|
<br />Interdisciplinary paleoflood assessment of questionable extreme rainstorms and flooding and
<br />their implications to flood-hazard management
<br />
<br />Robert D. Jarrett
<br />&
<br />Joseph P. Capesius
<br />U.S. Geological Survey
<br />P.O. Box 25046, MS 412
<br />Denver, CO 80225
<br />
<br />INTRODUCTION
<br />
<br />The environmental and economic importance of major floods emphasizes the need for a better
<br />understanding of hydrometeorologic processes and related climatic fluctuations or variability. In the United
<br />States, the average annual flood damage is $2.4 billion and the average annual number of deaths is 95
<br />(Jarrett, 1991). The economic cost of floods does not include the societal costs associated with
<br />overdesign of hydraulic structures in floodplains, and overestimated magnitude-frequency relations in
<br />floodplain management and related flood studies. Overdesign and overestimated flood estimates are
<br />related to a conservative approach used in engineering design and insufficient or questionable
<br />hydroclimatic data, particularly extreme-flood data. Proper design of structures and flood-plain
<br />management requires balancing potential flood losses with the potential costs of the overdesign. The cost
<br />of overdesign often is an unrealized economic loss to society in that it utilizes limited National resources
<br />that have competing uses. These uses include, but are not limited to, construction and maintenance of the
<br />Nation's infrastructure, health care, education, defense, and social programs. A critical component of
<br />research, science, engineering, and management of floods is to determine the best estimates of the
<br />magnitude, frequency, and risks of flooding for the optimal benefit of society.
<br />
<br />One of the basic premises of hydroclimatic investigations is that data are accurate and representative
<br />(Jarrett, 1988). Accurate precipitation data are essential to meteorologic, climatologic, hydrologic, and
<br />related water-resources studies (Winter, 1981; Lindsley and others, 1982; National Research Council,
<br />1985; Jarrett, 1986; Cudworth, 1989). Accurate data are particularly necessary for estimating the
<br />frequency of extreme events, rainfall-runoff modeling, flood-plain management, and for engineering design
<br />(Jarrett, 1986, 1987b, 1988; 1989; Jarrett and Crow, 1987; Hansen and others, 1988; Levish and others,
<br />1994; House and Pearthree, 1995). Jarrett (1988) also indicated that it would be difficult to discem the
<br />effect of climatic change on hydrology because of errors and uncertainty in hydroclimatic data. The
<br />importance and value of historic-flood data is directly proportional to its age and magnitude of the event. In
<br />the past two decades, several investigators have begun to question the accuracy of historical flood data
<br />and proposed approaches reassess and thus better utilize these invaluable historic extreme-value data
<br />(Randle and Humphrey, 1984; Quick, 1991; Jarrett, 1986, 1994; Trieste and Jarrett, 1987; Jarrett and
<br />Crow, 1987; Grimm, 1993; Levish and others, 1994; House and Pearthree, 1995).
<br />
<br />Insufficient data on extreme flooding hinders improving the knowledge of the hydrometeorology of
<br />flooding. Much of the questionable data is historical data, although questionable and erroneous rainfall and
<br />flood data have been documented recently. Use of questionable and erroneous data in research studies
<br />effects the understanding of hydrometeorological processes of floods. A poor understanding about floods
<br />contributes to unnecessary loss of life and increasing flood damages or costly overdesign and construction
<br />of hydraulic structures located in the floodplain (Jarrett, 1993; Baker, 1994). Before a hydroclimatic
<br />investigation is started, it is essential to ascertain that the theory used to describe hydrologic processes,
<br />data, methods, and equations used are valid and applicable to a study area being evaluated (Jarrett,
<br />1988). Therefore, it is important to identify erroneous hydroclimatic data because use of these data may
<br />result in invalid conclusions in subsequent hydroclimatic investigations. An understanding of the source,
<br />types, and magnitude of hydroclimatic-data errors is essential to mitigate its effect on these investigations.
<br />
|