Water Column Sound Speed Research Articles

Source depth estimation with feature matching using convolutional neural networks in shallow water

A feature matching method based on the convolutional neural network (named FM-CNN), inspired from matched-field processing (MFP), is proposed to estimate source depth in shallow water. The FM-CNN, trained on the acoustic field replicas of a single source generated by an acoustic propagation model in a range-independent environment, is used to estimate single and multiple source depths in range-independent and mildly range-dependent environments. The performance of the FM-CNN is compared to the conventional MFP method. Sensitivity analysis for the two methods is performed to study the impact of different environmental mismatches (i.e., bottom parameters, water column sound speed profile, and topography) on depth estimation performance in the East China Sea environment. Simulation results demonstrate that the FM-CNN is more robust to the environmental mismatch in both single and multiple source depth estimation than the conventional MFP. The proposed FM-CNN is validated by real data collected from four tracks in the East China Sea experiment. Experimental results demonstrate that the FM-CNN is capable of reliably estimating single and multiple source depths in complex environments, while MFP has a large failure probability due to the presence of strong sidelobes and wide mainlobes.

Joint trans-dimensional inversion for water-column sound speed and seabed geoacoustic models.

This letter considers joint estimation of the water-column sound-speed profile (SSP) and seabed geoacoustic model through Bayesian inversion of ocean-acoustic data. The inversion is formulated in terms of separate trans-dimensional models for the water column (as an unknown number of nodes of a piecewise-continuous SSP) and seabed (as an unknown number of uniform layers) to intrinsically parameterize each according to the information content of the data. The inversion estimates marginal posterior probability profiles, quantifying the resolution of water-column and seabed structure. To validate the proposed method, modal-dispersion data from the New England Mud Patch, collected using hand-deployable systems, are considered.

Experimental demonstration of low signal-to-noise ratio matched field processing with a geoacoustic model extracted from noise.

Passive localization of a low signal-to-noise ratio (SNR) source in a shallow water waveguide without prior geoacoustic information is accomplished by combining the mode-extraction method modal-MUSIC (multiple signal classification) with range-coherent matched field processing (MFP). Range-coherent MFP coherently combines snapshots from different resolution cells to obtain gain over noise. Modal-MUSIC uses knowledge of the water column sound speed profile (no bottom information) to extract noisy estimates of modal wavenumbers from ship noise recorded on a partially spanning vertical line array (VLA). A geoacoustic model is then fit to the wavenumber estimates extracted from noise with modal-MUSIC and used to compute replicas for range-coherent MFP. The combination of these two methods applied to a 21-element VLA achieves successful source localization at SNR levels as low as -20 dB using ten tonals transmitted during the SWellEx96 experiment.

Geoacoustic Inversion Using Simple Hand-Deployable Acoustic Systems

This article proposes the use of a simple, low-cost, hand-deployable pair of experimental assets to conduct geoacoustic inversion at sea. The system consists of an expendable, fully mechanical acoustic source called a rupture induced underwater sound source (RIUSS) and a new ropeless passive acoustic mooring called a TOSSIT (not an acronym). Used together, RIUSS and TOSSIT enable the collection of acoustic data suitable to perform single-hydrophone geoacoustic inversion. The method is illustrated using data collected on the New England Mud Patch in May 2021 from a relatively small (22 m) and inexpensive chartered fishing vessel. Modal time-frequency dispersion from 15 to 385 Hz is extracted from the TOSSIT/RIUSS data using warping, and used as input for Bayesian transdimensional geoacoustic inversion. The inversion results compare favorably to results obtained with data collected on the same track with traditional assets (e.g., a vertical line array) during the 2017 Seabed Characterization Experiment, even when jointly inverting for the water-column sound speed profile and seabed geoacoustic parameters. This further demonstrates inversion repeatability in a given location using data sets collected years apart, and under different (and potentially unknown) oceanographic conditions.

Spatial and temporal variation of three-dimensional ship noise coherence in a submarine canyon.

