thumbnail

Comparison of methods for computing streamflow statistics for Pennsylvania streams

Water-Resources Investigations Report 99-4068
Prepared in cooperation with the Pennsylvania Department of Transportation
By:  and 
https://doi.org/10.3133/wri994068

Links

Abstract

Methods for computing streamflow statistics intended for use on ungaged locations on Pennsylvania streams are presented and compared to frequency distributions of gaged streamflow data. The streamflow statistics used in the comparisons include the 7-day 10-year low flow, 50-year flood flow, and the 100-year flood flow; additional statistics are presented. Streamflow statistics for gaged locations on streams in Pennsylvania were computed using three methods for the comparisons: 1) Log-Pearson type III frequency distribution (Log-Pearson) of continuous-record streamflow data, 2) regional regression equations developed by the U.S. Geological Survey in 1982 (WRI 82-21), and 3) regional regression equations developed by the Pennsylvania State University in 1981 (PSU-IV). Log-Pearson distribution was considered the reference method for evaluation of the regional regression equations. Low-flow statistics were computed using the Log-Pearson distribution and WRI 82-21, whereas flood-flow statistics were computed using all three methods. The urban adjustment for PSU-IV was modified from the recommended computation to exclude Philadelphia and the surrounding areas (region 1) from the adjustment. Adjustments for storage area for PSU-IV were also slightly modified.

A comparison of the 7-day 10-year low flow computed from Log-Pearson distribution and WRI-82- 21 showed that the methods produced significantly different values for about 7 percent of the state. The same methods produced 50-year and 100-year flood flows that were significantly different for about 24 percent of the state. Flood-flow statistics computed using Log-Pearson distribution and PSU-IV were not significantly different in any regions of the state. These findings are based on a statistical comparison using the t-test on signed ranks and graphical methods.

Suggested Citation

Ehlke, M.H., Reed, L.A., 1999, Comparison of methods for computing streamflow statistics for Pennsylvania streams: U.S. Geological Survey Water-Resources Investigations Report 1998-4068, 80 p., https://pubs.er.usgs.gov/publication/wri994068.

Table of Contents

  • Abstract
  • Introduction
  • Methods for computing streamflow statistics
  • Comparison of Log-Pearson distribution to regional regression equations for gaged locations on Pennsylvania streams.
  • Summary and conclusions
  • References cited
  • Appendix 3. 7-day 10-year low-flow statistic (Q7,10) computed from Log-Pearson distribution of streamflow data and WRI 82-21 regional regression equations for gaged locations on streams in Pennsylvania unaffected by carbonate bedrock, extensive mining, or regulation
  • Appendix 4. Comparison of streamflow statistics computed using Log-Pearson distribution and regression equations
  • Appendix 5. Flood-flow statistics computed from Log-Pearson distribution of streamflow data and WRI 82-21 regional regression equations for gaged locations on streams in Pennsylvania
  • Appendix 6. Flood-flow statistics computed from Log-Pearson distribution of streamflow data and PSU-IV regional regression equations for gaged locations on streams in Pennsylvania
Publication type Report
Publication Subtype USGS Numbered Series
Title Comparison of methods for computing streamflow statistics for Pennsylvania streams
Series title Water-Resources Investigations Report
Series number 99-4068
DOI 10.3133/wri994068
Year Published 1999
Language English
Publisher U.S. Geological Survey
Publisher location Reston, VA
Contributing office(s) Pennsylvania Water Science Center
Description vi, 80 p. :ill., col. maps ;28 cm.
Google Analytic Metrics Metrics page
Additional publication details