FLOOD CHARACTERISTICS OF URBAN WATERSHEDS IN THE UNITED STATES by V.B. Sauer, W.O. Thomas, Jr., V.A. Stricker, and K.V. Wilson Prepared by the U.S. Geological Survey in cooperation with the U.S. Department of Transportation, Federal Highway Administration 1983 Water-Supply Paper 2207 TABLE OF CONTENTS Glossary Abstract Introduction Literature review Data base     Topographic and climatic variables     Land-use variables     Indices of urbanization     Flood-frequency estimates Estimating procedures for ungaged urban sites     Selection of data     Seven-parameter estimating equations     Three-parameter estimating equations     Seven-parameter alternate estimating equations     Correlation of significant variables     Limitations of significant variables     Other independent variables     Other methods and models         Ratio method         Difference method         Method of moments         Harley method     Verification and testing of regression equations         Split-sample analysis         Bias testing         Sensitivity testing     Urban peaks less than equivalent rural peaks     Effects of detention storage     Estimating basin lagtime Estimating procedures for gaged sites Summary References Appendix I-Selected data for stations used in nationwide urban flood-frequency study Appendix II - List of reports for estimating equivalent rural discharges for urban watersheds FIGURES 1. Map showing location of metropolitan areas included in the nationwide urban flood-frequency study 2. Schematic of typical drainage basin shapes and subdivision into thirds 3-28. Graphs showing: 3. Comparison of observed 2-year urban peak discharge to peak discharge estimated from equation 3 4. Comparison of observed 10-year urban peak discharge to peak discharge estimated from equation 5 5. Comparison of observed 100-year urban peak discharge to peak discharge estimated from equation 8 6. Comparison of observed 2-year urban peak discharge to peak discharge estimated from equation 10 7. Comparison of observed 10-year urban peak discharge to peak discharge estimated from equation 12 8. Comparison of observed 100-year urban peak discharge to peak discharge estimated from equation 15 9. Comparison of observed 2-year urban peak discharge to peak discharge estimated from equation 17 10. Comparison of observed 10-year urban peak discharge to peak discharge estimated from equation 19 11. Comparison of observed 100-year urban peak discharge to peak discharge estimated from equation 22 12. Comparison of equivalent 2-year rural peak discharge to peak discharge estimated from equation 24 13. Comparison of equivalent 10-year rural peak discharge to peak discharge estimated from equation 25 14. Comparison of equivalent 100-year rural peak discharge to peak discharge estimated from equation 26 15. Relation of urban/rural 2-year peak-flow ratio (UQ2/RQ2) to basin development factor and impervious area 16. Relation of urban/rural 10-year peak-flow ratio (UQ2/RQ10) to basin development factor and impervious area 17. Relation of urban/rural 100-year peak-flow ratio (UQ2/RQ100) to basin development factor and impervious area 18. Comparison of urban/rural 2-year peak-flow ratio (UQ2/RQ2) to Leopold (1968) curves 19. Relation of full-development urban/rural ratio (UQ2/RQ2) to drainage area size 20. Relation of full-development urban/rural ratio (UQ2/RQ2) to channel slope 21. Relation of full-development urban/rural ratio (UQ2/RQ2) to rainfall intensity 22. Relation of full-development urban/rural ratio (UQ2/RQ2) to storage 23. Relation of full-development urban/rural ratio (UQ2/RQ2) to equivalent rural discharge 24. Average relations of urban/rural ratios to basin development factor for 2-year and 100-year floods 25. Comparison of estimated 2-year urban peak discharge to 2-year equivalent rural peak discharge 26. Comparison of estimated 10-year urban peak discharge to 10-year equivalent rural peak discharge 27. Comparison of estimated 100-year urban peak discharge to 100-year equivalent rural peak discharge 28. Average relations between urban peak discharges estimated by seven-parameter equations and observed urban peak discharges affected by temporary detention storage TABLES 1. Metropolitan areas included in nationwide urban flood-frequency study 2. Correlation matrix for significant variables 3. Comparison of standard error or regression and average standard error of prediction 4. Sensitivity of 100-year computed urban peak discharge to errors in independent variables 5. Sensitivity of 100-year computed urban peak discharge to errors in the basin development factor, BDF 6. Compound error resulting from interrelation of drainage area size and 100-year rural peak discharge ABSTRACT A nationwide study of flood magnitude and frequency in urban areas was made for the purpose of reviewing available literature, compiling an urban flood data base, and developing methods of estimating urban floodflow characteristics in ungaged areas. The literature review contains synopses of 128 recent publications related to urban floodflow. A data base of 269 gaged basins in 56 cities and 31 states, including Hawaii, contains a wide variety of topographic and climatic characteristics, land-use variables, indices of urbanization, and flood-frequency estimates. Three sets of regression equations were developed to estimate flood discharges for ungaged sites for recurrence intervals of 2, 5, 10, 25, 50, 100, and 500 years. Two sets of regression equations are based on seven independent parameters and the third is based on three independent parameters. The only difference in the two sets of seven-parameter equations is the use of basin lag time in one and lake and reservoir storage in the other. Of primary importance in these equations is an independent estimate of the equivalent rural discharge for the ungaged basin. The equations adjust the equivalent rural discharge to an urban condition. The primary adjustment factor, or index of urbanization, is the basin development factor, a measure of the extent of development of the drainage system in the basin. This measure includes evaluations of storm drains (sewers), channel improvements, and curb-and-gutter streets. The basin development factor is statistically very significant and offers a simple and effective way of accounting for drainage development and runoff response in urban areas. Percentage of impervious area is also included in the seven-parameter equations as an additional measure of urbanization and apparently accounts for increased runoff volumes. This factor is not high significant for large floods, which supports the generally held concept that imperviousness is not a dominant factor when soils become more saturated during large storms. Other parameters in the seven-parameter equations include drainage-area size, channel slope, rainfall intensity, lake and reservoir storage, and basin lag time. These factors are all statistically significant and provide logical indices of basin conditions. The three-parameter equations include only the three most significant parameters: rural discharge, basin-development factor, and drainage-area size. All three sets of regression equations provide unbiased estimates of urban flood frequency. The seven-parameter regression equations without basin lag time have average standard errors of regression varying from plus/minus 37 percent for the 5-year flood to plus/minus 44 percent for the 100-year flood and plus/minus 49 percent for the 500-year flood. The other two sets of regression equations have similar accuracy. Several tests for bias, sensitivity, and hydrologic consistency are include which support the conclusion that the equations are useful throughout the United States. All estimating equations were developed from data collected on drainage basins where temporary in-channel storage, due to highway embankments, was not significant. Consequently, estimates made with these equations do not account for the reducing effect of this temporary detention storage. LizardTech's Djvu plug-in is needed to view these reports. CLICK HERE TO VIEW THE ENTIRE REPORT For questions or comments, contact K. Van Wilson.