Risk-informed decision framework for built environment: The incorporation of epistemic uncertainty

Abstract

Managing a risk to a built environment from natural and man-made hazards is an important issue for the prosperity of a nation. Risk assessment forms the basis for the risk management, which often involves uncertainties that arise from our ignorance about the risk, such as lack of data, errors in collected data, and assumptions made in the modeling and analysis. This uncertainty that arises from imprecise information is referred to as epistemic uncertainty, as opposed to the aleatory uncertainty that arises from the variability of possible outcomes. If epistemic uncertainty prevails, assessing and managing a risk rely on risk perception of a decision maker. Studies have suggested that the risk of low-probability high-consequence events tends to be overestimated by the public. Thus, the role of risk perception in risk management of civil infrastructure exposed to natural and man-made hazards becomes significant because of the potential catastrophic consequences of such risks (e.g. casualties, functional and economic losses of the built environment, etc.) to the public. The consideration of epistemic uncertainty and risk perception in risk assessment of a built environment may lead to a risk management solution that is different from what is obtained when it is not incorporated. In this study, we present a risk-informed decision-making framework that can assist decision makers, including governmental agencies that allocate limited resources to enhance the safety and security of the civil infrastructure. In this framework, epistemic uncertainty is incorporated by utilizing generalized interval probability theory and cumulative prospect theory. The framework is illustrated with an example of regional hurricane risk management for residential buildings located in Miami-Dade County, Florida, considering the effect of changing climate.