Lessons from Grand Forks: Planning Structural Flood Control Measures

Abstract

In the spring of 1997, the Red River Valley experienced a traumatic flood event that inundated 80% of Grand Forks, N.D., and arguably inflicted the largest economic damage per capita ever suffered by an American city and certainly far greater losses that cannot be measured in dollars. By probing what happened, this study found that the losses were made much larger by sequential surprises. For reducing surprises, integrated physically based weather-hydrologic-hydraulic modeling can quantify maximum flows and stages as well as extreme scenarios during floods in real time to provide “worst case” as well as “most probable” stages. It can also be used to structure forecasts around thresholds used in flood fighting decisions and to structure hazard mapping around optional configurations of structural measures and building patterns. It is particularly important to implement contingency planning for levee failure. Stochastic hydrologic modeling can be used to stimulate preparedness and reduce surprise.