Submitted to: Transactions of the ASAE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: July 1, 2002
Publication Date: December 1, 2002
Citation: SHERIDAN, J.M. PEAK FLOW ESTIMATES FOR COASTAL PLAIN WATERSHEDS. TRANSACTIONS OF THE AMERICAN SOCIETY OF AGRICULTURAL ENGINEERS. 45(5):1329-1326. 2002. Interpretive Summary: Computer models are increasingly being used to compute storm peak flows and to estimate the quantities of pollutants carried in storm runoff from agricultural areas. For increased confidence in the reliability of predictions made by these computer models, the adequacy of the equations used in these models should be tested by comparing model predictions to experimental field data when such data are available. In this study, an equation used by a number of current computer models to estimate storm peak flows was tested for applicability in the Coastal Plain region of the southeastern US. Peak flows computed using the equation were compared to storm peaks measured from 58 storms occurring on seven gaged, experimental areas in the Coastal Plain. The equation was found to over predict storm peaks from the study areas by an average of over 400%. Improved peak flow equations were developed for use in the southeastern Coastal Plain region. These equations will provide more accurate estimates of storm peaks and, a a result, will permit development of computer models that make more accurate estimates of pollutant movement in storm runoff from agricultural areas in the Coastal Plain. The improved capability for modeling storm runoff and pollutant movement is needed for natural resource and environmental planning and management that is required to improve the quality of the nation's streams, rivers, lakes and groundwater supplies.
Technical Abstract: Computer models are increasingly being used to predict hydrologic and water quality responses from ungaged watersheds as required to resolve critical water resource and quality issues and meet the goals of the Clean Water Act. However, computer models often rely on empirical algorithms to represent complex processes that are not well-understood or are not easily represented. For greater confidence in model predictive capabilities, empirical algorithms should be tested before applying in regions where conditions vary from those conditions for which the algorithms were developed. An empirical equation developed based on data from the midwestern US has been used in a number of water resource and quality models for estimating storm event peak flows. Limited testing of this equation has shown a tendency to overpredict storm peaks in regions of low topographic relief. In this paper, the equation was evaluated using 58 storm events on seven experimental watersheds in the southeastern US Coastal Plain, a region characterized by low-gradient drainage networks and extensive riparian storage. Results indicate that the equation overpredicts peak flows on Coastal Plain watersheds, with an average error of over 400%. Improved regional peak flow equations were developed relating to peak flows to drainage area, runoff amount and the Melton ratio, a drainage network parameter indicating relative stream density. The Coastal Plain peak flow equations provide improved capabilities for estimating storm peaks and pollutant transport from ungaged watersheds in the southeastern Coastal Plain, which is needed for accurate, reliable natural resource and environmental quality modeling.