Sensitivity analysis and uncertainty quantification of a seismic risk model for road networks

Abstract

Natural hazards may cause significant disruptions to road infrastructure, subsequently affecting road agencies, users, and productive activities. Despite the existence of infrastructure fragilities to seismic hazard and some operational consequences on network mobility, previous research has not modeled risk in terms of traffic disruptions and consequent travel time delays in subduction environments, analyzing the sensitivity to model parameters and quantified model uncertainty. This study proposes a risk framework to evaluate operational consequences in interurban road networks exposed to seismic hazard using travel time delays and propagate uncertainty in the model. Risk values are evaluated using Monte Carlo simulations, and uncertainty is propagated using a polynomial chaos expansion meta-model. The framework was applied to a very critical interurban network in central Chile. Results demonstrate that the parameters that most significantly influence risk are fragility, loss of road capacity, and traffic volume.