Using decision analysis to include climate change in water resources decision making

Is the prospect of possible climate change relevant to water resources decisions being made today? And, if so, how ought that prospect be considered? These questions can be addressed by decision analysis, which we apply to two investments in the Great Lakes region: a regulatory structure for Lake Erie, and breakwaters to protect Presque Isle State Park, PA. These two decisions have the elements that potentially make climate change relevant: long lived, ''one shot'' investments; benefits or costs that are affected by climate-influenced variables; and irreversibilities. The decision analyses include the option of waiting to obtain better information, using Bayesian analysis to detect whether climate change has altered water supplies. The analyses find that beliefs about climate change can indeed affect optimal decisions. Furthermore, ignoring the possibility of climate change can lead to significant opportunity losses-in the cases here, as much as 10% or more of the construction cost. Yet the consequences of climate uncertainty for Great Lakes management do not appear to be qualitatively different from those of other risks, and thus do not deserve different treatment. The methods of sensitivity analysis, scenario planning, and decision analysis, all of which are encouraged under US federal guidelines for water planning, are applicable. We recommend increased use of decision trees and Bayesian analysis to consider not only climate change risks, but also other important social and environmental uncertainties.