A multisensor geophysical investigation of portions of the Pine Tree Mound site (41HS15) located in Harrison County, Texas, was conducted prior to large-scale block excavations there. The investigation included electrical resistance, magnetic field gradient, and ground-penetrating radar surveys. The geophysical survey and subsequent excavations were undertaken in an effort to mitigate the effects of coal mining operations on the site. This article presents the results of our geophysical investigations and describes an online data analysis tool created with the intention of simplifying interpretation of these complex multivariate data and making these data available to a broad audience. The article also compares the geophysical signal response with excavation results from the site. Sources of uncertainty in both data sets are discussed, and possible reasons for observed incongruities between the geophysical and excavation data are explored. We end by making recommendations for innovative ground truthing methods that might allow us to more definitively interpret geophysical data in the future.
Early in 2006 Archaeo-Physics, LLC was contracted by Prewitt and Associates, Inc. to conduct a geophysical investigation at the Pine Tree Mound site (41HS15), a Caddo archaeological site located in Harrison County, Texas. Geophysical methods were used to evaluate two areas of the site that were slated for archaeological data recovery prior to portions of the site being destroyed by coal (lignite) mining operations. The objective of the investigation was to identify geophysical anomalies and anomalous patterning indicative of the presence of archaeological features and /or landscape modifications associated with the Caddo occupation.
Multiple geophysical methods were employed during the investigation, including magnetic field grathent, electrical resistivity, and ground-penetrating radar (GPR). Geophysical fieldwork was conducted between February 10-14, 2006. Parts of the Pine Tree Mound site were subsequently excavated during 2006 by Prewitt and Associates, Inc. Because each of the geophysical methods applied during this investigation responds to contrasts in different material properties, each individual method provides only a partial or limited understanding of the subsurface archaeological record. The view provided by each method is limited further when additional issues inherent to most real world geophysical survey situations are considered. For example, signal noise caused by system electronics, environmental conditions, and operator error can limit the visibility of buried archaeological resources. Signal clutter is often an even more significant source of uncertainty. It can be caused by modern trash (e.g., metal pin flags), modern disturbance (e.g., plowing, terracing, trenching), natural horizontal or vertical heterogeneity in the site's soils, and the presence of natural inclusions within the soil (e.g., stones with significant magnetic remanence). These three issues - limited signal contrast, signal noise, and signal clutter - often result in significant uncertainties when interpreting geophysical data from pre-contact Native American sites.
In a similar manner, there is often considerable uncertainty associated with the excavation and mapping of archaeological features. This is especially true in salvage archaeology situations where mechanical methods are utilized to explore large areas and time and budget constraints must be considered. Although many buried archaeological features are highly visible and easy to interpret, we believe most field archaeologists would agree that features often grade into ambiguity, and the mapping and recording of such features can be very subjective and challenging. With these many sources of uncertainty in mind, the archaeological record at a large and complex site such as the Pine Tree Mound site is best understood by carefully comparing all available sources of data. …