Archaeological Geophysics III Posters
Presiding: M E Everett, Department of Geology and Geophysics, Texas A&M University; J A Aitken, Dept. of Geology and Geophysics, University of Calgary
NS41A-01 0830h
Archaeological Feedback as a Research Methodology in Near-Surface Geophysics
A unique characteristic of archaeological geophysics is to present the researchers in applied geophysics with the opportunity to verify their interpretation of geophysical data through the direct observation of often extremely detailed excavations. This is usually known as archaeological feedback. Archaeological materials have been slowly buried over periods ranging from several hundreds to several thousands of years, undergoing natural sedimentary and soil-forming processes. Once excavated, archaeological features therefore constitute more realistic test subjects than the targets artifically buried in common geophysical test sites. We are presenting the outcome of several such verification tests aimed at clarifying issues in geometry and spatial resolution of ground penetrating radar (GPR) images. On the site of a Roman villa in SE Portugal 500 Mhz GPR images are shown to depict very accurately the position and geometry of partially excavated remains. In the Maya city of Palenque, Mexico, 900 Mhz data allows the depth of tombs and natural cavities to be determined with cm accuracy. The predicted lateral extent of the cavities is more difficult to match with the reality due to the cluttering caused by high frequency. In the rainforest of Western Africa, 500 MHz GPR was used to prospect for stone tool sites. When very careful positioning and high density data sampling is achieved, stones can be accurately located and retrieved at depths exceeding 1 m with maximum positioning errors of 12cm horizontally and 2 cm vertically. In more difficult data collection conditions however, errors in positioning are shown to actually largely exceed the predictions based on quantitative theoretical resolution considerations. Geophysics has long been recognized as a powerful tool for prospecting and characterizing archaeological sites. Reciprocally, these results show that archaeology is an unparalleled test environment for the assesment and development of high resolution geophysical methods.
NS41A-02 0830h
Preliminary Results of GPR Surveys in two Mesoamerican Archaeological Sites: Ixcaquixtla, Puebla and El Opeño, Michoacán, México
We present the results of Ground Penetrating Radar surveys conducted during two field seasons in 2002 and 2004 on the archaeological sites of San Juan Ixcaquixtla, Puebla, in Central Mexico and El Opeño, Michoacán, in the western part of the country. In both sites a SIR-2 system was used with 300 MHz and 900 MHz fixed antennas. Radan software was used for data processing with 3D QuickDraw and Interactive 3D modules. The first site corresponds to the Classic Period and is located in a carbonate environment with caliche. The second site is from the Early Formative Period and is found in volcanic tuffs. In both cases the main objective was the detection and recognition of buried archaeological remains, particularly tombs. Data processing including spatial 2D filtering, and the display of three-dimensional data volumes and time slices allowed us to identify two major anomalies in each of the sites that could correspond to tombs. These preliminary results will be verified when archaeological excavations are conducted.
NS41A-03 0830h
GPR Experience at Mesoamerican Archaeological Sites in Honduras: Puerto Escondido and Copan
A series of preliminary GPR surveys were conducted during the summer of 2000 at an Early-Middle Formative Mesoamerican archeological site at Puerto Escondido near San Pedro Sula, Honduras, and at the Classic Mayan site of Copan. Both studies were carried out using a Sensors and Software Pulse EKKO 100, equipped with 50, 100 and 200 MHz antennae. Most of the data were collected at 200 Hz on both reconnaissance profiles and 2D grids. Soils at the Puerto Escondido site were dominantly fluvial sediments associated with the Sula River. Likewise, most of the surveys at Copan were probing fluvial deposits of the Copan River. Penetration was generally limited at both sites to about 1 meter. The terrain at the Puerto Escondido site was generally flat and open, with no tree cover. Surveys at the Copan site were largely beneath tree cover with attendant problems of canopy backscatter. Three-dimensional surveys from the Puerto Escondido site were successful in detecting a stone wall and firepit that were confirmed by subsequent archaeological excavations, as well as similar features that have not yet been excavated. Surveys at Copan sampled a variety of landscape features, including the top of a major monumental structure. Interpretation of the 2D data was seriously compromised by what is interpreted as sidescatter from overhead tree canopies and lateral root systems. Three-dimensional data, however, were successful in delineating the internal structure of a road system (sacbe) linking Copan to the nearby satellite center of Sepulturas. Together these results indicate that GPR can be useful in delineating near surface archeological features in the fluvial sediments that dominate at these Honduras sites, but that 3D coverage is needed to discriminate buried archeological features in the presence of dense forest cover.
