Modeling the propagation from a horizontally directed high-frequency source in shallow water in the presence of bubble clouds and sea surface roughness

Among the many factors affecting the propagation of sound in shallow water, surface-generated microbubbles have remained virtually unexplored. The collection of microbubbles, bubbles which usually do not result in a uniform layer, presents a complex structure that varies not only in depth but also in range, and can be characterized as a collage of bubble clouds. A numerical procedure is developed in which the bubble clouds are modeled following a classification scheme proposed by Monahan [Natural Physical Sources of Underwater Sound, edited by B. V. R. Kerman (Kluwer Academic, Dordrecht, 1993), pp. 503–517]. An effective complex index of refraction of the bubble mixture is calculated for each point of the resulting range-dependent environment. The combined effect that the sea surface roughness and the bubbly environment have on forward propagation is then modeled through a high fidelity model [Norton et al., J. Acoust. Soc. Am. 97, 2173–2180 (1995)] which combines a finite element Parabolic Equation model with conformal mapping to handle surface roughness. The case of a horizontally directed source, operating at 20 kHz for a single realization of a shallow water environment is analyzed in detail. The presence of the bubble clouds severely affects the amplitude near the surface, but their influence rapidly diminishes in depth. For the case considered, the surface roughness, which has little effect on the transmission loss of the propagating field, is, however, responsible for interference effects of the forward field observed throughout the water column, while ignoring the rough surface and bubble clouds leads to a 12–15 dB error in the acoustic field near the surface, at a range of 150 m. To assume a uniformly stratified (range independent) bubble layer results in large errors near the surface at the location of the high void fraction packets, but is an acceptable approximation away from these features.