Seafloor image annotations from the Sabrina upper slope, East Antarctica

• Other Authors: POST, ALIX (hasPrincipalInvestigator), POST, ALIX (processor), O'BRIEN, PHILIP EDWARD (hasPrincipalInvestigator), ARMAND, LEANNE KAY (hasPrincipalInvestigator), CARROLL, ANDREW (hasPrincipalInvestigator), Australian Antarctic Data Centre (publisher)
• Format: Dataset
• Language: unknown
• Published: Australian Antarctic Data Centre
• Subjects: oceans; EARTH SCIENCE > BIOSPHERE > ECOSYSTEMS > MARINE ECOSYSTEMS > BENTHIC; CAMERAS; SHIPS; GEOGRAPHIC REGION > POLAR; CONTINENT > ANTARCTICA; OCEAN > SOUTHERN OCEAN; Southern Ocean; East Antarctica; Kongsberg; Antarc*; Antarctica
• Online Access: https://researchdata.ands.org.au/seafloor-image-annotations-east-antarctica/1385779; https://doi.org/10.26179/5caed60a7b076; https://data.aad.gov.au/metadata/records/AAS_4333_IN2017_V01_seafloor_imagery_annotations; https://data.aad.gov.au/metadata/records/AAS_4333_AAS_4333_IN2017_V01_seafloor_imagery_annotations; http://nla.gov.au/nla.party-617536

Four camera tow transects were completed on the upper slope during survey IN2017_V01 using the Marine National Facility’s Deep Tow Camera. This system collected oblique facing still images with a Canon – 1DX camera and high definition video with a Canon – C300 system. Four SeaLite Sphere lights provided illumination and two parallel laser beams 10 cm apart provided a reference scale for the images. This dataset presents results from the analysis of the still imagery. All camera tows were run at a ship speed over the ground of approximately 2 knots. Several sensors were attached to the towed body, including a SBE 37 CTD for collection of salinity, temperature and pressure data, a Kongsberg Mesotech altimeter and a Sonardynne beacon to record the location of the towed body. Transects were run downslope from the continental shelf break, with images analysed over a depth range of ~495 m to 670-725 m. Biota and substrates were characterised for every fifth image according to the CATAMI image classification scheme (Collaborative and Automated Tools for Analysis of Marine Imagery, Althaus et al., 2015). Images were loaded into the online platform SQUIDLE+ for analysis. Biota were counted as presence/absence of all visible biota for each image. Percent biological cover and substrate type for the whole image was calculated based on analysis of 30 random points across each image. Percent cover calculations were standardised according to the proportion of scored points on each image, excluding those that were too dark to classify. A total of 203 images were analysed. Images are available from: http://dap.nci.org.au/thredds/remoteCatalogService?catalog=http://dapds00.nci.org.au/thredds/catalog/fk1/IN2017_V01_Sabrina_Seafloor/catalog.xml