Acoustic mapping and monitoring of the seabed: From single-frequency to multispectral multibeam backscatter

With the increasing human activities in the marine environment, such as fisheries, dredging, coastal protection or construction of marine infrastructure, seabed sediment and habitat mapping have become highly relevant for the development of sustainable marine management strategies. Compared to traditional mapping methods, primarily based on bed sampling, multibeam echosounding belongs to the cutting-edge technology to time-efficiently acquire high-resolution bathymetric and backscatter (BS) data over large areas. Using classification methods to combine the acoustic data with ground-truthing, large-scale maps can be automatically and objectively produced, that enables to describe the distribution of benthic habitats or quantify marine resources. However, acoustic sediment classification still does not allow to discriminate between the entire heterogeneity of the seabed and is generally applied to a single multibeam echosounder dataset by means of revealing the seabed state only at a given time instant. Two challenging issues addressed within the scope of this thesis are summarized as: (1) Investigation on the applicability of repetitive multibeam (single-frequency) BS measurements for monitoring the seabed; and (2) Evaluation of the potential of multispectral BS to increase the acoustic discrimination between different seabed environments.