Modeling epistemic uncertainty in resilience assessment for water transportation systems

The study of resilience has aroused significant interest in the literature recently. The domain-specific applications of the concept have led to diverse definitions and characteristics described therein. In general, resilience is measured with respect to a system performance or figure-of-merit. Three foundational aspects have to be taken into consideration for this. Firstly, the description of the pre-, during and post- disruptive event should be built upon a model that can account for the evolution and intervention of the dynamic trajectory of the potential adverse event. Secondly, the subjective and epistemic nature of resilience analysis is to be reflected in the family of curves, which convey the knowledge of the analysts. Thirdly, resilience is dependent upon the foreseen adverse scenarios, especially the stress of the disruptive event: thus, a spectrum of scenarios with varying profiles should be scrutinized.