Acoustic Reverberation Research Articles

Acoustic reverberation from a laboratory model of a shelf break

Continental boundaries have a shallow sloping bottom that changes to a steeper slope. Our laboratory acoustic models (simulations) have (1) a 11° slope that changes to a 50.5° slope and (2) a 35° slope change to a 59° slope. The surface and bottom of the acoustic model are dry wall construction board. The source is a spark and the receiver is a small microphone. The Biot–Tolstoy exact time-domain solution was used. [I. Tolstoy and C. S. Clay, Ocean Acoustics (American Institute of Physics, New York, 1987), 2nd ed., Appendix 5, and C. Feuillade and C. S. Clay, J. Acoust. Soc. Am. 96, 501–514 (1994)]. The image-reflection and diffraction components depend on source and receiver locations. In model (1), source and receiver were within the 11° wedge and gave reverberations that were mainly from the 11° wedge. Source and receiver beyond the shelf break gave reverberations that were mainly from the 50° part of the model. In model (2), the source was up slope and receiver was near the break to observe diffractions. The finite number of arrivals (images) obeyed the Biot–Tolstoy rule. Experimental signals matched the theory in amplitudes and arrival times. Most reverberations were caused by multiple reflections in the wedge waveguide. The amplitudes of the image-reflection paths were an order of magnitude larger than the diffraction arrivals.

A simple approach for modeling acoustic scattering from rocks and sea shells lying on the ocean floor

Acoustic reverberation in the ocean is due to collective scattering from a large number and diverse range of distributed objects. In this work, a model for predicting the sonar target strengths of an important class of clutter objects, i.e., rocks and sea shells lying on the ocean floor, is presented. The model incorporates a simplified approach, in which rocks are represented by either rigid movable spheres or elastic spheres, a technique which has been previously used to model aqueous suspensions of sand over the frequency range of interest. Sea shells are represented by water-filled spherical shells. The model predicts scattering strength values for areas of the ocean floor covered by randomized arrangements of rocks and shells, using sizes, numbers, geologic type, and material compositions typically found on the bottom. The scattering as a function of azimuth is determined by coherent addition, and allows both monostatic and bistatic geometries to be investigated. [Work supported by ONR/NRL]

Sound-speed measurements in the surface-wave layer

Wave breaking at the surface of the ocean entrains bubbles, significantly modifying the phase speed and attenuation of acoustic waves propagating through the resulting two-phase medium. An autonomous buoy system was developed that directly measures sound speed at 3.33, 5, and 10 kHz at seven depths ranging from 0.7 to 7 m through the use of a travel-time technique. Simultaneous measurements at each depth are obtained at a 2-Hz rate, allowing observation of the unsteady sound-speed field from individual bubble injection events, as well as the calculation of mean sound speeds. The travel-time technique allows a direct measurement of the sound speed, eliminating the uncertainties common with inferring sound speeds from bubble population data. The sound speed buoy was deployed in the North Atlantic during the winter of 1993–94 as part of the Acoustic Surface Reverberation Experiment (ASREX). Our aim was to characterize the highly variable near-surface sound-speed field under varying environmental conditions. Forty-three days of data were obtained spanning several storm cycles with wind speeds and significant wave heights reaching 20 m/s and 8 m, respectively. During periods of intense wave breaking, average sound speeds below 1000 m/s were observed at the 0.7-m measurement depth while instantaneous sound speeds during individual events approached values as low as 300 m/s. Furthermore, the data suggest that the dispersive effects of bubbles may extend to frequencies as low as 5 kHz near the surface during storms. Strong correlations of the mean and rms sound speed with the overlying wind and wave fields were found.

Implications of nonfractal seafloor stochasticity on acoustical scattering from the Mid-Atlantic Ridge

Acoustical and bathymetric data were collected near the Mid-Atlantic Ridge as part of the Acoustical Reverberation Special Research Program (ARSRP) in 1993. The wideband time-domain envelope statistics of backscatter from prominent bathymetric features exhibit a non-Rayleigh character which previous researchers have described as event-like. The envelope probability density functions (pdf’s) show enhanced tails at high levels which can be a source of active sonar clutter. It is proposed that the origin of such clutter is single-scale, nonfractal roughness. High-resolution bathymetry reveals a power spectral density (PSD) with power-law form. Wavelet analyses reveal a single-scale character to the roughness. In order to test the hypothesis, multiscale and single-scale realizations of the measured PSDs are used in numerical simulations of time-domain backscatter. At the range and azimuthal resolutions in the experiment, the envelope pdfs for multiscale surfaces are Rayleigh, while the single-scale surfaces lead to enhanced tails as observed in the data. The conclusions drawn are that along with an interface’s rms height, correlation length, and power spectral shape, one should also be concerned with its scale structure because it plays an important role in the physics of wave interaction at random interfaces.

