Fast-marching methods to model the acoustic responses of biological seafloors

Abstract

Biologically colonized seafloors have been difficult to model for their acoustic interaction. The distributions of geoacoustic properties above and beneath the nominal bottom can be extremely complex. We have implemented workflows in software which: (i) efficiently construct many patterns of growth forms, bioturbation, organism aggregations, and flesh-skeleton morphologies, and (ii) submit them to “Eikonal” wavefront acoustic modeling. Ray-based finite-element and statistical methods can, in theory, be applied but they run into problems of the strongly inhomogeneous geoacoustics, ray-chaos (of several causes), absence of a clear division into the usual rough surface/inhomogeneous volume halfspaces, and heavy computational costs. Wavefront methods, namely, Eikonals implemented via Fast-Marching Algorithms, are able to deal with complex situations—as has been realized lately in deep seismic geophysics, in moving object tracking, and interestingly, in urban settings, for locating gunshots and mapping Wi-fi fields. They also to offer a new way to understand the acoustic backscatter results from multibeam, sidescan and single beam sonar systems over complex biologic seafloor types. This direction of research requires close knowledge of both the “bio” and the “geo” of seafloors.