Thomas R. Hahn

PROPELLER NOISE FROM A LIGHT AIRCRAFT FOR LOW-FREQUENCY MEASUREMENTS OF THE SPEED OF SOUND IN A MARINE SEDIMENT

The sound from a light aircraft in flight is generated primarily by the propeller, which produces a sequence of harmonics in the frequency band between about 80 Hz and 1 kHz. Such an airborne sound source has potential in underwater acoustics applications, including inversion procedures for determining the wave properties of marine sediments. A series of experiments has recently been performed off the coast of La Jolla, California, in which a light aircraft was flown over a sensor station located in a shallow (approximately 15 m deep) ocean channel. The sound from the aircraft was monitored with a microphone above the sea surface, a vertical array of eight hydrophones in the water column, and two sensors, a hydrophone and a bender intended for detecting shear waves, buried 75 cm deep in the very-fine-sand sediment. The propeller harmonics were detected on all the sensors, although the s-wave was masked by the p-wave on the buried bender. Significant Doppler shifts of the order of 17%, were observed on the microphone as the aircraft approached and departed from the sensor station. Doppler shifting was also evident in the hydrophone data from the water column and the sediment, but to a lesser extent than in the atmosphere. The magnitude of the Doppler shift depends on the local speed of sound in the medium in which the sensor is located. A technique is described in which the Doppler difference frequency between aircraft approach and departure is used to determine the speed of sound at low-frequencies (80 Hz to 1 kHz) in each of the three environments, the atmosphere, the ocean and the sediment. Several experimental results are presented, including the speed of sound in the very fine sand sediment at a nominal frequency of 600 Hz, which was found from the Doppler difference frequency of the seventh propeller harmonic to be 1617 m/s.

Low frequency sound scattering from spherical assemblages of bubbles using effective medium theory

The determination of the acoustic field scattered by an underwater assembly of gas bubbles or similar resonant monopole scatterers is of considerable theoretical and practical interest. This problem is addressed from a theoretical point of view within the framework of the effective medium theory for the case of spherically shaped assemblages. Although being valid more generally, the effective medium theory is an ideal instrument to study multiple scattering effects such as low frequency collective resonances, acoustically coupled breathing modes of the entire assembly. Explicit expressions for the scattering amplitude and cross sections are derived, as well as closed form expressions for the resonance frequency and spectral shape of the fundamental collective mode utilizing analytical S-matrix methods. This approach allows, in principle, a simultaneous inversion for the assembly radius and void fraction directly from the scattering cross sections. To demonstrate the validity of the approach, the theory is applied to the example of idealized, spherically shaped schools of swim bladder bearing fish. The analytic results of the theory are compared to numerical first-principle benchmark computations and excellent agreement is found, even for densely packed schools and frequencies across the bladder resonance.

Acoustic resonances in the bubble plume formed by a plunging water jet

Experiments were performed to investigate the near–field sound from an axisymmetric conical bubble plume formed by a continuous vertical freshwater jet as it penetrates the surface of a pool of fresh water. The volume fluxes of the air and water entering the pool were carefully controlled and monitored during the experiments and a hydrophone detected the acoustic pressure field adjacent to the plume at frequencies between 100 Hz and 1 kHz. Up to five non–uniformly spaced peaks were observed in the pressure spectrum. These spectral peaks are due to coherent collective oscillations of the bubbles within the plume, that is to say, the biphasic bubbly medium behaves as a continuum, acting as a resonant conical cavity beneath the jet. All the eigenfrequencies were found to exhibit inverse–fractional power–law scalings of the same form, fm∂uj–1/2q–1/4, where fm is the frequency of the mth spectral peak, uj is the jet velocity, q is the air entrainment ratio, that is, the ratio of the air–to–water volume fluxes in the jet, and the unspecified constant shows a nonlinear dependence on m. A two–component theoretical model has been developed for the eigenfrequencies of the plume. From a fluid–dynamics argument based on the conservation of momentum flux in the two–phase flow, the speed of sound within the bubbly medium is shown to increase as the square root of depth in the plume. This is incorporated into an acoustic analysis in which the wave equation is solved analytically, taking account of the cone–like geometry of the bubble–plume cavity, including the near–rigid boundary condition at the penetration depth, where the bubbly region ends abruptly. The resultant expression for the frequencies of the lowest–order longitudinal modes of the bubbly cavity exhibits the inverse–fractional power–law scalings observed in the experiments. The experiments and theory are consistent with the conclusion that the scaling of the eigenfrequencies with the inverse square root of the jet velocity stems from the square–root sound–speed profile within the biphasic plume.

Modeling the sound levels produced by bubble release of individual herring.

Herring (Clupea pallasii and C. harengus) are known to release gas from their swim bladder to assist a number of complex behaviors, such as buoyancy adjustments and predator avoidance. The noise associated with the release has recently been reported in the literature and related to oscillating bubbles. Average source levels (SLs) of 73 dB with regard to microPa rms reference 1 m have been reported for bubbles produced by herring in the laboratory. A model is provided for predicting the SL in terms of the gas flow rate from the swim bladder into the bubbles. Based on these laboratory conditions, an inversion yields a rate of 0.9 (0.3-3.2) ml/min. Furthermore, the model predicts an acoustic SL of 89 (79-99) dB with regard to microPa rms reference 1 m for pulses emitted by herring in a natural shallow water environment at unknown distance corresponding to a flow rate of 2.5 ml/min. An analysis of published acoustic data suggests that herring is capable of controlling the gas flow and the corresponding acoustic levels over a wide range according to different behavioral needs. The proposed model allows an extrapolation of the laboratory results to situations that are relevant for bubble release of herring schools in the ocean.

