Seafloor characterization from inversion of high-frequency backscatter

Abstract

Geoacoustic inversion of high-frequency backscatter data is presented for characterization of the seafloor. The APL-UW (Applied Physics Laboratory at the University of Washington) backscattering model is used to model the grazing-angle dependence of the backscattering strength for a number of seafloor parameters. Backscattering strength versus grazing angle data sets (with frequencies 12–35 kHz) are provided for three sites along with sediment ground truth. Roughness measurements are also available for two of the sites. Inversion of these data sets is performed via simulated annealing, with some of the parameters constrained by empirical relationships with mean grain size. From the inversion, estimates are obtained for mean grain size, roughness, and volume interation. Inversion of data from a smooth, silty site in Arafura Sea yields estimates in good agreement with ground truth. Results also compare well with ground truth in a rough, sandy site (Quinault) and in Onslow Bay (sand). [Work supported by SPAWAR.]