Evaluation of Methods for the Estimation of Tributary Mass Loading Rates

摘要

Tributary loading estimation methods were evaluated for application to a range of tributary types including highly event responsive systems. Evaluations were performed by conducting retrospective studies with comprehensive sets of field data for flow rates and total phosphorus concentrations. Four estimators from three broad categories were selected for study based on their performance in previous research. Estimators were evaluated using Monte Carlo sampling studies in which random subsamples of complete loading records were used to estimate annual mass loads. These estimates were then compared to 'true' loads determined by calculations using the entire record. Beale's Ratio Estimator consistently provided unbiased estimates, although stratification was necessary under event sampling. Differences in hydrograph characteristics played a strong role in the overall level of estimate error and in the amount of bias in averaging and regression estimators. Characteristics that are hypothetically responsible for the level of estimate error in different tributaries include frequency, magnitude and duration of high flow events. Daily flow simulation models to be used in sensitivity studies are described.