Using singing whales to track nonsingers

Abstract

The ability to track vocalizing whales over large distances (≳100 km) using horizontal arrays has recently received much attention in the underwater acoustics community [‘‘Use of Naval Facilities for Ocean Acoustic Research,’’ J. Acoust. Soc. Am. 95, 2851–2854 (1994)]. The high source levels of the whales, up to 190 dB re: 1 μPa at 1 m, make this possible. However, tracking vocal whales may provide an incomplete behavioral picture because a significant proportion of whales may be nonvocal at any given instance. (Visual observations are limited to the small proportion of time that whales surface.) Alternatively, the high source levels invite the possibility of treating the vocalizing whale as an active source. Sound from this source of opportunity that is scattered by other members of the herd may be sufficient to locate nonvocal whales with a towed array. Historical data are used to model the spatial distribution of vocal and nonvocal humpback whales migrating off the Eastern Continental Shelf of Australia. A spectral model for scattering from an object in a waveguide based on Ingenito’s method is used to determine the scattered field from the whales. Waveguide noise is modeled via the Kuperman and Ingenito approach. Simulations indicate that localization of nonvocal whales is plausible for towed arrays within roughly 10 km of the herd in typical ambient noise conditions. This range may be extended significantly with time-domain matched filtering.