Modeling high-frequency reverberation data from coastal waters

High-frequency (40–180 kHz) acoustic reverberation data have been taken from a region 26 miles south of Panama City, Florida. The data sets were obtained from a series of bottom scattering measurements taken in an area where the bottom was covered with coarse sand and shell fragments. Two different high-frequency scattering models designed for high-frequency reverberation predictions had previously failed to match the data. These data have been reexamined using highly accurate ocean acoustic computer models based on the finite-element solution to the Helmholtz wave equation. While these models were designed for lower frequency predictions, the availability of supercomputers allowed us to apply these models at these high frequencies. In this study, new comparisons are given to the data and to the previous modeling predictions. [Work supported by ONR/NRL and a High Performance Computing DoD grant.]

A Preliminary Investigation of the Perceived Smallest Sound Duration Change in a Room

The present work aims at producing a test, similar to a speech intelligibility test, which can be used to rate the acoustic quality of rooms for music. Listeners' perceptions of the duration of sounds were investigated in an attempt to find a useful indicator of the acoustics of a room for music and speech. The research design was based on a discrimination task and a 2AFC experimental procedure. Two experiments were carried out in which two acoustic variables, reverberation time and background noise level, were varied and listeners' smallest perceivable sound duration changes were measured. It was found that the listeners' duration perceptions were significantly influenced by the reverberation time and background noise level of the listening environment, and that these discrimination procedures may form the basis for room acoustics assessments.

Detection of migrating salmon in the Fraser River using 100-kHz sidescan sonars

During a 10-day period in September 1995, field trials were conducted in the Fraser River near Mission, B.C., to evaluate 100-kHz side-looking sonars for salmon detection. A two-element sidescan installation looking transverse to the river flow detected echo traces due to salmon at ranges up to 250 m in water 6-13 m deep. The salmon target trajectories were identified and counted against a background of acoustic reverberation from the river surface and bottom sediments. Occasional, strong interference from boat wakes and wind-induced surface waves and bubbles was observed. Fish echo trajectory analyses allowed extraction of upstream and cross-stream swimming speeds. For some targets acoustic multipaths from surface and bottom reflections were distinguishable from direct echoes at ranges up to 150 m. Acoustic models of surface and bottom boundary backscattering, combined with bubble layer scattering and attenuation under breaking wave conditions, were able to match the observed background echo levels. Comparisons of model and data predict that under ideal conditions fish detection should be possible at ranges up to 250 m, but under windy conditions detection was limited to less than 70 m.

Multivariate Fourier and wavelet analysis of acoustic backscattering and environmental data from the acoustic surface reverberation experiment

Data from the acoustic surface reverberation experiment (ASREX) [Williams et al., J. Acoust. Soc. Am. 97, 3404 (1995)], is analyzed using multivariate Fourier and wavelet techniques to investigate relationships between environmental conditions and observed acoustic surface backscatter at 200, 400, and 800-Hz frequencies. Environmental variables observed included wind speed and direction, wave height, air and sea temperatures, current speed and direction, ambient noise, and various measurements of bubble activity characterized by depth-integrated air volume and e-folding depth. The relative contribution of the environmental parameters to observed backscattering strengths at the various frequencies will be discussed. The analysis techniques will be described, compared, and contrasted. [Work supported by ONR.]

