Collaborative Research: CAIG: Artificial Intelligence-Driven Optimization of Coastal Morphodynamics and Infrastructure Resilience

Barrier islands support coastal communities and infrastructure exposed to sea-level rise and storm-driven hazards. To adapt, communities both respond to and alter the processes that shape coastal landscapes. The long-term effects of human interventions on barrier islands remain uncertain. This project addresses the relative importance of uncertainties in human actions and environmental forcing in the evolution of barrier islands over multiple decades. By developing an adaptive decision making framework that accounts for cascading uncertainties in barrier island evolution, this project will inform improved coastal management across the United States. Existing approaches to modeling barrier island futures largely treat human decisions as static. In this project, a geomorphic model is combined with reinforcement learning (RL) to create an adaptive decision making framework. The framework is used to examine how interactions among uncertainties manifest as changes in hazard risk and landscape shape and function through time. The project team will: (1) characterize how different sources of uncertainty influence projected barrier island futures under static decision making; (2) employ RL to guide adaptive decision making in the geomorphic model under uncertainty; and (3) develop methods to quantify the relative importance of uncertainty sources through analysis of the RL models. The methods developed in this project can be generalized to address management of any coupled human-natural system that can be structured as a sequential decision-making problem. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.