Date of Award


Degree Type


Degree Name

Master of Science


Environmental Engineering

Major Professor

James L. Smoot

Committee Members

R. Bruce Robinson, Bruce A. Tschantz


A long-term water quality program has been established at the Noland Divide Watershed (NDW) in the Great Smoky Mountains National Park. The Noland Divide Watershed is a spruce-fir forested catchment which has been shown in previous research to receive some of the largest fluxes of atmospherically-deposited nitrogen and sulfur compounds in the world. Streamwater chemistry data from November 1991 through August 1998 for two streams, the "northeast" (NE) and "southwest" (SW) stream, were examined to note results of this deposition on water quality. Automatic monitoring equipment on both streams measure and record pH, conductivity, and temperature readings every 15 minutes, and Stevens recorders in 3-foot H-flumes record stage height which corresponds to a flow rate every 15 minutes. In addition, grab samples were collected weekly and analyzed for pH, conductivity, acid neutralizing capacity (ANC), major anions and cations, aluminum, and silica. Experimental analysis was conducted to describe conditions, detect long-term and/or seasonal trends in water quality, and to relate water quality constituents with the watershed hydrology. In addition, parametric regression models were formed to note influence of several variables such as flow, time, seasonality, pH, and conductivity on analyte loads and concentrations and to test several sampling scenarios that may more-efficiently represent the water quality at NDW. It was determined from the analysis that high flow events are not well represented by the weekly grab samples and therefore water quality during these flow conditions is not fully understood. The SW stream is controlled more by groundwater inputs than is the NE stream, and the water quality characteristics of the two streams are statistically different (p< 0.05) with respect to all analyte concentrations except ammonium. Increased sulfate concentrations (+1.08 μeq/L in SW, +1.32 μeq/L in NE) were observed in the streams for each 1-inch increase of precipitation that occurs since the previous sampling visit. Decreased sulfate concentrations (-0.65 μeq/L in SW and -0.67 μeq/L in NE) were observed in the streams for each 1-day increase in consecutive dry days prior to sampling. Nitrate concentrations observed in the streams were not significantly influenced by precipitation prior to sampling, but decreased concentrations (-0.05 μeq/L in SW and -0.54 μeq/L in NE) were observed for each 1-day increase in consecutive dry days prior to sampling. Parametric regression models show that chloride, sodium, aluminum, and ammonium loads and concentrations are increasing over time, nitrate and silica loads and concentrations are decreasing over time, and sulfate, potassium, and hydrogen ion loads and concentrations are not changing over time. All analyte loads and concentrations except silica are significantly (p<0.10) influenced by seasonality. Parametric regression models also show that grab samples collected on a bi-weekly or tri-weekly frequency would be as statistically adequate for characterizing water quality concentrations and loads as are samples collected on a weekly basis.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Included in

Engineering Commons