See the summary of all GPR data collected at both sites with collection parameters:

Overall, a total of 24 grids were collected at the CATS and Hammer sites. Over 280 maps were created.

In summary, the parameters that seemed to provide the best results based on what was known in the ground, regardless of the environmental conditions, include:

In Field Collection:

-Using the smallest distance between transects in a grid possible.

400 MHz: 50 cm spacing was good, 25 cm spacing was better.

900 MHz: 25 cm spacing was optimal.

In Lab Processing:

-Processing data to remove background noise and hyperbola tails, and collapsing reflections to point sources using migration programs.

-When using imaging program to create maps (these were generated using Surfer 8), interpolate and smooth data.

-The program Surfer 8 proved better than Idrisi (a GIS-based program).

-Creating amplitude slice maps no thinner than 2 nanoseconds. Slicing maps thinner cuts waveforms, which results in very poor maps.

General antenna frequency results:

-Using a higher frequency antenna (such as 900 MHz) worked best in clay. This provides the best resolution at 10-50 cm in depth.

-900 MHz worked at all depths in dry sand at Hammer.

-400 MHz and 450 MHz antennas worked well at both sites. Optimum resolution occurred at 20-90 cm in depth.

-300 MHz antennas were poor at both sites. It provided poor resolution at shallow depths and is more suitable for penetrating beyond one meter.

Dr. Conyers doing a GPR survey at the site of El Brujo, Peru.

Areas for Future Study:

As with all research, new questions arose that could greatly benefit from future attention. These new areas of future research include:

-What types of clay produce what types of data?

-Comparing data from grids with parallel transects (one orientation) to grids with perpendicular transects (two orientations).

-The wet conditions at CATS (a clay soil) produced less "clutter" than wet conditions at Hammer (a sandy soil). This is counterintuitive, and deserves further study.