Areas contributing recharge to selected production wells in unconfined and confined glacial valley-fill aquifers in Chenango River Basin, New York

Abstract

In the Chenango River Basin of central New York, unconfined and confined glacial valley-fill aquifers are an important source of drinking-water supplies. The risk of contaminating water withdrawn by wells that tap these aquifers might be reduced if the areas contributing recharge to the wells are delineated and these areas protected from land uses that might affect the water quality. The U.S. Geological Survey, in cooperation with the New York State Department of Environmental Conservation and the New York State Department of Health, began an investigation in 2019 to improve understanding of groundwater flow and delineate areas contributing recharge to 16 production wells clustered in three study areas in the basin as part of an effort to protect the source of water to these wells. Areas contributing recharge were delineated on the basis of numerical steady-state groundwater-flow models representing long-term average hydrologic conditions. In the Cortland study area, four water suppliers operate 10 production wells that withdraw a total average rate of 2,480 gallons per minute from an unconfined aquifer consisting of well-sorted sand and gravel deposits. Simulated areas contributing recharge to these wells at their average pumping rates covered a total area of 6.93 square miles. Simulated areas contributing recharge extend upgradient from the wells to upland till deposits and to groundwater divides. Some simulated areas contributing recharge include isolated areas remote from the wells. Short simulated groundwater traveltimes from recharging locations to discharging wells indicated that the wells are vulnerable to contamination from land-surface activities; 50 percent of the traveltimes were 10 years or less. Land cover in some of the areas contributing recharge included a substantial amount of urban and agriculture land use. The groundwater-flow model of the Cortland study area was calibrated to available hydrologic data by inverse modeling using nonlinear regression. The parameter variance-covariance matrix from model calibration was used to create parameter sets that reflect the uncertainty of the parameter estimates and the correlation among parameters to evaluate the uncertainty associated with the single, predicted contributing areas to the wells. This analysis led to contributing areas expressed as a probability distribution. Because of the effects of parameter uncertainty, the size of the probabilistic contributing areas was larger than the size of the single, predicted contributing area for the wells. Thus, some areas not in the single, predicted contributing area might actually be in the contributing area, including additional areas of urban and agriculture land use that have the potential to contaminate groundwater. Additional areas that might be in the contributing area included recharge originating near the pumping wells that have relatively short groundwater-flow paths and traveltimes. In each of the Greene and Cincinnatus study areas, one water supplier operates three wells that are screened near the top of the bedrock surface in a confined aquifer consisting of poorly to well-sorted sand and gravel deposits. This confined aquifer is overlain by a lacustrine confining unit of very fine sand, silt, and clay, which in turn is overlain by a thin unconfined aquifer of sand and gravel. The groundwater-flow models for these two areas were manually calibrated because of the limited hydrologic data. Simulated areas contributing recharge to the Greene study area wells covered a total area of 0.35 square mile for the average pumping rate of 170 gallons per minute. The contributing areas extended southeastward of the wells to the groundwater divide in the till uplands. The contributing areas also included remote, isolated areas on the opposite side of the Chenango River from the wells primarily in the till uplands. For the Cincinnatus study area wells, which have a low average pumping rate (34 gallons per minute), the simulated contributing areas totaled 0.06 square mile and were on the same side of the river as the wells, but they are isolated areas remote from the wells primarily in the till-covered bedrock uplands. Land cover in these contributing areas for both study areas is primarily agriculture and forested, with the contributing areas to the Greene study area wells also including some urban land uses. Because the Greene and Cincinnatus study area wells are screened relatively deep and some flow paths to the wells partly travel through the confining unit, which impedes the connection with surface sources of recharge, overall groundwater traveltimes are greater than for wells in the Cortland study area. Fifty percent of Cortland study area wells, but only 9 and 44 percent of Greene and Cincinnatus study area wells, respectively, have groundwater traveltimes of 10 years or less.