ASSESSMENT OF HOUSING DENSITY IMPACTS ON GROUNDWATER QUALITY: INTEGRATION OF WATER QUALITY DATA INTO A GIS-BASED MODEL FOR ESTIMATING GROUNDWATER NITRATE CONCENTRATIONS

Groundwater is the sole source of drinking water in the Quonochontaug headland in Charlestown, Rhode Island. Management and preservation of this water resource and the surrounding coastal ponds is of increasing importance as population increases and seasonal homes are converted to year-round residences. The potential for saltwater intrusion in the fractured bedrock aquifer, and pollution of groundwater and the coastal salt ponds from anthropogenic sources of nitrate are addressed in this study. Water samples were collected from 47 private wells with a variety of depths, including shallow wells constructed in the stratified glacial material and deep wells drilled into the underlying fractured granitic bedrock. The sampling locations allow for analysis of spatial and depth related trends in water chemistry. This area has many seasonal residents, resulting in greater stress on the groundwater system in the summer. In order to evaluate temporal trends, samples were collected in three rounds; early summer, late summer, and fall. Chloride concentrations, an indicator of saltwater intrusion, indicate that there is no significant spatial distribution of elevated concentrations and no temporal trend in chloride concentrations observed. Although not widespread, some isolated incidents of saltwater intrusion (up to 7500 mg/L Cl^-) were found, indicating that the area is susceptible to saltwater intrusion, likely facilitated by transmission of brackish/saline water along fractures in the bedrock. Nitrate concentrations were above background levels (0-1 mg/L NO₃-N) in 92% of samples collected, and 27% of samples reached 5 mg/L, indicating widespread anthropogenic impact from septic systems, cesspools and fertilizers. The overall distribution of nitrate remained consistent throughout the three sampling rounds, indicating no temporal trend. Data analysis revealed a strong association between housing density and observed nitrate concentrations. This positive correlation was used to develop a GIS-based model for predicting nitrate concentrations throughout the salt pond region. The resulting model provides a management tool that can be used for targeting areas of concern and to better constrain the magnitude of groundwater-derived nitrogen loading to the nearby salt ponds.