Assessing Robustness in Railway Networks

The role of network in transportation systems has gained significant and increasing relevance in the literature especially focusing on vulnerability and resilience.The structure of a transport system can positively or negatively influence the geographical areas it serves. A disruption of the network causes certain socio-economic problems to all the users involved, whether they are commuters or occasional users.In this paper an empirical application, exploring the behavior of a local robustness measures, the so-called \lq\lq Effective Resistance Centrality \rq\rq (ERC), is provided and it is based on railway network of Lombardy that is the Italian region with the most diffused rail network of the country and has the highest number of train journeys in Italy per day.The effect in the changes of the topological properties of the network is studied in order to analyze the impact of the deactivation of each station or journey at a time on the robustness of the network, thus simulating potential external disturbances and shocks. Moreover, the impact of deleting the most connected stations is analyzed and it is provided a consistent comparison with several well-known topological measures that assess the node importance (betweenness, eigenvector centrality and residual closeness).Two clear findings emerge in the results, the first is represented by the consistency between the classification provided by the network manager and the importance of the stations found by the analysis. Furthermore, the geographical position of some stations, even not relevant, is critical for the connection of the overall network and the disconnection of these nodes can compromise the stability and communication between a great part of the Lombard railway stations