Maximum entropy inference of seabed properties using waveguide invariant features from surface ships

Area of Science:

• Underwater acoustics
• Geophysics
• Oceanography

Background:

• Acoustic data were collected using vertical line arrays (VLAs) in the New England Mud Patch.
• The experiment focused on characterizing the seabed in approximately 75 meters of water.
• Merchant ship passages provided acoustic signals for analysis.

Purpose of the Study:

• To develop a geoacoustic model (M) of the seabed using geophysical information and acoustic data.
• To infer statistical properties of geoacoustic parameters, specifically sound speeds in mud and sand layers.
• To investigate the waveguide invariant (β) for mode pairs in the 15-80 Hz band.

Main Methods:

• Development of a geoacoustic model M using acoustic data from the merchant ship KALAMATA.
• Application of a feature-ensemble maximum entropy method to infer statistical properties from multiple ship data.
• Analysis of singular points for waveguide invariant β for mode pairs (1,n).

Main Results:

• The geoacoustic model M successfully reproduced observed β≈-1 for frequencies below 20 Hz and mode cutoff features around 15 Hz.
• Statistical inference provided evidence of an angle of intromission at the water-sediment interface.
• Average sound speed ratio at the interface was found to be approximately 0.986, with an average sound speed of 1775 m/s in the deeper sand layer.

Conclusions:

• The study successfully developed and validated a geoacoustic model for the New England Mud Patch.
• The findings provide insights into the sound speed structure of the seabed, including mud and sand layers.
• The research highlights the utility of merchant ship acoustic data for geoacoustic modeling and understanding underwater sound propagation.