Passive acoustic detection of schools of herring

Herring (Clupea pallasii and C. harengus) have been observed to release gas from their bladders during vertical migration likely to adjust buoyancy and also when under strong predation pressure. Based on recently measured and modeled sound for individual fish, spectral levels are estimated for entire herring schools in the ocean for both scenarios, and the feasibility of passive detection is explored. For a typical school of migrating herring near-surface spectral levels of about 50 dB rel., 1 microPa radical Hz at 3-7 kHz are predicted. If wind conditions are calm where migrating herring are found, such as for Pacific herring in Prince William Sound, Alaska, passive detection is very likely. For an exemplary 10 metric ton compact school, peak spectral source levels of about 80-90 dB rel. 1 microPa radical Hz ref. 1 m are predicted, yielding a range of detection against calm wind background of about 1000 m. Field measurements of potential gas-release events agree with the predictions for the compact school scenario with regard to levels and spectral shape and indicate that passive acoustic monitoring is feasible and could be a prime tool to study predator-prey interactions.

Integrating passive and active acoustics for the assessment of fish stocks

The single most important information for the conservation of exploited marine fish stocks are precise measurements of their biomass so that harvest rates can be established that do not deplete the stock. However, the measurement of marine fish stocks is difficult due to the size, structure, and composition of the ocean, and the highly dynamic movements of the fish. Furthermore, traditional, discrete net sampling approaches have lacked sampling power to assess single fish stocks in time and space [2]. Without the ability to independently measure fish stock biomass with precision, managers have instead relied upon the commercial catch and deterministic indices as a primary source of empirical data. Also, without precise empirical data on stock biomass, the models used to make predictions are unverifiable and highly uncertain. Despite the severe management risks, this is the status quo, and it greatly confounds our efforts to sustain our fisheries, conserve exploited fish stocks and understand the dynamics of population response to natural and anthropogenic changes in the environment. High frequency active acoustics has been used to assess fish stocks for over four decades. When first introduced in the 1970s, there were hopes that acoustics would overcome the marine fish stock measurement problem because of a 105 increase in sampling power. However, prior to the introduction of acoustics the management agencies had already chosen large ocean areas to survey fish stocks, specifically in the summer months when the weather was good. In doing so, the agencies had assumed that surveys in large ocean areas would allow a representative assessment of single stocks of fish.

Acoustic cross sections and resonance frequencies of large fish schools

A prerequisite for stable inversion of bioacoustic parameters of fish schools from large-scale broadband acoustic observations is a theoretical model that permits fast and accurate calculations of acoustic cross sections and school resonance frequencies based on realistic geometrical models of fish schools. The schools of commercially important species, such as sardines, anchovies, and herring, may be characterized by dense nuclei which contain tens of thousands of individuals (N) with separations (S) on the order of one fish length, and diffuse “fuzz” regions with fish at significantly larger separations. Numerical computations of cross sections and school resonances based on the fundamental equations of multiple scattering for point scatterers for these fish school geometries will be presented. Initial results indicate that bubble cloud frequencies of large schools depend primarily on the average spacing between fish in the nuclei and are essentially independent of school size and shape. It will be show...

Passive acoustic determination of herring size.

Over the last decade, acoustic signatures of a variety of fishes have been recorded and analyzed. More recently, these vocalizations have been used to passively detect aggregations of fish, demonstrating the potential to supplement and enhance traditional active acoustic surveys. Based on previously published work on acoustic emissions of herring, this paper discusses the possibility of not merely passively detecting absence or presence of aggregations of herring but, additionally, assessing abundance and size. Theoretical considerations as well as data collected in Prince William Sound, Alaska, on Pacific herring (Clupea Pallasii) are presented.

The effects of multiple scattering on bubble cloud and fish school cross sections.

Scattering or transmission of sound from or through clouds of bubbles or schools of swim bladder bearing fish is a problem of considerable practical importance in acoustical oceanography that has received ongoing interest from the community for many years. Part of this interest is rooted in the fact that mutual acoustic interactions of the scatterers in the form of multiple scattering complicate the physics of the problem. In practice, this can be both a limitation, or, if physically well understood, an opportunity that can lead to more powerful inversion schemes. In this paper, several multiple scattering effects, such as collective resonance modes and frequency shifts are explored using an effective medium approach as well as direct forward numerical computations. While conceptually of great utility, analytically treated effective medium techniques are mostly limited to simple geometries. On the other hand, direct numerical computations from first principles can accommodate complex, realistic configurations of a large number of scatterers. [Work supported by the Office of Naval Research.]

Passive Acoustic Detection and Monitoring of Schools of Herring

Passive acoustic detection and monitoring of various marine fishes has recently received much attention in the literature. It has been recognized that passive acoustic techniques have the potential to complement traditional active acoustic surveys and to significantly increase their overall efficiency, if the acoustic signatures of the considered species are well understood. In this paper, the potential of passive acoustic techniques is explored for the specific case of Pacific herring (Clupea palassii). It is demonstrated that schools of herring can acoustically be detected by observing the sound of coordinated bubble release, triggered, e.g., by predator activity. This sound not only has identifiable features that can be exploited for determining the presence or absence by simple means, but could also carry abundance and size information. Work supported by ONR and the NMFS via the PWSSC.