LESSONS LEARNED: APPLYING GPR TO 17TH AND 18TH CENTURY HISTORICAL ARCHAEOLOGICAL SITES IN COASTAL NEW ENGLAND

Applying GPR to early colonial archaeological sites (1607-1690) requires knowledge of both the Holocene geology and bedrock geology of the region as well as early settlement building techniques. This presentation reports on the results of collecting GPR data at these sites. Many early sites are built on low headlands, with a shallow bedrock interface covered by .5-4 meters of well sorted sand overlying marine clays. Initial building construction may not have used foundation stones, probably had a shallow sill trench, may not have had a cellar hole, could have a palisade trench, and probably had a packed clay floor. An original wattle and daub chimney may be replaced by a stone or brick chimney. Many structures were destroyed by fire during various frontier wars. These conditions create subtle features not easily detected and the records tend to need extensive postprocessing to identify valid targets.

The GPR is site calibrated using a 2cm hand corer to develop site stratigraphy, and moisture content is estimated with a simple agricultural probe. If possible, depth to bedrock and major lithologic boundaries are calculated with a quick seismic refraction survey. Data acquisition is along tapes and preexisting archaeological grids, therefore GPS is only used to georeference grid corners, not during data collection. Gain points are set with the first at the base of the plowzone. Post processing includes removal of surface and cross-talk effects and range gain in the upper 20ns of the record. Migrating the data rarely improves record clarity. Display and data (esp. Hilbert) transforms are effective when searching for graves, wells, privies and small cellar holes such as root cellars. Brick and stone chimneys, provide excellent datum targets for a site. More primitive wattle and daub chimneys are recognizable by a clay hearth. Clay floors present a good anomalous horizon with a break indicating the location of the cellar hole. Foundation walls present special problems with upper layer stones disturbed by plowing. The resulting shallow rock scatter, if dense enough, can obscure the location of the lower wall. The best all around antenna for these sites is medium frequency (350-500 mhz). Analysis beneath the bottom of archaeological test pits provides useful data with higher frequency antennae (900 mhz - 1.5 ghz).