Bayesian geoacoustic parameters inversion for multi-layer seabed in shallow sea using underwater acoustic field

Abstract

Seabed geoacoustic parameters play an important role in underwater acoustic channel modeling. Traditional methods to determine these parameters, for example, drilling, are expensive and are being replaced by acoustic inverse technology. An inversion method based on Bayesian theory is presented to derive the structure and geoacoustic parameters of a layered seabed in a shallow sea. The seabed was considered a layered elastic medium. The objective of this research was to use the sound pressure detected by underwater acoustic sensors at different positions and to use nonlinear Bayesian inversion to estimate the geoacoustic parameters and their uncertainties in the multi-layer seabed. Specifically, the thickness, density, compression wave speed, shear wave speed, and the attenuation of these two wave speeds were determined. The maximum a posterior (MAP) model and posterior probability distribution of each parameter were estimated using the optimized simulated annealing (OSA) and Metropolis-Hastings sampling (MHS) methods. Model selection was carried out using the Bayesian information criterion (BIC) to determine the optimal model that thoroughly explained the experimental data for different parameterizations. The results showed that the OSA is much more capable of delivering high-accuracy results in multi-layer seabed models. The compression wave speed and shear wave speed were less uncertain than the other parameters, and the parameters in the upper layer had less uncertainty than those in the lower layer.

A back propagation neural network-based approach for inverting layered seabed acoustic parameters in shallow waters

“…With the development of acoustic technology and the popularity of machine algorithms, it is convenient and efficient to in-vert underwater acoustic parameters using acoustic methods (Li and Zhang, 2017). Aiming at parameters inversion, a large number of underwater acoustic parameters inversion works have been carried out by predecessors, and many methods for underwater acoustic parameters in-version have been developed, such as the acoustic parameters inversion method based on transmission loss (Dragna and Blanc-Benon, 2017), the seafloor sediment parameters were estimated using the genetic algorithm and bottom reflection loss curve with wide grazing angles (Wang et al, 2023), a solution to the problem of stratified seabed parameters estimation using the sound field matching meth-od with vertical angle spectra (Xue et al, 2023),three-dimensional sediment modelling and inversion of geoacoustic parameters using a cross-dimensional Bayesian approach with wideband acoustic sources (Ke et al, 2013), the acoustic signal arrival time (Wang et al, 2023), and the waveguide dispersion characteristics (Kerzhakov and Kulinich, 2016). Among them, the most widely used methods can be summarized as the matching field inversion of underwater acoustic parameters by using the physical characteristics of underwater acoustic signals combined with the global optimization algorithm (Potty et al, 2017).…”

Bottom Multi-Parameter Bayesian Inversion Based on an Acoustic Backscattering Model

“…Among these steps, the selection of the forward model is an important prerequisite for inversion. In most studies on acoustic inversion, underwater acoustic propagation models are typically used as the forward model [8,9]; the matching physical quantities used for inversion include normal mode dispersion characteristics, the bottom reflection coefficient, sound propagation loss, and so on [10][11][12]. Under this condition, the parameters obtained via acoustic inversion are usually sediment acoustic parameters (sound speed and attenuation coefficient), sediment density, and the layered structure.…”

Direction-of-Arrival Estimation Based on Sparse Representation of Fourth-Order Cumulants

“…In ocean acoustic research, underwater direction of arrival (DOA) estimation has been a hot topic for scholars [1], [2], [3], [4]. In the current study, most of the algorithms are effective only in the white noise environment of the array element noise, however, the actual underwater environment is affected by the noise of distant ships, wind noise, biological sounding, etc., the background noise is often colored, and the performance of the DOA estimation methods applicable to white noise will be seriously degraded [5], [6], [7], [8], [9].…”