Skip Links

USGS - science for a changing world

Scientific Investigations Report 2011–5227

Prepared in cooperation with the Maine Geological Survey

Simulation of Groundwater Conditions and Streamflow Depletion to Evaluate Water Availability in a Freeport, Maine, Watershed

By Martha G. Nielsen and Daniel B. Locke

Thumbnail of and link to report PDF (7.69 MB)Abstract

In order to evaluate water availability in the State of Maine, the U.S. Geological Survey (USGS) and the Maine Geological Survey began a cooperative investigation to provide the first rigorous evaluation of watersheds deemed “at risk” because of the combination of instream flow requirements and proportionally large water withdrawals. The study area for this investigation includes the Harvey and Merrill Brook watersheds and the Freeport aquifer in the towns of Freeport, Pownal, and Yarmouth, Maine. A numerical groundwater- flow model was used to evaluate groundwater withdrawals, groundwater-surface-water interactions, and the effect of water-management practices on streamflow. The water budget illustrates the effect that groundwater withdrawals have on streamflow and the movement of water within the system.

Streamflow measurements were made following standard USGS techniques, from May through September 2009 at one site in the Merrill Brook watershed and four sites in the Harvey Brook watershed. A record-extension technique was applied to estimate long-term monthly streamflows at each of the five sites.

The conceptual model of the groundwater system consists of a deep, confined aquifer (the Freeport aquifer) in a buried valley that trends through the middle of the study area, covered by a discontinuous confining unit, and topped by a thin upper saturated zone that is a mixture of sandy units, till, and weathered clay. Harvey and Merrill Brooks flow southward through the study area, and receive groundwater discharge from the upper saturated zone and from the deep aquifer through previously unknown discontinuities in the confining unit. The Freeport aquifer gets most of its recharge from local seepage around the edges of the confining unit, the remainder is received as inflow from the north within the buried valley.

Groundwater withdrawals from the Freeport aquifer in the study area were obtained from the local water utility and estimated for other categories. Overall, the public-supply withdrawals (105.5 million gallons per year (Mgal/yr)) were much greater than those for any other category, being almost 7 times greater than all domestic well withdrawals (15.3 Mgal/yr). Industrial withdrawals in the study area (2.0 Mgal/yr) are mostly by a company that withdraws from an aquifer at the edge of the Merrill Brook watershed. Commercial withdrawals are very small (1.0 Mgal/yr), and no irrigation or other agricultural withdrawals were identified in this study area.

A three-dimensional, steady-state groundwater-flow model was developed to evaluate stream-aquifer interactions and streamflow depletion from pumping, to help refine the conceptual model, and to predict changes in streamflow resulting from changes in pumping and recharge. Groundwater levels and flow in the Freeport aquifer study area were simulated with the three-dimensional, finite-difference groundwater-flow modeling code, MODFLOW-2005. Study area hydrology was simulated with a 3-layer model, under steady-state conditions.

The groundwater model was used to evaluate changes that could occur in the water budgets of three parts of the local hydrologic system (the Harvey Brook watershed, the Merrill Brook watershed, and the buried aquifer from which pumping occurs) under several different climatic and pumping scenarios. The scenarios were (1) no pumping well withdrawals; (2) current (2009) pumping, but simulated drought conditions (20-percent reduction in recharge); (3) current (2009) recharge, but a 50-percent increase in pumping well withdrawals for public supply; and (4) drought conditions and increased pumping combined. In simulated drought situations, the overall recharge to the buried valley is about 15 percent less and the total amount of streamflow in the model area is reduced by about 19 percent. Without pumping, infiltration to the buried valley aquifer around the confining unit decreased by a small amount (0.05 million gallons per day (Mgal/d)), and discharge to the streams increased by about 8 percent (0.3 Mgal/d). A 50-percent increase in pumping resulted in a simulated decrease in streamflow discharge of about 4 percent (0.14 Mgal/d).