Ship noise recorded by a vertical hydrophone array in the Mississippi Canyon region of the Gulf of Mexico is shown to contain the acoustic influence of bathymetric features, seabed properties, and water column sound speeds. Noise coherence is demonstrated to be an effective metric not just for identifying ship traffic in recorded data but also for "fingerprinting" the environment. A three-dimensional acoustics model adopting automatic identification system ship track information and realistic oceanographic conditions is used to compute noise coherence distributions across the canyon area and enables numerical study of the water column sound speed effects that can lead to temporal changes in noise coherence. The study shows the importance of including in situ sound speed measurements or constraints in passive ship noise localization from coherence measurements. Seasonal variability is also examined with models suggesting a strong influence of seasonal changes.

Mapping of surface-generated noise coherence

The performance of a hydrophone array can be evaluated by its coherent gain, which depends on the spatial correlation of both the signal of interest and the background noise between different array elements, where one hopes to maximize the former while minimizing the latter with array signal processing. In this paper, a computational vertical noise coherence map of the first zero-crossing is generated near Alvin Canyon, south of Martha’s Vineyard, Massachusetts, to study its dependence on the spatial variation in bathymetry, water column sound speed and sediment type. A two and three-dimensional Parabolic Equation propagation model based on reciprocity theory were used for the simulation. The results showed that the seabed parameters have the greatest impact on vertical noise coherence at the array location in the Alvin Canyon area, when compared to 3-D bathymetric and water column sound speed profile variability, especially in the shallower water. The analysis reveals the ideal spacing for a vertical hydrophone array for better signal detection in acoustic experiments. In the continental shelf and slope regions, the ideal spacing lies between 3λ⁄8 in deep water and λ⁄2 in shallow water, and for areas with strong bathymetric variations the ideal spacing can be determined by comprehensive numerical models.

Modal-MUSIC: A passive mode estimation algorithm for partially spanning arrays.

State-of-the-art mode estimation methods either utilize active source transmissions or rely on a full-spanning array to extract normal modes from noise radiated by a ship-of-opportunity. Modal-MUSIC, an adaptation of the MUSIC algorithm (best known for direction-of-arrival estimation), extracts normal modes from a moving source of unknown range recorded on a partially spanning vertical line array, given knowledge of the water column sound speed profile. The method is demonstrated on simulations, as well as on data from the SWellEx-96 experiment. Extracted normal modes from ship noise during the experiment are used to successfully localize a multitone source without any geoacoustic information.

Improved techniques to resolve the water column sound speed structure for multibeam ray-tracing

Multibeam bathymetry surveys conducted in highly stratified environments are routinely affected by sound speed errors. The approach proposed here aims to combine measured with synthetic sound speed profiles in order to minimize sound speed errors. The adequacy of synthetic profiles derived from a regional hydrodynamic model and spatiotemporal interpolation is therefore investigated. Ray-tracing comparisons at the 65° launch angle between measured and synthetic profiles demonstrates that the expected depth bias for modeled profiles will be in excess of 1% of water depth. Spatiotemporally interpolating hourly sampled profiles decreases the depth bias up to a factor of three for higher beam angles. Residual sounding depth biases are observed and ascribed to a residual surface sound speed error.

Joint Tracking of Source and Environment Using Improved Particle Filtering in Shallow Water

Shallow water is a complex sound propagation medium, which is affected by the varying spatial–temporal ocean environment. Taking this complexity into account, the classical processing techniques of source localization and environmental inversion may be improved. In this work, a joint tracking approach for the moving source and environmental parameters of the range-dependent and time-evolving environment in shallow water is presented. The tracking scheme treats both the source parameters (e.g., source depth, range, and speed) and the environmental parameters (e.g., water column sound speed profile (SSP) and sediment parameters) at the source location as unknown variables that evolve as the source moves. To counter sample impoverishment and robustly characterize the evolution of the parameters, an improved particle filter (PF), which is an extension of the standard PF, is proposed. Two examples with simulated data in a slowly changing environment and experimental data collected during the ASIAEX experiment are utilized to demonstrate the effectiveness of the joint approach. The results show that we were able to track the source and environmental parameters simultaneously, and the uncertainties were evaluated in the form of time-evolving posterior probability densities (PPDs). The performance comparison confirms that the improved PF is superior to the standard PF, as it can reduce the parameter uncertainties. The tracking capabilities of the improved PF were verified with high accuracy in real-time source localization and well-estimated rapidly varying parameters. Moreover, the influence of different particle numbers on the improved PF tracking performance is also illustrated.

