Multi-criteria robust optimization framework for bridge adaptation under climate change

In order to adapt civil infrastructure to changing climate conditions, quantifiable and deep uncertainties must be integrated into the decision-making process. The quantifiable uncertainties, i.e. variability for which a likelihood can be defined, are typically integrated into the management process by considering the reliability or risk level of a structure. The deep uncertainties, i.e. the variability for which a likelihood cannot be defined, are beginning to be integrated in the decision making process as a few robust decision making procedures have been proposed. However, the deep uncertainty associated with the multiple feasible future climate scenarios also provokes a “wait and see” mentality for some decision makers, causing the flexibility of a strategy to be valued. This paper introduces the Gain-Loss Ratio (GLR) as a metric that systematically quantifies what may be gained by postponing adaptation while also considering what is lost with the delay. Additionally, bi-objective optimization models for optimizing bridge adaptation strategies under deep uncertainties are proposed; the advantages and disadvantages of each are highlighted as they pertain to the management of a typical riverine bridge. Two rivers are considered that have comparable climate change trends as those predicted for the Columbia and Mississippi Rivers. It is demonstrated that the desire for flexibility may be justified for certain locations, but may be detrimental in others.