FLOOD CHARACTERISTICS OF MISSISSIPPI STREAMS by M.N. Landers and K.V. Wilson, Jr. Prepared by the U.S. Geological Survey in cooperation with the Mississippi State Highway Department 1991 WATER-RESOURCES INVESTIGATIONS REPORT 91-4037 TABLE OF CONTENTS Abstract Introduction     Purpose and scope     General description of study area     Acknowlegments Records of flooding Statistical characteristics of annual peak flow Flood-frequency estimates from streamflow records     Weighted flood-frequency estimates     Ungaged sites on gaged streams     Accuracy of flood-frequency estimates for gaged stream sites     Historical record evaluation     Flood-frequency of the Pearl River main stem Regional flood-frequency estimates for rural streams     Generalized-least squares regression     Explanatory variables     Regional boundaries     Selection of the appropriate flood-frequency equation         Delta         GT800         East         West Regional flood-frequency estimates for urbanized streams Limitations of regional flood-frequency estimates     Rural streams     Urbanized streams Summary References Appendix--Regional skew coefficients FIGURES 1. Map showing location of streamflow-gaging stations and flood-frequency regions 2-14.--Graphs showing: 2. Distribution of systematic peak-flow record lengths 3. Relation of basin shape coefficient to drainage area for the Pearl River main stem 4. Relation of flood-frequency discharge to drainage area for the Pearl River main stem 5. Relation of flood-frequency discharge for an average-shaped basin to drainage area for the Pearl River main stem 6. Relation of the 10-year discharge to basin characteristics 7. Characteristics of the 10-year flood residuals for the drainage basin subgroups and the whole-sample group of sites outside the Delta region 8. Characteristics of the 10-year flood residuals for the whole-sample (statewide) group and for the Delta region 9. Distribution of drainage area for 282 gaging stations outside the Delta region 10. Characteristics of the 10-year flood residuals for drainage area subgroups and the whole-sample group of sites outside the Delta region 11. Characteristics of the 10-year flood residuals for drainage area subgroups and the whole-sample group of sites outside the Delta region having drainage areas less than or equal to 800 square miles 12. Characteristics of the 10-year flood residuals for the whole- sample group of sites outside the Delta region and for the principal subgroup regions 13. Relation of observed 2-year, 10-year, and 100-year urban peak discharge to peak discharge estimated from equations 39, 41, and 44 14. Skew coefficient characteristics and boundaries of homogeneous skew regions in Mississippi 15. Map showing weighted-grid unbiased regional skew of log- transformed annual peak flow 16. Graph showing error of regional skew from weighted-grid and Interagency Advisory Committee on Water Data (IACWD) contour maps, where error of weighted-grid map is measured as the means sum of squared prediction errors (MPRESS) and error of IACWD map is measured as mean square error (MSE) TABLES 1. Flood quantiles, standard error of estimate, and other information for gaging stations 2. Standard error of prediction for each flood-frequency regression equation 3. Additional basin characteristics for urban streams in Mississippi 4. Errors of prediction using the seven-parameter nationwide equations for urbanized streams in Mississippi 5. Characteristics of explanatory variables used in regression calculations for basins in the East and West regions with areas less than or equal to 800 square miles, basins in the Delta, and basins in the East or West regions with areas greater than 800 square miles (GT800) 6. Gaging stations used in skew coefficient analysis ABSTRACT Flood magnitudes for selected recurrence intervals from 2 to 500 years were determined for 330 gaged sites in the study area where annual peak- flow records have been collected. The principal study are is Mississippi; however, selected data collected in adjoining States on streams draining into or from Mississippi are also included. Flood frequency at a gaged stream site is defined by fitting the Pearson Type III probability dis- tribution to the log-transformed annual peaks. The accuracy of the flood frequency determined for a gaged site is determined primarily by the number of years of annual peak-flow record (the sample size). Greater accuracy is achieved in the current analysis than in previous analyses because of the additional years of annual peak-flow record. Flood-fre- quency and basin characteristics at gaged sites were used to developed regression equations for estimating flood frequency where annual peak-flow records are not available. Flood frequency for ungaged stream sites in Mississippi may be estimated using basin characteristics in regression equations. Regression equations were computed using the generalized-least-squares procedure rather than the ordinary-least-squares procedure used in previous regional hydrologic analyses. The generalized-least-squares procedure considers the variable error of the gaging station flood frequencies and corrects for the cross- correlation of concurrent annual peaks. When the gaging stations in the sample for regression analysis have widely varying record lengths and concurrent peak flows, which are correlated between sites, the generalized- least-squares procedure provides more accurate estimates of the regression coefficients and model error than does the ordinary-least-squares procedure. These flood-frequency equations provide managers with improved tools for estimating flood frequencies for purposes of management and design. 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.