Purpose and Scope

The purpose of this report is to present simulated changes to the water-table altitude and streamflow in Falmouth, Cape Cod, relative to current (2019–23) average conditions for two wastewater-return flow scenarios requested by the Town of Falmouth. This report describes the groundwater flow model that was used for these simulations and the results of those simulations, as well as discusses the limitations of the modeling analysis. All numerical model input and output files are available in the USGS data release that accompanies this report (Goldstein and McCobb, 2025).

Geographic and Hydrologic Setting

The aquifer system underlying Falmouth is composed of highly permeable unconsolidated glacial sediments, which are about 300 feet (ft) thick in some areas (Fairchild and others, 2013) and is bounded below by relatively impermeable bedrock, above by the water table, and laterally by saltwater (LeBlanc and others, 1986). The sediments underlying Cape Cod were deposited in environments marginal to continental ice sheets during the Wisconsinan glaciation about 18,000 years ago and consist of gravel, sand, silt, and clay deposited in moraines and coalescing deltaic systems (Giese and others, 2015; Larson, 1982). Glacial deltaic sediments generally have a fining-downward sequence with coarse-grained fluvial outwash underlain by finer grained glaciolacustrine sediments (Masterson and others, 1997).

There is a groundwater divide to the northeast of Falmouth from which groundwater flows radially outward from the high point at the divide towards the coast (Masterson and Walter, 2009). Groundwater in Falmouth generally flows in a southerly to southwesterly direction and ultimately discharges to coastal waters as direct-groundwater discharge or streamflow (Savoie, 1995). The water table in Falmouth ranges in altitude from 0 ft above the National Geodetic Vertical Datum of 1929 (NGVD 29; referred to as the sea level in this report) at the coast to about 50 ft above sea level near the town’s northeastern border (Savoie, 1995).

Groundwater and surface water (such as ponds and streams) in the hydrologic system on Cape Cod are hydraulically connected, composing a single freshwater resource (Winter and others, 1998) called the Cape Cod aquifer system. Streams in Falmouth and the bordering Town of Mashpee (listed from west to east) include the Coonamessett River, Backus River, Bourne River, Childs River, Quashnet River, and Red Brook (fig. 1). These streams are baseflow dominated, meaning they receive most of their flow from direct-groundwater discharge across the streambeds (LeBlanc and others, 1986). Numerous freshwater ponds in Falmouth are within glacial kettle hole depressions (formed by retreating glaciers), the largest of which is Coonamessett Pond (Oldale, 2001; Giese and others, 2015; fig. 1). Kettle ponds are flow-through ponds that receive groundwater discharge in upgradient areas and contribute water back to the aquifer in downgradient areas or are drained by streams (Hull and others, 2019). Long Pond, another kettle pond in southwestern Falmouth, is used as a water source by the town.

Recharge at the water table is the sole source of water for the aquifer. Natural recharge from precipitation is augmented in some areas by infiltration of wastewater-return flow from septic systems and at centralized WWTFs. These WWTFs have the potential to locally raise the water table in their vicinities, where water is returned to the aquifer away from the area it was pumped for municipal supply (Masterson and Walter, 2009). The average natural recharge to the Sagamore flow lens from 1995–2000 was about 27 inches per year (Walter and others, 2019). Groundwater discharges directly to coastal waters as submarine discharge and indirectly as streamflow, and is pumped from production wells for public supply. Discharge to coastal waters accounts for about 65 percent of total outflow, and discharge to streams accounts for about 28 percent of total outflow (Walter and Whealan, 2005). About 7 percent of total outflow is withdrawn from public-supply wells (Walter and Whealan, 2005).

The water supply for the Town of Falmouth is withdrawn from Long Pond, four production wells located within Falmouth, and three production wells operated by the Upper Cape Regional Water (UCRW) Cooperative located to the north of Joint Base Cape Cod (fig. 1). From 2019–23, the average withdrawal from Long Pond was 382,170 cubic feet per day (ft³/d), which accounted for about 64 percent of Falmouth’s municipal supply (table 1; fig. 2A). Two-thirds of annual groundwater withdrawals from 1995–2000 occurred between the months of May and September (Walter and Whealan, 2005).

