Evaluation of Snowmelt Flood Forecasting Techniques for the Upper Midwest

Abstract

Certain areas of the midwestern United states are normally blessed with sufficient quantities of water to adequately meet the needs of agriculture, urban population centers, industry, recreation, navigation, and wildlife. Within recent years, however, melt water from the winter snowpack and early spring rains have combined to produce excessive volumes of runoff, causing extensive flooding. This overabundance of water during the springs of 1965 and 1969 caused a combined total of more than $300 million in damage. The monetary figure includes damage to urban areas, agricultural losses, and disruption of transportation and communication networks as well as municipal, lock and dam, and other unclassified losses. This figure can in no way indicate the great personal anguish of the thousands of people who were driven from their homes by rising flood waters. Both in 1965 and in 1969, early warnings of the snowmelt floods greatly reduced the amount of physical damage that occurred. These flood forecasts allowed preparations to be made which substantially reduced losses. In 1969 Operation Foresight, a joint flood relief effort spearheaded by the Corps of Engineers, reduced damages by more than $100 million at a cost of $15 million. Accurate forecasts of floods are necessary not only to permit general warnings to the public, but also to facilitate proper design of emergency dikes and pumping stations, floodproofing, personnel evacuation, and proper operation of existing flood control structures and navigation facilities. Although emergency flood relief measures based on flood forecasts greatly reduce flood damage, they are only a temporary solution to flood problems. However, even with permanent flood control structures, accurate foreoasts will oontinue to be needed for proper operation of reservoirs and interior drainage facilities and for floodproofing and other operations. The purpose of this study was to evaluate the performance of some of the available oomprehensive mathematioal models with respect to their usefulness in predicting snowmelt and rainfall floods in the Upper Midwest.