thumbnail

Verification of regression equations for estimating flood magnitudes for selected frequencies on small natural streams in Georgia

Water-Resources Investigations Report 86-4337
By:  and 
https://doi.org/10.3133/wri864337

Links

Abstract

In 1976 the U.S. Geological Survey, in cooperation with the Georgia Department of Transportation, began a program to monitor small natural streams in Georgia to verify the accuracy of the flood frequency estimating equations for the five flood frequency regions that were published in a previous study. Data collection consisted of obtaining an additional 10 yr of annual peak flow records at 24 gaging stations and establishing and collecting annual peak flow records at 15 additional gaging sites in areas of the State where data were unavailable. Data also were collected for an additional 10 yr at four gaging stations that were converted to continuous record gaging stations in 1976. The flood frequency equations were verified by comparing the observed and regression equation estimated discharges for the 2-, 25-, and 100-yr floods: (1) for the 28 gaging stations continued an additional 10 yr; (2) for the 15 gaging stations that have about 10 yr of record where data were unavailable; and (3) for all gaging stations on drainage areas of < 50 sq mi for which data were available in all five flood frequency regions. The rainfall-runoff model simulated discharges from the previous study also were verified by comparisons of the observed and the rainfall-runoff model simulated discharges for the 2-, 25-, and 100-yr floods for gaging stations calibrated in the previous study. These comparisons, based on student 's t-test statistics at the 0.05 level of significance, indicated that all the flood frequency equations computed in the previous study are valid and unbiased except for Regions 2 and 3, where the equations appear to be slightly biased. The comparison of the discharges simulated by the rainfall-runoff model with the observed discharges for the gaging stations that have 20 yr of record were unbiased, but simulated discharges for stations having 10 yr of record were biased, probably because of ' loss of variance ' in the averaging procedures of the rainfall-runoff model and the short length of record. The flood-frequency estimating equations computed in the previous study have been verified and are considered to be valid for natural streams in Georgia that have drainage areas of 0.1 to 20 sq mi. (Author 's abstract)

Suggested Citation

Price, M., and Hess, G.W., 1986, Verification of regression equations for estimating flood magnitudes for selected frequencies on small natural streams in Georgia: U.S. Geological Survey Water-Resources Investigations Report 86-4337, Report: vi, 39 p.; 1 Plate: 19.27 x 23.55 inches, https://doi.org/10.3133/wri864337.

Study Area

Publication type Report
Publication Subtype USGS Numbered Series
Title Verification of regression equations for estimating flood magnitudes for selected frequencies on small natural streams in Georgia
Series title Water-Resources Investigations Report
Series number 86-4337
DOI 10.3133/wri864337
Year Published 1986
Language English
Publisher U.S. Geological Survey
Publisher location Reston, VA
Contributing office(s) South Atlantic Water Science Center
Description Report: vi, 39 p.; 1 Plate: 19.27 x 23.55 inches
Country United States
State Georgia
Additional publication details