Tonight reporter Gary Chitten and cameraman Pete Cassam from King 5 Television broadcast a nice story about the successful first test of a FLIR camera for detecting killer whales at night. The pilot study was designed by Jim Thomson of the UW Applied Physics Lab, his Master’s student Joe Graber, and his other staff. In collaboration with Jason Wood, Research Curator for The Whale Museum, and with the cooperation of Whale Watch State Park, the camera was deployed atop the Lime Kiln lighthouse where orcas commonly pass by close to shore during the summer months.
In just a week of effort, a suite of useful data were collected that are allowing Joe to assess to what extent automated signal processing algorithms can reliably detect orcas at night when they are up to 300 yards from the camera. Preliminary results suggest that a human can identify whales among the noise generated by waves at 300 yards, but the computer isn’t very reliable much further out than 100 yards. Perhaps a higher-resolution, top-of-the-line camera could improve the detection range?
The King 5 piece did an admirable job of conveying the importance of passive acoustic monitoring when attempting to detect approaching orcas. While infrared cameras may ultimately be able to detect killer whales up to a kilometer away, the loud calls and clicks made by the whales can be detected at 10 times that range when background noise levels are low. Together, infrared video cameras and real-time hydrophone networks constitute our best nighttime chance of knowing whether and when to shut down a turbine to mitigate it’s potential impacts on the fragile southern resident killer whale population. Sighting networks like Orca Network and human eyes at tidal turbine sites will be critical supplemental techniques for tracking whales during the day.