Spatial analysis of multibeam acoustic data for the predictiion of marine substrates and benthic communities in temperate coastal waters

The combined use of acoustic backscatter and bathymetric data generated by multibeam echo-sounders provides a powerful tool to investigate substrate characteristics at fine spatial scales across large areas. New methods are now required to assess the links between the identification of physical substrate classes from multibeam data with visually classified classes from AUV data. In this paper, we use statistical algorithms to describe the patterns between the physical covariates derived from both the multibeam data, and substrate type observed in AUV images. We then use these relationships to predict substrate type and describe the biological biotopes across the mapped extent of south eastern Tasmania.