Streamflow depletion in Harvey Brook was evaluated by use of the numerical groundwater-flow model and an analytical model. The analytical model estimated negligible depletion from Harvey Brook under current (2009) pumping conditions, whereas the numerical model estimated that flow to Harvey Brook decreased 0.38 cubic feet per second (ft3/s) because of the pumping well withdrawals. A sensitivity analysis of the analytical model method showed that conducting a cursory evaluation using an analytical model of streamflow depletion using available information may result in a very wide range in results, depending on how well the hydraulic conductivity variables and aquifer geometry of the system are known, and how well the aquifer fits the assumptions of the model. Using the analytical model to evaluate the streamflow depletion with an incomplete understanding of the hydrologic system gave results that seem unlikely to reflect actual streamflow depletion in the Freeport aquifer study area.

In contrast, the groundwater-flow model was a more robust method of evaluating the amount of streamflow depletion that results from withdrawals in the Freeport aquifer, and could be used to evaluate streamflow depletion in both streams. Simulations of streamflow without pumping for each measurement site were compared to the calibratedmodel streamflow (with pumping), the difference in the total being streamflow depletion. Simulations without pumping resulted in a simulated increase in the steady-state flow rate of 0.38 ft3/s in Harvey Brook and 0.01 ft3/s in Merrill Brook. This translates into a streamflow-depletion amount equal to about 8.5 percent of the steady-state base flow in Harvey Brook, and an unmeasurable amount of depletion in Merrill Brook. If pumping was increased by 50 percent and recharge reduced by 20 percent, the amount of streamflow depletion in Harvey Brook could reach 1.41 ft3/s.

First posted March 15, 2012

For additional information contact:
U.S. Geological Survey
Maine Water Science Center
196 Whitten Road
Augusta, ME 04330
(207) 622-8201

Part or all of this report is presented in Portable Document Format (PDF); the latest version of Adobe Reader or similar software is required to view it. Download the latest version of Adobe Reader, free of charge.

Suggested citation:

Nielsen, M.G., and Locke, D.B., 2012, Simulation of groundwater conditions and streamflow depletion to evaluate water availability in a Freeport, Maine, watershed: U.S. Geological Survey Scientific Investigations Report 2011– 5227, 72 p., at




Purpose and Scope

Description of Study Area

Hydrogeologic Framework


Surficial Geology and Mapped Soils

Glacial Geology at Depth


Groundwater Resources

Hydraulic Properties

Groundwater Flow


Groundwater Levels

Surface-Water Resources

Streamflow Measurements in Harvey and Merrill Brooks

Conceptual Model of the Groundwater Flow System

Water Use and Withdrawals

Reported Withdrawals from the Freeport Aquifer

Estimated Withdrawals from Groundwater

Simulation of Groundwater Flow and Discharge to Streams

Steady-State Numerical Groundwater-Flow Model

Spatial Discretization of the Model

Boundary Conditions


Hydraulic Properties

Model Calibration Using Parameter Estimation and Observations



Changes to the Conceptual Model

Model Fit to Observations

Simulated Groundwater Levels and Flow Under Steady-State Conditions

Model Sensitivity Analysis and Parameter Uncertainty

Limitations of the Model

Model-Calculated Water Budget for Harvey and Merrill Brooks and the Buried Freeport Aquifer

Evaluation of Streamflow Depletion in Harvey Brook

Calculation of Instream Flow Requirements for Harvey Brook

Streamflow Depletion Estimates Based on STRMDEPL08

Simulation of Streamflow Depletion Based on the Steady-State Groundwater-Flow Model

Comparison of Methods Used to Evaluate Streamflow Depletion

Comparison of Streamflow Depletion Estimates to Instream Flow Requirements

Suggestions for Improving Methods of Study for Water Availability

Summary and Conclusions

References Cited

Appendix 1. List of Wells and Observations Used in the Freeport Aquifer Study

Appendix 2. Details of Groundwater Model Calibration

Accessibility FOIA Privacy Policies and Notices logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: Contact USGS
Page Last Modified: Thursday, January 10, 2013, 08:14:29 PM