Modeling low-frequency reverberation near the Mid-Atlantic Ridge and comparison with ARSRP data

The Acoustic Reverberation Special Research Program (ARSRP) of the Office of Naval Research conducted low-frequency acoustic reverberation experiments just west of the Mid-Atlantic Ridge in 1993. Analyses of these data are supported by a bathymetric survey that produced nearly full-coverage bathymetry gridded to a resolution of 200×200 m. At this scale direct-path reverberation data has been successfully modeled using Lambert’s law applied to the local grazing angle determined from the high-resolution bathymetry. The analysis suggests that there could be a slightly weaker dependence on grazing angle than sine squared. The good agreement of this simple model with data shows that, for this region, two-dimensional seafloor morphology at the proper scale determines the ability to predict reverberation. For a model, such as Lambert’s law, that glosses over the details of seafloor microroughness or texture at the scale the acoustic wavelength (λ=6 m), the seafloor morphology scale that is critical is shown to be an order of magnitude or two greater than an acoustic wavelength. An average Lambert coefficient for the region is between −14 and −17 dB, but it is less than −20 dB for deeply sedimented areas and around −12 dB for rugged exposed rock areas.

On the use of acoustic ambient noise to characterize backscattering strength

Strong correlation between acoustic ambient noise and backscattering strength was observed during analysis of data from the acoustic surface reverberation experment (ASREX) [Williams et al., J. Acoust. Soc. Am. 97, 3404 (1995)]. Traditional empirical formulas for acoustic backscattering strength, such as those of Chapman and Harris [R. P. Chapman and J. H. Harris, J. Acoust. Soc. Am. 34, 1592–1597 (1962)] and Ogden and Erskine [P. M. Ogden and F. T. Erskine, J. Acoust. Soc. Am. 95, 746–761 (1994)], utilize wind speed, frequency, and grazing angle as their input variables. An alternative approach which utilizes ambient noise in place of wind speed is presented here. Unlike wind speed, ambient noise measurements are easily made from subsurface platforms and this technique may therefore be advantageous for certain applications (i.e., submarines and covert operations). [Work supported by ONR.]

Microphone array systems for hands-free telecommunication

Microphone array systems can be effective in combating the detrimental effects of acoustic noise and reverberation in hands-free telecommunication. This paper discusses classical delay-sum beamformers as well as the more general filter-sum beamformers. Filter-sum beamformers add the ability to control the array beampattern as a function of frequency and a new design method for a constant beamwidth filter-sum beamformer is presented. The delay-sum and filter-sum beamformers require array sizes that are comparable to the acoustic wavelength. These designs can result in arrays that are large in size. For applications that are space constrained, differential microphone array systems are presented. Finally, two types of adaptive beamformers are presented: a broadside array and a two-element differential microphone.

Surface wave parameters for the prediction of breaking waves and acoustic backscatter

Results from the acoustic surface reverberation experiment (ASREX) and critical sea test 7 (CST-7 phase 2) are examined to determine if surface wave parameters can be used to augment wind speed in air–sea process models. The wave parameters are extracted from directional wave spectra using a fully automated spectral partitioning approach. Isolated directional wind–sea spectra provide the basis for estimating the total rate of energy dissipation by breaking waves. The method employs Phillips’ equilibrium range theory and extends Felizardo and Melville’s recent work obtaining wave dissipation estimates from one-dimensional wave spectra. Preliminary results are encouraging; wave dissipation reduces the scatter in whitecap fraction observations by several orders of magnitude when wind speed is replaced with wave dissipation as the independent variable in the standard power-law model for whitecap fraction. It will be demonstrated that the success of this approach is partially attributed to a wind history bias that is automatically compensated for in the wave parameters. Results of similar on-going investigations using an extensive ambient noise and surface scatter observation set, collected during ASREX by the University of Miami, will also be discussed. [Work supported by ONR Code 32OA.]

Fluctuations of the acoustic field Bragg-backscattered from rough surfaces

Abstract A rigorous theory of backscattering from slightly rough, statistically non-stationary surfaces is developed. The development proceeds through the application of two ensemble averages: the first is applied to all realizations of the surface having fixed non-stationary features, and the second over all non-stationary surface characteristics. The resulting theory is applied to acoustic sea-floor reverberation, where it is found that non-stationarity must be assumed to obtain agreement with observed reverberation fluctuations. The analysis is further extended to the time domain where it is demonstrated that, for Bragg scatter, no reduction in the predicted fluctuations occurs for frequency diverse waveforms, thus providing a method of differentiating Bragg scatter from other non-resonant scattering processes.

Low cost methods for mitigating hands‐free acoustics problems.

