An evaluation of water-quality records for Texas streams

Abstract

Chemical data for 12 streamflow-sampling stations in Texas, each having from 8 to 24 years of available historical records, were analyzed to determine functional relationships between concentrations of the major inorganic constituents and specific conductance or stream discharge. Regression equations for each station were derived by using data for an identical 5-year period; the remaining record served as independent test data for checking the adequacy of the functions on the basis of continued specific-conductance determinations as an index variable. Daily solute concentrations and loads at a sampling site were simulated, and monthly and annual averages were computed by using daily records of specific conductance in conjunction with the regression equations. Simulated annual mean concentrations of most individual solutes and dissolved solids averaged within 10-20 percent of comparable concentrations determined from the composited chemical analyses, although systematic biases were detected in results for some of the relationships used in the simulation. Results of the evaluation of the 12 stations indicated that regression equations, using specific conductance as the independent variable, can be used as an alternative to composite analyses for calculation of solute concentrations and loads in Texas streams and is a flexible system that can be applied with sufficient accuracy to satisfy most anticipated needs for data.