Salvatore Piro

Discovery of a 1st Century AD Roman Amphitheater and Other Structures at the Forum Novum by GPR

A Roman marketplace and town called the Forum Novum, initially began construction in the 1st century BC and flourished well into the 4th century AD. At present most of the town remains is below ground. The site is situated next to a completely restored 1st century AD church and a partially reconstructed marketplace. Ground penetrating radar surveys have been conducted at the Forum Novum during 1998 through 2001 field seasons to remotely detect buildings of this ancient town. Using 300 and 500MHz radar antennas and closely spaced radar profiles, time slice and 3-D volume analysis of the reflection data were computed. GPR surveys in the area detected a large oval-shaped structure 45meter along its major axis. This structure is from a 1st century AD Roman amphitheater. Subsequent excavations from one of the eight entrances reveal that the internal amphitheater oval is nearly a meter thick and is composed of a cemented stone wall buried 60cm below the ground surface. The time slices revealed completely differ...

ADVANCES IN GPR IMAGING WITH MULTI-CHANNEL RADAR SYSTEMS FROM ENGINEERING TO ARCHAEOLOGY

Advances in ground penetrating radar imaging with multi-channel systems have greatly improved the speed and areal coverage of the ground. Along with improved imaging software, datasets recorded with multi-channel systems can be processed at similar speeds to coarsely spaced single channel data that would normally require additional time for interpolation processes to fill in the gaps between lines. With the cross-line spacing approaching a 1/4 wavelength of the transmitted microwaves into the ground, multi-channel systems have the advantage of complete coverage of a site with no need for interpolation in most cases except to fill in the gaps between adjacent tracks if so desired. Multi-channel systems do require additional RSP (radagram signal processes) in order to balance the channels and to condition the data prior to imaging. Spectral whitening and several other RSP methods are shown with their application to imaging of sites from bridge decks for deterioration to the discovery of subsurface archaeological remains. Data processed from several different multi-channel GPR systems are shown.

Forum Novum–Vescovio: Studying urbanism in the Tiber valley

The Roman town of Forum Novum lies in the Sabine hills to the northeast of Rome. Its study forms part of the British School at Romes Tiber Valley Project, a collaborative research initiative which studies the Tiber valley as the hinterland of Rome, tracing the impact of Romes development on the history of its settlement, economy, and cultural identity from 1000 B.C. to A.D. 1300 (Patterson and Millett 1999; Patterson et al. 2000) (fig. 1). The project draws on the extensive work carried out in this area to produce a new, material-based history of the valley. While the project seeks to re-evaluate past survey material, a vital contrast is provided by the development of new field projects to fill the gaps in settlement knowledge. Three main lacunae have been identified: the study of urban centres; the dearth of data from the E bank of the Tiber; and the poor understanding of the late-antique and early Mediaeval landscape. Forum Novum offers an opportunity to address all these lacunae. Urbanism forms a key research theme for the Tiber Valley Project. In marked contrast to the intensity of archaeological work on rural settlement in this area, there has been little systematic research on towns. Study has tended to concentrate on the excavation of monumental structures or, more rarely, the investigation of single and exceptional towns such as Ostia and Rome itself. Surprisingly little is known of the organization of the smaller towns and knowledge of their history is based largely on the epigraphic and documentary evidence.

GPR Image Construction and Image Processing

Often the most valuable information contained in GPR radargrams is not from examination of individual radargrams and their vertical profile of the ground, but from the generation of images which connect anomalies from closely spaced profiles. Images that look across horizontal slices in the ground and changes in recorded reflection amplitudes are referred to as time slices (Goodman et al. 1995; Malagodi et al. 1996). Time slicing is an essential image process which can often contain the most important information on the subsurface archaeology at a site. They are called time slices since the vertical axis of the radargrams – before they are converted to depth with knowledge of the ground microwave velocities – is the travel time of reflections. In 1994, the authors had seen unpublished reports with images and photos that date back from as early as 1981 of a digital GPR system at Battelle National Laboratories in the US, with the antenna and electronic equipment being deployed by a tractor. In these reports, there are also the first versions of a time slice maps made at an archaeological site and the authors believe that these are probably the earliest time slices ever generated from GPR (?).

GPR Imaging on Historical Buildings and Structures

GPR technology is capable of providing information on fractures, previous reconstructions, material integrity and a variety of the characteristics of building materials. For this reason GPR has found a multitude of applications in studying the integrity of historical buildings(Barone et al. 2010; Cosentino et al. 2011; Kadioglu et al. 2010; Perez-Gracia et al. 2009; Sambuelli et al. 2011). GPR surveying can effectively identify critical structural supports such as rebar and tension ties in walls that help to constrain the horizontal load forces (Bartroli et al. 2011; Ranalli et al. 2004). After earthquakes or landslides the changes and hidden damage to structures can be detected using radar (Orlando 2007). In the case when GPR is used to monitor a building, recordings at time 1 and time 2 can be differenced to indicate only those areas which have undergone changes since the last survey. GPR imaging for evaluating building safety are common enough in 2012 that most geotechnical companies advertise this service on their websites.

