GPR Archaeometry

Solving subsurface problems in the field of archaeology without destructively intervening with the buried materials has become a prime focus of the archaeological community. The science to study, measure and quantify archaeological structures remotely has been designated as the field of Archaeometry. Remotely detecting archaeological structures is very important because excavation of a site can inadvertently destroy essential archaeological evidence which can then never be recovered. Because of the successful application of a variety of geophysical tools, and in particular GPR to probe beneath the ground, many archaeologists now regularly initiate geophysical surveys before studying or excavating potential sites. The application of GPR in archaeology has ranged from studying protected sites which can never be excavated, to using GPR to quickly and cost-effectively plan and carry out mitigation projects. GPR surveys at sites that are impacted by development fall under the category of rescue archaeology. This is the largest growing segment of the GPR applications in archaeology, and these kinds of surveys are expedited by a growing number of geotechnical consulting firms. The first application of GPR in archaeology was initiated soon after the first commercial equipment became available in the 1970s. One of the earliest documented uses of GPR for archaeological prospection occurred in the mid-1970s when Bevan and Kenyon (1975) and Bevan (1977) used GPR to look for radar reflections from buried walls and variety of other historic structures; and Vickers and Dolphin (1975) used GPR to look for radar reflections from suspected buried walls associated with the native American Indian structures at Chaco Canyon. Dolphin et al. (1978) applied GPR in the successful search for caves on Victorio Peak in New Mexico. A variety of GPR case histories were published in the 1980s and 1990s. Vaughn (1986) used GPR to discover a sixteenth century Basque whaling station and to locate the graves of fisherman. In Japan, Imai et al. (1987) applied GPR to discover pit house floors buried in volcanic soils with great precision. DeVore (1990) used GPR for investigations at the Fort Laramie National Historic Site. Other notable early GPR surveys for archaeological prospection includes studies by Bevan (1991), Sheets et al. (1985), Fischer et al. (1980) and Batey (1987). These early GPR studies were primarily concerned with the discovery of buried features within known sites rather than imaging them. The early surveys used only paper records of the real-time GPR survey and never had the ability to perform any post processing on the data. Nonetheless, great efforts were used by the early investigators to create maps of subsurface anomalies by hand-contouring locations of continuous anomalies mapped in the field. These early crude maps have since been replaced by computer-generated time slice images. It is not known when the first computer-generated images that mapped horizontal changes in recorded reflections at constant time intervals were investigated. The authors have had access to unpublished reports (furnished by Bruce Bevan) showing as early as 1980, GPR surveys were done by researchers at Batelle National Laboratories in which a tractor-mounted digital-recording radar was used and computer-generated amplitude time slice maps were created from the recorded radargrams at an archaeological site. A crude form of time slice analysis was also available in 1986 from Geophysical Survey Systems Inc., the company that developed the first commercial GPR system. Nishimura and Kamei (1990) and Milligan and Atkin (1993) were among the first to employ a basic form of time slice analysis, in which radar reflections were mapped horizontally for archaeological applications. The early time slice softwares created very pixilated maps where the profile spacing represented the width of the pixels displayed on the computer screen. These early maps were difficult to interpret and use because the line density remained fairly coarse. The early method has recently been explored again where a very fine line density recorded in the field has been applied in the data collection. Grasmueck and Weger (2002) have applied this older method to map marine sediments; however, their line density of 10 cm or less is rarely applied in archaeological environments. Goodman and Nishimura (1993) and Goodman et al. (1995) refined the GPR time slice method by binning data and applying interpolation procedures to estimate inter-profile locations. Interpolated time slice maps have helped to create more useful images for archaeological applications, particularly at sites where clutter is a problem that masks the continuity of features at a site. Several investigators have emulated the inter-line interpolation method for time slice analysis in radar, also with successful imaging results (e.g., Conyers and Goodman, 1997; Conyers and Cameron, 1998; Neubauer et al., 1999; Kvamme, 2001; Conyers and Connell, 2007; Piro and Goodman, 2008). In this chapter we will briefly introduce the GPR field methods for archaeological investigation and look at a few GPR case histories that involve GPR imaging techniques.