At the Main and New Silver Beach WWTFs, treated wastewater recharges the aquifer through open sand beds. The majority of wastewater-return flow at the WWTFs occurs at the Main WWTF, which, from 2019–23, had an annual average return-flow rate of 39,375 ft³/d at beds 9–13 and 20,199 ft³/d at beds 14–15 (table 2; fig. 2B). The New Silver Beach WWTF had an annual average return-flow rate of 2,133 ft³/d from 2019–23 (table 2; fig. 2B). Most wastewater (443,811 ft³/d or 88 percent of the total) recharges the aquifer through individual septic systems (table 2; fig. 2B). Untreated septic system effluent has substantially higher concentrations of nitrogen than treated effluent and is a major source of excess nitrogen in coastal waters (U.S. Environmental Protection Agency, 1977).

Existing Groundwater Flow Model

The existing USGS MODFLOW-2005 groundwater flow model simulates steady-state flow in the Sagamore flow lens of the Cape Cod aquifer (Walter and others, 2019). The model is a three-dimensional grid divided into even cells that are 400 ft on each side and consists of 32 layers of varying thickness (Walter and others, 2019). The upper boundary is the simulated position of the water table (Masterson, 2004; Walter and Whealan, 2005; Masterson and others, 2009), the altitude of land surface at streams, or the altitude of the seabed at coastal boundaries. The lower boundary is the bedrock surface or the freshwater-saltwater interface. The position of the interface was derived from a dynamic interface model developed by Walter and others (2016); this model used the Seawater Intrusion (SWI2) package (Bakker and others, 2013), which balances flow and two-solution densities to estimate the freshwater-saltwater interface position beneath the Sagamore and Monomoy flow lenses (Walter and others, 2016). The interface position from that analysis was mapped to the groundwater model grid and is represented as a fixed position in the simulations.

Inflows to the groundwater system in the model include natural recharge from precipitation and recharge from wastewater-return flow from both WWTFs and septic systems. Initial rates of natural recharge from precipitation were estimated using a Soil-Water Balance model and adjusted during calibration, which resulted in spatially variable recharge with a mean of 27.3 inches per year (Walter and others, 2019). Wastewater-return flows from both WWTFs and septic systems are accounted for as separate components of the total simulated recharge in the model. Recharge from onsite septic systems was derived from parcel-scale water-use data (Massachusetts Estuaries Project, 2024) that were mapped to the groundwater-model grid (Walter and others, 2019).

Modeled outflows include specified groundwater withdrawals from production wells, surface water withdrawals from Long Pond, and discharge to coastal waters and streams. Ponds in the model, including Long Pond, are represented as areas with essentially infinite horizontal and vertical hydraulic conductivities of 100,000 and 10,000 feet per day (ft/d), respectively, to allow the pond-water levels to respond uniformly to changes in hydraulic stresses across the pond area (Walter and others, 2019). Streams in the model are represented as head-dependent flux boundaries using the Drain Package in MODFLOW-2005, where flow into a model cell representing a stream reach is proportional to the simulated hydraulic gradient across the streambed. As a result, the model can only simulate discharge from the groundwater system to streams, and not potential inflows from streams to the groundwater system. This simplified approach is appropriate for this model scale and for predominantly gaining streams, such as those that generally occur on Cape Cod.

Current (2019–23) Condition

The model described in Walter and others (2019) was calibrated using average hydrologic data, pumping stresses, and recharge data from 1995–2000; this calibrated model was then used by Walter and others (2019) to simulate hydrologic conditions in 2010 and “future” conditions in 2030. For the analyses in this report, the “current” hydrologic condition was updated using average pumping and recharge stresses from 2019–23 and represents average conditions for that period. Recharge from precipitation, parcel-scale distribution of septic-system wastewater-return flow in Falmouth, and pumping and wastewater-return-flow rates for towns outside of Falmouth (other than production wells located within 2 miles of the Falmouth town border) were not updated from the average 2010 conditions simulated in the model described in Walter and others (2019) as “current,” since the focus of this study was on the Town of Falmouth.

