Date of Award
Master of Science in Civil and Environmental Engineering
Civil and Environmental Engineering
Wet weather has a significant influence on the water quality in the Narragansett Bay. While dry weather contaminant loads have been determined for a large number of watersheds draining to the Bay, information on wet weather loads is not as widely available. Prediction of wet loads from easily obtainable watershed and rainfall characteristics would therefore be desirable. Wet weather water quality data from nine major watersheds in Rhode Island was collected and processed, including total suspended solids (TSS), chloride, sodium, ammonia, nitrate, orthophosphate, and six heavy metals. The wet loads of these constituents were used as response variables in a multiple linearized regression analysis with watershed characteristics (drainage area, land use, and CN curve number) and rainfall characteristics (antecedent dry period, total rainfall, average rainfall, peak rainfall) as predictor variables. The best regression equation to predict wet volume (R2 = 0.788, C.V. = 11.1) contains drainage area and total rainfall as predictor variables. Conservative constituent wet loads are predicted better than solids and nitrogen and phosphorus compounds. The equations generally include drainage area, total rainfall, CN, and percentage of forested lands as predictor variables. For the modeling of heavy metal wet loads further predictor variables seem to be necessary, as patterns are found in the residual analysis. The overall uncertainty in the regression equations is high. Additional predictor variables, data classification and model expansions are recommended to model heavy metals and increase the predictive quality of regression equations for common constituents.
Haefner, Christian, "A Regression Analysis of Wet Loads in Rhode Island Watersheds" (2004). Open Access Master's Theses. Paper 2074.