Index of content:
Volume 125, Issue 3, March 2009
- UNDERWATER SOUND 
125(2009); http://dx.doi.org/10.1121/1.3075765View Description Hide Description
Modal structure of the wave field in a deep ocean environment with sound speed fluctuations induced by random internal waves is considered. An approximate analytical description of the modal structure at megameter ranges is derived by combining two known results: (i) relations expressing mode amplitudes through parameters of ray paths and (ii) stochastic ray theory. For a monochromatic wave field, a simple analytical estimate has been obtained for a coarse-grained distribution of acoustic energy between normal modes. Significant attention has been paid to the investigation of the mode pulses, that is, sound pulses carried by individual modes. Analytical estimates for the spread of mode pulse and bias of its mean travel time in the presence of internal waves are derived.
Prediction of acoustic radiation from axisymmetric surfaces with arbitrary boundary conditions using the boundary element method on a distributed computing system125(2009); http://dx.doi.org/10.1121/1.3056467View Description Hide Description
This paper presents a computational technique using the boundary element method for prediction of radiated acoustic waves from axisymmetric surfaces with nonaxisymmetric boundary conditions. The aim is to predict the far-field behavior of underwater acoustic transducers based on their measured behavior in the near-field. The technique is valid for all wavenumbers and uses a volume integral method to calculate the singular integrals required by the boundary element formulation. The technique has been implemented on a distributed computing system to take advantage of its parallel nature, which has led to significant reductions in the time required to generate results. Measurement data generated by a pair of free-flooding underwater acoustic transducers encapsulated in a polyurethane polymer have been used to validate the technique against experiment. The dimensions of the outer surface of the transducers (including the polymer coating) were an outer diameter of with an wall thickness and a length of . The transducers were mounted coaxially, giving an overall length of . The cylinders had resonance frequencies at 13.9 and , and the data were gathered at these frequencies.
125(2009); http://dx.doi.org/10.1121/1.3076202View Description Hide Description
The possibility of exploiting low-resolution acoustic signals used for the navigation of Lagrangian floats to simultaneously estimate the speed of sound is studied. Acoustic navigation of Lagrangian floats is regularly carried out by measuring travel times from three fixed stations assuming a known value for the speed of sound. The sound speed is considered here as a variable of the problem to be estimated from the travel-time data simultaneously with the horizontal location of the float. The estimation problem is linearized and solved analytically, and closed-form expressions for the sound-speed estimation errors are derived. Typical acoustic navigation (RAFOS) signals are characterized by limited time resolution challenging the accuracy of sound-speed estimation, depending on the location of the float with respect to the fixed stations. By exploiting travel-time data from multiple floats, the sound-speed estimation accuracy can be increased, which reflects in higher localization accuracy as well. In the case of a single float improved sound-speed estimates and localization results can be obtained by combining travel-time data from different float locations. Numerical results verify the theoretical error estimates and demonstrate the efficiency of the method.
125(2009); http://dx.doi.org/10.1121/1.3075558View Description Hide Description
Although a significant amount of theoretical and numerical modeling effort has been put into the study of three-dimensional (3D) acoustic propagation on a coastal wedge, including the development of the ASA 3D benchmark problem set, there have been few observations of the predicted 3D propagation effects. Significant horizontal multipath arrivals were observed in a pair of acoustic transmission tests on the continental shelf off the east coast of Florida in September 2007 and February 2008. For many transmissions, arrivals were received coming from nearly the global positioning system(GPS) bearing of the ship, as well as up to 30 deg inshore of the true bearing. The inshore path was up to 25 dB stronger than the direct path in some cases. The experimental waveforms transmitted included continuous-wave transmissions ranging in frequency from 24 to 415 Hz as well as wideband linear frequency modulation pulses (20–420 Hz). Horizontal multipath arrivals were observed for source ranges from 10 to 80 km, source depths of 20 and 100 m, and along several different bearings (inshore and along the 250 m isobath). It is a conclusion of this paper that the bearing bias and multiple horizontal arrivals are the result of 3D propagation due to the local shoaling bathymetry.
125(2009); http://dx.doi.org/10.1121/1.3076201View Description Hide Description
This paper describes a method of estimating hydrophone positions in a receiver array using the noise from a passing ship. Relative arrival times of the ship-noise signal between pairs of hydrophones are obtained from several time windows of data (corresponding to different ship locations) by cross-correlating the band-pass filtered time series. The relative arrival times are used as data in an array element localization inversion to estimate both the hydrophone and ship locations based on iterated linearization of the acoustic ray equations. The inversion applies the method of regularization to include prior information such as approximate location estimates and uncertainties for the source and receivers and the expectation that the array shape and∕or source tracks are smooth functions of position. Linearized and nonlinear (Monte Carlo) estimates of the position errors are in good agreement and indicate a high degree of confidence in the receiver positions (relative uncertainties of approximately in the horizontal and in the vertical). The ability to improve upon the initial source position estimates depends on the geometry of the problem, as investigated with simulations.