Measurements, Metrics, and Modeling of Normal-Incidence Acoustic Interaction With Ocean Sediment

Abstract

Normal-incidence echosounder signals are attractive for sediment characterization since they can be done easily and remotely over a large area. However, in addition to being sensitive to sediment type, these measurements are also sensitive to other factors such as layers, scattering, and gradients, which make sediment characterization much more difficult. In this study, a well-characterized set of normal-incidence bottom loss measurements are compared with a finite element model using measured environmental parameters such as sound speed, attenuation, and interface roughness. The acoustic measurements were taken in the midfrequency range (5–10 kHz) during the Target and Reverberation Experiment near the Florida panhandle in May 2013. To make the comparison, two sets of metrics were considered. The first was based on the statistical moments of the acoustic signal, a method commonly used for sediment characterization. The second set of metrics was based on features of the return identified by physical mechanisms. It was determined that the physically derived metrics were more robust for model-data comparison. The three physically derived metrics used were bottom loss, full-width-at-half-maximum (FWHM), and the ratio of the half-width-at-half-maximum (HWHM) before the peak to the HWHM after the peak (i.e., HWHM ratio). The finite-element model allowed for effects such as layering, gradients, and interface roughness to be disambiguated from the bulk sediment parameters, thus identifying the influence of these factors on the normal-incidence echo return. Considering all measured environmental and sediment properties, the model could not account for all of the bottom loss without a substantially lower value for the density than measured. Thus, the sediment was modeled using an effective density, a proxy for poroelastic effects. Although the low density model was consistent for most of the data for all three metrics, several areas of outliers in both FWHM and HWHM ratio indicated that other mechanisms such as volume scattering may be important in some areas.