Geoacoustic Inversion Research Articles

Source localization at low frequency using deconvolution of the waveguide impulse response

Matched-mode processing (MMP) methods have been widely investigated for source localization and geoacoustic inversion. For horizontal line arrays (HLA), it aims at retrieving the horizontal wavenumbers and thus at inferring source range and bearing. In this work, the matched-mode is performed by applying the deconvolution of the waveguide response in a narrowband context in order to improve the localization accuracy. Deconvolution usually refers to a sparse framework when the number of sources is smaller than the number of sensors of the HLA. The Orthogonal Matching Pursuit algorithm is then used in this work. First results from numerical simulations for Pekeris waveguide highlight better localization accuracy than MMP. The deconvolution shows as well stronger robustness to mismatch in the sound speed values of the seabed. Results from measurement campaign, involving large HLA, low to ultra-low frequency and water depth up to 1500 m are investigated as well.

Small grazing angle reflection and the sound siphon effect over a low velocity layer of sediments.

The low-order normal modes with small grazing angles (SGA) often control long-range sound field characteristics in shallow water. The SGA reflection loss from a half-space low-velocity bottom (LVB) is independent of the sound attenuation, except around the angle of complete transmission; the SGA bottom reflection loss (BRL) from a seafloor with a top low-velocity layer is very insensitive to the LVB attenuation also, except around a few selected frequencies. Thus, the "seafloor velocity-attenuation coupling" problem will be more fatal for LVB geo-acoustic inversions. The dispersion equation of the normal modes in the LVB layer is coincidentally the same as the singularity expression of the SGA reflection coefficient in the water column, resulting in a sound siphon effect that causes the abnormally high SGA BRL and transmission loss in the water at the siphon frequencies. The siphon effect is very sensitive to seafloor acoustic parameters, might offer a physical base for geo-acoustic inversion, and show a dim light in a "gray area" for inverting the LVB sound attenuation at low to mid frequencies. As an example, the acoustic siphon effect and related seafloor geophysical parameters forming it in the Yellow Sea are reported in this paper.

Geoacoustic inversion using Bayesian optimization with a Gaussian process surrogate model.

Geoacoustic inversion can be a computationally expensive task in high-dimensional parameter spaces, typically requiring thousands of forward model evaluations to estimate the geoacoustic environment. We demonstrate Bayesian optimization (BO), an efficient global optimization method capable of estimating geoacoustic parameters in seven-dimensional space within 100 evaluations instead of thousands. BO iteratively searches parameter space for the global optimum of an objective function, defined in this study as the Bartlett power. Each step consists of fitting a Gaussian process surrogate model to observed data and then choosing a new point to evaluate using a heuristic acquisition function. The ideal acquisition function balances exploration of the parameter space in regions with high uncertainty with exploitation of high-performing regions. Three acquisition functions are evaluated: upper confidence bound, expected improvement (EI), and logarithmically transformed EI. BO is demonstrated for both simulated and experimental data from a shallow-water environment and rapidly estimates optimal parameters while yielding results comparable to differential evolution optimization.

Integrated methodology for gas content assessment and prediction in shallow muddy lake sediments: acoustic mapping and correlation analysis

This paper provides a step-by-step description of integrated methodology for quantification and prediction of gas (methane, CH4) content dynamics in shallow aquatic sediments under changing spatial and temporal conditions. Presence of gas bubbles even in small concentrations significantly affects sediment compressibility, which in turn decreases sound speed in sediment. Our integrated methodology consists of two basic steps. In the first step, free gas content is evaluated by acoustic applications based on the sound speed inferred from the reflection coefficient from gassy bottom. The experimental bottom reflections are registered and compared to the simulated ones, using a geoacoustic inversion technique. The best match between the model and the experiment provides sediment sound speed estimate, which is converted into free gas content using a basic relation. In the second step, a multivariate linear regression is fitted for gas content and closed form expression of gas content dependence on the following predictors, which change spatially and temporally over the aquatic ecosystem, is obtained: 1) water depth, 2) short-leaving CH4 production rate peaks fueled by punctuated organic matter deposition; and 3) CH4 bubble dissolution rates.•Gas content and sound speed in the sediment are estimated via the geoacoustic inversion technique by matching the experimentally recorded and simulated bottom reflections•Only single source and receiver are required for the acoustic methodology•A multivariate linear regression is fitted for gas content to indicate its dependence on various predictors that change spatially and temporally over the lake

Using machine learning techniques in inverse problems of acoustical oceanography

AbstractThe goal of the work presented here is to study a novel approach for inverting acoustic signals recorded in the marine environment for the estimation of environmental parameters of the water column and/or the seabed. The proposed approach is based on signal feature extraction using a discrete wavelet packet transform, applied to the measured signal, and hidden Markov models that exploit the sequential patterns of the signals. The signal feature is thereafter used in the framework of a mixture density network, which, after training with sets of simulated signals calculated within a predefined search space, provides conditional posterior distributions of the recoverable parameters. The technique is tested with two test cases corresponding to different types of inverse problems. The first case corresponds to a simple problem of geoacoustic inversion, while the second is referred to a, rather unusual, still interesting problem of recovering the shape of a seamount using long‐range acoustic data. Both test cases are based on simulated experiments. The inversion results obtained using the proposed scheme are compared with inversion results using statistical features of the acoustic signal, which is another inversion approach well documented in the literature and is also based on the wavelet packet transform of the measured signal.