Precursor arrivals of explosive sounds during the Seabed Characterization Experiment 2021

Travel time and amplitude of precursor arrivals of explosive sounds in shallow water contain information of marine sediments and could be used for geoacoustic inversion. The precursor arrivals have been identified in the received signals from explosive SUS charges detonated during the Seabed Characterization Experiment (SBCEX2021) conducted in the New England continental shelf and slope region in the summer of 2021. These SUS charges were deployed at various ranges along and across 200-m and 250-m bathymetry contours and recorded by two vertical line arrays (VLAs) positioned at 200-m and 250-m water depths. In addition to the two acoustic VLAs equipped with environmental sensors, there were three environmental moorings simultaneously capturing the water column fluctuation of temperature and salinity. In this research, measured precursor arrivals are first extracted and analyzed. Simulated results are then obtained and compared with the measured ones. Finally, a sensitivity study of the precursor arrivals on water column sound speed and sediment parameters (e.g., sound speed and attenuation) is performed. [Work supported by ONR Ocean Acoustics.]

Marine Acoustic Zones of Australia

Underwater sound is modelled and mapped for purposes ranging from localised environmental impact assessments of individual offshore developments to large-scale marine spatial planning. As the area to be modelled increases, so does the computational effort. The effort is more easily handled if broken down into smaller regions that could be modelled separately and their results merged. The goal of our study was to split the Australian maritime Exclusive Economic Zone (EEZ) into a set of smaller acoustic zones, whereby each zone is characterised by a set of environmental parameters that vary more across than within zones. The environmental parameters chosen reflect the hydroacoustic (e.g., water column sound speed profile), geoacoustic (e.g., sound speeds and absorption coefficients for compressional and shear waves), and bathymetric (i.e., seafloor depth and slope) parameters that directly affect the way in which sound propagates. We present a multivariate Gaussian mixture model, modified to handle input vectors (sound speed profiles) of variable length, and fitted by an expectation-maximization algorithm, that clustered the environmental parameters into 20 maritime acoustic zones corresponding to 28 geographically separated locations. Mean zone parameters and shape files are available for download. The zones may be used to map, for example, underwater sound from commercial shipping within the entire Australian EEZ.

Estimation of three-dimensional water column sound speed profiles and sediment compressional wave speed and density profiles using a distributed network of buoys.

Broadband data acquired during the Modal Mapping Experiment (MOMAX) V experiment are used to invert simultaneously for the three-dimensional (3D) water column sound speed profiles and the compressional wave speed and density profiles of the seabed in shallow waters off the coast of New Jersey. Linear Frequency Modulation sweep signals in the band 50-300 Hz are transmitted from a nearly stationary source at several discrete positions to a set of freely drifting receivers. Mode travel times are estimated from the signals acquired by the drifting buoys, and these are then used as input data in an inversion algorithm that estimates the acoustic properties of the water column and sediments. The resulting 3D compressional wave speed profiles in the seabed are generally consistent with the one-dimensional profile obtained during the narrowband component of MOMAX V, as well as the results from other experiments in the same area. The validity of the inversion results has also been assessed by the ability of the inverted model to predict the fields measured during the narrowband experiments.

3-D acoustic propagation through an estuarine salt wedge

The estuarine salt wedge presents a dynamic and highly refractive waveguide, the acoustic propagation characteristics of which are controlled by the water column sound speed gradient and boundary interactions. Acoustically, the salt wedge consists of two isospeed layers separated by a thin, three-dimensional, high-gradient layer. The behavior of a broadband (500–2000 Hz) acoustic field under the influence of an estuarine salt wedge in the Columbia River estuary is explored using two 3-D acoustic propagation models: 3-D rays and 3-D parabolic equation (3DPE). These model results are compared to data collected during a field experiment in 2013. Results demonstrate that the dominant physical mechanism controlling acoustic propagation in this waveguide shifts from 3-D bottom scatter in a non-refractive waveguide (before the entrance of the salt wedge) to 3-D acoustic refraction with minimal bottom interaction in a refractive waveguide (when the salt wedge occupies the acoustic transect). Vertical and horizontal refraction in the water column and out-of-plane scattering by the bottom are clearly evident at specific narrowband frequencies; however, these mechanisms contribute to, but do not account for the total observed transmission loss.

