A twentieth century history of flooding in the U.S. and what the future might hold

The National Weather Service has collected information on flood damages and fatalities for the period from 1903 to the present. Data collection procedures and data quality have varied over time and introduce some uncertainty in the interpretation of the data. Due to the high degree of year-to-year variability in the time series of both fatalities and damages, combined with the time variation in data collection methods, formal trend analyses are questionable. However, when adjusted to constant dollars, there is an indication of a greater frequency of large flood losses in the more recent record. Possible explanations will be considered.

More detailed examination of the data reveal that most deaths are due to flash flood events, which usually do not make a substantial contribution to the annual damage totals. A consistent feature in the data is that in most years a few large flood events account for the vast majority of the total flood losses. Several specific flood events will be examined as examples of the impacts flooding can have.

Finally, how floods might impact the U.S. into the next decade will be examined. This will include consideration of both forecast improvements and actions society might take to mitigate the consequences of flooding. Growing skill in climate forecasting can be combined with hydrologic forecast schemes to produce better flood forecasts. But there are currently some limitations that will be discussed. Climate-hydrological forecast systems also hold promise for major improvements in water supply management, benefitting agricultural as well as municipal and industrial users of water, as well as helping optimize the efficiency of hydro power generation.