A multi-hypothesis tracker for clicking whales

Abstract

This paper describes a tracker specially designed to track clicking beaked whales using widely spaced bottom-mounted hydrophones, although it can be adapted to different species and sensors. The input to the tracker is a sequence of static localization solutions obtained using time difference of arrival information at widely spaced hydrophones. To effectively handle input localizations with high ambiguity, the tracker is based on multi-hypothesis tracker concepts, so it considers all potential association hypotheses and keeps a large number of potential tracks in memory. The method is demonstrated on actual data and shown to successfully track multiple beaked whales at depth.

Changes in the Movement and Calling Behavior of Minke Whales (Balaenoptera acutorostrata) in Response to Navy Training

“…maximum change in latitude or longitude was 0.06 • , and the least squares error of the actual and modeled times had to be under 0.075 s. A minimum of 12 calls were required in a track, which is based on one animal at the nominal call rate (1 call/∼5 min) over the course of an hour. A multi-hypothesis tracker (Baggenstoss, 2015) was used to remove any outlying localizations. The dataset used in the next stage of the analysis was made up of 62,323 localizations associated into 629 tracks.…”

“…The methodological development that allows us to do this is the generation of acoustic tracks for individual whales from the localizations (Helble et al, 2015, 2020a,b Klay et al, 2015Henderson et al, 2018;Harris et al, 2019a;Guazzo et al, 2020). In addition, we assessed whether minke whales were more likely to cease calling during sonar exposure, as changes in calling behavior have been reported in response to anthropogenic noise for a number of other baleen whale species [blue whales -Melcon et al, 2012;bowhead whales (Balaena mysticetus, Blackwell et al, 2013, 2015; humpback whales- Risch et al, 2012;Cerchio et al, 2014]. of Kauai, Hawaii, and separated from one another by 5-7 km at depths of 650-4750 m (Figure 1).…”

Tracking time differences of arrivals of multiple sound sources in the presence of clutter and missed detections

“…These multiple TDOA tracks have been previously estimated using various approaches such as the manual analysis of bearing-time scatterplots; 22 automated analysis based on properties of the signals of interest; 17 and a traditional MTT approach based on multiple hypothesis tracking. 23 Although the automated methods have advantages over the manual analysis, such as being less timeconsuming and producing more objective results, they can suffer from computational limitations when dealing with high clutter density and many closely spaced targets. 24 We propose to cast the TDOA estimation problem into a more general statistical MTT framework, where clutter (false TDOAs), missed detections (when a target is present but no measurement is collected), and sources' appearance and disappearance are incorporated in the problem formulation.…”