Using the coherence of the frequency-difference autoproduct to facilitate rough surface information identification

Abstract

Acoustic waves, with wavenumber k and incidence angle θ, forward scattered from a randomly rough surface, with root-mean-square roughness height h, lose coherence with the incident field as khcos θ increases. Recovering reflected field-coherence is possible via the frequency-difference autoproduct, a quadratic product of complex field amplitudes at nearby frequencies within the signal bandwidth. By downshifting recorded frequencies, the apparent roughness of a surface is reduced and coherence can be regained. An additional consideration exists, however, as the nonlinearity of the frequency-difference autoproduct introduces a significant dependence on the surface autocorrelation function, and consequently, the surface power spectrum by Fourier transform. The relationship between coherence recovery, surface autocorrelation function, and surface power spectrum is discussed and results are shown. Furthermore, this presentation investigates the potential utility afforded by these relationships for environmental characterization in rough surface scattering. By employing coherence-based frequency-difference autoproduct methods, remote identification of the spectral content, lateral statistics, and vertical statistics of a rough surface are explored. The work uses acoustic data collected at sea during SW06 (off New Jersey, depth 80 m), where the geometry, signal bandwidth, and anisotropic ocean surface conditions put khcos θ > 2.5. [Work supported by ONR and by the US DoD through an NDSEG Fellowship.]