GROUND-TRUTHING 6.5-KHZ SIDE SCAN SONOGRAPHS - WHAT ARE WE REALLY IMAGING

A 1000-km2 area on the distal lobe of Monterey Fan shows a digitate pattern of juxtaposed high and low backscatter on GLORIA side scan sonographs. This area was investigated using stereo photography, high-resolution seismic profiles, and measurements of physical properties of cores to quantitatively evaluate the causes of backscatter from the 6.5-kHz side scan sonar. Stereo photography and bottom video were used to determine that the sediment-water interface typically has a bed roughness less than 10 cm over the entire ground truth area; consequently, bed roughness is not a significant contributor to the sonar backscatter. Vertical-incidence 3.5-kHz profiles reveal that high-backscatter areas allow less penetration and have slightly more relief than low-backscatter areas. Closely spaced measurements of p wave velocity, density, and grain size made on transponder-navigated cores are used to investigate the geoacoustic properties of the sediment with the aid of a numerical model. The model results demonstrate that the sediment-water interface is, in most cases, acoustically transparent to the sonar energy and that most or all of the energy is refracted into the sediment to depths of at least a few meters rather than scattered from the surface. The variability of the dominant lithofacies (sands with minor silty clays or silty clay with minor sands) accounts for a regional correlation with backscatter when viewed over hundreds to thousands of square kilometers, but at the detailed scale of hundreds to a few square meters, we have not yet been able to identify a correlation of backscatter intensity with the angle of incidence, surface lithology, measured acoustic impedance, or attenuation. However, there is a qualitative, but unexpected, correlation with lithostratigraphic variability. In this area, thick (up to 50 cm) sand deposits with thin interbeds of silty clay correlate with lower backscatter than do silty clay deposits with thin interbeds of sand. This suggests that volume inhomogeneities and complex constructive and destructive interferences caused by the subsurface volume inhomogeneities within the top few meters of the sediment ultimately modulate the intensity of backscatter. The ground-truthing effort was further complicated by trying to compare the very large sonar footprint of a pixel (approximately 70,000 m2) relative to the area of an individual core (0.006 m2), bottom photograph (approximately 5 m2), or seismic profile. Although 6.5-kHz sonographs appear easy to interpret in a conventional and simplistic manner, caution should be used when interpreting lithofacies from backscatter intensities.