Abstract:
Making deep-sea benthic habitat maps requires expensive and complex technologies to acquire, process and interpret remotely-sensed data. Recent technologies that have advanced the field significantly include side-scan sonar, multibeam echosounders, and Autonomous Underwater Vehicles. Acquisition of bathymetric and backscatter data using these tools underpins modern habitat mapping, but requires carefully considered survey strategies and end-product development. Practical issues such as launch and recovery operations, navigation, accuracy and acquisition of additional data are highly dependent on the type of equipment used and can affect data quality. Processing of bathymetry and backscatter data can yield quantitative information about the seafloor, which in turn may be used as proxies to characterise the habitat. Data for proxies such as bathymetry, slope, rugosity or reflectivity can be partitioned into multiple segments with similar multivariate values (segmentation), which can be further grouped spatially to form a connected region of similar habitat (classification). Finally, the acquisition of ground-truthing data is critical to validate any benthic habitat map generated using such means. We illustrate the application of habitat mapping methodologies through examples selected primarily from studies conducted within the New Zealand region.