Purpose and Scope

This report documents the collection of synoptic groundwater level data in Fauquier County, Virginia in late October and early November 2018. These field measurements were used in a geographic information system (GIS) to develop a county-wide groundwater-level contour map. A basic error analysis was performed on the estimated groundwater-levels. Antecedent conditions that influenced the field measurements, and the limitations of the model, also are discussed.

Description of Study Area

Fauquier County consists of 650 square miles (mi²) in north-central Virginia. The county straddles the drainage divide between the north-flowing tributaries of the Potomac River and the south-flowing tributaries of the Rappahannock River (fig. 1). The county includes parts of the Piedmont and Blue Ridge Physiographic Provinces, two geographic regions with large topographic differences. In the northwestern part of Fauquier County, the Blue Ridge Physiographic Province has steep slopes with altitudes up to 2,400 feet. In the southeastern part of the county, the Piedmont Physiographic Province is characterized by rolling terrain and lower altitudes, from 160 to 800 feet. These altitude differences account for the slight variation in mean annual precipitation, which averaged between 40 and 43 inches, across the county for the period from 1981 to 2010 (PRISM Climate Group, 2019).

Geologically, Fauquier County is underlain by crystalline rocks of the Blue Ridge Physiographic Province and the northwestern part of the Piedmont Physiographic Province. These rocks act as one principal aquifer and are referred to in this study as the Blue Ridge crystalline aquifer (fig. 1). The remainder of the Piedmont Physiographic Province is further divided by altitude and geologic structure. The central lowlands of the Piedmont Physiographic Province are underlain by Mesozoic sedimentary rocks deposited in a northeast-southwest trending rift basin, referred to as the Mesozoic basin aquifer for this study. The far southeastern corner of Fauquier County, part of the Piedmont Physiographic Province uplands, is underlain by rocks of multiple island-arc and volcanic origins and is referred to as the Piedmont crystalline aquifer in this study (Davis and others, 2001).

The City of Warrenton is the seat of local government in Fauquier County and the largest municipality. Nine service districts organized along major highways provide water and sewer utilities to low- and medium-intensity development areas (fig. 2). Most of the land area of Fauquier County, and 45 percent of its population, lie outside the nine service districts. Most of the county is rural, with major land cover consisting of deciduous forest or pasture (McCoy and Ladd, 2019).

As part of ongoing USGS investigations, groundwater levels are being continuously monitored at locations in and around Fauquier County. Continuous water-levels have been recorded in well 48U 26 since October 2007 and in well 49V 1 since September 2003. Continuous water-levels have been recorded in monitoring wells 48S 38, 48T 7, 48U 32, and 49U 81 since April 2018 (fig. 1). Water-levels are uploaded in near real time automatically by satellite telemetry to the USGS National Water Information System (NWIS) and verified with field measurements on a 6- to 8-week interval.

Collection of Groundwater-Level Data

A synoptic survey of groundwater levels in 129 wells across Fauquier County was conducted from October 29 through November 2, 2018 (appendix 1; fig. 2). Field personnel requested access to the wells and any available well information from private-domestic well owners. Water levels were measured manually with a calibrated steel tape to the nearest one-hundredth of a foot. A calibrated electric tape was substituted for the steel tape in wells where the depth to water could not be determined with the steel tape, most often because of moisture inside the well casing. At each visit, multiple measurements were made, at least five minutes apart, until a pair of measurements were within two one-hundredths of a foot. This standard procedure reduces error from any short-term groundwater-level fluctuations attributed to recent well pumping (Cunningham and Schalk, 2011). The water-level measurements and pertinent well information were entered, reviewed, and approved, then stored in the USGS National Water Information System (NWIS) database.

Development of Groundwater-Level Contour Map

The synoptic groundwater-level data were used as one input in the development of a groundwater-level contour map for Fauquier County. Assuming that streams are a surficial expression of groundwater, the altitudes of key points in the stream network were selected to provide additional spatial coverage and help confine groundwater levels below land surface. These key stream points were defined as the intersections of streams from the National Hydrography Dataset (NHD) with a Strahler stream order of two or greater, with the intent to eliminate ephemeral headwaters that do not reliably indicate the presence of groundwater. Altitudes for the key stream points were obtained from a three-meter (9.84 ft) digital elevation model (DEM) of Fauquier County (U.S. Geological Survey, 2011). To avoid representing groundwater levels beyond the scope of field-verified measurements, key stream points were further restricted to those with an altitude less than the highest measured groundwater level. Stream points within 3 miles (mi) of the Fauquier County boundary were included in the analysis to avoid problematic interpolation along the county line.

