A CHANGING LANDSCAPE
Chapter Three (Part B)
GROUND PENETRATING RADAR SURVEY
Ground Penetrating Radar (GPR) is a remote sensing technique commonly used by archeologists, geologists and engineers to determine the underlying stratigraphy of the ground surface unobtrusively. The GPR equipment sends an electromagnetic pulse into the ground and records the speed of the signal returning to the surface. The materials the energy passes through, or is reflected from, determines the return speed of the signal. All materials react differently to electromagnetic energy. Materials are assigned a measurement called a dielectric constant; the approximate time (in nanoseconds) for the energy to pass through the material and reflect back to the surface is calculated using these constants.
Interpretation of electromagnetic energy is based on the fact that energy passes more quickly through dry air and soil than through water or rock. This characteristic of electromagnetic energy allows GPR to locate buried disturbances or objects. As the surface GPR antenna receives return signals, the GPR recorder records all differences in subsurface materials. The result is a printout recording a sub-surface view of the soil stratigraphy. The GPR operator then interprets this printout for areas of sharp contrast called anomalies. The type of equipment used and the makeup of the soils tested determines the accuracy and level of detail of recorded results (Conyers and Goodman 1997:11 - 55; Heimmer 1992:41).
Successful GPR surveys depend on several environmental factors and on operator technique. Moisture plays an important role in the success of GPR. A small amount of moisture can improve return qualities of the GPR, but too much water will scatter the energy so much that GPR will not provide an accurate reading of the soil. High moisture clays can also scatter radar returns and pose a problem to GPR surveys. The presence of natural features, such as tree roots and other plantings, produces anomalies in the returns. Segregating such natural anomalies from cultural anomolies without ground truthing is very difficult. Even close proximity to power lines and magnetic materials can skew data returns. With so many environmental variables affecting data results, it is imperative to employ careful methods, avoid error sources, and know the soil makeup of the site prior to survey (Conyers and Goodman 1997:11 - 55; Heimmer 1992:38)
Several natural subsurface factors in the Manassas National Battlefield area cause anomalies in the data returns. A local clay formation called "blackjack" lies underneath the soil in linear rows, often producing radar returns that appear to be foundation walls, graves and other linear features. Excavators at both Brawner farm and the Robinson House sites have encountered this material, noting that dry silt and rock runs in lines alternating with either hard pack or very moist clay (Creveling 1998; Parsons et al, in prep). Despite these error-inducing factors, GPR can still be a useful tool, but should be used in combination with other remote sensing equipment such as proton magnetometer and electrical resistivity (Creveling 1998:2).
Claude Petrone, Matthew Reeves and Jason Rust conducted the GPR survey on September 9, 1997. The GPR survey used a Subsurface Interface Radar #8 (SIR), 500 MgHz Geophysical Survey Systems, Inc., fiberglass encased antenna sled. The operator pulled the antenna across the surface at a constant speed along a predetermined line of measuring tape laid out across the ground surface. Data collection was set to occur continuously with reference tick marks placed every four feet on the data return.
Survey lines were laid out north, south and east of the structure. The stp grid served to align these survey lines. The lines ran north - south on the main site grid and east-west in the area west of the structure (Figure 3.7). The antenna was dragged from north to south along survey lines, except in the area west of the house where the antenna was dragged west to east. Survey lines were spaced three feet apart in the areas adjacent to the house and six feet apart in the area to the south of the structure. Distance traveled by the sled was marked at four foot intervals. The four foot intervals were denoted by a single click on the survey printouts. A double click along with a written notation demarcated natural features and stps. The antenna was picked up and carried around areas of unavoidable obstruction, such as trees and rocks. A double click on the printout also demarcated these areas.
A test run was conducted to calibrate the antenna and sample the radar return potential of the soils at Sudley Post Office. The test involved sinking a lead brick one foot into the soil and then running the antenna over the area. The distance of the lead brick to the surface on the printout allowed for a determination of depth of the signal on the printouts (Figure 3.8).
Claude Petrone analyzed all printed data returns after the survey was complete, and labeled anomalies in terms of their depth from the surface as shallow (S), medium (M) or deep (D) (Figure 3.9-3.11). At the Sudley Post Office Site, shallow anomalies occur between zero and one foot, medium anomalies are one to two foot down and deep anomalies exceed two feet in depth.
The GPR survey revealed five concentrations of anomalies (Figure 3.12). The first GPR concentration lay to the north of the structure near metal detector concentration M4. The second GPR concentration is north of the structure near the extant outhouse. The third GPR concentration is to the west of the structure, in metal detector concentration M2. The fourth and fifth GPR concentrations were to the south of the structure in metal detector concentration M1. Due to time constraints on the field excavation, many of these anomalies were not ground- truthed. Instead, anomalies detected by the GPR survey were interpreted using data from the metal detector and stp finds.
