Uncertainty Analysis of Interdependencies in Dynamic Infrastructure Recovery: Applications in Risk-Based Decision Making

Abstract

The development of preparedness strategies for natural and malevolent man-made hazards and approaches with which to compare such investments is fraught with uncertainty. The dynamic inoperability input-output model (DIIM) quantifies the inoperability that propagates through interdependent sectors following a disruptive event and then diminishes with time. This approach has been shown to quantify the efficacy of preparedness strategies for interconnected sectors of the economy. Work presented in this paper strengthens the DIIM with a multiobjective approach—the uncertainty DIIM—that evaluates the inherent uncertainty in the parameters of interdependency and its impact on projected economic loss calculated using the DIIM. Preparedness strategies can then be compared based on projected economic loss and on their sensitivity to changes in the interdependent nature of infrastructure sectors. Additionally, key sector analyses are discussed, where sectors are ranked according to their sensitivity to changes in interdependent relationships. Such enumeration of key sectors allows decision makers to focus on certain sensitive infrastructure sectors for the development of preparedness strategies. The models developed in this paper are illustrated with numerical examples.