Information theory application to inversions of acoustic data from a continental shelf environment.
David P. Knobbles, Robert A. Koch, and Jason D. Sagers


This research examines the uncertainty in waveguide parameter values obtained from acoustic inversion of data collected in range-dependent environments. To account for range-dependence in the environmental parameters, the number of inversion parameters must be increased. Generally, from a fixed data sample the uncertainty in the solution for individual parameter values increases as the number of inversion parameters increases, especially if the structure of the propogated acoustic field is sensitive to the additional parameters. This effect can be offset by prior information that reduces the volume of the parameter search space and to a limited extent by larger inversion data samples. How an inversion algorithm accounts for prior information is thus an essential issue to be addressed for range-dependent waveguides. Measurements made on the New Jersey shelf in 70-90 m of water provide an opportunity to study the effects of environmental range dependence on sound propagation and inversion processing. Inversions based on pre-existing geophysical information are combined with a maximum entropy approach to quantify the amount of information on the waveguide parameters that can be extracted from measured propagation over range-dependent tracks. Both towed sources emitting CW lines and impulsive sources are employed.

Back to the main WHOI SW06 home


WHOI logo    © Woods Hole Oceanographic Institution
                        © University of Washington
University of Washington Seal
             Last updated June 2009

ONR logo