Modeling the effect of random roughness on synthetic aperture sonar image statistics.

A model has been developed to predict the effect of random seafloor roughness on synthetic aperture sonar (SAS) image statistics, based on the composite roughness approximation-a physical scattering model. The continuous variation in scattering strength produced by a random slope field is treated as an intensity scaling on the image speckle produced by the coherent SAS imaging process. Changes in image statistics caused by roughness are quantified in terms of the scintillation index (SI). Factors influencing the SI include the seafloor slope variance, geo-acoustic properties of the seafloor, the probability density function describing the speckle, and the signal-to-noise ratio. Example model-data comparisons are shown for SAS images taken at three different sites using three different high-frequency SAS systems. Agreement between the modeled and measured SI show that it is possible to link range-dependent image statistics to measurable geo-acoustic properties, providing the foundation necessary for solving problems related to the detection of targets using high-frequency imaging sonars, including performance prediction or adaptation of automated detection algorithms. Additionally, this work illustrates the possible use of SAS systems for remote sensing of roughness parameters such as root mean square slope or height.

Resolution dependence of rough surface scattering using a power law roughness spectrum.

Contemporary high-resolution sonar systems use broadband pulses and long arrays to achieve high resolution. It is important to understand effects that high-resolution sonar systems might have on quantitative measures of the scattered field due to the seafloor. A quantity called the broadband scattering cross section is defined, appropriate for high-resolution measurements. The dependence of the broadband scattering cross section, σbb, and the scintillation index, SI, on resolution was investigated for one-dimensional rough surfaces with power-law spectra and backscattering geometries. Using integral equations and Fourier synthesis, no resolution dependence of σbb was found. The incoherently averaged frequency-domain scattering cross section has negligible bandwidth dependence. SI increases as resolution increases, grazing angle decreases, and spectral strength increases. This trend is confirmed for center frequencies of 100 and 10 kHz, as well as for power-law spectral exponents of 1.5, 2, and 2.5. The hypothesis that local tilting at the scale of the acoustic resolution is responsible for intensity fluctuations was examined using a representative model for the effect of slopes (inspired by the composite roughness approximation). It was found that slopes are responsible in part for the fluctuations, but other effects, such as multiple scattering and shadowing may also play a role.

Direct inference of first-year sea ice thickness using broadband acoustic backscattering.

Accurate measurements of sea ice thickness are critical to better understand climate change, to provide situational awareness in ice-covered waters, and to reduce risks for communities that rely on sea ice. Nonetheless, remotely measuring the thickness of sea ice is difficult. The only regularly employed technique that accurately measures the full ice thickness involves drilling a hole through the ice. Other presently used methods are either embedded in or through the ice (e.g., ice mass balance buoys) or calculate thickness from indirect measurements (e.g., ice freeboard from altimetry; ice draft using sonars; total snow and ice thickness using electromagnetic techniques). Acoustic techniques, however, may provide an alternative approach to measure the total ice thickness. Here laboratory-grown sea ice thicknesses, estimated by inverting the time delay between echoes from the water-ice and ice-air interfaces, are compared to those measured using ice cores. A time-domain model capturing the dominant scattering mechanisms is developed to explore the viability of broadband acoustic techniques for measuring sea ice thickness, to compare with experimental measurements, and to investigate optimal frequencies for in situ applications. This approach decouples ice thickness estimates from water column properties and does not preclude ice draft measurements using the same data.

Scattering statistics of rock outcrops: Model-data comparisons and Bayesian inference using mixture distributions.

The probability density function of the acoustic field amplitude scattered by the seafloor was measured in a rocky environment off the coast of Norway using a synthetic aperture sonar system, and is reported here in terms of the probability of false alarm. Interpretation of the measurements focused on finding the appropriate class of statistical models (single versus two-component mixture models), and on appropriate models within these two classes. It was found that two-component mixture models performed better than single models. The two mixture models that performed the best (and had a basis in the physics of scattering) were a mixture between two K distributions, and a mixture between a Rayleigh and generalized Pareto distribution. Bayes' theorem was used to estimate the probability density function of the mixture model parameters. It was found that the K-K mixture exhibits a significant correlation between its parameters. The mixture between the Rayleigh and generalized Pareto distributions also had a significant parameter correlation, but also contained multiple modes. It is concluded that the mixture between two K distributions is the most applicable to this dataset.

Measurements of two-dimensional spatial coherence of normal-incidence seafloor scattering.

