Ground-water budgets were developed for 32 small basin-based zones in the Greenwich area of southwestern Connecticut, where crystalline-bedrock aquifers supply private wells, to determine the status of residential ground-water consumption relative to rates of ground-water recharge and discharge. Estimated residential ground-water withdrawals for small basins (averaging 1.7 square miles (mi²)) ranged from 0 to 0.16 million gallons per day per square mile (Mgal/d/mi²). To develop these budgets, residential ground-water withdrawals were estimated using multiple-linear regression models that relate water use from public water supply to data on residential property characteristics. Average daily water use of households with public water supply ranged from 219 to 1,082 gallons per day (gal/d).

A steady-state finite-difference ground-water- flow model was developed to track water budgets, and to estimate optimal values for hydraulic conductivity of the bedrock (0.05 feet per day) and recharge to the overlying till deposits (6.9 inches) using nonlinear regression. Estimated recharge rates to the small basins ranged from 3.6 to 7.5 inches per year (in/yr) and relate to the percentage of the basin underlain by coarse- grained glacial stratified deposits. Recharge was not applied to impervious areas to account for the effects of urbanization. Net residential ground-water consumption was estimated as ground-water withdrawals increased during the growing season, and ranged from 0 to 0.9 in/yr.

Long-term average stream base flows simulated by the ground-water-flow model were compared to calculated values of average base flow and low flow to determine if base flow was substantially reduced in any of the basins studied. Three of the 32 basins studied had simulated base flows less than 3 in/yr, as a result of either ground-water withdrawals or reduced recharge due to urbanization. A water-availability criteria of the difference between the 30-day 2-year low flow and the recharge rate for each basin was explored as a method to rate the status of water consumption in each basin. Water consumption ranged from 0 to 14.3 percent of available water based on this criteria for the 32 basins studied.

Base-flow water quality was related to the amount of urbanized area in each basin sampled. Concentrations of total nitrogen and phosphorus, chloride, indicator bacteria, and the number of pesticide detections increased with basin urbanization, which ranged from 18 to 63 percent of basin area.