Sound Speed Profile Inversion Research Articles

Modeling sound speed profile based on ocean normal mode

IntroductionStatistical methods such as empirical orthogonal functions (EOFs) are often used to model the sound speed profile (SSP). However, their statistical nature often leads to the sample dependence and physical fuzziness.MethodThis study proposes a technique for modeling the SSP from the perspective of ocean dynamics. It employs the ocean normal mode, which is the mode of fluid particles motion, to deduce perturbations in the SSP, which is called the ocean mode basis (OMB).ResultThe results of SSP reconstruction of in-situ samples showed that a few leading orders of the OMB can provide a compact representation of the SSP. Oscillations of the contours and gradient of the sound speed in thermocline were analyzed by using the first two orders of the projection coefficients of the relationship between the OMB and the baroclinic mode. As a physical model, this technique can also be used to characterize the dynamics of internal solitary waves. Furthermore, the OMB derived from archival date was used for SSP inversion. The results showed that the OMB can reconstruct SSP of a reasonable resolution without requiring in-situ samples.DiscussionCompared with statistical models, the OMB can better explain the ocean dynamics underlying variations in the SSP while requiring fewer samples.

Sound Speed Inversion Based on Multi-Source Ocean Remote Sensing Observations and Machine Learning

Ocean sound speed is important for underwater acoustic applications, such as communications, navigation and localization, where the assumption of uniformly distributed sound speed profiles (SSPs) is generally used and greatly degrades the performance of underwater acoustic systems. The acquisition of SSPs is necessary for the corrections of the sound ray propagation paths. However, the inversion of SSPs is challenging due to the intricate relations of interrelated physical ocean elements and suffers from the high costs of calculations and hardware deployments. This paper proposes a novel sound speed inversion method based on multi-source ocean remote sensing observations and machine learning, which adapts to large-scale sea regions. Firstly, the datasets of SSPs are generated utilizing the Argo thermohaline profiles and the empirical formulas of the sound speed. Then, the SSPs are analyzed utilizing the empirical orthogonal functions (EOFs) to reduce the dimensions of the feature space as well as the computational load. Considering the nonlinear regression relations of SSPs and the observed datasets, a general framework for sound speed inversion is formulated, which combines the designed machine learning models with the reduced-dimensional feature representations, multi-source ocean remote sensing observations and water temperature data. After being well trained, the proposed machine learning models realize the accurate inversion of the targeted ocean region by inputting the real-time ocean environmental data. The experiments verify the advantages of the proposed method in terms of the accuracy and effectiveness compared with conventional methods.

Inversion of Sound Speed Profile in the Luzon Strait by Combining Single Empirical Orthogonal Function and Generalized Regression Neural Network

Inversion of Sound Speed Profile in the Luzon Strait by Combining Single Empirical Orthogonal Function and Generalized Regression Neural Network

Experimental Results of Underwater Sound Speed Profile Inversion by Few-Shot Multi-Task Learning

Underwater Sound Speed Profile (SSP) distribution is crucial for the propagation mode of acoustic signals, so fast and accurate estimation of SSP is of great importance in building underwater observation systems. The state-of-the-art SSP inversion methods include frameworks of matched field processing (MFP), compressive sensing (CS), and feed-forward neural networks (FNNs), among which the FNN shows better real-time performance while maintaining the same level of accuracy. However, the training of FNN needs quite a lot historical SSP samples, which is difficult to satisfy in many ocean areas. This situation is called few-shot learning. To tackle this issue, we propose a multi-task learning (MTL) model with partial parameter sharing among different training tasks. By MTL, common features could be extracted, which accelerates the learning process on given tasks, and reduces the demand for reference samples, enhancing the generalization ability in few-shot learning. To verify the feasibility and effectiveness of MTL, a deep-ocean experiment was held in April 2023 in the South China Sea. Results show that MTL outperforms the other mainstream methods in terms of accuracy for SSP inversion, while inheriting the real-time advantage of FNN during the inversion stage.

