Statistical Ocean-Acoustics and Environmental Inversion After Stochastic Propagation and Scattering
Abstract
The long term goal of this research is to provide a unified statistical theory for ocean-acoustic measurements made after stochastic propagation and scattering that accounts for both the temporal and spatial coherence of the received field. Such a theory must be in place before acoustic measurements can be effectively used as a tool to either probe the marine environment or communicate and autonomously navigate within it. To understand why this is the case, one must first realize that for many diverse realizations of stochastic wave propagation in ocean-acoustics, from transmission through a fluctuating waveguide to reverberation from the oceanic boundaries or volume to acoustic imaging of submerged surfaces and objects, the field received by a sonar system is found to undergo statistical fluctuations even in the absence of additive noise. In stochastic propagation, these fluctuations are caused by natural disturbances in the marine environment such as turbulence or passing surface gravity and internal waves. In stochastic scattering, the fluctuations arise from slight variations in the volume inhomogeneities, surface roughness or object structure interrogated by the acoustic field. Most inverse methods in ocean-acoustics, whether their purpose is to probe the marine environment, image or locate submerged objects, require the nonlinear inversion of fluctuating acoustic field data. However, the nonlinear inversion of this random data leads to inherent statistical biases and variances. These biases and variances are a significant problem because they not only depend upon the deterministic marine environmental or target parameter to be estimated, but also on the way the measurements are made and fluctuate.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 30, 1998
- Accession Number
- ADA574861
Entities
People
- Nicholas C. Makris
Organizations
- Massachusetts Institute of Technology