Date of Award
Master of Science in Ocean Engineering
Lora Van Uffelen
The Seaglider, a buoyancy driven Autonomous Underwater Vehicle (AUV) equipped with oceanographic sensors that glides through the water in a sawtooth pattern, shows promise as a quieter alternative to propelled AUVs for deployable autonomous acoustic sensing. However, Seaglider functions such as pitch, roll, buoyancy adjustments and oceanographic sensing still produce sounds which impact acoustic recordings. The goal of this research is to identify and acoustically characterize sounds produced by the Seaglider by description of signal temporal and spectral characteristics with absolute sound pressure levels using spectrograms, power spectral density, and 1/3 octave analysis.
Acoustic data collected during two separate deployments of a Seaglider with an integrated onboard passive acoustic monitoring system, which supported acoustic analysis up to 64 kHz and sound pressure levels up to 164.08 dB re 1 μPa, are aligned with vehicle data logs to draw relationships between Seaglider functions and acoustic events. During alignment, a clock drift was identified in the passive acoustic monitor resulting from an integration issue which required a software update to enable proper clock sync in future missions. Using Seaglider self-noise signals, the offset was rectified for acoustic analysis.
Analysis revealed the guidance and control functions including pitch, roll, and buoyancy changes, which primarily occur at the surface and near apogee, were the predominant sources of noise from the Seaglider, generating broadband noise spanning up to the observable maximum 64 kHz with the greatest impacts below 10 kHz, producing average sound pressure levels ranging from 143.5 dB re 1 μPa to 120.8 dB re 1 μPa. The pumps used to increase buoyancy produced the longest and loudest guidance and control noise, generating up to 3.2 minutes of full spectrum broadband noise while shorter, lower frequency, roll events were the most abundant guidance and control function occurring an average of 85 times per dive. Other sources of noise on the Seaglider include a microprocessor clock which produced a continuous tonal at 32.76 kHz, the SBECT which produced a 52.5 kHz tonal, ± 0.5 kHz, lasting approximately 0.75 seconds and the altimeter which produced a short broadband impulse that exceeded the 164.08 dB threshold of the Seaglider’s passive recording system followed by a reverberation of the user selected peak frequency in the water column.
Each of these functions produce a uniquely identifiable acoustic signal. These noise characteristics are essential for researchers to develop filtering techniques, identify vehicle noise in data, and understand the Seaglider as an acoustic platform.
Ross, Melodie, "A SELF-NOISE ANALYSIS OF A SEAGLIDER AUV USING A PASSIVE ACOUSTIC MONITOR" (2022). Open Access Master's Theses. Paper 2128.
Available for download on Friday, May 17, 2024