A meta-deep-learning framework for spatio-temporal underwater SSP inversion

Sound speed distribution, represented by a sound speed profile (SSP), is of great significance because the nonuniform distribution of sound speed will cause signal propagation path bending with Snell effect, which brings difficulties in precise underwater localization such as emergency rescue. Compared with conventional SSP measurement methods via the conductivity-temperature-depth (CTD) or sound-velocity profiler (SVP), SSP inversion methods leveraging measured sound field information have better real-time performance, such as matched field process (MFP), compressed sensing (CS) and artificial neural networks (ANN). Due to the difficulty in measuring empirical SSP data, these methods face with over-fitting problem in few-shot learning that decreases the inversion accuracy. To rapidly obtain accurate SSP, we propose a task-driven meta-deep-learning (TDML) framework for spatio-temporal SSP inversion. The common features of SSPs are learned through multiple base learners to accelerate the convergence of the model on new tasks, and the model’s sensitivity to the change of sound field data is enhanced via meta training, so as to weaken the over-fitting effect and improve the inversion accuracy. Experiment results show that fast and accurate SSP inversion can be achieved by the proposed TDML method.

Performance of single empirical orthogonal function regression method in global sound speed profile inversion and sound field prediction

The sound speed profile (SSP) reflects the change in sound speed from the surface to the bottom of the seawater, which will have an important influence on underwater acoustic detection and communication. Depending on the inversion of sea surface data sets measured by satellite to obtain near-real-time and high-precision SSPs has become a research hotspot. In this paper, based on the Argo data and the sea surface data sets, we use the single empirical orthogonal function regression (sEOF-r) method to inverse the global SSPs and evaluate the performance of this method. There are also differences in the performance of inversed results at different sites and in different months at the global scale of the ocean. The eddy kinetic energy (EKE) affects the accuracy of inversion, we propose to add the EKE into the inversion formula at the same time to further optimize the inversed results and improve the accuracy, the global average inversion error after optimization reduces by 20%. Finally, we verify the effectiveness of the optimized method from the perspective of sound field prediction. In most cases, the correlation coefficient of the transmission loss (TL) calculated through the inversion SSP and Argo-SSP exceeds 0.7 and the error is controlled below 6 dB, which will have important implications for the actual sound field prediction and hydroacoustic communication.

Wide-area sound speed profile estimation based on a pre-classification scheme for sound speed perturbation modes

IntroductionThe trend of sound speed profile (SSP) inversion is towards wide-area sound speed estimation. However, the traditional inversion method of dividing the latitude and longitude grids has limitations in terms of significantly lower accuracy when samples are lacking. k-means clustering algorithm (K-means) can divide the training class to achieve high accuracy estimation.MethodThis paper proposes a grid-free pre-classification inversion scheme based on empirical orthogonal function (EOF) vectors. The scheme is based on the K-means to classify the samples according to the perturbation mode of the SSP. After classification, the SSP inversion is carried out using the self-organizing map algorithm (SOM). The experimental sea area is selected from the South China Sea, and the inversion results are evaluated using root mean square error (RMSE) as the criterion.ResultThe inversion results show that the inversion error is 2.1 m/s for the pre-classification solution and 2.7 m/s for the solution without pre-classification, a steady improvement of more than 20% in the inversion error. Accuracy is also improved by 2.14 m/s in the depth range where the sound speed perturbance is greatest.DiscussionThis pre-classification scheme has smaller inversion errors and the classification results are reasonable in terms of distribution in time and space. It provides a feasible solution for SSP inversion in sea areas where samples are lacking.

Estimation of Sound Speed Profiles Using a Random Forest Model with Satellite Surface Observations

Sound speed profile (SSP) inversion is usually performed by linear statistical regression, such as the single empirical orthogonal function regression (sEOF-r) model. However, due to the complex dynamic activities of the ocean, the relationship between parameters is not strictly linear, often resulting in an unsatisfactory inversion result. In this study, an algorithm based on the random forest (RF) integrated learning model, for SSP inversion, was proposed. Using the sea surface temperature anomaly (SSTA) and sea surface height anomaly (SSHA) data, the sound speed profile of the upper 1000 m layer in the South China Sea was reconstructed, and its accuracy was evaluated through the root mean square error (RMSE). The accuracy of the evaluation demonstrated that the RF model proposed here could reconstruct the SSP in the upper 1000 m layer better than the sEOF-r can. Compared with the latter, the average reconstruction accuracy of the RF model was improved by 0.56 m/s. The linear regression of the sEOF-r model fell short of expectations in the regression between surface and subsurface parameters. By removing the constraints of linear inversion, the nonlinear regression of the RF model showed a smaller RMSE and better robustness in the reconstruction process and was superior to the sEOF-r model at all depths. As a result, it provided an effective integrated learning model for SSP reconstruction.