Environmental information content of ocean ambient noise.

In recent years, methods have been developed to estimate a variety of environmental parameters based on measurements of the ocean ambient noise. For example, noise has been used to estimate water depth using the passive fathometer technique and bottom loss estimated and used to invert for seabed parameters. There is also information in the noise about the water column sound speed, volume attenuation, and the sea-state. The Fisher information can be used to quantify the basic information available in the noise measurements and its inverse, the Cramér-Rao lower bound (CRLB), provides the lower limit on the variance of an unbiased estimator of a particular parameter. The CRLB can be used to study the feasibility of various measurement configurations and parameter sensitivities. In this paper, the CRLB is developed for ocean ambient noise and the environmental information contained in the measurements is determined. The CRLBs provide an estimate of the underlying information in the data, however, it is independent of the estimation methodology. This is useful to determine if a given estimation method is reaching the lower bound. Results illustrating the bounds as well as sensitivities and performance of estimators are demonstrated using both simulations and data.

3D acoustic propagation through an estuarine salt wedge at low-to-mid-frequencies: Modeling and measurement

The estuarine salt wedge presents a dynamic and highly refractive waveguide, the acoustic propagation characteristics of which are controlled by the water column sound speed gradient and boundary interactions. Acoustically, the salt wedge consists of two isospeed layers separated by a thin, three-dimensional (3D), high-gradient layer. The behavior of a broadband (500-2000 Hz) acoustic field under the influence of an estuarine salt wedge in the Columbia River estuary is explored using two 3D acoustic propagation models: 3D rays and 3D parabolic equation. These model results are compared to data collected during the field experiment. Results demonstrate that the dominant physical mechanism controlling acoustic propagation in this waveguide shifts from 3D bottom scatter in a non-refractive waveguide (before the entrance of the salt wedge) to 3D acoustic refraction with minimal bottom interaction in a refractive waveguide (when the salt wedge occupies the acoustic transect). Vertical and horizontal refraction in the water column and out-of-plane scattering by the bottom are clearly evident at specific narrowband frequencies; however, these mechanisms contribute to, but do not account for, the total observed broadband transmission loss.

Multipath Broadband Localization, Bathymetry, and Sediment Inversion

Transmission of linearly frequency modulated pulses generates receptions at a vertical line array that can be cross correlated with the source signal to provide estimates of the oceanic waveguide impulse response. For short ranges, distinct path arrivals can be identified including the direct, surface reflection, bottom reflection, and sediment reflection. Accurate estimation of arrival times of such paths is tightly related to successful inversion for source location and water column depth and sound speed and, subsequently, estimation of sediment sound speed and thickness. To achieve accurate estimation, particle filtering is applied to the received time series at 16 phones combined with a simple cross-correlation method. Using linearization, arrival time probability density functions are connected to the geometry and water column sound-speed parameters, providing point estimates as well as probability densities. These are then employed in sediment sound speed and thickness estimation. The results, obtained from the application of the method to data collected during the Seabed Characterization Experiment, are consistent with prior information on the site.

Bayesian inference for water column and sediment inversion in ocean acoustics

Bayesian methods provide inference on unknown parameters given prior beliefs and a statistical model for the observations in a noisy environment. It was shown in “Sequential filtering and linearization for inversion in the Seabed Characterization Experiment,” J. Acoust. Soc. Am. 144, 1913 (2018) how geometry and sediment parameter estimation in an oceanic experiment can be achieved with a two-stage Bayesian approach. Here, we demonstrate the potential of the method for inversion with a different data set presenting challenges in identifying features of the acoustic field necessary for the second stage of the estimation process. The water column sound speed is complex in this experiment, necessitating a complex inverse model. The transmitted signals are lfm pulses and probability densities of multipath arrival times are calculated at the receivers. For the forward and inverse calculations, we compute the Jacobian matrix for sound speed estimation using EOF coefficients. The inversion, employing the direct and surface and bottom paths, leads to inference on the geometry and the water column sound speed profile. Using the computed probability densities and introducing one more arrival, sediment property densities are calculated. Results are in excellent agreement with expected values for the site. [Work supported by ONR.]

