Measuring network interdependency between dependent networks: A supply-demand-based approach

•A supply-demand-based network interdependency approach is proposed and applied.•Interdependency increases with the increase of similarity in topology and function.•Functional correlation between networks is dominating network interdependency.•The heterogeneity of network flows has great impacts on network interdependency.•The larger transfer capacity of stations the more interdependency with bus stops. Evidences demonstrate that real-world networks are becoming more and more interacted with each other. Most studies have contributed to the analysis of interdependent networks in topology. However, existing researches usually address the impacts of topology connectivity, and ignore the function interdependency between networks underestimating the performance of interdependent networks under failures. Therefore, the function interdependency is left unknown and unexplored. This study presented a supply-demand-based interdependency approach integrating topology and function interdependency. The proposed approach not only captures the function interdependency, but also explores bidirectional dependency between interdependent networks. The methodology was applied to the interdependent metro and bus networks of Xi'an, China. Results demonstrate that topology and function characteristics have joint effects on network interdependency. The similarity in network topological connectivity and function would increase network interdependency. Furthermore, function dependency has a dominant impact on interdependent networks. Even if the topology dependency is significantly decreased under failures, the performance of interdependent networks would depend largely on the bidirectional function dependency. Improving the network interdependency could increase the capacity of both networks against attacks. Findings of this work would have implications for the planning and emergency responses of interdependent infrastructure systems building resilient network of networks under disruptive events.