Caitlin Spence, Ph.D Candidate
Environmental and Water Resources Engineering Program, Department of Civil Engineering, University of Massachusetts Amherst
Hydrologic alteration caused by water management operations and nonstationary climate threatens freshwater ecosystems, while flood management systems’ performance may also change under nonstationary climate. Ecosystem and flood risk managers need tools to design systems which efficiently achieve flood risk reduction objectives while sustaining ecosystem functions and services into an uncertain hydrologic future. Using principles introduced by Eco-Engineering Decision Scaling (EEDS), we extend robust optimization techniques to design flood management systems that meet both economic and ecological goals across a broad range of potential climate trajectories. The approach is illustrated through an example application to the Iowa River flood management system. Iowa City has experienced multiple severe flooding episodes in the recent decades, raising concerns that the flood management system provides an insufficient degree of protection. The robust optimization search seeks a design based on the existing system which will lower the combined cost of flood damage and management while increasing ecologically beneficial bank overflow events downstream of the city. Results stemming from parallel design searches which each seek to maximize an alternate measure of robustness reflect implicit assumptions about risk preferences and present stakeholders with a range of adaptation options.