Sound attenuation at low to mid frequencies in low velocity seabottoms.

Attenuation is the most difficult seafloor acoustic property to get, particularly at low to mid frequencies. For low velocity bottoms (LVB), it becomes even more challenging, due to its small attenuation and lower velocity (relative to the velocity of the adjacent water). The latter one causes a fatal "seafloor velocity-attenuation couplings" in geo-acoustic inversions. Thus, attenuation inversions for the LVB require an accurate seafloor velocity profile, especially the velocity in the LVB layer. The propagation of explosive sound in the Yellow Sea with a strong thermocline and a top LVB layer exhibits many prominent characteristics: modal dispersion (the ground wave, water wave, Airy phase), two groups of water waves at high frequencies, and the siphon effect which causes abnormally large sound transmission loss at selected frequencies, etc. These observations are used to precisely measure the critical frequency, the Airy frequency, Airy wave velocity, 1st mode group velocity, and to derive the velocities in the LVB layer and in the basement. Using inverted seafloor parameters, the source level-normalized transmission loss and the first mode decay rate in ranges up to 27.66 km, the sound attenuations in the LVB are derived for a frequency range of 13-5000 Hz.

On the Dependence of Acoustic Modes on Media Parameters in the Pekeris-Airy Waveguide

A waveguide consisting of an isospeed water layer overlying a halfspace with a sound speed gradient is considered. The dependence of horizontal wavenumbers of normal modes in such a waveguide on the sound frequency, water depth and geoacoustic parameters of the bottom are investigated. It is shown that these dependencies are described by ordinary differential equations that can be solved numerically at very low computational cost providing a substantial increase in the efficiency of broadband modelling of sound propagation. The explicit formulae for the derivatives of horizontal wavenumbers with respect to the bottom parameters can be also used in dispersion-based geoacoustic inversion methods. This approach can be extended to the case of a waveguide consisting of several layers of Airy-type media. We also propose a new and concise proof that the spectrum of the Sturm-Liouville problem from which the modes are found for this waveguide is purely discrete.

Meso-scale seabed quantification with geoacoustic inversion

Knowledge of sub-seabed geoacoustic properties, for example depth dependent sound speed and porosity, is of importance for a variety of applications. Here, we present a semi-automated geoacoustic inversion method for autonomous underwater vehicle data that objectively adapts model inference to seabed structure. Through parallelized trans-dimensional Bayesian inference, we infer seabed properties along a 12 km survey track on the scale of about 10 cm and 50 m in the vertical and horizontal, respectively. The inferred seabed properties include sound speed, attenuation, density, and porosity as a function of depth from acoustic reflection coefficient data. Parameter uncertainties are quantified, and the seabed properties agree closely with core samples at two control points and the layering structure with an independent sub-bottom seismic survey. Recovering high resolution seabed properties over large areas is shown to be feasible, which could become an important tool for marine industries, navies and oceanic research organizations.

End-to-End Geoacoustic Inversion With Neural Networks in Shallow Water Using a Single Hydrophone

End-to-End Geoacoustic Inversion With Neural Networks in Shallow Water Using a Single Hydrophone

Effect of oceanographic fluctuations on geo-acoustic inversions and source localization using ships of opportunity

Several merchant ship-radiated noise events were recorded in two separate seabed characterization experiments in the New England Mudpatch region in 2017 (SBCEX17) and 2022 (SBCEX22). These shallow water acoustic experiments were conducted under different oceanographic conditions, resulting in fluctuations in sound propagation. Several statistical inference approaches are used to invert geo-acoustic parameters such as sound speed and density in a mud over sand sediment (See https://doi.org/10.1121/10.0008419). In this work, we explore the effect of sound speed profile fluctuations in the water column on geo-acoustic inversions in the 360–1100 Hz frequency band. Inversions based on measured and simulated data using acoustic models are provided to support our findings. The results demonstrate that sediment parameters become more sensitive to oceanographic variations as the frequency increases and, therefore, sound speed profile fluctuations in the water column need to be taken into close consideration for more accurate geo-acoustic inversions. [Work supported by ONR].

