Dependence of modal attenuation coefficient frequency variation on upper sediment attenuation

Abstract

The range-averaged transmission loss increase in shallow water propagation depends critically on the intrinsic attenuation of the upper sediment. The attenuation coefficients of low-frequency (<1 kHz) propagating modes determine the frequency dependence. Ingenito [J. Acoust. Soc. Am. 53, 858–863 (1973)] showed with measurements and theory that while individual mode attenuation coefficients decrease with frequency f, the sediment attenuation coefficient increases proportional to f1.75. When results from many other shallow-water transmission experiments (broadband and narrowband) over sandy-silty sediments are compared to numerical calculations, it is found that a nonlinear-frequency dependent attenuation is required with an exponent between 1.5 and 2. The question considered here is how the intrinsic upper-sediment attenuation produces such behavior. A recent simplification of the Biot model [A. D. Pierce et al., J. Acoust. Soc. Am. 114, 2345 (2003)] has a power-law exponent of two. With this frequency-dependent bottom attenuation, a two-layer Pekeris waveguide yields modal attenuation coefficients that decrease with frequency as observed by Ingenito. However, a depth-dependent attenuation profile or a third near-surface layer with requisite properties can reverse this behavior. This suggests why higher-frequency numerical computations may require nonlinear frequency-dependent sediment profiles to calculate sound transmission accurately. [Work partially supported by ONR.]