ABSTRACT
Sound propagation in the ocean is strongly influenced by the fact that sound speed is inhomogeneous, leading to significant refraction effects (Tolstoy and Clay 1966; Urick 1983; Leighton 1998). The variations in sound speed may be relatively small, of the order of 3 per cent, but in most applications these variations have a major effect. There are several reasons why one may seek methods to compute the sound field resulting from an acoustic source in the ocean. The most obvious reason is to predict the performance of sonar systems, a task that depends critically on one’s ability to estimate the transmission loss (TL) between a source and a receiver. This probably represents the greatest source of error in performance calculations and is strongly dependent on the spatial variability of the sound speed. A second use for propagation models is in inversion schemes, where one makes measurements in the ocean and then uses the propagation model to construct estimates of the environmental parameters. This approach is attractive to oceanographers since it offers the potential to survey large tracts of ocean at one time, a goal which is unrealisable using conventional surveying techniques.
