Abstract
Recent Littoral Acoustic Demonstration Center (LADC) multi-mooring Environmental Acoustic Recording System (EARS) data from the northern Gulf of Mexico are analyzed to deduce identifications of individual beaked whales. Procedures are built on previously applied self-organizing map techniques for clustering sperm whale clicks and beaked whale clicks from workshop data. Associating the clicks from beaked whales is difficult because, compared to sperm whales, beaked whales have lower source level, a narrower beam, and a faster rate of turning. Recordings of individual clicks can be clustered according to their time domain signal, frequency spectrum, or wavelet spectrum. For example, clicks clustered according to the magnitude of their frequency components show similarities for all the clicks in a class (representing an individual), but significant differences from class to class, suggesting that this approach has promise for identifying individuals. Recent work by Baggenstoss (J. Acoust. Soc. Am. 130, 102–112 (2011); J. Acoust. Soc. Am. 133, 4065–4076 (2013)), who has used cross correlations of clicks to assist in associating beaked whale clicks into trains, reinforces the idea that single click properties can be associated with individuals. [Research supported by SPAWAR and ONR.]
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call