Total annual water withdrawal data for 2019–23 were provided by the Town of Falmouth Water Division for the town’s production wells and surface-water supply source at Long Pond, by water suppliers in the towns of Bourne (Bourne Water District) and Mashpee (Mashpee Water District) for production wells in those towns that are within 2 miles of the Falmouth border, and by the Upper Cape Regional Water (UCRW) Cooperative (table 1; fig. 2A; Goldstein and McCobb, 2025). The UCRW Cooperative provides drinking water to towns on Cape Cod, including Falmouth, to meet shortfalls in supply. The UCRW Cooperative has three production wells located to the north of Joint Base Cape Cod in the Town of Sandwich. The total average 2019–23 withdrawal at production wells used for Falmouth public supply was 598,259 ft³/d, a 7-percent increase over the total average withdrawal in 2010 (table 1; fig. 2A).

Wastewater-return-flow rates for 2019–23 at the Main WWTF and New Silver Beach WWTF in Falmouth were obtained from the Town of Falmouth (table 2; fig. 2B; Goldstein and McCobb, 2025). Overall, the trend of wastewater-return-flow rates at the WWTFs increases over time between 2010 and 2023, but the averages for 2010 and 2019–23 were relatively similar: the average total WWTF return-flow rate for Falmouth was 61,668 ft³/d in 2010 and 61,707 ft³/d in 2019–23 (table 2; fig. 2B). One significant difference between these data from 2010 and 2019–23, however, was the addition of open sand beds 14–15 at the Main WWTF (fig. 3A, 3B), which distributed discharge across beds 9–13 and 14–15.

The portion of total wastewater-return flow recharged to the aquifer through individual septic systems was estimated in areas served by the water-supply system using the following steps: (1) the total volume pumped for public supply was reduced by 15 percent to account for consumptive water use (for example, water used for irrigation that is lost to evapotranspiration); (2) the adjusted volume was further decreased by subtracting water discharged at WWTFs; and (3) this final volume of water was distributed over residential areas that were not sewered and relied on septic systems for wastewater management.

The adjusted volume was distributed spatially in proportion to parcel-scale water use based on 2010 water-use rates as discussed in Walter and others (2019). Therefore, return-flow rates were updated based on data from 2019–23, but the relative spatial distribution remained the same as 2010, which is in general agreement with recent (since 2010) development. The model-grid cells where return flow is recharged through septic systems are documented in the model application data release (Goldstein and McCobb, 2025). The volume of wastewater-discharge to septic systems was assumed to equal the volume of withdrawals from wells in areas served by domestic wells and septic systems, thus resulting in no-net hydrologic stress in these areas. Additional descriptions of this approach are included in Walter and others (2016) and Walter and others (2019).

In addition to updating the model to simulate average 2019–23 pumping and wastewater-return-flow rates, the locations of sewered areas, production wells, and open sand beds at the WWTFs in Falmouth were updated. Between the 2010 and 2019–23 periods, the neighborhoods in the areas of Falmouth Heights and the Maravista Peninsula near Little Pond were connected to municipal sewers that discharge at the town’s Main WWTF (fig. 3A, 3B), pumping was initiated at the Mares Pond well (which was inactive in 2010), the location of the intake at Long Pond was changed to its current position (fig. 1; Massachusetts Bureau of Geographic Information, 2024), and open sand beds 14–15 were added to existing open sand beds 9–13 at the Main WWTF (fig.  3A, 3B). Additionally, an error was identified in the model by Walter and others (2019) that placed the intake at Long Pond in layer 2, which represents lower hydraulic conductivities characteristic of aquifer sediments, instead of in layer 1. As part of updating the model for these simulations, the horizontal location of the Long Pond intake was updated by changing the model cell that the intake was assigned to in the Well Package, and the error in the original model was remedied by assigning the intake to layer 1, where higher hydraulic conductivity values are used to represent the pond.

