Evaluating ground-penetrating radar antenna performance for investigating Mississippian mound construction compared with data from solid-earth cores and magnetometry

Mississippian mounds in the south-eastern and mid-continental United States often contain multiple construction stages and a diverse array of features. Investigating mound construction with excavation is expensive, logistically challenging, and sometimes politically impossible. Three mounds at the Angel site (12VG1), a Mississippian village (ca ad 1100-1450) in the lower Ohio River Valley, were investigated with non-invasive geophysical and minimally invasive geoarchaeological methods. We compare the efficacy of two different ground-penetrating radar (GPR) antenna frequencies (250 and 500 MHz) for detecting elements of mound construction previously identified with solid-earth cores, magnetometry, and electrical resistance tomography. Solid-earth cores provide stratigraphic information about mound layers and construction soils and materials with low horizontal spatial resolution. Surface magnetometry provides information on shallow features, such as buried building walls and floors at high resolution, but little information about depth. Radar profiles taken with 500 MHz and 250 MHz antennae were evaluated to determine which antenna frequencies are most suitable for detecting features and strata that had been documented using the other methods. Utilizing the solid-earth cores for velocity correlations, the 500 MHz antenna provided the most useful information about prehistoric structures near the mound surface, but was limited by its relatively shallow penetration depth. The 250 MHz antenna was not able to resolve shallow features as clearly as the 500 MHz antenna. However, it was able to detect possible interior mound platforms and features, although without providing many details about possible buried structures. In combination, GPR, solid-earth cores, and magnetometry provided new information about the construction of mounds at the Angel site.