterintuitively, it is found that transfer stations with large numbers of connections are not the most crucial, but the stations and lines with large betweenness centrality are essential, if subway networks are being attacked. It is also found that cycles reduce such correlations due to the existence of alternative paths. To simulate the data-based observations, a network model is proposed to characterize the dynamics of subway systems under various intensities of attacks on stations and lines. This study sheds some light onto risk assessment of subway networks in metropolitan cities.