Rapid environmental assessment in the South China Sea: Improved inversion of sound speed profile using remote sensing data

Rapid environmental assessment in the South China Sea: Improved inversion of sound speed profile using remote sensing data

Passive acoustic characterization of sub-seasonal sound speed variations in a coastal ocean

Acoustic noise interferometry is applied to retrieve empirical Green's functions (EGFs) from the ambient and shipping noise data acquired in the Shallow Water 2006 experiment on the continental shelf off New Jersey. Despite strong internal wave-induced perturbations of the sound speed in water, EGFs are found on 31 acoustic paths by cross-correlating the noise recorded on a single hydrophone with noise on the hydrophones of a horizontal linear array about 3.6 km away. Datasets from two non-overlapping 15-day observation periods are considered. Dispersion curves of three low-order normal modes at frequencies below 110 Hz are extracted from the EGFs with the time-warping technique. The dispersion curves from the first dataset were previously employed to estimate the seabed properties. Here, using this seabed model, we invert the differences between the dispersion curves obtained from the two datasets for the variation of the time-averaged sound speed profile (SSP) in water between the two observation periods. Results of the passive SSP inversion of the second dataset are compared with the ground truth derived from in situ temperature measurements. The effect of temporal variability of the water column during noise-averaging time on EGF retrieval is discussed and quantified.

Underwater Target Ranging and Sound Speed Profile Inversion via Eigenray Propagation Time-Delay Matching

An underwater target ranging and sound speed profile (SSP) inversion algorithm for synchronous beacon is proposed. In the condition of actual SSP unavailable, based on the multipath propagation time-delay estimation, a genetic algorithm is employed to estimate the target range and SSP by eigenray propagation time-delay matching. Lake trial results show the algorithm could achieve target ranging and SSP inversion simultaneously.

Acoustic sound speed profile inversion based on orthogonal matching pursuit

The estimation of ocean sound speed profiles (SSPs) requires the inversion of an acoustic field using limited observations. Such inverse problems are underdetermined, and require regularization to ensure physically realistic solutions. The empirical orthonormal function (EOF) is capable of a very large compression of the data set. In this paper, the non-linear response of the sound pressure to SSP is linearized using a first order Taylor expansion, and the pressure is expanded in a sparse domain using EOFs. Since the parameters of the inverse model are sparse, compressive sensing (CS) can help solve such underdetermined problems accurately, efficiently, and with enhanced resolution. Here, the orthogonal matching pursuit (OMP) is used to estimate range-independent acoustic SSPs using the simulated acoustic field. The superior resolution of OMP is demonstrated with the SSP data from the South China Sea experiment. By shortening the duration of the training set, the temporal correlation between EOF and test sets is enhanced, and the accuracy of sound velocity inversion is improved. The SSP estimation error versus depth is calculated, and the 99% confidence interval of error is within ±0.6 m/s. The 82% of mean absolute error (MAE) is less than 1 m/s. It is shown that SSPs can be well estimated using OMP.

Application of iterative hill climbing to the sound speed profile inversion in underwater acoustics

An application of the iterative hill climbing algorithm to the solution of inverse problems of underwater acoustics is discussed. Modal dispersion data extracted from a recording of a pulse acoustical signal is used as the input for the geoacoustic inversion procedure. The mismatch of the dispersion curves extracted from experimental data and computed from a given set of waveguide parameters is minimized over the parameters space. The solution of two test inversion problems is considered. It is shown that the iterative launches of the hill climbing algorithm allow to locate either the global minimum or a local minimum which is reasonably close to the former. In the latter case a reasonable estimate of the waveguide parameters is obtained.

Investigating sound speed profile assimilation: An experiment in the Philippine Sea

In this paper, a single datum assimilation method is described for constructing the sound speed profile (SSP) in the ocean. Our method's key is assimilation radius, where various radii provide different SSPs. By means of the Bellhop acoustic model and acoustic data, we calculated the optimal SSP assimilation radius. Due to the lack of proper in-situ observations, an SSP inversion method is proposed. Inverted SSP compared to assimilated SSP indicates close correlation, showing that both methods yield reliable results. Historical data analysis indicates a stable correlation structure under certain conditions, implying the optimal assimilation radius would measure SSP effectively under similar conditions.

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.

