Modelling meter‐scale acoustic intensity fluctuations from oceanic fine structure and microstructure

Abstract

Using model spectra of sound speed fluctuations, we have calculated the intensity variance and the wave number spectrum of intensity for acoustic propagation from a point source. We consider solutions in the unsaturated (Rytov) approximation, which is accurate when the intensity variance, normalized to unity mean intensity, is less than about 0.3. Our model medium spectra are consistent with observed vertical profiles of sound speed. Model fine‐scale sound speed inhomogeneities are characterized by the Garrett‐Munk (GM) internal wave spectrum or a simpler anisotropic (A) inverse power law spectrum. Microscale sound speed structure is modelled by a band‐limited isotropic Kolmogorov inertial subrange spectrum. With the GM model medium, the Rytov approximation is typically valid when the combination σR2 of frequency σ and range R is less than that given by 90 kHz and 1.1 km. Experiments within this window can determine characteristics of the fine scale medium spectra. In order to probe microstructure effects, frequencies near the high limit are necessary so that the Fresnel radius is 4 m or less. Previously published unsaturated vertical wave number intensity spectra recorded in the Kane Basin northwest of Greenland are compared with theory. The intensity spectral shapes from the GM and A fine structure models correspond well with those from the experiment. GM models give spectra low in level, by up to a factor of 20. Predictions from A type models can match the experimental spectral level if they are given an anisotropy ratio of sound speed inhomogeneities of 0.15, compared with the GM value of ƒ/N ≈ 0.02. Calculations show that microstructure may have intermittently influenced the Kane Basin transmission at small Fresnel radius Rƒ and that high intensity variance at large Rf must result from fine‐structure variability.