Dig Diary: Scanning for Clovis

by Dr. Meg Watters

The survey techniques we use in geophysics follow fundamental rules in relation to the physical properties of the earth and objects buried with in it. When we're preparing to conduct geophysical surveys for any archaeological site, we first ask a lot of questions about the conditions at the site. This helps us determine what geophysical techniques and parameters would be most appropriate to use for effective data collection. But sometimes, what you think you know when you first arrive at a site can change significantly once you start investigating and collecting data.

The geophysics team's job at the Topper site was to use geophysical surveys to try to map evidence of the Clovis to help researchers identify potential areas for excavation. Our initial reaction was that it could not be done. The information we had before we arrived led me to believe that the only thing we should be able to map would be changes in soil type that may relate to the Clovis occupation rather than specific objects.

However, once on site, we spoke with site director Al Goodyear and learned that the survey area was clean sand from the ground surface to the known Clovis layer and that there was a 10 cm thick "floor" of Clovis artifacts at approximately 0.75 m depth, also described as a "carpet of debris."

Based on this information, we thought we might be able to use ground penetrating radar (GPR) to pick up the thick "floor" of Clovis artifacts, thinking it would appear similar to band of gravel across the site. We used the GPR gear we brought with us, the 400 MHz antenna, and collected 1 scan per centimeter along transects spaced every 20 centimeters. We felt this should pick up a change in stratigraphy at the given depth.

Once data collection started, I looked into some of the excavation units more closely. It turns out the "floor" or "carpet of debris" was not what I envisioned. Instead, the Clovis artifacts were scattered across the area, not in a single solid layer, but randomly spaced. These small, widely-spaced objects are not the kind of targets that I would have dreamed could be mapped with GPR, which is normally used to identify larger features such as stone walls or buried cellars.

To add to this surprise, the "clean" sand from the ground surface down to the "floor" of artifacts was, in fact, chock full of tree roots from the dense woods through which the firebreak is cut.

So, in the end, our initial perception of a gravel layer in clean sand was in no way the reality of the site. Unfortunately, this happens from time to time with the type of work that we do.

Bryan and I re-grouped and decided to take advantage of the opportunity of being on a known Clovis site to see if we could actually map traces of the occupation with GPR. With two days left, we changed tactics and had a 900 MHz antenna shipped to us overnight. This higher frequency antenna focuses on detecting smaller targets closer to the surface. The next day, we returned to the site with a new outlook. No one has ever mapped anything like this before. Given that we are on a known Clovis site with artifacts at a known depth, our precise data collection over the next two days would attempt to develop a method for mapping the remains of the Clovis occupation.

In the end, although neither of us would stake our lives on it, we may have mapped traces of the Clovis layer, seen as a faint interface in the northeastern corner of the survey area. Within this interface are a number of single point reflectors that, when carefully examined, do not appear to be tree roots. Compared to the trenches on either side of the GPR survey area, these point reflectors may very well be Clovis artifacts. However, proving this through excavation is another story and would take a combined effort on the part of the archaeologists and geophysics team to confirm our results.

Visit the Incident Room to take a closer look at the GPR data we collected at Topper and learn more about our analysis.