Modelling intra-dependencies to assess road network resilience to natural hazards

Estimating the resilience of a road network (one of the essential critical infrastructures in times of crisis) to natural hazards is crucial in achieving the goals of disaster risk reduction (DRR). This study proposes a new predictive method to implement, in an operational way, the concept of resilience by exploring the robustness of the road network in Baalbek-Hermel Governorate (Lebanon) in order to predict its future behavior in response to natural hazards occurrence. The proposed methodology consists of a predictive-spatial-analytic approach based on geospatial numerical models combined with an R-NetSwan function for modeling and simulating critical infrastructures. The results show that Baalbek-Hermel's road network is moderately resilient since it reaches a total loss of connectivity when nearly 60% of its critical nodes are blocked or damaged. Earthquakes proved to be the most disruptive hazards of this network, threatening the connectivity, starting its first damaged nodes, and causing the highest percentages of connectivity loss (70%). The novelty of this method lies in utilizing network analysis to reveal roads resilience to different natural hazards and serve several operational targets: revealing the defects of the road network for improvement or the construction of new detours, as well as allowing the first aid services to better visualize these weaknesses and to better prepare themselves. This study facilitates the implementation of a proactive approach to DRR and the protection of CI networks for better crisis response and much more effective evacuation plans.