Data from the synoptic groundwater-level survey and key stream points were merged into a geographic information system (GIS), which was used to create a spatially referenced raster surface of groundwater-level altitude using a natural-neighbor interpolation. A GIS contouring tool converted the interpolated raster surface into a line-based dataset. These contour lines represent an equal altitude of estimated groundwater levels above the NGVD 29 datum. To reduce errors where the interpolated groundwater levels were greater than three meters (9.84 feet) above the DEM land surface, contours were manually adjusted where steep hydraulic gradients are assumed to mimic the steep topography. Adjustments to groundwater-level contours also were required to depict groundwater discharge in stream bottoms. Gaining streams were an assumption embedded in this approach using stream altitude and are consistent with findings elsewhere in the Blue Ridge and Piedmont physiographic provinces (McCoy and others 2015).

The edited groundwater-level contours served as the primary dataset for the Fauquier County groundwater-level contour map, alongside the measured synoptic groundwater-level measurements and a background of land-surface altitude, also referenced to the NGVD 29 datum (fig. 3).

A final GIS contour to raster tool was used to convert the estimated groundwater-level contours to a continuous, thirty-meter (98.4 ft) resolution, estimated groundwater-level altitude raster surface for subsequent error analysis. Both the contour lines and raster surface are available as a USGS data release (Kearns, 2022).

Antecedent Conditions

Although a synoptic survey of groundwater levels is an attempt to provide a snapshot in time, it is important to recognize that groundwater is a dynamic resource, responsive to changes to inputs to, or demands on, the hydrologic system. Groundwater levels can fluctuate on a seasonal basis or deviate dramatically from average conditions during exceptionally wet periods or drought. Groundwater withdrawals also impact the water levels, particularly during drought conditions or anytime withdrawals exceed recharge.

Precipitation and Groundwater Levels

The synoptic survey period was selected to occur around the time of the usual seasonal minimum in the annual cycle of groundwater fluctuations, which – due to factors including precipitation, temperature, and growing season – typically includes a period of recharge in the winter and spring followed by a general decrease throughout the summer and fall (McCoy and others, 2015). Conditions prior to the October-November 2018 synoptic survey, however, were abnormally wet. Monthly mean PRISM precipitation data for Fauquier County from June to September (fig. 4A) were twice the 30-year monthly mean for the period 1981 to 2010 (PRISM Climate Group, 2019). This likely contributed to above-normal groundwater levels for most of the summer, as shown by data from the two USGS monitoring wells in the region with records long enough to compute a 10-year daily median from data published in NWIS for the period 2007 to 2017 (fig. 4B,C).

Monitoring well 48U 26 is in the Blue Ridge crystalline aquifer of northern Fauquier County. Water levels in 48U 26 were generally elevated during the summer, decreased throughout October 2018, and were nearing long-term median values during the period of the synoptic survey. Consistent with findings elsewhere in the crystalline aquifers of the Blue Ridge physiographic province (McCoy and others, 2015), well 48U 26 is highly sensitive to fluctuations in groundwater level; for wells less sensitive to short-term conditions, water levels that rose in response to the extreme summer precipitation would likely take even longer to return to normal. Monitoring well 49V 1 is completed in the Mesozoic basin aquifer approximately two kilometers east of Fauquier County. Water levels in well 49V 1 were consistently higher than 10-year daily median values for most of the summer of 2018. Although the groundwater level did decline throughout October, it rose again immediately before and during the period of the synoptic survey. Compared to well 48U 26, well 49V 1 appears less sensitive to fluctuations in water level and has a smaller range of median values throughout the year.

Evaluation of the combination of county-wide precipitation data and a monitoring well record in each of the two major aquifer systems indicates that it would be reasonable to conclude that groundwater levels in Fauquier County were likely at or above long-term median values during the October-November 2018 synoptic survey.

Groundwater Withdrawals and Use

The USGS, in cooperation with programs like the VADEQ’s Virginia Water Use Data System, provides national water-use estimates summarized by county every five years. Data for 2015 (U.S. Geological Survey, 2020; the most recent available at time of publication) show that nearly 99 percent of the estimated 4.32 million gallons per day (Mgal/d) of groundwater withdrawn in Fauquier County was used for public- or private-water supply. Most (1.76 Mgal/d) of the public-water supply in Fauquier County and all (2.51 Mgal/d) private-domestic-water supply comes from groundwater wells (table 1). Private-domestic supply is the largest categorical use of groundwater in the county. The remaining 1 percent (0.05 Mgal/d) of the estimated groundwater withdrawal supports livestock and the irrigation of approximately 200 acres on golf courses. From 1985 to 2015, the estimated public-supply demand for groundwater in Fauquier County rose from 0.6 to 1.76 Mgal/d—a nearly threefold increase. During the same period, estimated withdrawal from private-domestic wells rose 18 percent, from 2.13 to 2.51 Mgal/d (Dieter and others, 2018).

Table 1.    

Estimated groundwater withdrawals in Fauquier County, Virginia, 1985–2015. Water volumes in million gallons per day (Mgal/d). Data from Dieter and others (2018) and U.S. Geological Survey (2020).

