thumbnail

Methods for estimating annual exceedance probability discharges for streams in Arkansas, based on data through water year 2013

Scientific Investigations Report 2016-5081
Prepared in cooperation with the Arkansas State Highway and Transportation Department and the U.S. Army Corps of Engineers, Little Rock District
By: , , and 
https://doi.org/10.3133/sir20165081

Links

Abstract

In 2013, the U.S. Geological Survey initiated a study to update regional skew, annual exceedance probability discharges, and regional regression equations used to estimate annual exceedance probability discharges for ungaged locations on streams in the study area with the use of recent geospatial data, new analytical methods, and available annual peak-discharge data through the 2013 water year. An analysis of regional skew using Bayesian weighted least-squares/Bayesian generalized-least squares regression was performed for Arkansas, Louisiana, and parts of Missouri and Oklahoma. The newly developed constant regional skew of -0.17 was used in the computation of annual exceedance probability discharges for 281 streamgages used in the regional regression analysis. Based on analysis of covariance, four flood regions were identified for use in the generation of regional regression models. Thirty-nine basin characteristics were considered as potential explanatory variables, and ordinary least-squares regression techniques were used to determine the optimum combinations of basin characteristics for each of the four regions. Basin characteristics in candidate models were evaluated based on multicollinearity with other basin characteristics (variance inflation factor < 2.5) and statistical significance at the 95-percent confidence level (p ≤ 0.05). Generalized least-squares regression was used to develop the final regression models for each flood region. Average standard errors of prediction of the generalized least-squares models ranged from 32.76 to 59.53 percent, with the largest range in flood region D. Pseudo coefficients of determination of the generalized least-squares models ranged from 90.29 to 97.28 percent, with the largest range also in flood region D. The regional regression equations apply only to locations on streams in Arkansas where annual peak discharges are not substantially affected by regulation, diversion, channelization, backwater, or urbanization. The applicability and accuracy of the regional regression equations depend on the basin characteristics measured for an ungaged location on a stream being within range of those used to develop the equations.

Suggested Citation

Wagner, D.M., Krieger, J.D., and Veilleux, A.G., 2016, Methods for estimating annual exceedance probability discharges for streams in Arkansas, based on data through water year 2013: U.S. Geological Survey Scientific Investigations Report 2016–5081, 136 p., http://dx.doi.org/10.3133/sir20165081.

ISSN: 2328-0328 (online)

Study Area

Table of Contents

  • Abstract
  • Introduction
  • Methods
  • Estimating Annual Exceedance Probability Discharges for Streams in Arkansas
  • Accuracy and Limitations of Regression Equations
  • Applications of Regional Regression Equations
  • Summary
  • References Cited
  • Appendixes 1-3
Publication type Report
Publication Subtype USGS Numbered Series
Title Methods for estimating annual exceedance probability discharges for streams in Arkansas, based on data through water year 2013
Series title Scientific Investigations Report
Series number 2016-5081
DOI 10.3133/sir20165081
Year Published 2016
Language English
Publisher U.S. Geological Survey
Publisher location Reston, VA
Contributing office(s) Lower Mississippi-Gulf Water Science Center
Description Report: ix, 136 p.; Appendix Tables
Country United States
State Arkansas
Online Only (Y/N) Y
Additional Online Files (Y/N) Y
Google Analytic Metrics Metrics page
Additional publication details