Item talk:Q68989

Add topic
Revision as of 02:09, 30 July 2023 by Sky (talk | contribs) (Added abstract and other texts to publication item's discussion page for reference)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Hydrogeology and simulation of source areas of water to production wells in a colluvium-mantled carbonate-bedrock aquifer near Shippensburg, Cumberland and Franklin Counties, Pennsylvania

This report presents the results of a study by the U.S. Geological Survey in cooperation with the Shippensburg Borough Authority to evaluate the source areas of water to production wells in a colluvium-mantled carbonate-bedrock aquifer in Cumberland and Franklin Counties, Pa. The areal extent of the zone of contribution was simulated for three production wells near Shippensburg, Pa. by use of a ground-water-flow model. A 111-square-mile area was selected as the model area and includes areas of the South Mountain Section and the Great Valley Section of the Valley and Ridge Physiographic Province. Within the model area, the geologic units in the South Mountain area are predominantly metamorphic rocks and the geologic units in the Great Valley are predominantly carbonate rocks. Hydrologic and geologic information were compiled to establish a conceptual model of ground-water flow. Characteristics of aquifer materials were determined, and streamflow and water levels were measured. Streamflow measurements in November 2003 showed all streams lost water as they flowed from South Mountain over the colluvium-mantled carbonate aquifer into the Great Valley. Some streams lost more than 1 cubic foot per second to the aquifer in this area. The Shippensburg Borough Authority owns three production wells in the model area. Two wells, Cu 969 and Fr 823, are currently (2004) used as production wells and produce 500,000 and 800,000 gallons per day, respectively. Well Cu 970 is intended to be brought on line as a production well in the future. Water levels were measured in 43 wells to use for model calibration. Water-level fluctuations and geophysical logs indicated confined conditions in well Cu 970. Ground-water flow was simulated with a model that consisted of two vertical layers, with five zones in each layer. The units were hydrostratigraphic units that initially were based on geologic formations, but boundaries were adjusted during model calibration. Model calibration resulted in a root mean square error of 9.8 feet. A parameter-estimation package was used during model calibration to estimate three parameters. The parameter estimation resulted in a value of 233 feet per day for horizontal hydraulic conductivity of the highly fractured carbonate rocks and sandy colluvium in layer 1; 3.97 feet per day for horizontal hydraulic conductivity of the ridge-forming unit in layer 1; and a value of 1.73 for horizontal anisotropy in both layers. The calibrated model was used to delineate the areal extent of the zone of contribution for wells Cu 969 and Fr 823. Although well Cu 970 is not currently (2004) being used, the areal extent of its zone of contribution also was simulated without additional model calibration. The shape of the areal extent of the zone of contribution was similar for each well and included an area that extended from the well southwest along the Tomstown Formation, and then extended southeast into the metamorphic rocks of South Mountain. The contributing areas from the watersheds of losing streams were also delineated because losing stream reaches bisect the areal extent of the zones of contribution. Spatial uncertainty of the areal extent of the zone of contribution was illustrated using a Monte-Carlo analysis. The model was run 1,000 times using randomly generated parameter sets that were normally distributed within the confidence interval around the optimal values for the three estimated parameters. The model converged and had a reasonable water budget for 980 of the model runs. For each of those 980 model runs, the recharge area was determined, and the results for all runs were compiled and contoured. The results of the Monte-Carlo analysis were compared to the results of the deterministic model, illustrating that the deterministic model has the greatest certainty in the area closest to each well in the Tomstown Formation. The areas farther from the well, upgradient, and in the metamorphic rocks have a higher degree

Return to "Q68989" page.