[—, no data]

Table 1.    Estimated groundwater withdrawals in Fauquier County, Virginia, 1985–2015. Water volumes in million gallons per day (Mgal/d). Data from Dieter and others (2018) and U.S. Geological Survey (2020).

Year	Population, public supply groundwater	Public supply, groundwater	Population, domestic self-supplied	Domestic self-supplied	Industrial self-supplied	Thermoelectric power self-supplied	Livestock self-supplied	Irrigation, golf, self-supplied	Irrigation, crop, self-supplied
1985	4,750	0.6	28,350	2.13	0	0	—	—	—
1990	14,340	1.45	27,400	2.06	0	0	—	—	—
1995	11,300	1.69	32,170	2.41	0.02	0	—	—	—
2000	13,280	0.69	33,889	2.54	0	0	0.06	0	0.06
2005	22,418	1.77	31,472	2.36	0	0.01	0.05	0	0.16
2010	22,703	1.49	31,207	2.34	0	0.01	0.04	0.01	0
2015	26,229	1.76	31,328	2.51	0	0	0.04	0.01	0

Water-use information is available on smaller and more recent spatio-temporal scales from VADEQ, but only for entities meeting reporting thresholds (withdrawing an average of 10,000 gal/day). This threshold would exclude private-domestic use but does capture public-water suppliers, golf course irrigation, and a small thermoelectric plant in Fauquier County. VADEQ data from 2015 to 2019 indicate an average annual reported groundwater withdrawal of 675 Mgal, or 1.85 Mgal/day—slightly greater than corresponding 2015 USGS estimates (Virginia Department of Environmental Quality, 2020). A monthly breakdown of VADEQ data for 2018 shows a clear seasonal fluctuation in groundwater withdrawals, with the highest demand from May to October, a peak in July, and the lowest demand in February (fig. 5). This seasonal demand contrasts with the annual pattern of groundwater availability in Fauquier County (as shown by the median conditions [water levels] in wells 48U 26 and 49V1 in figure 4B), in that demand is often highest when water levels typically are at their lowest. The groundwater withdrawal in October 2018 of 58.31 Mgal was 0.5 percent below the average October withdrawal and 3.6 percent above the average monthly withdrawal for 2015 to 2019 (Virginia Department of Environmental Quality, 2020). These values indicate that groundwater demand in Fauquier County was normal prior to the synoptic survey.

Precipitation and groundwater-levels prior to the synoptic survey suggest a slightly above-normal supply following a very wet summer. Estimated and reported groundwater withdrawals suggest normal seasonal demand. Determining the dominant force on groundwater levels from a single set of synoptic measurements is unlikely; however, the review of the data on antecedent precipitation, water levels, and withdrawals indicates the groundwater-level contour map was not developed from an extreme mismatch in groundwater conditions.

Error Analysis and Limitations of Estimated Groundwater Levels

Groundwater levels estimated from the contour map were compared to the levels measured during the synoptic survey. Root-mean-squared error (RMSE) of the estimated groundwater levels was 33.52 feet (eq. 1).

Root-mean-squared error

A plot of estimated groundwater levels against measured levels (fig. 6A) and residual error (fig. 6B) indicated that the estimated groundwater-level altitudes generally were lower than the measured altitudes. This bias generally was consistent across all groundwater levels, with a slight increase at greater altitudes. A GIS comparison of the groundwater-level raster surface and the DEM of Fauquier County indicated that the greatest error in estimated altitude of the groundwater-level surface was on hillsides and in stream valleys in steep terrain. On hillsides in particular, the depth-to-water and residual error are sensitive to small shifts of groundwater-level contour lines. Additionally, the horizontal resolutions of the raster and DEM used to estimate groundwater-level and land-surface altitudes are less accurate in high-relief terrain.

As discussed in the study methods, the groundwater-level contours were adjusted in areas of steep topographic gradient. Adjustments to contours resulted in lowering of estimated groundwater-level altitudes in comparison to actual water-level observations, increasing error in these steep areas. Thus, the development of a county-wide groundwater-level contour map is a balance between reducing local error between estimated and measured water levels while maintaining large-scale topographic and hydrologic integrity.

The Fauquier County groundwater-level contour map was developed with the intent to represent steady-state groundwater conditions when recharge and groundwater levels were near their long-term averages. With only 129 observations across the 650-mi² county, it was necessary to interpolate groundwater-level contours from limited data. The interpolations assumes that the groundwater occurs and moves within a single, unconfined system; however, gradients and pressures within the system are regionally controlled by surface topography and geology. Therefore, estimated groundwater levels differed from those measured in wells, as indicated by the RMSE and residuals. Extrapolation of the results of this study to smaller-scale investigations would require more hydrogeologic data than were available at the time of publication. Documenting changes in groundwater levels due to seasonal variation, extreme conditions, climate change, or groundwater withdrawals also are outside the scope of this report. Despite these limitations, the groundwater-level map presented here represents the best-available science concerning county-wide groundwater levels and serves as a solid basis for future modeling of Fauquier County groundwater resources.