|
<br />Flood Characteristics of Urban Watersheds
<br />in the United States
<br />
<br />By V. B. Sauer, W. O. Thomas, Jr., V. A. Stricker, and K. V. Wilson
<br />
<br />Abstract
<br />
<br />"
<br />
<br />A nationwide study of flood magnitude and frequency in
<br />urban areas was made for the purpose of reviewing available
<br />literature, compiling an urban flood data base, and developing
<br />methods of estimating urban f1oodflow characteristics in
<br />ungaged areas. The literature review contains synopses of 128
<br />recent publications related to urban floodflow. A data base of
<br />269 gaged basins in S6 cities and 31 States, including Hawaii,
<br />contains a wide variety of tOfX)graphic and climatic character-
<br />istics, land-use variables, indices of urbanizat~onJ and flood.
<br />frequency estimates.
<br />Three sets of regression equations were developed to
<br />estimate flood discharges for ungaged sites for recurrence
<br />intervals of 2, 5, 10, 25, 50, 100, and 500 years. Two sets of
<br />regression equations are based on seven independent param-
<br />eters and the third is based on three independent parameters.
<br />The only difference in the two sets of seven-parameter equa-
<br />tions is the use of basih lag time in one and lake and reservoir
<br />storage in the other. Of primary importance in these equations
<br />is an independent estimate of the equivalent rural discharge
<br />for the ungaged basin. The equations adjust the equivalent
<br />rural discharge to an urban condition. The primary adjustment
<br />factor, or index of urbanization, is the basin development fac-
<br />tor, a measure of the extent of development of the drainage
<br />system in the basin. This measure includes evaluations of
<br />storm drains (sewers), channel improvements, and curb-and-
<br />gutter streets.
<br />The basin development factor is statistically very signifi-
<br />cant and offers a simple and effective way of accounting for
<br />drainage development and runoff response in urban areas.
<br />Percentage of impervious area is also included in the seven-
<br />parameter equations as an additional measure of urbanization
<br />and apparently accounts for increased runoff volumes. This
<br />factor is not highly significant for large floods, which supports
<br />the generally held concept that imperviousness Is not a domi-
<br />nant factor when soils become more saturated during large
<br />storms. Other parameters in the seven-parameter equations
<br />include drainage area size, channel slope, rainfall intensity,
<br />lake and reservoir storage, and basin lag time. These factors
<br />are ali statistically significant and provide logical indices of
<br />basin conditions. The three-parameter equations include only
<br />the three most significant parameters: rural discharge, basin-
<br />development factor, and drainage area size.
<br />All three sets of regression equations provide unbiased
<br />estimates of urban flood frequency. The seven-parameter
<br />regression equations without basin lag time have average stan-
<br />dard errors of regression varying from :t 37 percent for the
<br />5-year fiood to :t 44 percent for the l00-year flood and :t 49
<br />percent for the SOD-year flood. The other two sets of regression
<br />equations have similar accuracy. Several tests for bias, sensi-
<br />tivity, and hydrologic consistency are Included which support
<br />the conclusion that the equations are useful throughout the
<br />
<br />United States. All estimating equations were developed from
<br />data collected on drainage basins where temporary in-channel
<br />storage, due to highway embankments, w~s not significant.
<br />Consequently, estimates made with these equations do not
<br />account for the reducing effect of this temporary detention
<br />storage.
<br />
<br />INTRODUCTION
<br />
<br />The U.S. Geological Survey, in cooperation with'
<br />State, local, and other Federal agencies. conducts pro-
<br />grams to collect and analyze flood-runoff data in numer-
<br />ous cities throughout the United States to provide
<br />hydraulic and hydrologic data needed for zoning, plan-
<br />ning, and designing. Most of these urban programs were
<br />started during the past 10 or IS years, but some data are
<br />available for longer periods. Analyses of the data have
<br />been made mostly for individual cities and metropolitan
<br />areas and have provided those areas with valuable plan-
<br />ning and design information.
<br />With urban growth and development, there is an
<br />ever-increasing need for flood information and estimat-
<br />ing techniques in areas where little or no data exist. In
<br />1978 the Federal Higbway Administration. Department
<br />of Transportation (FHW A), contracted with the Geo-
<br />logical Survey to make a nationwide study of urban
<br />flood frequency. The purposes of the study were (1) to
<br />review tbe literature of urban flood studies; (2) to com-
<br />pile a nationwide data base of flood-frequency charac-
<br />teristics; topographic, physical, and climatic character-
<br />istics; and land-use variables for as many urbanized
<br />watersheds as possihle; and (3) to define estimating
<br />techniques that could be used in ungaged urban areas.
<br />This report describes the results of that study.
<br />The authors wish to acknowledge the Federal
<br />Highway Administration. which provided financial sup-
<br />port, and Dr. Roy Trent, FHWA, who provided the
<br />leadership and guidance to initiate the project. Special
<br />assistance from Dr. Walter J. Rawls, Department of
<br />Agriculture. Science and Education Administration, is
<br />also acknowledged. Dr. Rawls acquired and provided to
<br />the Geological Survey a large part of the data used in the
<br />study. specifically data on land use and soils. He also
<br />collaborated with the Geological Survey to compile and
<br />publish the literature review. Finally. special acknowl-
<br />edgment is given to the many Geological Survey person-
<br />nel in district offices throughout the nation who assisted
<br />in compiling the gaging-station data used in this study.
<br />
<br />Introduction 1
<br />
|