Bayesian optimization for geoacoustic inversion

Geoacoustic inversion of high-dimensional parameter spaces is a computationally intensive procedure, often necessitating thousands of forward model evaluations to accurately estimate the geoacoustic environment, such as Markov chain Monte Carlo sampling. This study introduces Bayesian optimization (BO), an efficient global optimization technique, to estimate geoacoustic parameters with significantly fewer evaluations, typically on the order of hundreds. BO involves an iterative search within the parameter space to locate the global optimum of an objective function; in this study, the Bartlett power is used. BO consists of fitting a Gaussian process surrogate model to existing evaluations of the objective function, followed by selecting a new data point for evaluation using a heuristic acquisition function. The effectiveness of BO is showcased through its application to both simulated and real-world data from a shallow-water environment for multidimensional parameter space encompassing source location, array tilt, and seabed properties.

Model manifold of transmission loss from a Pekeris waveguide using VGS parameters

In geoacoustic inversion, selecting an appropriate seabed parametrization, especially with an unknown number of sediment layers, is a challenge that is compounded by potential bias when establishing bounds in the parameter search space. One approach to addressing these issues is rooted in the techniques of Information Geometry. Information Geometry informs model selection and parameterization by quantifying which model parameters are informed by observational data. This paper provides an information geometric analysis of the Pekeris waveguide, where the acoustic properties of the lower half-space are derived from the viscous grain-shearing (VGS) model. Specifically, we consider single frequency transmission loss (TL) across a wide range of VGS parameters. By exploring the limits and boundaries of the geometric manifolds, particularly as parameters approach both low and high extremes, this approach provides indications of parameter hierarchies and correlations. Results will include slices of the model manifold and an evaluation of the boundary structure, providing insight into the relative impact of VGS parameters and the delineation of limiting regions. In doing so, this paper seeks to inform model selection and parameterization in geoacoustic inversion studies. [Work supported by the Office of Naval Research. Grant N00014-21-S-B001.]

Solutions to muddy (geoacoustic inversion) problems

Waveguide propagation is a ubiquitous topic in ocean acoustics. This presentation focuses on low-frequency (f < 500 Hz) sound propagation in coastal oceans (water depth shallower than 500 m). These environments are highly dynamical and complex, and act as dispersive waveguides, bounded by the sea-surface and the seafloor. The underwater acoustic field can thus be described by a set of modes that propagate with frequency-dependent speeds. Propagation of transient acoustic signals to a distant receiver (range > 1 km) accrues multi-modal dispersion information about the propagation medium. To extract relevant information about the oceanic environment from the acoustic recording, physics-based processing methods must be used. In this presentation, I will briefly review modal propagation and time-frequency analysis. I will then show how these approaches can be combined into a non-linear signal processing method dedicated to extracting modal information from a single receiver: this information is the foundation of a transdimensional inversion method used to characterize the oceanic environment. This method will be used to perform geoacoustic inversion on the New England Mud Patch. Using several datasets collected under different oceanographic conditions, I will notably demonstrate a successful estimation of the seafloor properties, consistent with geophysical core measurements and other inversion studies, even when the water column is dynamical and mostly unknown.

Deconvolution of the waveguide impulse response for source localization at low frequency

Matched-mode processing (MMP) methods have been widely investigated for source localization and geoacoustic inversion. For horizontal line arrays (HLA), it aims at retrieving the horizontal wavenumbers and thus at inferring source range and bearing. In this work, the matched-mode is performed by applying the deconvolution of the waveguide response in a narrowband context in order to improve the localization accuracy. Deconvolution usually refers to a sparse framework when the number of sources is smaller than the number of sensors of the HLA. The Orthogonal Matching Pursuit algorithm is then used in this work. First results from numerical simulations for Pekeris waveguide highlight better localization accuracy than MMP. The deconvolution shows as well stronger robustness to mismatch in the sound speed values of the seabed. Results from measurement campaign, involving large HLA, low to ultra-low frequency and water depth up to 1500 m are investigated as well.

Inference of geoacoustic model parameters from acoustic field data: Perspectives on Geoacoustic Inversion

Estimation of parameters of geoacoustic models from acoustic field data has been a central research theme in acoustical oceanography and ocean acoustics. During the past several decades, highly efficient numerical inversion techniques have been developed that provide model parameter estimates and their uncertainties based on statistical inference methods. However, the methods are model-based and the inversions are prone to errors related to model mismatch. In any event, the inversions can generate only effective models of the true structure of the ocean bottom, which is generally highly variable over relatively small spatial scales in range and depth. There are also questions about the theory for modelling sound propagation in porous sediment media that raise doubt about the validity of inversion results. In most inversions, a visco-elastic theory is used, but is this the most appropriate propagation model? Another question is about the impact of neglecting shear waves in geoacoustic models. Most inversions assume a fluid model of the ocean bottom. This paper revisits issues that have raised questions about limitations of geoacoustic inversion methods, and discusses the impact of various mitigation measures that have been applied. The paper concludes with musings about new inversion techniques based on machine learning.

