Optimal Design and Life-Long Adaptation of Civil Infrastructure under Climate Change and Uncertain Demands

Various climate change effects pose increasing risks to the nation’s infrastructure. Available methodologies address the risk-management problem primarily through cost-benefit analysis frameworks, which evaluate a comprehensive set of protection strategies against a wide range of simulated possible future scenarios. However, due to the substantial climate model uncertainties present over the future planning horizon, such strategies can often lead to less informed policies that might be optimal in an average sense, over the mean of anticipated future scenarios, but cannot offer adaptive solutions based on the actual climate effects evolving in time. To address these limitations, in this research, climate risk mitigation is instead formulated as a decision-making problem within a closed-loop stochastic control-based framework using Markov decision processes (MDP), taking real-time data into account, for evaluating the evolving conditions, and selecting the best possible, most informed life-cycle actions in time. Although broadly applicable, the merit of the framework will be illustrated through coastal risk mitigation against storm surge and sea-level rise in an idealized coastal city setting.