SEARCH FOR GLOBAL f -MODES AND p -MODES IN THE 8 B NEUTRINO FLUX

The impact of global acoustic modes on the 8B neutrino flux time series is computed for the first time. It is shown that the time fluctuations of the 8B neutrino flux depend on the amplitude of acoustic eigenfunctions in the region where the 8B neutrino flux is produced: modes with low n (or order) that have eigenfunctions with a relatively large amplitude in the Sun's core, strongly affect the neutrino flux; conversely, modes with high n that have eigenfunctions with a minimal amplitude in the Sun's core have a very small impact on the neutrino flux. It was found that the global modes with a larger impact on the 8B neutrino flux have a frequency of oscillation in the interval 250 \mu Hz to 500 \mu Hz (or a period in the interval 30 minutes to 70 minutes), such as the f-modes (n=0) for the low degrees, radial modes of order n smaller or equal to 3, and the dipole mode of order n=1. Their corresponding neutrino eigenfunctions are very sensitive to the solar inner core and are unaffected by the variability of the external layers of the solar surface. If time variability of neutrinos is observed for these modes, it will lead to new ways of improving the sound speed profile inversion in the central region of the Sun.

Inversion of sound speed profile in three-dimensional shallow water based on transmission time

Horizontal deflection often occurs when a ray propagates in shallow water with an irregular sea bottom, which makes it difficult to determine eigen-ray trajectory and transmission time. A new method for seeking eigen-ray in three dimensions was proposed and sound speed profile (SSP) was inverted in three-dimensional shallow water using the fastest eigen-ray transmission times to form the cost function. In order to satisfy long-time monitoring of SSP, a kind of sonobuoy was developed with the function of DGPS, short wave radio controlling and dada transmission and experiment was implemented in the South China Sea. It is indicated that the sonobuoy has high reliability and stability in rough sea condition. The precision of SSP inversion can be enhanced dramatically by considering horizontal deflections of eigen-ray. In addition, the influence of small mismatch of seabed depth on precision of inversion can be neglected.

Sound speed profile inversion by matching the acoustic intensity striations in shallow water

Estimating the sound speed profile by matching the acoustic intensity striation using a mounted horizontal line array is investigated. Due to constructive and destructive interference between normal modes, the acoustic intensity manifests range-frequency striation structure in shallow water. Using the environment and acoustic data measured during a tomography experiment conducted in the South China Sea in June, 2010, the possibility and validity of inverting for the water column sound speed profile by matching the acoustic intensity striations are studied. The results show that the fluctuation of the sound speed profile has obvious influence on the acoustic intensity striations. The acoustic data recorded in an experiment conducted in the South China Sea in June 2010 are analyzed to invert for the sound speed profiles. The sound speed profile is represented by Empirical Orthogonal Functions (EOFs) to reduce the unknown parameters. The estimates agree well with the measured sound speed profiles and most of the standard deviations are less than 1 m/s.

Matched-field inversion of sound speed profile in shallow water using a parallel genetic algorithm

A sound speed profile plays an important role in shallow water sound propagation. Concurrent with in-situ measurements, many inversion methods, such as matched-field inversion, have been put forward to invert the sound speed profile from acoustic signals. However, the time cost of matched-field inversion may be very high in replica field calculations. We studied the feasibility and robustness of an acoustic tomography scheme with matched-field processing in shallow water, and described the sound speed profile by empirical orthogonal functions. We analyzed the acoustic signals from a vertical line array in ASIAEX2001 in the East China Sea to invert sound speed profiles with estimated empirical orthogonal functions and a parallel genetic algorithm to speed up the inversion. The results show that the inverted sound speed profiles are in good agreement with conductivity-temperature-depth measurements. Moreover, a posteriori probability analysis is carried out to verify the inversion results.

Modal wave number tomography and bottom parameter dependence

A shallow water tomography scheme based on the modal wave number inversion technique is considered in this paper. The scheme is based on the assumption that modal wave number for transpped mode can be measured in a suitable way. The tomographic inversion is accomplished into steps; firstly, the bottom parameter are inverted by using the bottom reflection phase shift with the known sound speed profile; secondly, the variation of sound speed profile at different time is inverted provided the bottom parameters are known. A numerical simulation shows that the proposed scheme works well, and the sensitivity analysis of sound speed profile inversion is performed, for shallow water environmental parameters: sound speed, density and attenuation coefficient of the bottom.