| Nitrate contamination in shallow ground-water systems is a significant resource-management concern where such systems are (1) tapped extensively by potable water-supply wells and (2) incised by baseflow-dominated streams. The "lag" time required for nitrate to flush out of such systems, however, is often not well understood. In this thesis, a three-dimensional, steady-state, ground-water-flow model and particle-tracking analyses were used to assess implications for nitrate transport in the watershed of Phillips Branch, an Atlantic Coastal Plain stream in eastern Sussex County, Delaware (USA). The particle-tracking analyses assume nitrate is a conservative contaminant at the study site (i.e., advection is the main transport mechanism, diffusion and dispersion are ignored, and nitrate is neither produced nor consumed in the aquifer). This assumption is justified due to predominantly oxic ground-water conditions and a relatively non-reactive aquifer matrix (mostly quartz sand).; The MODFLOW-based model was calibrated with respect to average hydraulic head and stream discharge under baseflow conditions, and further evaluated with respect to ground-water age determined by chlorofluorocarbon (CFC) dating. Horizontal and vertical hydraulic conductivities of 40 and 4 meters per day (m/day), respectively, and a Phillips Branch drain conductance per unit length of 2 m/day, yielded the best calibration. Specifically, the root-mean-squared error (RMSE) and the normalized RMSE (NRMSE) between simulated head and head observed on July 19, 1989 were 0.127 m and 1.81%, respectively, and the absolute difference between simulated and observed baseflow was 0.8%. The RMSE and NRMSE between CFC-modeled age and flow-modeled age were 5.2 yrs and 20.2%, respectively. Based on an extensive analysis, the model was found to be more sensitive to recharge than to hydraulic conductivity.; Results of particle-tracking analyses indicate the source area for ground water discharging to the stream is markedly different from its surface watershed counterpart in terms of both area and appearance. The source area (5.51 km 2), which has the shape of an elongated ellipse that is warped due to regional ground-water flow, is 20% smaller than the surface watershed (6.85 km2). Areas where the surface watershed extends beyond the source area suggest that large quantities of recharge (approximately 2,296 m3/day) discharge to boundaries other than Phillips Branch. Ground-water residence time in the source area ranges from a few days to almost 100 yrs; however, residence time is highly skewed. The mean and median residence times are 15 and 10 yrs, respectively, while 95% are less than 50 yrs. Ground water discharging to Phillips Branch becomes older and more diverse in age with distance downstream; this trend indicates the degree of mixing along the stream channel. Particle-tracking data for well clusters with CFC-modeled ages imply that small vertical differences in screen midpoints (approximately 5.5 to 6 m) translate to very large (>100 m) horizontal differences in recharge location. The particle data further indicate that ground water entering wells screened near the base of the surficial aquifer was recharged as far as 1-4 km from the wellhead.; Simulation of hypothetical wastewater disposal systems within the stream's source area indicates that such systems create artificial flow boundaries, which can have considerable effects on the ground water-surface water system and, consequently, contaminant transport. Increased recharge from wastewater disposal causes (1) increased contaminant spreading both normal and parallel to regional ground-water flow, (2) increased contaminant residence time, (3) increased stream discharge, and (4) decreased stream source area. With regard to (4), wastewater recharge essentially displaces ground water, which would otherwise discharge to Phillips Branch under pre-disposal conditions. Under the highest disposal scenario evaluated (1 million... |
PDF Full Text Request