Understanding GPR via a Simulator

GPR radargrams often have no resemblance to the subsurface structures over which the profile was recorded. Various factors including the innate design of the survey equipment and the complexity of electromagnetic propagation in the ground can disguise complex near surface earth structures recorded on GPR reflection profiles. A very useful way to understand the nature and content of GPR radargrams is to understand what components are needed to develop a basic simulator. The simulator would provide a means to predict radargrams made across candidate models of the ground and help to explain how buried target structures get translated into recorded reflection profiles. The simulator can help to better comprehend the limitations of this remote sensing method, but also to highlight the capabilities of GPR and where this exploration tool and in what subsurface environments that it can and cannot function optimally.

Integrated GPR and archaeological investigations to study the site of Aquinum (Frosinone Italy)

To enhance the knowledge finalised to the location and conservation of the unknown buried structures below the actual studied levels, in the territory of the Ancient Aquinum (Frosinone, Italy), an integrated archaeological and ground remote sensing study has been developed during 2008–2009 and it is still in progress. Analysis of the historical and oblique aerial photographs, combined with topographical and archaeological field-walking surveys, allowed the preliminary interpretation of the main town-planning of Aquinum. To verify this preliminary interpretation an extensive geophysical surveys, employing Ground Penetrating Radar (GPR) method, has been made during 2008–2009. The obtained results indicate the good matching between the interpretation of aerial photographs and GPR images at different depths. The location, depth and size of the individuated archaeological structures were effectively estimated. Archaeological excavations made during the summer 2009, in a southern portion of the investigated area, have confirmed the results obtained with GPR method.

Basic GPR Signal Processing

Most GPR data analysis requires some conditioning of the raw recorded pulses before construction of images from closely spaced profiles is implemented. Radargram signal processing (RSP) involves using specialized filters to enhance and adjust the digitized reflections, as well as to remove noises contained in the raw recorded radar pulses. RSP can be applied to the radar pulses prior to constructing images from the radargrams. What signal processes are needed will depend on a variety of factors observed in the raw recorded radargrams. There are a variety of RSP that are essential and some are only used if certain noises are found to exist in the data. The basic RSP often used in GPR surveys are discussed:

North America: GPR Surveying at Historic Cemeteries

GPR remote sensing has found one its most successful archaeological applications in the discovery and mapping of historic burial sites in North America (Doolittle et al. 2010; Bevan 1991; Conyers 2006; King et al. 1993). One of the first applications of GPR in cemeteries for discovering unmarked graves were initiated nearly four decades ago by Bevan and Kenyon 1975. The primary motivation for GPR surveys has often been to assist in the mitigation of sites by helping to detect the presence of unmarked burials. Cemeteries have come under constant pressure from development projects and knowing the areal extent of cemeteries in prime land areas has been crucial. Many cemeteries which are not under developmental pressures may have been abandoned or lacked funds to maintain the sites. Often these cemeteries are missing headstones or site markers after being vandalized or from just plain neglect over time. There are new and growing cemetery management efforts to find unmarked burials so that vacant spaces in these cemeteries can be sold as new burial plots.

Imaging Over Sites with Topography and Vector Imaging

Correcting the GPR dataset for topography is often necessary to properly image and detect the subsurface structures beneath a site. For instance, archaeological sites that were once built on level ground but have since been subject to variable deposition and thus have variable surface topography, may need to have time/depth slice images created at level elevation across the data to properly reveal the subsurface structures. Goodman et al. (1995) made level plane depth slices across the Spiro House mound sites in Oklahoma. The level plane slices revealed square pithouse floor foundations from Native American Indians which only show up on (horizontal) slices made from topography adjusted radargrams. Standard topographic adjustments to the GPR involve assuming a velocity model for the radargrams, and then shifting the scans vertically to account for the change in topography (Tanaka et al. 2009). Some problems with this method, such as the “peeling effect” caused by differential slicing across the strong ground wave reflection can occur. An effective method to completely remove the peeling effect without having to apply drastic measures such as background filtering to remove the groundwave is possible. The method involves generating a volume from time slices that are independently normalized at each level and then graphically warping these volumes into their topography. Topographically adjusted volumes can then be sliced level plane without any influence of strong ground wave reflection. The details of this method are presented in the section on Level Plane Time Slices in this chapter.