Fast computation of seabed spherical-wave reflection coefficients in geoacoustic inversion

Area of Science:

• Oceanography
• Geophysics
• Computational Acoustics

Background:

• Geoacoustic inversion relies on accurate forward modeling of seabed reflectivity.
• Calculating spherical-wave reflection coefficients (SWRCs) for layered seabeds is computationally intensive.
• Existing methods often require significant computation time and memory.

Purpose of the Study:

• To develop a computationally efficient numerical approach for SWRCs.
• To improve the speed and reduce memory requirements of forward modeling for geoacoustic inversion.
• To assess the impact of using SWRCs versus plane-wave reflection coefficients (PWRCs) in geoacoustic inversion.

Main Methods:

• Exploited the Sommerfeld-integral representation of SWRCs as a Hankel transform.
• Applied Levin integration to a rapidly oscillating integrand.
• Utilized trans-dimensional Bayesian inversion for parameter estimation and uncertainty quantification.

Main Results:

• Achieved speed-up factors of an order of magnitude or more compared to Simpson's rule for SWRC computation.
• Demonstrated reduced memory requirements, beneficial for GPU parallelization.
• Quantified the impact of SWRCs vs. PWRCs on geoacoustic parameter estimation and model selection.

Conclusions:

• The developed Levin integration method offers substantial computational savings for SWRCs.
• Efficient SWRC computation enhances geoacoustic inversion accuracy and reduces uncertainty.
• This approach is valuable for analyzing broadband seabed reflectivity data.