Comparison of 2010 and 2019–23 Simulated Conditions

The difference in the simulated water table between average 2010 hydrologic conditions and average 2019–23 hydrologic conditions is shown as lines of equal change (contours) in simulated water-table altitude (fig. 4). Differences were calculated by subtracting simulated water-table altitudes for 2019–23 conditions from the 2010 conditions; a positive value indicates a decrease in water-table altitude and a negative value indicates an increase.

Changes between the 2010 and 2019–23 hydrologic conditions occur where hydrologic stresses in the model have been modified. There was a decrease in the simulated water-table altitude (indicated by contours with a positive value) at or near production wells that had larger withdrawal rates in 2019–23. Conversely, the simulated water-table altitude increased at or near wells with withdrawal rates in 2019–23 lower than rates in 2010 (fig. 4). Between 2010 and 2019–23, the Mares Pond well (which was inactive during the 2010 period) was added as a source for municipal supply in Falmouth, resulting in a simulated decrease in the water-table altitude of about 1.2 ft near the well (fig. 4).

The change in the simulated water table is also a function of the changes in the locations of pumping wells and wastewater-return flow. Two additional open sand beds (14–15) were added after 2010 at the Main WWTF (fig. 3B). Wastewater-return flow, which had a similar long-term average value for 2010 and 2019–23 (table 1; fig. 2A), was distributed across the new (14–15) and existing (9–13) open sand beds. This redistribution of discharge caused a decline in the simulated water table at the existing open sand beds (9–13) and a rise in the simulated water table at the new open sand beds (14–15; fig. 4). Pumping from Long Pond for public water supply is simulated as withdrawal occurring from a single model cell at approximately the intake location for the raw-water pump station. The intake location at Long Pond was moved northward between 2010 and 2019–23. There was an increase (about 0.6 ft) in the simulated water-table altitude near the former location of the intake and a decrease (about 0.9 ft) near the new location (fig.  4).

Simulated Water-Table Responses to Changes in Wastewater-Return Flow

The simulated regional water-table altitudes were found to be similar for the current (2019–23) condition and for simulated conditions in scenarios 1 and 2 (fig. 5). In scenario 1, the simulated water table decreases in altitude and contours shift upgradient south of Route 28 and near the Main WWTF, where wastewater-return flow was removed (fig. 5A). The regional water table altitude decreases over a larger area in scenario 2, particularly in southern Falmouth where there is more residential development and the hydrologic responses to sewering would be expected to be larger (fig. 5B). The direction and magnitude of regional hydraulic gradients remain similar to the current (2019–23) conditions for both scenarios (fig. 5A, 5B).

In scenario 1, all residential properties in the water-system service area south of Route 28 in Falmouth are connected to municipal sewers and wastewater-return flow at the Main WWTF is removed (fig. 3C), as would occur with the installation of an ocean outfall pipe. This scenario simulates the removal of 180,343 ft³/d of wastewater-return flow from properties in the southern part of Falmouth and at the Main WWTF, which was a decrease of 35.7 percent from current (2019–23) wastewater-return flows (table 2). Decreases in the simulated water-table altitude for scenario 1 are observed near the Main WWTF and on the coastal necks south of Route 28, where septic wastewater-return flow was removed following proposed sewering of those areas (fig. 6). The decrease in simulated water-table altitude is largest (about 1.1 feet) near the Main WWTF, because treated wastewater-return flow at the open sand beds at the Main WWTF site (which may have created a locally elevated water table) was removed (fig. 6). Decreases of less than 0.2 ft occur in the areas in the vicinity of the production wells and Long Pond (fig. 6). Decreases of about 0.1 ft occur south of Route 28 (fig. 6) where properties were converted to sewers, including on the coastal necks (fig. 6).