A modular neural network to quickly approximate the modal dispersion in coastal waters

Low-frequency acoustic propagation modeling in coastal waters usually relies on numerical models based on modal theory such as Kraken and Orca. These models compute the modal parameters (e.g., modal wavenumbers and depth functions) that can be used in the calculation of the acoustic field. Their repeated use in broadband applications, or for inversion purposes, comes with a notable computational cost. To mitigate this, a modular neural network (NN) was trained to approximate modal parameters for varying modes and frequencies, across diverse environments, with variable water sound speed profile and variable seabed geoacoustic parameters. The training dataset is generated using Kraken and the NN is evaluated on many environments not seen during training. Once trained, the NN can make broadband predictions without prior knowledge on the number of modes, even when the number of modes changes over the frequency-band of interest. This approach reduces computation time compared to the original forward propagation model, while maintaining high precision. The effectiveness of our method is demonstrated through transmission loss calculations and a simulated geoacoustic inversion scenario.

Augmenting sparse arrays in ocean acoustics: A Gaussian Process approach

Source localization and geoacoustic inversion performance depends on the location of receiving phones, their spacing, their number, and the array aperture. Performance suffers when practical design issues limit the capabilities of the array. In this work, sparse arrays are augmented leading to the generation of virtual arrays for better sampling of the ocean and, thus, improved estimation performance. To that effect, Gaussian Processes are employed, which are shown to create high-fidelity field “measurements” at virtual, densely spaced hydrophones. Kernel functions are key building blocks in the implementation of Gaussian Processes, as they quantify the field coherence at neighboring spatial points. Functions of interest are the squared exponential, Matern, and plane wave kernels. We validate our method with application to synthetic data as well as data collected during the Seabed Characterization Experiment conducted in 2022. [Work supported by ONR.]

Joint geoacoustic inversion based on Pearson correlation coefficient constraints.

This Letter proposes a joint geoacoustic inversion method for modal group velocity dispersion and amplitudes of waveform by incorporating a Pearson correlation constraint. Numerical simulations show that this joint inversion leads to improved geoacoustic inversion performance with smaller uncertainties compared to separate inversion methods when applied to data from a single receiver. Additionally, the effective use of the Wasserstein metric from optimal transport theory is explored and compared to the more-common L2 norm misfit measure. The Letter also presents a qualitative representation of joint inversion convergence obtained through multiple independent runs of genetic algorithms. The algorithm is applied to simulated data.

Estimates of geoacoustic parameters and source range using airgun sound in the East Siberian Sea, Arctic Ocean

In a shallow-water waveguide, low-frequency sound propagating over several kilometers has dispersive properties and thus the dispersion curves can be used for the inversion of geoacoustic parameters. In September 2019, seismic survey was conducted by the icebreaking research vessel Araon, operated by Korea Polar Research Institute. A single hydrophone was moored at East Siberian Shelf, a nearly range-independent shallow water (<70 m), to receive seismic airgun sounds generated from the R/V Araon. It was observed that the arrivals corresponding to the first two modes were clearly dispersed in the spectrogram of the received signal. In order to use these dispersion curves for the geoacoustic inversion, the dispersion curves corresponding to the first two modes were extracted using the warping transform combined with the wavelet synchrosqueezing transform. Geoacoustic parameters including sound speed and density in the sediment and the distance between the source and receiver were then estimated by comparing the extracted dispersion curves to the model replicas predicted by the KRAKEN normal-mode acoustic propagation model. Finally, the inversion results and their error analysis will be discussed in this talk. [This research was a part of the project titled ‘Korea-Arctic Ocean Warming and Response of Ecosystem (K-AWARE, KOPRI, 1525011760)’, funded by the Ministry of Oceans and Fisheries, Korea and supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (2020R1A2C2007772).]

Estimation of seabed properties from ocean ambient noise in shallow water: A review

Ocean ambient noise is the sound persisting in the ocean, which includes contributions from natural and anthropogenic sources, such as rain, wind, shipping, and marine mammals. Multiple interactions of the sound with the surface and bottom in shallow water affect the spatial characteristics of ambient noise. Thus, the spatial noise properties, coherence, and directionality of ambient noise depend on the characteristics of the shallow water environment, such as depth, sound speed profile, and seabed sediment properties. Acoustic properties of seafloor sediments can be extracted from ambient noise based on these noise properties using various types of noise data and geo-acoustic inversion methods. This paper presents a review of literature concerning investigations into available methods, the implementations of these methods, and their verification and validation. The efficacy and the applicability of these inversion methods for a wide range of ocean noise data are also examined in this review.