Modelling Acoustic Transmission Loss Due to Sea Ice Cover

The propagation of underwater acoustic signals in polar regions is dominated by an upward refracting sound speed environment and the presence of a dynamic highly variable ice canopy. This paper provides an overview of the acoustic properties of sea ice and assesses the influence of ice canopy and water column properties on acoustic transmission loss for propagation within 20 km of a sound source at 20 m depth. The influence of the ice canopy is assessed first as a perfectly flat surface, and then as a statistically rough surface. A Monte Carlo method is used for the inclusion of ice deformation and roughness. This involves the creation of sets of synthetic ice profiles based on a given sea ice thickness distribution, followed by statistical methods for combining the output of individually evaluated ice realisations. The experimental situation being considered in the framing of this problem is that of an Autonomous Underwater Vehicle (AUV) operating within 50 m of the surface. This scenario is associated with a frequency band of interest of 9-12 kHz and a horizontal range of interest up to 20 km. The situation has been evaluated for a set of typical ice statistics using Ray and Beam acoustic propagation techniques. The sound speed profile (based on real data) results in a strong defocussing of direct path signals at ranges from 9-20 km and depths shallower than 50 m. This reduction in the signal strength of the direct path creates areas where the influence of surface reflected paths becomes significant. The inclusion of a perfectly flat ice layer reduces the transmission loss between 9-20 km by 15-50 dB. When the ice layer is included as a rough surface layer the results show a boost to signal strength of up to 8 dB in the small areas of maximum defocussing. Sea ice is a strongly time and space varying sea surface and exists in areas where defocussing of the direct path due to the sound speed profile reduces the range of direct path dominated transmission. This work presents methods for including a statistically relevant rough surface through a technique for generation of sets of surfaces based on ice deformation statistics. It outlines methods for including ice in acoustic modelling tools and demonstrates the influence of one set of ice statistics on transmission loss.