Date of Award

Fall 2025

Project Type

Thesis

Program or Major

Ocean Engineering

Degree Name

Master of Science

First Advisor

Gabriel R Venegas

Second Advisor

Jenna Hare

Third Advisor

Anthony P Lyons

Abstract

This thesis describes the development of a geoacoustic inversion technique to estimate seafloor sediment characteristics that is designed for use with existing fisheries survey single beam echosounder (SBES) datasets. The geoacoustic inversion technique consists of a forward model and inversion algorithm. The forward model is a high frequency, physics-based, near normal incidence acoustic backscatter model enhanced with a density depth gradient model and geoacoustic sediment regressions. The inversion algorithm is a parallel tempered Markov-chain Monte Carlo Bayesian inversion. The geoacoustic inversion produces posterior probability distributions for mean grain size, density, porosity, sound speed, seafloor roughness characterization, and sediment volume scattering description.

Two datasets are used in this study: a calibrated multi-frequency EK60 scientific SBES dataset collected by the National Oceanographic and Atmospheric Administration and a seafloor sediment sample dataset collected by the United States Geological Survey and the University of New Hampshire. The combination of these two datasets provides thousands of geographically co-located pairs of acoustic and sediment data. The geoacoustic inversion uses the 18, 38, and 120 kHz frequency data and is validated by comparison to sediment sample mean grain size.

The inverted mean grain size results are shown to agree with ground truth samples for sandy sediments, while predicting a finer mean grain size than ground truth samples for muddy sediments. The geoacoustic inversion presented in this thesis is the foundational work for inversion application to fisheries survey calibrated SBES datasets for mass seafloor characterization with existing data.

Download

Share

COinS