Key problems in hands‐free acoustics for teleconferencing are acoustic echo cancellation, reverberation, noise, varying amplitude, and restricted audio bandwidth. None of these problems can be completely solved via low‐cost methods, but their nuisance can be significantly reduced. For acoustic echo cancellation, a sub‐band approach has been found to work well, where the speech is divided up into many bands and echo cancellation occurs individually on each band. Reverberation in speech pickup may be significantly reduced via a simple two dipole, automatically steered array which forms a dipole pattern which may be electronically rotated a full 360 deg. Background noise from fans may be suppressed using spectral subtraction techniques. Varying amplitudes of speech may be addressed by means of a voice‐activated AGC (automatic gain control), where the voice detector eliminates pumping artifacts usually found in agressive AGC schemes. Finally the use of 7‐kHz digital audio compression schemes alleviates the problems of muffled sounding speech due to the 3.4‐kHz restricted audio bandwidth found on telephones.

Acoustical investigation of phytoplankton

The existence of gas vacuoles and the dispersion of sound in water containing phytoplankton is demonstrated under natural conditions by measurement of acoustic reverberation and sound velocity, using phase, resonance, and other methods. The volume of vacuoles ranges from 0.1-30% of the cell body volume. Two types of gas cavities are observed: one is diminished elastically under compression, while the other is decreased irreversibly. However, the volume is restored in a few hours with exposure to light. The sound velocity has been measured over the frequency range 10-30 kHz, in water containing phytoplankton cells at a volume concentration of 10 −4 . The sound velocity increased on average by 0.1% over this frequency range.

Seafloor acoustic backscattering from different geological provinces in the Atlantic Natural Laboratory.

The characteristics of acoustic signals backscattered from inside‐corner oceanic crust and outside‐corner crust are investigated using acoustic reverberation data from the 1993 ARSRP Acoustics Cruise. Specifically, the seafloor dip distribution is compared between areas of each crustal type and correlations are sought between true grazing angle and backscattered signal strength. Beamformed and match filtered acoustic data from the monostatic, wideband, LFM experiment at Site A are used to find the scattering strength corresponding to specific areas of the seafloor. Scattering strength determined from intersecting beams of different segments are summed to reduce left–right ambiguity. At the scale of analysis, high scattering strengths are found to correspond to steep flanks of seafloor features and can be used to map their shape and orientation. Some of these features are characteristic of specific crustal regions. Scattering strength is found to increase with increasing effective grazing angle at a rate of about 0.1±0.01 dB/degree. However dip values are so scattered that the seafloor dip, on the scale of a few hundred meters cannot be used to predict backscatter intensity. Some other characteristics of steeply dipping areas, such as subsurface heterogeneities or smaller scale surface features, strongly influence the level of backscattered signals. [Work supported by ONR.]

Backscattering mechanisms for reverberation from deep ocean features using ARSRP data taken at the Mid‐Atlantic Ridge.

In July 1993 SACLANTCEN participated in an experiment for the acoustic reverberation special research program (ARSRP). The primary objective was to take high‐resolution measurements to determine the detailed physical processes dominating the low‐frequency scattering from rough topographic features and from deep sediment pond areas. A very detailed set of monostatic and bistatic scattering experiments were conducted (using low frequencies from 200–375 Hz) just west of the Mid‐Atlantic Ridge near 26° N and 47° W. This paper presents towed array data recorded by SACLANT Centre’s R/V ALLIANCE for comparison with FEPE two way parabolic equation model estimates of selected ARSRP geometries. The receivers were horizontal arrays of 128 elements spaced at 0.5, 1, and 2 m. Source/receiver depths were ≊130 and 450 m, respectively. Inferences for monostatic bottom scattering mechanisms are discussed based on the FEPE model from Collins to interpret the data. The modeling work isolates different scattering mechanisms to assess their relative contributions to the overall reverberation levels observed in the data.

Vertical spatial and temporal coherence of ocean bottom reverberation.

An experiment to study acoustic backscattering from deep‐ocean sediments was conducted in July 1993 as part of the Acoustic Reverberation Special Research Program (ARSRP). An acoustic source transmitting chirp signals in the frequency range 250–650 Hz and a 24‐element vertical receiving array attached to the source were suspended near the seafloor over a sediment pond in the vicinity of the Mid‐Atlantic Ridge. In a previous paper [Tang et al., J. Acoust. Soc. Am. 98 508–516 (1995)], a study on bottom scattering due to sediment volume inhomogeneities in this area was presented. Here, the spatial and temporal coherence of the scattered field using the data collected on the vertical array will be examined. The scatterers are modeled as point scatterers. It is shown that when the coherently reflected signals due to sediment layering are removed from the measurements, the results of the point scatter coherence model agree satisfactorily with data. [Work supported by ONR.]

