93cdf2b7-b6a4-4d26-ad71-70b4c2edd20c urn:uuid eng UTF-8 dataset Dataset 2022-05-10T00:54:22 creation ISO 19115 2016 point 7279 NESP Biodiversity Hub Hunter Marine Park Towed Video Imagery 93cdf2b7-b6a4-4d26-ad71-70b4c2edd20c Raw downward facing still imagery of the seabed for Hunter Marine Park acquired using NSW Department of Planning Industry and Environment towed video system aboard RV Bombora for the National Environmental Science Program (NESP Biodiversity Hub). Fieldwork was funded, both cash and in-kind, by NSW DPIE and NESP in Year 5 (2019) of the program for 8-10 days of towed video over areas surveyed earlier in the program using a multi-beam echosounder (MBES) in 2015-16 and 2018-19 (NSW DPIE). A report titled 'Mapping and characterising the rocky reef habitat, fish and sessile invertebrate assemblages of the Hunter Marine Park' is available on the NESP website. Sites were randomly selected across mapped areas of the park's inner shelf, >3NM offshore of Seal Rocks, Broughton Island and Outer Gibber and covering depths of ~25-110m. Georeferenced (XYZ)and time-stamped (UTC) still imagery is accessible via DPIE Information Asset Register (IAR) and NSW Sharing and Enabling Environmental Data (SEED) website and linked to the online annotation platform Squidle+. MARINE-Biology MARINE-Coasts MARINE-Reefs Marine conservation PHOTOGRAPHY-AND-IMAGERY-Remote-Sensing baseline monitoring dataset 2020-07-16T00:00:00 Water column imagery has been edited out of the dataset provided here but available upon request. dataset 2020-07-16T00:00:00 Where drop-outs of GPS positioning of the tow fish occurred, an interpolated value was calculated as described in the Lineage statement above. Forward looking video and recorded tow-fish parameters (i.e. roll/pitch/heave) are also available but not provided with this digital still dataset. dataset 2020-05-18T00:00:00 As the imagery is acquired using a moving platform some images may or may not capture part of the same section of seabed in subsequent images. Survey speed of ~1kt (0.51m/s); image capture rate 1 image every 2 seconds; and approximate area of seabed captured in each image (0.5 - 2m2); 2 laser pointers denote a horizontal distance of 10cm where visible. nb: tow-fish roll, pitch and heave values are recorded by the video camera and saved within the video data file. dataset 2020-05-18T00:00:00 XY position of fish is determined by relative position of USBL transponder (fish) to transceiver (vessel). GPS position is G2 quality (<0.4-0.5m XYZ in realtime equiv. RTK) and is relative to the vessels Centre of Mass (reference frame 0:0:0) with a 2.4m forward offset (X = -1.5) from transceiver to COM entered in the USBL software (Y = 0). USBL heading offset is checked following each installation (nominally 23 degrees from centreline) and USBL (Tracklink, USA) precision is 3 degrees of slant range value. Original still image time-stamping is synched with computer UTC (<1s) but image naming convention utilises USBL/Video text feed based on POS MV time ss:mm:hh dd:mm:yyyy (UTC) output. The Tracklink USBL system suffers from drop outs (noise, power) for fish positioning and so when dropouts occurred the following protocol was followed when processing and interpolating positioning data: 1. Extract raw USBL data and partition into transects: interpolate (running average) where gaps are <10s: marked as FLAG 1 in the metadata file; 2. Extract GPS and Fish position (and other data) aligned with timestamps for each image (image to USBL time offset <1s); 3. Import ship and fish positions into Arc, then identify outlier points using 4m radius filter for successive UBSL fish position fixes; delete values and interpolate (running average) where gap is <10s; mark as FLAG 1 in metadata file. 4. Determine a median and SD of layback for true fish fixes and generate new fish position values over periods of <120s; mark as FLAG 2; 5) Use difference in GPS ship position (no gaps) from one point to the next and apply the same offset to the position of the last known USBL fish-position fix; Where gaps are >10s and as calculation of the new fish position relies on the last fish position there is likely to be an increase in positional error for the assumed fish-position over time. To rectify this, we applied an incremental (linear) offset/adjustment to ensure that the end of the ‘drop-out’ period aligned with next available USBL fish position in the sequence (i.e. when USBL positioning returned); marked as FLAG 3. dataset 2020-07-16T00:00:00 No non-quantitative attributes are provided here