Doppler spectral moments of multibeam echosounder data to infer the ascent velocity of methane bubbles emanating from natural seafloor seeps

Abstract

A multibeam echosounder operating at 500 kHz was deployed on a remotely operated vehicle to detect methane bubble streams emanating from natural seafloor seeps in 1400-m water depths. The sonar transmitted 10-ms tone bursts every 100 ms over a 7°×120° beam, and formed 128 receive beams regularly spaced 0.94° apart over the 120° sector. The sonar head faced upward and signals backscattered from ascending bubbles were sampled at about 6 kHz between 10 m and 50 m from the head. Doppler spectral moments were computed for each ping with the conventional data autocovariance method. For unimodal and symmetric Doppler spectra, the mean power is related to the backscattering strength of the sampled volume, the mean Doppler frequency yields the mean radial velocity of the bubbles, and the spectral width measures the mean velocity uncertainty with some indication of bubble oscillation. Mean ascent velocities of 0.2 m/s matched those obtained from concurrent optical measurements. However, beams free of bubbles showed mean descent velocities of about 1.5 cm/s, most likely due to marine snow, which could bias mean ascent velocity estimates for bubbles at greater distance from the sonar because the corresponding Doppler spectra might no longer be unimodal.