Figure 1: Metal Detecting and Mapping Finds at Patriot Park North Site.

There was a slave quarters at one time where Patriot Park North now sits. Fairfax County Park Authority archaeologists found some of its features the same way some beach walkers find things – with a metal detector.

During recent excavations, archaeologists identified the site where enslaved persons lived from approximately 1760 to 1820. They discovered features indicating there was a cellar, a storage pit and something else they are still trying to characterize. There were gaps where relatively few artifacts were recovered during a process called tight interval shovel testing or unit excavation. These methods generally are very effective at finding larger, domestic structures where there are substantial amounts of material left behind by the people who lived there. However, one goal of this investigation was to learn more about the built landscape, including all the man-made structures within the site and secondary structures, such as a smokehouse or corncrib. Those usually leave a relatively light archaeological signature and can be hard to find with standard practices. So out came the metal detectors.

Archaeologists have long used metal detection when working on certain sites, such as military sites. Other archaeological programs, including those at James Madison’s Montpelier, have proven that metal detection helps identify secondary structures at domestic sites. With the help of Montpelier-trained volunteers, Park Authority archaeologists went metal-detecting in two areas where little had been found. The areas were divided into 10-meter square blocks. Within each block, metal detectorists would walk rows spaced approximately two meters apart. When all the rows in a block were scanned in one direction, it would be investigated again in perpendicular rows.

Metal detector hits were marked with non-metallic flags. Each hit was then excavated, and anything found was bagged according to its location. The location of an

Figure 2: Bagging Finds.

archaeological find is critical to learning its use, and moving a find greatly diminishes its historic value and the lessons it can teach us. When multiple finds at Patriot Park North were identified within a half-meter of each other, they were placed in the same bag. FCPA archaeologists then mapped the location of the discoveries using surveyor tools, logged the bags, and collected them for transport and analysis in the lab.

Once the metal detection finds were cataloged, archaeologists needed a meaningful way to interpret the data. They turned to a Geographic Information System project. Because they were looking for structures, they only considered nails in their analyses. They started with methods commonly used (Kriging and Inverse Distance Weighted (IDW) analyses) to create nail distribution maps. The system maps areas according to the number of artifacts found, and then it projects expected results across areas from which there are no data. Of the two methods, the IDW analysis is more visually appealing and seemingly more useful (Figure 3). However, staff immediately discovered that these tools

Figure 3: Inverse Distance Weight (IDW) Analysis of Nails Found During Metal Detection.

didn’t answer the questions they were asking. Even lumping artifacts found within a half-meter of each other, the maximum number of nails recovered from any single location was four. FCPA archaeologists wanted to know if there were any significant clusters of nails in the area.

Onwards to another analysis, Nearest Neighbor 5 (NN5) function within the Optimized Hot Spot. This approach considers not only the number of artifacts (nails) recovered during metal detection but also the proximity of hits to each other. Using both of these factors, the analysis looks for meaningful clusters. This analysis revealed one cluster west of the storage pit features with another, less intense cluster northwest of the main cluster and, possibly, another minor cluster east of a cellar feature (Figure 4). It is possible that these clusters represent the locations of support structures.

Figure 4: Nearest Neighbor 5 (NN5) Analysis of Nails Found During Metal Detection. (Note: The number generated by NN5 analysis reflects the strength of relationship rather than the number of finds.)

The next step will be to look at the artifacts recovered from excavations near the clusters. These data may help archaeologists pinpoint how certain structures were used. Hopefully, this will lead to better reconstruction of the cultural landscape and

Author Christopher Sperling is the Senior Archaeologist with the Fairfax County Park Authority.