Interactions among flood predictions, decisions, and outcomes: Synthesis of three cases

To complement other flood mitigation measures, hydrometeorological predictions are often used in decisions leading up to and during floods. Understanding the role played by predictions in flood events can help forecasters provide more useful information, and it can help decision makers use this information more effectively as part of broader flood loss-reduction strategies. This article examines the interactions among predictions, decisions, and flood-related outcomes by analyzing three cases of severe flooding in the United States: the Red River basin flood of April 1997 in Grand Forks and Fargo, N.D.; the Fort Collins, Colo. flood in July 1997; and the Pescadero Creek basin, California flood in February 1998. The floods occurred in different hydrometeorological and societal circumstances, had different types of predictive information available, and had different societal impacts, providing an opportunity to compare and contrast lessons learned. Issues explored include the interplay between the floods and their hydrometeorological and societal context and the roles of predictions and predictive uncertainty in decisions and outcomes.