Abstract
Multibeam imaging sonars have application to monitoring fish and marine mammal presence and behaviours in the near field of tidal turbine installations, including evaluating avoidance, evasion, and potential blade strikes. SOAR conducted field experiments to help reduce uncertainty in performance of the Tritech Gemini 720is and Teledyne Blueview M900-2250 multibeam imaging sonars for identifying and tracking discrete targets in high-flow environments. This information will help inform the Department of Fisheries and Oceans Canada, tidal energy developers, and other stakeholders in the design and implementation of effective monitoring systems for tidal energy projects in the Bay of Fundy and beyond. These two imaging sonars were the technologies recommended for testing by the subject matter expert for imaging sonars during the first phase (Global Capability Assessment) of the Pathway Program. The Tritech Gemini 720is operates at 720 kHz and has a maximum effective sampling range of approximately 50 m. The Teledyne Blueview M900-2250 has operating frequencies of 900 or 2250 kHz, with a 10 m range for the high frequency transducer head. As per the recommendation from the Global Capability Assessment, this report focuses on the Blueview’s capabilities while operating at 2250 kHz, for which the effective sampling range is 10 m.
Field trials were conducted in Grand Passage aboard research vessel Grand Adventure. The two sonars and a camera were mounted on a pole which could be lowered over the vessel’s port side and fixed in position. The deployed sonars were oriented such that the top of their ensonified areas extended behind the boat approximately parallel with the water surface and extended downward at a 20 degree angle. The Grand Adventure was anchored in mid-channel during ebb and flood tide flow conditions, such that current velocities ranged from approximately 1 to 2.5 m/s with the instruments oriented downstream. Targets were suspended approximately 2 m beneath a 3 m long surfboard (SciBoard) and included a 2.54 cm (1 inch) diameter tungsten carbide sphere, 0.45 kg (1 lb.) (9.5 cm long x 3.8 cm max diameter) lead fishing weight, approx. 12 cm diameter basalt rock in a lobster bait bag, and a V-Wing glider (approx. 52 cm wing tip to tip and 46 cm nose to tail) from Dartmouth Ocean Technologies. During data collection the SciBoard and suspended target were held at constant ranges from the sonars along the port side and downstream of the Grand Adventure, and also released to freely drift downstream with increasing range.
The visualization and organization of the data was conducted using the industry standard software for each sonar: Gemini SeaTec and Teledyne ProViewer. Data were exported to video and organized into training and test data sets, which were shared with 9 sonar observers who conducted the manual analysis for target detection, identification, and tracking. Links to the training and test data sets for each sonar are provided below. The data are best viewed in video form. As such, readers of this report are encouraged to watch these data videos for better understanding of the results and conclusions discussed in this report.
Gemini training data (https://vimeo.com/473580369)
Gemini test data with 50m range (https://vimeo.com/473665614)
Gemini test data with 10m range (https://vimeo.com/473688042)
Blueview training data (https://vimeo.com/473964794)
Blueview test data (https://vimeo.com/474025663)
The Gemini 720is and Blueview M900-2250 multibeam imaging sonars were both found to be useful for detection and tracking of all target sizes used in our experimentation. However, differentiation of similar targets such as the 2.54 cm (1 inch) tungsten carbide sphere (Target 1) and 0.45 kg (1 lb.) lead fishing weight (Target 2) proved difficult. The sonars performed best for detecting, identifying, and tracking the V-Wing. This is an expected result as it was the largest target and had the most recognizable backscatter signature due to its characteristic shape. Entrained air from turbulence, waves, and the vessel/pole wake made tracking targets more difficult, but target persistence allowed them to be effectively detected and tracked by eye for all target types tested.
SOAR recommends use of the Tritech Gemini 720is for application to monitoring interactions between marine animals and tidal turbines. With the 10 m range setting, the Gemini demonstrated comparable ability to the Blueview to identify targets and outperformed the Blueview in average target detection and tracking scores. At 50 m range, the Gemini still demonstrated a high level of utility for target detection, tracking, and presence/absence, though was less effective (ca. 50%) for target identification. It is likely that this technology will contribute significantly to effective monitoring and advancing knowledge of importance to regulators and other stakeholders. The Blueview M900-2250 was included in testing due to its higher frequency output, which is better suited for close range target detection and tracking. The Blueview is an impressive technology and offered the ability to resolve finer scale features of the targets and their movements in some cases. However, the MKI model of the Blueview M900-2250 has a hardware limitation which results in multiple high-noise bands in the output data, which limited our ability to detect and track targets considerably. We conclude that data from the Blueview did not add substantial value or insight to the target analysis when used in conjunction with the Gemini. This should not rule out potential use of other MHz frequency multibeam sonars for monitoring the 10 m range in a combined sonar approach, including MKII of the Blueview.
We evaluated the effects of acoustic interference (cross talk) between the Gemini and Blueview based on the ability of manual observers to detect, track, and identify targets through repeat collections of data with the sonars running both concurrently and independently. In general, the acoustic interference can be described as distracting, but tolerable. We observed no relationship between flow speed and observers’ abilities to detect and track targets with testing up to approximately 2.5 m/s. Tidal flows are faster at the FORCE site in the Minas Passage, with flow speeds exceeding 2.5 m/s 30 to 40% of the time.
The project addressed the objective of assessing the performance of surface deployed multibeam imaging sonars for target detections, including the extent of signal interference from waves/turbulence, and entrained air. Further testing of and research into multibeam sonar usage from a vessel mounted (near surface) position would be useful in four focus areas, including:
- fish and other marine animals in locations and seasons (times) with high levels of animal abundance and variety,
- evaluating the most effective sonar orientations for monitoring the near field of tidal turbines,
- flow speeds that exceed 3 m/s, and
- increasing efficiency in data assessment, including reliable automation.
This work should build upon success in Grand Passage to conduct next steps in stronger flow conditions present in Petit Passage and Minas Passage. The report titled “Field Assessment of Multi-beam Sonar Performance in Bottom Mount Deployments” (Trowse et al. 2020) provides similar analysis for the case of seabed mounted Gemini 720is and Blueview M900-2250, including comparison of results and further recommendations for next steps.