Bayesian inference for the propagation of measurement-based model parametrizations of bottom scattering strength to reverberation prediction uncertainty

Measurements conducted on the Malta Plateau south of Sicily and at a site north of Elba in the Tyrannian Sea are used as the basis for the construction of bottom-scattering parameter probability density functions using Bayesian inference. The volume scattering and rough surface scattering prediction capability of the OASES model [Schmidt and Kuperman, J. Acoust. Soc. Am. (1995), and LePage and Schmidt, J. Acoust. Soc. (2003)] are used to model the bottom-scattering measurments in order to facilitate the extraction of these paremeter density functions. Issues regarding the resolution and resolvability of the bottom-scattering parameters are discussed, and comparisons between Bayesian probability density functions for reverberation predictions based on the scattering measurements and measured reverberation data are made. [Work supported by the ONR Capturing Uncertainty DRI.]