Detection of resonant knapped lithics using chirp sonar

Sub-bottom profiling (SBP) systems have been used to detect lithics on the seafloor, where their presence has been inferred from a convex pattern of reflections in the water-column portion of the profile, called haystacks. These signals have been attributed to the resonant properties of human-knapped lithics, but the physical mechanism responsible for their unexpected appearance has not been explained. We investigate the origin of these signals through controlled experiments and modeling. In experiments conducted over lithic assemblages in an anechoic pool, we demonstrate that flaked lithics on the seafloor generate water-column signals in chirp SBP data, with the strongest response coming from concentrations of smaller flakes, potentially due to their greater numbers of resonators with resonance frequencies within the chirp bandwidth. We developed an idealized model of a chirp SBP insonifying resonant objects on the seafloor which reveals that the haystack is a processing artifact. Together, experimental and modeling results confirm that narrowband lithic resonance, combined with standard chirp matched-filter processing, is responsible for the haystack signal. These findings suggest that SBP chirp parameters, particularly bandwidth and frequency range, could be optimized to target lithic resonance, and that alternative processing approaches operating on pre-matched-filter data may offer improved detection capability.