Propagation and analysis issues in the prediction of long-range reverberation

Data collected from the Office of Naval Research-special research program (ONR-SRP) bottom reverberation research cruises consist of various environmental measurements (sound-speed profiles, bottom properties, etc.), extensive bathymetric mapping of the 300 km×150 km natural laboratory, and high-quality acoustic reverberation data recorded in both monostatic and bistatic geometries. The analysis of the acoustic data and the effects of propagation are investigated. Accurate GPS tracking of the participating research vessels provided the precision necessary to attempt to correlate ‘‘spiky’’ reverberation events with bathymetric features. Understanding the limitations of our ability to resolve such features—due to imperfect signal processing, environmental variability, and complex multipath structures—is the main objective of this work. Using a simplified time-to-range conversion, the measured reverberation data can be displayed over the local bathymetry. This process shows good correlation between large-scale [O(km)] high-level returns and bathymetric ridge structures. Employing a PE propagation model in a quasi-cw manner, many of the large-scale reverberation features can be shown to correlate well with predicted high-level ensonification of the total field [low transmission loss (TL)]. Extending the model predictions to include time domain broadband pulse propagation, it can be shown for ranges greater than ∼20 km that multipaths and forward scatter produce complex ensonification patterns which are unresolvable in time and/or range. The effect of this on the ability to correlate finer scale returns and bathymetric features will be discussed. Preliminary analysis of data from a bottom mounted vertical array in the vicinity of 1/2 CZ exhibits strong evidence for the existence of these predicted multipaths and provides a means of confirming the relative intensities predicted by propagation models.

Modeling of the Acoustic Reverberation Special Research Program deep ocean seafloor scattering experiments using a hybrid wave propagation simulation technique

Quantitative modeling of bottom‐interacting ocean acoustic waves is complicated by the long propagation ranges and by the complexity of the scattering targets. We employ a two‐dimensional (2‐D) hybrid technique combining Gaussian beam, finite difference, and Kirchhoff integral solutions of the wave equation to simulate ocean acoustic experiments within half of a convergence zone in the SOFAR channel. The 2‐D modeling approach is reasonable due to the one‐dimensional (1‐D) velocity distribution in the water column and the strong lineation of the seafloor morphology parallel to the mid‐ocean ridges. Full‐waveform modeling of ocean acoustic data requires that the topography and the material properties of the seafloor are available at scales that are several orders of magnitude smaller than typical bathymetric sampling rates. We have therefore investigated the effects on the ocean acoustic response of a stochastic interpolation scheme used to generate seafloor models. For typical grazing angles of the incident wave field (approximately 5°–20°), we found that different stochastic realizations of the same seafloor segment (sampled at 200 m) yield an intrinsic uncertainty of the order of 3–8 dB in amplitude and 0.1–0.3 s in time for individual prominent events in the reverberant acoustic field. Hybrid simulations are compared to beam‐formed ocean acoustic data collected during the Acoustic Reverberation Special Research Program (ARSRP) cruises. Side lobe noise in the observed acoustic data is simulated by adding band‐limited white noise at −30 dB relative to the maximum intensity in the synthetic data. Numerical simulations can be limited to the response of only one of the mirror azimuth beams provided that the experimental geometry is suitably chosen. For the 2‐D approximation to be valid, the cross‐range resolution of the observed data must be smaller than the characteristic scale of seafloor lineations, and the beams of interest must be approximately perpendicular to the dominant structural grain. Under these premises, the arrival time and maximum intensity of the observed backscattered acoustic events can be modeled within 0.3 s and 5–10 dB, respectively. The mismatch in arrival time is interpreted to be due predominantly to intrinsic uncertainties in the stochastic interpolation of the seafloor profiles, whereas the fact that the intensity of the backscattered events is systematically overestimated is attributed to 3‐D effects within the “footprint” of the beam and/or to underestimated noise levels.