Regional analysis of streamflow chemical quality in Texas

Publisher Website
Google Scholar
Abstract
Thirty‐one streamflow water quality stations in Texas, with long‐term historical records ranging from 7 to 28 years, were selected for statistical analyses of the annual mean concentrations of the major chemical constituents. A statewide multiple regression analysis using frequency characteristics of annual mean water quality constituents as dependent variables and selected basin characteristics as independent variables was investigated as a means of regionalizing information on annual water quality characteristics. For the two different regression models investigated, a single independent variable, generally stream discharge or average basin rainfall, explained a significant part of the variance of the dependent water quality variables. Areal analysis of residuals of the statewide regression models indicated two distinct regional patterns. Correlation coefficients for the concentration‐rainfall regional models ranged from 0.80 to 0.97 and the standard errors of estimate were between 26 and 99% of the means of the dependent variables. An analysis of various concentration‐conductance relationships for the statewide data also gave favorable results, and provided an alternate method for obtaining estimates of streamflow chemical quality characteristics.