Simple concentration-flow regressions may be inadequate as predictive models for water quality because of the generally small amount of variance explained by them. A more complex model, based on the separation of river flow into runoff, baseflow and effluent components, was used to describe the behaviour of chloride, orthophosphate, total hardness and alkalinity at a site in the lower reaches of the River Wye and explained between 42% (total hardness) and 83% (orthophosphate) of the variance of calibration data. Comparison of concentrations generated by the model with observations outside the calibration period indicated that between 59% (chloride) and 88% (orthophosphate) of the variance of such test data was explained. © 1981.
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below