Evaluation of River Load Estimation Methods for Total Phosphorus

Abstract Accurate estimates of pollutant loadings to the Great Lakes are required for trend detection, model development, and planning. On many major rivers, infrequent sampling of most pollutants makes these estimates difficult. However, most large rivers have complete daily flow records available from the U.S. Geological Survey. A number of different estimation methods have been developed to use these data in conjunction with available concentration data to calculate pollutant loadings. Several of these methods have been evaluated in this paper for a typical major tributary. Intensive measurements of pollutant loadings to Lake Michigan from the Grand River in 1976–1977 were conducted by USEPA, Region V. Daily concentrations of several chemical parameters, including total phosphorus, were determined. This information, together with daily river flow rates from USGS, provide an excellent data set for testing various load estimation methods. Ten different methods were tested using 680 randomly generated 25-sample subsets of the daily total phosphorus data set. The methods tested included several simple calculation schemes, as well as flowconcentration correlations and the stratified ratio estimator. The latter method is shown to be at least slightly better than the others for these data and for this type of sampling, including when sampling is concentrated during periods of high flow.