Quantifying ocean acoustic environmental sensitivity

This paper examines the sensitivity of acoustic propagation data required for sonar performance predictions to physical parameters of the ocean environment. Sensitivity is quantified here by an appropriate measure of the relative uncertainty of the acoustic data due to realistic uncertainties in environmental parameters. Within a linear approximation, Gaussian-distributed environmental parameter uncertainties lead directly to Gaussian data uncertainties which scale linearly with the parameter uncertainties. However, the actual data uncertainties, which can be estimated numerically via Monte Carlo methods, do not necessarily possess these desirable linear properties. The applicability of the linearized approximation and the relative magnitude of acoustic sensitivities are examined for realistic uncertainties in seabed geoacoustic parameters and oceanographic features of the water-column sound-speed profile. Sensitivities are considered as a function of source and receiver depth, range, and source frequency for various shallow-water environments. [Funding provided by DRDC-Atlantic Rapid Environmental Assessment Project.]