Water Resource Management: Planning for Uncertainty
Even the best climate modeling cannot predict exactly how climate change will affect regional water systems like the Great Lakes. Managing water resources in the twenty-first century will require strategic planning that is effective despite deep uncertainties about the degree and pace of environmental change at regional scales. Flexible adaptation pathways are carefully calibrated decision processes that offer a practical means of managing this uncertainty. By triggering the implementation of specific strategies as predetermined thresholds are met, flexible adaptation pathways guide resource management in a practical, cost-effective manner.
With this CAREER award, Scott Steinschneider, Biological and Environmental Engineering, is developing new machine-learning tools to improve our ability to model and forecast the effects of climate change on lakes, rivers, and other water systems, with the goal of optimizing flexible adaptation pathways. Using data from the Lake Ontario eco-hydrologic system as the initial test case, researchers will explore the hypothesis that the optimal design of flexible adaptation pathways depends on 1) the mechanisms of dynamic and thermodynamic climate change influencing the water system and 2) the degree of natural climate variability and predictability.
This project pursues innovations in physics-informed machine learning to support process-guided climate simulation, hydrologic prediction, and sub-seasonal-to-seasonal forecasting. The goal is to create a computationally efficient and probabilistic modeling framework for designing flexible adaptation pathways that respond dynamically to the trajectory of climate change as it unfolds.