Disrupting road networks: a case study on the vulnerabilities of the city of Messina

Robustness of complex systems, that is the capacity of a system to continue performing its functions after a major failure, is one of the most relevant topics of complex network analysis. This is studied by quantifying the impact of node and edge removal due to random and targeted attacks on network connectivity. The most effective strategies for network disruption can be identified to understand which network elements to protect. In this study, we develop a graph model of the road network of Messina (Italy) based on spatial data and filtered to include only roads accessible to motor vehicles. We apply various edge centrality measures, which are commonly used in spatial network analysis, alongside novel measures based on random walks, to perform targeted attacks and compare the results accordingly to random edge removals. The topology of road networks is constrained by the physical space in which they are embedded, which limits long-range connections and shapes node degree distributions. In this paper, we evaluate and compare methods for identifying critical roads whose removal significantly disrupts network connectivity. We quantify disruption effects using several metrics, such as the size of the largest connected component and the number of connected components. Our results demonstrate that current flow edge betweenness is the most effective centrality measure for simulating connectivity disruption in the proposed road network.