Dag Tollefsen

Bayesian geoacoustic inversion of ship noise on a horizontal array

This paper applies geoacoustic inversion to low-frequency narrow-band acoustic data from a quiet surface ship recorded on a bottom-moored horizontal line array in shallow water. A Bayesian matched-field inversion method is employed which quantifies geoacoustic uncertainties and allows for meaningful comparison of inversion results from different data sets. Geoacoustic inversion results for ship-noise data are compared with inversion results for multitone data from a towed controlled source collected in the same experiment, and with independent geophysical measurements. To increase the information content of low-level ship-noise data, the effect of including multiple, independent data segments in the inversion is investigated and shown to significantly reduce geoacoustic parameter uncertainties. Geoacoustic uncertainties are also shown to depend on ship range and orientation, with increased uncertainties for long ranges and for the ship stern oriented away from the array.

Matched-field geoacoustic inversion with a horizontal array and low-level source

This paper applies geoacoustic inversion to acoustic-field data collected on a bottom-moored horizontal line array due to a continuous-wave towed source at a shallow water site in the Barents Sea. The source transmitted tones in the frequency band of 30–160Hz at levels comparable to those of a merchant ship, with resulting signal-to-noise ratios of 9–15dB. Bayesian inversion is applied to cross-spectral density matrices formed by averaging spectra from a sequence of time-series segments (snapshots). Quantifying data errors, including measurement and theory errors, is an important component of Bayesian inversion. To date, data error estimation for snapshot-averaged data has assumed either that averaging reduces errors as if they were fully independent between snapshots, or that averaging does not reduce errors at all. This paper quantifies data errors assuming that averaging reduces measurement error (dominated by ambient noise) but does not reduce theory (modeling) error, providing a physically reasonable...

Three-dimensional source tracking in an uncertain environment

This paper develops an approach to three-dimensional source tracking in an uncertain ocean environment using a horizontal line array (HLA). The tracking algorithm combines matched-field focalization for environmental (seabed and water column) and source-bearing model parameters with the Viterbi algorithm for range-depth estimation and includes physical constraints on source velocity. The ability to track a source despite environmental uncertainty is examined using synthetic test cases for various track geometries and with varying degrees of prior information for environmental parameters. Performance is evaluated for a range of signal-to-noise ratios in terms of the probability of estimating a track within acceptable position/depth errors. The algorithm substantially outperforms tracking with poor environmental estimates and generally obtains results close to those obtained with exact environmental knowledge. The approach is also applied to measured narrowband data recorded on a bottom-moored HLA in shallow water (the Barents Sea) and shown to successfully track both a towed submerged source and a surface ship in cases where simpler tracking algorithms failed.

Estimates of geoacoustic model parameters from inversions of horizontal and vertical line array data

This paper describes results from geoacoustic inversion of low-frequency acoustic data recorded at a receiving array divided into two sections, a sparse bottom laid horizontal array (HLA) and a vertical array (VLA) deployed in shallow water. The data are from an experiment conducted by the Norwegian Defence Research Establishment (FFI) in the Barents Sea, using broadband explosives (shot) sources. A two-layer range-independent geoacoustic model, consistent with seismic profiles from the area, described the environment. Inversion for geoacoustic model parameters was carried out using a fast implementation of the hybrid adaptive simplex simulated annealing (ASSA) inversion algorithm, with replica fields computed by the ORCA normal mode code. Low-frequency (40-128 Hz) data from six shot sources at ranges 3-9 km from the array were considered. Estimates of sediment and substrate p-wave velocities and sediment thickness were found to be consistent between independent inversions of data from the two sections of the array

Three‐dimensional source tracking in an uncertain environment via Bayesian marginalization.

This paper develops a non-linear Bayesian marginalization approach for three-dimensional source tracking in shallow water with uncertain environmental properties. The algorithm integrates the posterior probability density via a combination of Metropolis-Hastings sampling over environmental and bearing model parameters and Gibbs sampling over source range/depth, with track constraints on source velocity applied. Marginal distributions for source range/depth and source bearing are derived, with source position uncertainties estimated from the distributions. The Viterbi algorithm is applied to obtain the most probable three-dimensional track. The approach is applied to experimental narrowband data recorded on a bottom-moored horizontal line array in the Barents Sea.

