Three-dimensional localization of sperm whales using a single hydrophonea)
Schematic of longline deployment plus instrumented anchor line, west of the longline, with acoustic recorders at depths 83 and .
Contour map (depth in meters) of the fishery near Sitka, AK, shows a steep continental shelf. The location of the array deployment on May 9, 2004, is marked with a star.
(a) Spectrogram of acoustic data from receiver at depth starting at 11:42:15 on May 9, 2004. Broadband sperm whale clicks appear as vertical stripes. (b) Spectrogram summed over frequency bins; broadband clicks appear as peaks in the normalized spectral sum time series. Arrival pattern from one click event is highlighted. (c) Highlighted arrival pattern is used to find similar arrival patterns in spectral sum which are extracted and time aligned. (d) Arrival patterns from (c) represented as a 2D surface that conveys relative arrival time and amplitude of arrivals via grayscale intensity.
(a) Arrival pattern surface for receiver at depth starting at 11:42 on May 9, 2004. (b) Relative arrival times are extracted from surface in (a) to make an arrival map. Arrival patterns evolve over time as the source (whale) changes position relative to the receiver.
Output of ray tracing model shows predicted arrival patterns at receiver at depth from sources at depth, range, over a range-dependent bathymetry radial on a bearing of from the receiver.
(a) Arrivals from a modeled source at range, depth, bearing overlaid on measured arrivals from 11:42:15 on May 9, 2004. Few arrivals overlap, so it is given a low “overlap score” indicating a low likelihood that the true source is at the modeled source location. (b) Arrivals from a modeled source at range, depth, bearing overlaid on measured arrivals from 11:42:15 on May 9, 2004. All measured and modeled arrivals overlap, so this is the best estimate of true source location.
An ambiguity surface showing overlap scores on a grid of ranges and depths along bearing from the receiver. The peak at range, depth is marked and indicates best estimate of source location for the click event recorded at 11:42:15 on May 9, 2004.
(a) Arrival patterns from receiver at depth starting at 11:18 on May 9, 2004. (b) Range and depth estimates along bearing as derived using arrivals data from receivers at 83 and independently. Both receivers track a whale diving from the surface in the absence of human fishing activity, consistent with divergence of direct and surface-reflected arrivals seen in (a) at 11:18.
(a) Arrival pattern surface for receiver at depth starting at 09:26:30 on May 9, 2004. (b) Range, depth, and bearing estimates as derived using arrivals data from receivers at 83 and jointly, plus fishing boat location per GPS data. Track follows a whale diving to an unusually shallow foraging depth as whale swims towards a fishing vessel conducting a longline recovery. Late arrivals in (a) disappear at 09:28 as depth track shows whale leveling from dive, implying directionality of sound source.
Range, depth, and bearing estimates of clicking sperm whale starting at 10:24 on May 9, 2004 as derived from receivers at 83 and independently, plus locations of ship position and visual sightings of sperm whale. Acoustic track matches visual observations of whale resting at surface and diving.
Whale tracking using the data from Sec. IV A is repeated for both receivers using an isovelocity sound speed environment and compared to localizations using a historic sound speed profile. Localizations from the different replicas agree closely until the source moves beyond in range.
Metrics from sperm whale tracks from May 9, 2004 derived using data from receivers at 83 and depth jointly.
Article metrics loading...
Full text loading...