Synthetic-array beamforming for bottom-loss estimation using marine ambient noise

Abstract

Previous studies have shown that, using a vertical line array, the bottom loss can be estimated in an ambient-noise field from the output power of beams steered toward the sea surface and beams reflected off the seabed. With short arrays, the low angular resolution of the bottom-loss estimate is one of the main limitations of the approach. Synthetic-array processing is proposed as a technique that can improve the angular resolution of the bottom-loss estimate to a level comparable to that of an array with twice as many physical sensors (at equal inter-sensor spacing). The proposed technique follows naturally from a new derivation in frequency-wavenumber domain of the bottom-loss estimation procedure. The conditions under which the approach can be successfully applied are analyzed, with particular regard to the need for the array cross spectral density matrix to be “close” to Toeplitz. A metric is proposed for assessing this particular circumstance in practice. The technique is illustrated using synthetic data from simulations and then applied to data from several experimental campaigns. Results show that a 16-element synthetic array can achieve an angular resolution comparable to that of a 32-element array.