Correcting Multibeam Echosounder Bathymetric Measurements for Errors Induced by Inaccurate Water Column Sound Speeds

In this contribution a method for correcting bathymetric measurements affected by inaccurate water column sound speed profiles (SSPs) is presented. The method exploits the redundancy in the multibeam echosounder measurements obtained from the overlap of adjacent swaths by minimizing the difference between depths along overlapping swaths. Two optimization methods are used, i.e., Differential Evolution (DE) and Gauss-Newton (GN). While DE inverts for the sound speed by minimizing the depth variation, GN inverts for both bathymetry and sound speed by minimizing the squared sum of the differences between the modeled and measured travel times. The inversion method assumes a constant SSP in the water column. Applying the method to a salt wedge survey area with large variations in the water column sound speed indicates a good agreement between the original depth measurements and those derived after the inversion with the mean and standard deviation of the depth differences equaling 0.009m and 0.024m, respectively. This indicates that even with a simple parametrization of the sound speed in the water column, the correct bathymetry can be derived from the inversion. The SSP inversion method is also applied to an area with existing refraction artefacts. It corrects the bathymetry and reduces the mean and standard deviation of the depth standard deviation by a factor of around 2.75 compared to the case where the measured SSPs were used. Furthermore, the SSP inversion method neither manipulates the existing morphology nor introduces artificial bathymetric features in the areas where such refraction artefacts are not present. Considering constant SSPs, both DE and GN give almost identical results with GN being faster. However, GN is less flexible with regards to varying sound speed parameterizations.

Quantifying seabed and other environmental information contained in ocean ambient noise

The ocean ambient noise field contains a surprising amount of information about the environment. Information about the seabed includes, layering structure, sound speed, density, and attenuation. There is also information about the ocean environment such as water column sound speed, volume attenuation, and the sea-state. Estimators for these quantities have been developed based mostly on vertical hydrophone arrays and beamforming. Although this has led to some simple and direct estimates for important quantities such as bottom loss, these estimates may be less than ideal in some cases. For example, when using inadequately sampled arrays or in certain unfavorable measurement geometries. The Fisher information can be used to quantify the basic information available in the noise measurements; and its inverse, the Cramer-Rao lower bound (CRLB), provides the lower limit on the variance of any unbiased estimator. The CRLB can be used to study the feasibility of various measurement configurations and parameter sensitivities. In this presentation, an overview will be given specifying the variety of environmental information contained in ocean noise. Parameter sensitivities and performance of beamforming estimators relative to the CRLB will also be discussed. [Work supported by the Office of Naval Research Ocean Acoustics Program.]

Inversion of the Sound Speed With Radiated Noise of an Autonomous Underwater Vehicle in Shallow Water Waveguides

In December 2013, an experiment using an autonomous underwater vehicle (AUV) as a moving source for water column sound speed profile (SSP) inversion was conducted in Mogan Lake, China. With in-situ conductivity, temperature, and depth (CTD) measurements, the empirical orthogonal functions were first constructed to represent the SSP. Radiated noises of the AUV were received on a vertical line array and processed via acoustic field matching to obtain estimates of the range-independent SSP, along with the position and velocity of the AUV and the water column depth. The frequency of the radiation noise of the AUV was approximately 14 kHz. At such high frequency, even though the source motion was slow, a significant Doppler shift/broadening was observed. To incorporate the Doppler effects to better match the measured data, a forward acoustic model is derived based on the waveguide Doppler and normal mode theory. An analytical solution of the forward model is obtained for arbitrary signal integration intervals with a monochromatic source, moving radially relative to the receiver. Through simulations and experimental data processing, the feasibility of using an AUV source for water column SSP inversion has been demonstrated; it is also shown that the waveguide Doppler model is more effective compared with the model that does not consider the waveguide Doppler effect.