FLOOD05191 - Laserfiche WebLink

Home Browse Search

LITERATURE REVIEW The first phase of the study was a major search of the literature to compile a bibliography of reports that describe urban runoff, primarily those concerning the magnitude and frequency of peak discharge. Shortly after the project began, it was learned that a similar lit- erature review was being made by the U . S. Department of Agriculture, Science and Education Administration (SEA); thereafter, the USGS and SEA worked together and published their reviews jointly (Rawls and others, 1980). That report contains synopses of 128 recent pub- lications on urban floodflow frequency that describe procedures ranging from simple statistical methods for estimating peak discharge and recurrence intervals, to procedures for estimating flood hydrographs, to sophis- ticated modeling procedures for estimating complete storm hydrographs. In the literature review, the follow- ing information is presented for each reference: I. Bibliographic citation. 2. Abstract, or synopsis, including a brief description of the procedure and data requirements for cali- brating and applying it. 3. General classification of the type of procedure. 4. Geographical location for which the procedure ap- pears applicable. In this review it was observed that many urban equations and models were derived for use in a specific geographical area. Although most of the models designed for f1ood-hydrograph and continuous-record synthesis could be applied regionally or nationally, statistical techniques for estimating the magnitude and frequency of instantaneous peak discharges are much more limited areally and generally cannot be transferred outside the specific area for which they were developed. Some of the widely applicable techniques described in the litera- ture review are highlighted in the following discussion. Leopold (1968) defined the ratio of the urban to equivalent rural mean annual flood for several metro- politan areas and graphically related this ratio to the percentage of the basin served by storm sewers and the percent of the basin covered by impervious surfaces. Sauer (1974) used the Leopold curves for mean annual floods, combined with a method suggested by Anderson (1970) to estimate peaks of any magnitude up to a 100- year event for urban sites in Oklahoma. Using local rainfall intensity data to define the slope of flood- frequency curves, Sauer estimate~ flood magnitude based on the mean annual flood for rural conditions adjusted by the Leopold ratio. A characteristic of the Sauer method is that the urban flood-frequency curve will always be greater than the rural curve for watersheds which do not have significant in-channel or detention storage. Espey and Winslow (1974) derived regression 2 Flood Characteristics of Urban Watersheds equations from data obtained for 60 urban watersheds located in cities along the East Coast and in Texas, Missis- sippi, Michigan. and Illinois. These regression equations relate flood peaks of various frequencies to drainage-area size, percent impervious area, channel slope, rainfall for 6-hour duration, and an index numerically describing the channel condition and the storm-sewer network. Harley (1978) proposed methods to evaluate the effects of urbanization on flood peaks. He concluded that with certain modifications, a combination of proce- dures described by Anderson (1970) and Carter (1961) offers a simplified and accurate approach to developing a nationally applicable technique. He proposed a regres- sion equation that included factors accounting for local runoff, imperviousness. drainage-area size, lagtime, and surface storage. Alternate procedures using modifi- cations of the proposed equation were employed to esti- mate either the ratio of urban to rural discharge or the difference between them. Harley tested his proposed methods on a small number of sites in a few cities and reported encouraging results. Among his recommenda- tions was the compilation of a large data base for use in testing and refining the proposed methods. The literature review supported the generally held concept that urbanizing a natural drainage basin usually causes runoff volume to increase and basin response time to decrease; it also found that peak discharges gen- erally increase for those watersheds which do not have significant in-channel or detention storage. These in- creases are usually most dramatic for low-order floods which occur frequently; they become less pronounced as flood magnitude increases. In a recent report not included in the literature review, Malcolm (1980) presents tbe results of modeling several basins in Charlotte, North Carolina. This report shows that temporary in-channel and detention storage can be highly effective in reducing peak discharges, and that much of this storage can be the result of uninten- tional in-channel storage behind undersized roadway culverts and bridges. The effect of such structures is sharply reduced at points further downstream, however, and when stream crossings are improved (enlarged), peak discharges increase. Malcolm's report nonetheless shows that because of in-channel and detention storage, urban peak discharges can be less than equivalent rural peaks in spite of other urbanization effects. In urbanizing a basin, naturally pervious surfaces are converted to impervious surfaces. Because infiltra- tion is reduced, such areas cause increased runoff; the usually smoother surface allows more rapid drainage; and depression storage usually is reduced. In addition, the drainage system is often altered by enlarging, straightening, and smoothing its channels and by install- ing storm sewers and curb-and-gutter systems. These alterations usually facilitate rapid runoff with a resultant