Seismic exploration noise reduction in the Marginal Ice Zone

A sonobuoy field was deployed in the Marginal Ice Zone of the Fram Strait in June 2011 to study the spatial variability of ambient noise. High noise levels observed at 10-200 Hz are attributed to distant (1400 km range) seismic exploration. The noise levels decreased with range into the ice cover; the reduction is fitted by a spreading loss model with a frequency-dependent attenuation factor less than for under-ice interior Arctic propagation. Numerical modeling predicts transmission loss of the same order as the observed noise level reduction and indicates a significant loss contribution from under-ice interaction.

Geoacoustic Information Content of Horizontal Line Array Data

This paper examines the effectiveness of horizontal line arrays (HLAs) for matched-field inversion (MFI) by quantifying geoacoustic information content for a variety of experiment and array factors, including array length and number of sensors, source range and bearing, source-frequency content, and signal-to-noise ratio (SNR). Emphasis is on bottom-moored arrays, while towed arrays are also considered, and a comparison with vertical line array (VLA) performance is made. The geoacoustic information content is quantified in terms of marginal posterior probability distributions (PPDs) for model parameters estimated using a fast Gibbs sampler approach to Bayesian inversion. This produces an absolute, quantitative estimate of the geoacoustic parameter uncertainties which can be directly compared for various experiment and array factors.

Passive acoustic source localization with multiple horizontal arrays in shallow water

This paper considers concurrent matched-field processing of data from multiple, spatially separated acoustic arrays with application to towed-source data received on two bottom-moored horizontal line arrays from the SWellEx-96 shallow water experiment. Matched-field processors are derived for multiple arrays and multiple-snapshot data using maximum-likelihood estimates for unknown complex-valued source strengths and unknown error variances. Starting from a coherent processor where phase and amplitude is known between all arrays, likelihood expressions are derived for various assumptions on relative source spectral information (amplitude and phase at different frequencies) between arrays and from snapshot to snapshot. Processing the two arrays with a coherent-array processor (with inter-array amplitude and phase known) or with an incoherent-array processor (no inter-array spectral information) both yield improvements in localization over processing the arrays individually. The best results with this data s...

Source Localization With Multiple Hydrophone Arrays via Matched-Field Processing

This paper considers approaches to combining information from multiple arrays in matched-field processing (MFP) for underwater acoustic source localization. The standard approach is to apply conventional MFP for each array independently, and sum the resulting Bartlett ambiguity surfaces computed for each array; this approach assumes that individual arrays comprise calibrated sensors which are synchronized in time. However, if the relative calibration and/or time synchronization is known between some or all arrays, more informative multiple-array processors can be derived using maximum-likelihood methods. If the relative calibration between arrays is known, the observed variation in received signal amplitude between arrays provides additional information for matched-field localization which is absent in the standard processor. If synchronization is known between arrays, phase variations provide additional localization information. Multiple-array processors accounting for different levels of interarray information are derived and evaluated in terms of the probability of correct localization from Monte Carlo analyses for a range of signal-to-noise ratios and the number of frequencies for simulated shallow-water scenarios with multiple horizontal and/or vertical arrays. The analysis indicates that, dependent on array configurations, significant improvements in source localization performance can be achieved when including relative amplitude and/or phase information in the multiple-array processor. The improvement is reduced by environmental and array (calibration and synchronization) mismatch; however, this degradation can be partially mitigated by including additional frequencies in the processing.

Three-dimensional multiple-source focalization in an uncertain ocean environment

This letter develops a Bayesian focalization approach for three-dimensional localization of an unknown number of sources in shallow water with uncertain environmental properties. The algorithm minimizes the Bayesian information criterion using adaptive hybrid optimization for environmental parameters, Metropolis sampling for source bearing, and Gibbs sampling for source ranges and depths. Maximum-likelihood expressions are used for unknown complex source strengths and noise variance, which allows these parameters to be sampled implicitly. An efficient scheme for adding/deleting sources is used during the optimization. A synthetic example considers localizing a quiet source in the presence of multiple interferers using a horizontal line array.