Integrated optical and acoustic monitoring of deep-sea bubble plume – A case study in Haima cold seep

Abstract

Monitoring underwater bubble plumes plays a vital role in understanding the plumes’ impacts on marine ecosystems. In this study, an integrated optical and acoustic deep-sea bubble plume observation system was designed and presented. The system was deployed in the Haima cold seep at a depth of 1,386 m, where a total of 4.5 h of optical and acoustic recordings were collected. Customized methods for processing both optical and acoustic data were also developed: a convolutional neural network (CNN) was used to identify individual bubbles from optical data stream, a passive acoustic bubble detection method integrating cell averaging constant false alarm rate (CA-CFAR) detector was applied to address human-induced interferences, and parabolic interpolation was employed to improve the resolution of time delay estimation for active acoustic observation. All the collected data were comprehensively processed and analyzed. The quantitative analysis of the video data yielded a mean bubble radius of 2.35 mm, and a gas flux ranging from 0.052 L/min to 0.086 L/min. The frequency-dependent sound speed variations obtained from the active acoustic data align well with existing theory on linear pressure wave propagation in bubbly liquid, suggesting the deep-sea bubbles influence the sound speed in predictable ways. This study represents an important early development of an integrated optical and acoustic system for in-situ deep-sea bubble plume monitoring, offering valuable insights and practical considerations for future iterations of related instrumentation and data analysis algorithms.