Trans-Dimensional Inversion of Modal Dispersion Data on the New England Mud Patch

Abstract

This paper presents single receiver geoacoustic inversion of two independent data sets recorded during the 2017 seabed characterization experiment on the New England Mud Patch. In the experimental area, the water depth is around 70 m, and the seabed is characterized by an upper layer of fine grained sediments with clay (i.e., mud). The first data set considered in this paper is a combustive sound source signal, and the second is a chirp emitted by a J15 source. These two data sets provide differing information on the geoacoustic properties of the seabed, as a result of their differing frequency content, and the dispersion properties of the environment. For both data sets, source/receiver range is about 7 km, and modal time-frequency dispersion curves are estimated using warping. Estimated dispersion curves are then used as input data for a Bayesian trans-dimensional inversion algorithm. Subbottom layering and geoacoustic parameters (sound speed and density) are thus inferred from the data. This paper highlights important properties of the mud, consistent with independent in situ measurements. It also demonstrates how information content differs for two data sets collected on reciprocal tracks, but with different acoustic sources and modal content. Index Terms-Geoacoustic inversion, New England mud patch, seabed characterization experiment (SBCEX), underwater acoustics, warping.

Nonlinear time-warping made simple: A step-by-step tutorial on underwater acoustic modal separation with a single hydrophone

“…For in situ signals, source deconvolution requires a good measurement (or model) of the source signal. In a warping context, it has been successfully applied on lightbulb (Duan et al, 2016) and CSS data (Bonnel et al, 2019;Bonnel et al, 2018), with 2 ½0:01; 0:1.…”

“…This is because in the considered frequency band and for the considered modes, the experimental environment can be well-approximated with a Pekeris waveguide. Actually, the experimental environment is more complicated than that (Bonnel et al, 2019), and such a match would have been impossible over a broader frequency band, and/or for a greater number of modes. Localizing the source using all the modes would have required the use of an environmental model much more complex than the Pekeris one.…”

“…Warping has notably been used for environmental estimation studies, mostly seabed geoacoustic inversion (Bonnel et al, 2013a;Bonnel et al, 2019;Bonnel et al, 2012;Dong et al, 2017;Duan et al, 2016;Feng-Hua et al, 2014;Petrov, 2014;Zeng et al, 2013), but also water column tomography (Ballard et al, 2014), as well as joint estimation of water column and seabed properties (Warner et al, 2015). On a more basic research point of view, it is interesting to note that warping has also been used to estimate modal depth functions Thode and Bonnel, 2015), as well as to filter modes from noise interferometry data (Sergeev et al, 2017;Tan et al, 2018).…”

An Experimental Benchmark for Geoacoustic Inversion Methods

“…Another simpler option, from the benchmark perspective, would be to use a single inversion method and to benchmark the performance of various experimental designs and associated data processing. This could already be done for subsets of inversion methods from SBCEX that use the same inverse algorithm, e.g., trans-dimensional inversion [42]- [44]. There is a clear advantage of this approach since it focuses the benchmark on experimental design and data processing, without dealing with optimization/inversion issues.…”

Bayesian Inversion for Geoacoustic Parameters in Shallow Sea

“…The inversion results of traditional optimize methods are all prone to being trapped by local minima, can only determine the best-fit model and not provide quantitatively nonlinear uncertainty estimation of the awaiting inversion parameters. In recent years, the Bayesian inversion method underwent considerable development and significantly contributed to estimating bottom properties and their uncertainties based on a Bayesian formulation [15][16][17][18][19]. The Bayesian inversion method is a global optimization algorithm based on the probability theory, applied mathematics, and optimization theory.…”