MODELING OF LOW-FREQUENCY TRANSMISSION LOSS IN THE CENTRAL ARCTIC

The magnitude of low-frequency (10-100 Hz) propagation loss in the central Arctic is known to be larger than predicted by most free surface scattering theories. This high loss is introduced primarily by scattering at the ice canopy, with which the acoustic path interfers regularly due to the presence of a strongly upward refracting surface duct. While distinct feature scattering theories such as boss or elastic keel models predict higher losses, these results are typically too frequency dependent to agree with the data over the entire band of interest. Perturbation theory gives close to the correct frequency dependence, but historically predicts too little loss when the dynamics of the ice canopy is neglected or the ice is assumed to be locally reacting. In this paper it is proposed that when perturbation theory is extended to include scattering of incident acoustic energy into propagating elastic modes in the ice, the resulting expressions correctly predict low-frequency losses measured in the Arctic.