Advancing active hydroacoustic methods with broadband echosounders for ecological surveys

Abstract

Marine ecosystems are undergoing increasing environmental pressure from climate change and industry developments. These environmental pressures are particularly impacting in the Arctic. Increased monitoring of Arctic marine ecosystems with active acoustic surveys could improve our understanding of the ecological impacts from climate change stressors. However, traditional research vessel-based surveys are very expensive, have limited availability and measure a disrupted pelagic environment by introducing light and noise. We propose that broadband echosounders mounted on uncrewed vehicles, moorings or lowered probes would provide complementary measurements to address these shortcomings. In the case of uncrewed vehicles, they could complement traditional vessel-based surveys by extending the temporal and spatial extent. Because broadband echosounders are relatively new technologies, methods and standard processing procedures need to be modified and tested. The present thesis studies three methods for broadband echosounders to increase the independence of acoustic data from research vessels for monitoring marine ecosystems, with the Arctic as a case study. The three methods are the inverse method, model-informed classification and mesocosm-informed classification. These methods aim to increase the information retrieved from acoustic data by maximizing the use of the broadband target spectra measurements. The methods are applied to sound scattering layers of zooplankton aggregations, individual zooplankton, shrimp, and fish to advance methods for analyzing broadband acoustic data from uncrewed vehicles or moorings where few or no direct sampling data are available. However, new challenges intrinsic to broadband echosounders and their signal processing arise, such as understanding the sensitivity of broadband measurements of targets as individuals and as aggregations. To conclude, I discuss the applicability of these methods for ecosystem-based fisheries management and commercial fishing. I conclude on the progress towards fulfilling the promise of broadband acoustics for species identification.