Scenario 2 represents the connection of all residential properties in Falmouth to municipal sewers and removal of wastewater-return flow at the Main WWTF with the installation of an ocean outfall pipe (fig. 3D). Decreases in simulated water-table altitudes resulting from removal of septic-return flow are observed over a larger area across Falmouth (fig. 7). In scenario 2, decreases in the water table altitude of about 0.1–0.2 ft occur throughout Falmouth and extend north into the Town of Bourne. The simulated water-table altitude generally decreased between about 0.1–0.2 ft on the coastal necks south of Route 28 and about 0.4–0.5 ft at Long Pond (fig. 7). The largest decrease of about 1.3 ft was near the Main WWTF, and decreases of less than 0.3 ft occurred at production wells in Falmouth. The town-wide connection to municipal sewers and removal of wastewater-return flow at the Main WWTF in scenario 2 represents a removal of 503,385 ft³/d of wastewater-return flow, or a 99.6 percent decrease from current (2019–23) flows (table 2).

The simulated hydrologic response of water-table altitudes in both scenarios was relatively small when compared to fluctuations that occur as a result of variations in natural recharge. For example, the average annual range in water-table altitude observed at USGS monitoring well MA-FSW 172 Falmouth, Mass. (413522070373601; fig. 1), from 2019–23 was 1.92 ft (U.S. Geological Survey, 2024), with a minimum annual range of 0.49 ft and a maximum annual range of 5.57 ft. Both the average and maximum annual ranges in water-table altitude at this well are larger than the long-term average change of 1.3 ft simulated near the Main WWTF in scenario 2 (fig. 7), which was the largest change simulated in either scenario. The decrease in total wastewater-return flow (from both the Main WWTF and septic systems) in each scenario is also relatively small when compared to the total flow throughout the system. The total decrease in wastewater-return flow in scenario 2 is 503,385 ft³/d (table 1), which is about 5 percent of the total recharge that occurs over the model area during current (2019–23) conditions. The relatively small changes to long-term average water-table altitudes are associated with the small percentage that the decrease in wastewater-return flow accounts for in total recharge over the study area. It is also possible that the small changes to water-table altitude on the coastal necks, in particular, may be associated with the closeness to coastal boundary, the small absolute hydraulic heads above the baseline of 0–4 ft above sea level, and the relatively thin freshwater lenses present on the coastal necks.

Simulated Streamflow Responses to Changes in Wastewater-Return Flow

Streamflow responses for six selected streams in Falmouth and Mashpee were evaluated using the simulated hydrologic budgets, which calculate inflows and outflows to the groundwater system, for each scenario. Streams on Cape Cod generally are gaining streams that accumulate groundwater discharge downstream (LeBlanc and others, 1986). Discharge to a stream is proportional to the hydraulic gradient between the stream and the water table; therefore, decreases in water-table altitudes are correlated with decreases in streamflows. The Coonamessett River, Backus River, Bourne River, Childs River, Quashnet River, and Red Brook were selected for this study (fig. 1).

Simulated average baseflow, which refers to the component of streamflow from groundwater discharge, decreased from current (2019–23) conditions for both scenarios (table 3; fig. 8). Streamflow decreased in all six simulated streams owing to the removal of wastewater-return flow, and the decreases in water-table altitudes and hydraulic gradients that occurred as a result. The decrease is greater in scenario 2, since a greater volume of septic-system wastewater-return flow was removed when representing town-wide sewering. The decreases in streamflow were greatest at the Coonamessett River (which is closest to the Main WWTF) in both scenarios. The average decrease in streamflow at the Coonamessett River was 0.1 cubic feet per second (ft³/s) in scenario 1 and 0.6 ft³/s in scenario 2, which represent decreases of 1.1 and 5.4 percent, respectively, from current (2019–23) conditions (table 3; fig. 8). The total decrease in streamflow relative to current (2019–23) conditions for scenarios 1 and 2 were 0.3 and 1.4 ft³/s, respectively (table 3; fig. 8). These decreases represent 0.7 and 3.0 percent, respectively, of total streamflow in scenarios 1 and 2. Both scenarios result in a relatively small change in long-term average streamflow, because the decrease in septic-system return flow and treated wastewater flow is a small percentage of the total recharge over the study area.