The first concentration revealed by GPR is a cluster of thirteen medium and 14 shallow anomalies. These anomalies are to the north of the structure centered at N725 E325, with an area occupying 30ft east -west and 50ft north-south. No metal detector hits are associated with this area, however nearby shovel test pits (N725 E300 and N725 E325) recovered ceramic fragments. Lack of clay in the subsoil (as evidenced in the stps excavated in the area) rules out the possibility that this concentration is a phantom anomaly. This anomaly might be the result of the Woodward=s planting activities. Another possible source for these analogies is the root systems of the trees located to the west.
Concentration A2 is a medium anomaly near to the extant outhouse. Excavation Unit 24 was placed near this anomaly and revealed a deep feature, most likely the remains of an earlier privy.
GPR concentration A3 is a large area to the west of the structure, measuring 35ft north-south by 100ft east-west. It contains nine deep, nine medium and 33 shallow anomalies. Deep anomalies may indicate a former structure in this area. Numerous metal detector hits were encountered in this area (Concentration M2). The proximity of this concentration to John Thornberry's wheelwright shop to the north suggests this area might have served as an extension of Thornberry's shop. Further ground-truthing measures are needed to test this interpretation.
Concentration A4 is a series of seven shallow, two medium and four deep anomalies discovered at N605-625 and E305-335. The anomalies form a 20-x 30ft rectangle. Numerous metal detector hits and nearby stps at N625 E300 and N625 E325 produced historic materials adjacent to Concentration A5. Combined evidence suggests the presence of a structure in this area.
Concentration A5 is a cluster of 22 shallow, nine medium and three deep anomalies. These anomalies form into a roughly rectangular area measuring 40 ft wide (N515 E235 - N515 E275) and 60 ft long (N515 E250 - 575 E 250). This area coincides with a concentration of historical artifacts located by stp survey. Artifacts recovered from this area include bottle, window and unidentified glass. Further ground truthing is needed to determine the nature of these anomalies.
Besides the noted concentrations of anomalies, many singular shallow anomalies lie outside these concentrations in all areas of the site. Many of these shallow anomalies can be dismissed as phantom readings caused by clayey subsoils, or remnants of scattered plantings. Shallow anomalies near the tree line at the southern and eastern boundaries of the site are probably root intrusions.
Overall, the GPR survey of the Sudley Post Office site uncovered a historic privy feature, and may have located two additional outbuilding structures, one to the north of the extant post office and one to the south. Two other possible activity areas were found, one that might be from activities associated with John Thornberry's wheelwright and another to the south near the driveway . Only future ground truthing will determine the actual existence and extent of these possible features.
The three survey methods used at Sudley Post Office allowed for the identification of several cultural resources. Stp findings, metal detector hits, and anomalies identified by gpr revealed three historic middens, three potential outbuildings, the location of John Thornberry's wheelwright shop, a privy hole, and a prehistoric site. Four of these areas, two outbuildings, a historic midden, and an activity area potentially associated with Thornberry's wheelwright shop, were independently verified by at least two of the three survey methods used at Sudley Post Office (Figure 3.13).These resources are categorized into three resources: outbuildings, historic middens, and prehistoric scatters. These resources are summarized below.
Five potential outbuildings were identified at Sudley Post Office. Gpr located a privy predating the extant privy to the north of the house. Metal detector sweeps found a concentration of materials potentially delineating the location of John Thornberry's wheelwright shop. This potential shop location is downslope and to the west of the house. Also, metal detector hits and GPR defined an area to the south that might be associated with Thornberry's shop. This area contained several deep anomalies and a scatter of metal scraps. Fifty feet to the south of the structure, stps and GPR readings highlighted an area that contained a high density of artifacts along with several medium to deep anomalies. This area has the potential to be an outbuilding with an associated downslope midden. One hundred and twenty-five feet to the southwest of the Sudley Post Office structure, stps and GPR indicated the potential location for another outbuilding. Stps revealed a low concentration of historic materials and GPR readings reflected the presence of some deep and medium anomalies in this area. To the north of the Sudley Post Office structure, metal detector sweeps uncovered a concentration of historic materials and GPR readings registered several medium and shallow anomalies. Local residents recalled mid-twentieth-century shed in the area. This structure is most likely the source of the metal detector hits and anomalies detected by GPR.
Shovel test pits revealed three historic middens at Sudley Post Office. All three of these middens are adjacent to the structure. The midden to the east is the heaviest concentration of materials. This midden contains late nineteenth to early twentieth-century materials. The second midden is 50 ft to the south of the structure. GPR readings picked up anomalies that suggest a potential structure in this area. The midden is just to the east and downslope of this area and is potentially associated with this potential structure. The third midden is 50 ft to the west of the structure. This midden contains late nineteenth to early twentieth-century century materials.
Stps unearthed two prehistoric scatters at Sudley Post Office. The first scatter is 125 ft south of the Sudley Post Office structure. This scatter was very light and contained no diagnostic materials. The second scatter is 250 ft southwest of the structure. Stps in this area revealed several diagnostic points dating to the Late Archaic Period.