Measurements have been made near normal incidence of the two-dimensional spatial coherence of the acoustic field scattered from the lakebed in Seneca Lake, New York. In the test region, the lakebed consists of a series of sediment layers created by a sequence of distinct depositional processes. The spatial coherence length of the scattered field is shown to be dependent on the structure of the underlying sediment sequences. Significant ping-to-ping variability in the spatial coherence surface was also observed for each sediment sequence. This variability is quantified by a two-dimensional spatial coherence metric that measures the coherence lengths and asymmetric coherence surface orientation. The ping-to-ping variation of the surface asymmetry appears to be linked to the spatial isotropy of the sediment scattering strength. The scattering strength of the deepest observed sequence in the sub-bottom is the most spatially isotropic and the ping-to-ping variability of the coherence lengths and surface orientations are random. The scattering strength of the shallower sequences is spatially anisotropic and the coherence lengths and surface orientations show intervals of non-random ping-to-ping behavior.

An Empirical Mode Decomposition-based detection and classification approach for marine mammal vocal signals.

Detecting marine mammal vocalizations in underwater acoustic environments and classifying them to species level is typically an arduous manual analysis task for skilled bioacousticians. In recent years, machine learning and other automated algorithms have been explored for quickly detecting and classifying all sound sources in an ambient acoustic environment, but many of these still require a large training dataset compiled through time-intensive manual pre-processing. Here, an application of the signal decomposition technique Empirical Mode Decomposition (EMD) is presented, which does not require a priori knowledge and quickly detects all sound sources in a given recording. The EMD detection process extracts the possible signals in a dataset for minimal quality control post-processing before moving onto the second phase: the EMD classification process. The EMD classification process uniquely identifies and labels most sound sources in a given environment. Thirty-five recordings containing different marine mammal species and mooring hardware noises were tested with the new EMD detection and classification processes. Ultimately, these processes can be applied to acoustic index development and refinement.

A metric for characterization of two-dimensional spatial coherence.

A metric is developed providing a quantitative measure of the two-dimensional spatial coherence of scattered fields. The metric is based on fitting a function similar to bivariate Gaussian to measured two-dimensional coherence surfaces. This function provides a robust fit to the measured data for a range of coherence lengths and surface asymmetries. Through an eigendecomposition of the bivariate Gaussian covariance matrix, it is possible to define surface orientation as well as coherence lengths along the major and minor axes. The metric is applied to normal-incidence scattering data collected in recent field trials at Seneca Lake, NY.

Measurements of high-frequency acoustic scattering from glacially eroded rock outcrops.

Measurements of acoustic backscattering from glacially eroded rock outcrops were made off the coast of Sandefjord, Norway using a high-frequency synthetic aperture sonar (SAS) system. A method by which scattering strength can be estimated from data collected by a SAS system is detailed, as well as a method to estimate an effective calibration parameter for the system. Scattering strength measurements from very smooth areas of the rock outcrops agree with predictions from both the small-slope approximation and perturbation theory, and range between -33 and -26 dB at 20° grazing angle. Scattering strength measurements from very rough areas of the rock outcrops agree with the sine-squared shape of the empirical Lambertian model and fall between -30 and -20 dB at 20° grazing angle. Both perturbation theory and the small-slope approximation are expected to be inaccurate for the very rough area, and overestimate scattering strength by 8 dB or more for all measurements of very rough surfaces. Supporting characterization of the environment was performed in the form of geoacoustic and roughness parameter estimates.

Consistency in statistical moments as a test for bubble cloud clustering.

Frequency dependent measurements of attenuation and/or sound speed through clouds of gas bubbles in liquids are often inverted to find the bubble size distribution and the void fraction of gas. The inversions are often done using an effective medium theory as a forward model under the assumption that the bubble positions are Poisson distributed (i.e., statistically independent). Under circumstances in which single scattering does not adequately describe the pressure field, the assumption of independence in position can yield large errors when clustering is present, leading to errors in the inverted bubble size distribution. It is difficult, however, to determine the existence of clustering in bubble clouds without the use of specialized acoustic or optical imaging equipment. A method is described here in which the existence of bubble clustering can be identified by examining the consistency between the first two statistical moments of multiple frequency acoustic measurements.

Normal incidence reflection loss from a sandy sediment.

Acoustic reflection loss at normal incidence from a sandy sediment, in the Biodola Gulf on the north side of the island of Elba, Italy, was measured in the frequency band 8-17 kHz, using a self-calibrating method. The water depth was approximately 11 m. The mean and standard deviation of the sand grain diameter were 2.25 (0.21 mm) and 0.6 phi, respectively. The reflection loss was measured using an acoustic intensity integral method, which is insensitive to roughness effects within the selected frequency band. The measured value of reflection loss was 11 dB, +/- 2 dB. The result is consistent with previous measurements in the published literature. The computed reflection loss for a flat interface between water and a uniform fluid or visco-elastic medium with the same properties is 8 dB, +/- 1 dB. The theoretical and experimental values do not significantly overlap, which leads to the conclusion that the visco-elastic model is inappropriate. The Biot model is suggested as a better alternative but more work is needed to ascertain the appropriate parameter values.