Inversion of ship-noise scalar and vector fields to characterize sediments: A critical review of experimental results (2007–2014)

Abstract

During geophysical and ecosystem surveys in Mediterranean and Atlantic shallow waters, research vessels, fishing boats, and passenger ferries were used as acoustic sources of opportunity to characterize bottom sediments of different types including mud, fluid mud, and sand. The ships sailed a straight course at constant engine speed while a compact vertical array of pressure or pressure-gradient sensors, deployed from a small boat, sampled the generated noise field over a broad frequency range. Different acoustic observables were investigated that do not depend on the knowledge of ship noise characteristics to mitigate the impact of such uncertainty. For sole pressure measurement, extraction and parameterization of striation structures in range-frequency spectrograms at different receiver depths allow to determine effectively compression wave speed and thickness of a sediment layer as compared to a reference. For additional pressure gradient measurement, vertical impedances are estimated at different depths whose range and frequency dependence are highly sensitive to bottom properties and, in particular, density. Global optimization via genetic algorithm and sequential Bayesian estimation via particle filtering maximize the degree of similarity between predicted and measured impedance data. The paper will critically review the experimental results and compare them with ground truth data. [Work supported by ONR, FNRS-CNPq, WBI-CAPES, PREFACE project EC DG Env FP7.]