NS41A-04 0830h
GPR Surveys at Los Naranjos, an Early Formative Mesoamerican Site in Central Honduras
Extensive GPR surveys were carried out in the summer of 2003 at the Early Formative Mesoamerican archeological site of Los Naranjos, on the shore of Lake Yojoa, in central Honduras. Sensors and Software's Pulse EKKO 100, equipped with 50, 100 and 200 MHz antennae and a Noggin 250 MHz were both used to collect 2D grids and linear profiles. The surveys were located on lacustrine and volcanic soils. Penetration was generally excellent even at the higher frequencies (eg. 200-250 MHz), and clutter problems from nearby canopy were minimal. Features were routinely mapped at 2-5 meters depth. The surveys covered both apparent archeological and natural features. Three-dimensional volume images reveal clear deformation of the youngest strata, including possible fault offsets that may mark paleoseismic events. Reflection amplitudes within the strata were highly variable; local radar bright spots remain to be explained. Numerous individual diffractors were mapped within the geological strata, especially at shallow depths. Some of these are likely to be generated from natural rockfall; however, some may well mark archeological artifacts or features. Identification remains to be provided by excavation. GPR surveys over the major monumental structure at this site encountered much more limited penetration than through the surrounding geological strata. This suggests the use of radar absorptive building materials or coatings. A GPR transect over a topographic high that appeared to be a manmade wall of volcanic rocks indicates that it is a natural rather than an artificial feature. In contrast, GPR profiles near a large isolated rock outcrop indicate that it is not in place, but was transported to its current location. GPR surveys at Lake Yojoa proper demonstrated excellent penetration, suggesting that GPR should be a useful tool in delineating quaternary stratigraphy, and perhaps paleoclimatology, recorded by this central water body. Future geological dating of the strata should help constrain the age of deformation of both natural and artificial markers associated with this early Mesoamerican settlement.
NS41A-05 0830h
Geoprospection of Mound A, Etowah Mounds State Park, Georgia
Mound A, located at Etowah Mounds State Park, Georgia, was the subject of a multi-sensor geoprospection study in 2001-02. Mound A, a late prehistoric mound, built by the Mississippian Culture, ca. 1250 - 1400 AD,is, due to its size, ~ 1 ha in area at the base and 20 m in height, a formidable subject for the use of shallow geoprospection techniques. Techniques used were ground radar (GPR), conductivity (EM) and electrical (resistivity) methods. Common Mid-Point (CMP) radar data produced detail on the mound interior from surface to base. Electrical pesudo-sections produced excellent detail of the mound's interior. The EM data appears relevant for only the upper half of the mound, perhaps to a depth of 10 m, and suggests significant heterogeneity in the sediment fill used in the mound's construction. Our results speak directly to the efficacy of shallow geophyscial techniques in exploring large archeological mounds and tells. Another important aspect of this study is the use of a geoprospection approach as a non-invasice methodology for characterizing culturally sensitive archaeological sites.
NS41A-06 0830h
Recent Geophysical Investigations in Iceland
Interest in the archaeology of the North Atlantic region has grown considerably over the last two decades but until very recently the application of geophysical methods had remained limited to a few isolated investigations. This poster presents results from a program of research to assess the potential of established archaeological prospection techniques in Iceland: an area that presents new archaeological contexts and natural environmental conditions. Throughout many parts of the world magnetic survey methods are the most widely employed geophysical techniques for the location and characterization of archaeological sites. However, they are often considered to be of little value over igneous bedrocks due to their intense thermoremanent magnetization. This poses a challenge in Iceland where 90% of the geology is igneous. In addition to basaltic lavas there are numerous glacial deposits, sandur plains and volcanic tephra layers. Electrical methods have long been used in archaeological investigations to accurately plan structural remains, however such features also present a problem in Iceland. The lack of good building stone and timber resulted in turf being the dominant building material from the Viking settlement in about 871AD until the beginning of the 20th century. High-resolution surveys have been undertaken at 40 sites across Iceland using the Geoscan Research FM36 fluxgate gradiometer (0.5m x 0.25m) and RM15 earth resistance meter (0.5m x 0.5m). The results presented here demonstrate that whilst the intense geological anomalies often dominate the magnetometer data, overwhelming more subtle anomalies due to pits and ditches, it is possible to identify areas of activity as clusters of magnetic bipoles due to near-surface rocks from archaeological contexts. Furthermore, turf remains and compacted floor layers have also been successfully identified in earth resistance surveys. Combining the results from these two complementary methods aid in the distinction between natural and anthropogenic sources and often allows accurate identification of archaeological features. Despite the limiting factors, this work demonstrates the potential of these techniques for archaeological prospection in Iceland and throughout the North Atlantic, especially when undertaken as part of an integrated approach.