Through the Marine Biodiversity Hub research program in southeastern Tasmania, we employed an autonomous underwater vehicle (AUV). The AUV is part of the Integrated Marine Observing System (IMOS) AUV Facility operated by The Australian Centre for Field Robotics (ACFR) at The University of Sydney, led by Dr Stefan Williams. The AUV Sirius is able to obtain high resolution multibeam sonar and spatially rectified fine-scale stereo still photographs of the seafloor, providing the opportunity to image fine-scale habitat features and count individual benthic fauna and flora. Data collected by AUV can therefore facilitate the examination of biological complexity and physical surrogacy at a scale finer than possible with towed-video.
During each mission, high-resolution multibeam and photographic imagery were collected 2-3 m above the seafloor along a 1-10 km pre-programmed grid path. Multibeam bathymetry was collected using a Imagenex 837 DeltaT Profiling 260 kHz system,flown at 2 m above the seabed providing a 4 m wide swath. Bathymetry and backscatter data were gridded to 10 cm resolution. Photographic images were taken every second along theseafloor using a stereo camera system with approximately >40% overlap to enable photographic images to be combined as a mosaic and provide a continuous interwoven 2-3 m wide image of the seafloor along the path of the mission. In order to correctly mosaic seabed imagery, geolocated photographs were processed using SLAM image-recognition software whereby identical features seen in sequential photographs were aligned and stitched together. The same technique was used to align the multiple cross tracks of the grid so that identical features seen on intersecting tracks were also aligned and stitched together to create a single mosaic image for the entire trackline.
For more information on the processing of the AUV data please refer to the publications website of the ARC Centre of Excellence for Autonomous Systems.
A SONY TRV900 digital camera is used to record underwater video footage at regular transects throughout our acoustic surveys. We do this to ensure accurate echo-sounder habitat boundary definitions and obtain more detailed information on habitat modifiers (Levels 3 and 4 of the classification table), particularly in relation to the characterisation of dominant seagrass and macroalgal species.
The video camera frame contains around 40 kg of weight and is towed at a speed less than 1 knot to ensure that the camera is positioned as close to the GPS antenna as possible (this is often visible directly under the vessel in depths less than 20 m). The camera is suspended approximately 1 m from the sea floor giving a field of view (FOV) that varies between 1-2 m wide. Differential GPS location, time, date and water depth are overlayed onto the video from the GPS sounder using a genlock device and logged into a txt file for GIS mapping.
The video footage is reviewed in the labratory, and in conjuntion with the field notes, is used to estimate the percentage cover of each of the major visually dominant macroalgal species for reef habitat attibution. This information is correlated against depth and exposure to determine characteristic biotic community types for combinations of each of the physical variables or habitats.