Evaluation of methods for comparing propagation models

Abstract

Comparisons of the various methods for computing underwater acoustic fields can serve several useful purposes. The accuracy of results obtained using a new method can be verified by comparisons with results from a previously validated method or by comparisons with an exact solution to a sample problem. Also, results from standard methods (ray, normal mode, FFP, Green's function, etc.) can be compared in order to evaluate the relative merits of each method for particular cases of interest. For purposes of discussion, we arbitrarily divide sound velocity profiles into one group consisting of single layer or single function profiles and another group consisting of arbitrary profiles. As examples of the first group, we discuss work by Pedersen and Gordon, who compared ray theory and normal mode theory; Wood, who compared ray and uniform asymptotic solutions with exact solutions; and others. As examples of the second group, we discuss work by DiNapoli (FFP and normal mode), Spofford and Tatro (ray theory with caustic correction, and normal mode theory), and Blatstein and Überall (ray theory, normal mode theory, and modified ray theory), The advantages and disadvantages of the various types of comparisons will be discussed.