Bayesian source localization with uncertain Green's function in an uncertain shallow water ocean.

Area of Science:

• Ocean acoustics
• Signal processing
• Geophysics

Background:

• Matched-field acoustic source localization is sensitive to environmental uncertainties.
• Environmental mismatch significantly degrades localization performance.

Purpose of the Study:

• Develop a Bayesian approach to enhance robustness against environmental mismatch in acoustic source localization.
• Quantify localization uncertainty by integrating over environmental variability.

Main Methods:

• Model the waveguide Green's function as an uncertain random vector.
• Employ Bayesian inference to obtain a joint marginal probability distribution for source range and depth.
• Approximate high-dimensional integration using a 1D integration over a correlation measure and modal analysis for covariance matrix approximation.

Main Results:

• The proposed method provides efficient and reliable source localization.
• Demonstrated improved robustness to environmental mismatch compared to other methods.
• Successfully quantified localization uncertainties.

Conclusions:

• The developed Bayesian approach offers a robust solution for acoustic source localization in uncertain oceanic environments.
• The method effectively mitigates performance degradation caused by environmental mismatch.
• Accurate localization and uncertainty quantification are achieved.