Bowhead whale localization using asynchronous hydrophones in the Chukchi Sea

Abstract

This paper estimates bowhead whale locations and uncertainties from Bayesian inversion of modally dispersed calls recorded on asynchronous recorders in the Chukchi Sea, Alaska. Bowhead calls were recorded on a cluster of seven asynchronous ocean-bottom hydrophones that were separated by 0.5–7.5 km. A warping time-frequency analysis is used to extract relative mode arrival times as a function of frequency for nine frequency-modulated whale calls that dispersed in the shallow water environment. Each call was recorded on multiple hydrophones and the mode arrival times are inverted for: the whale location in the horizontal plane, source instantaneous frequency (IF), water sound-speed profile, subbottom layering and geoacoustic parameters, relative recorder clock drifts, and residual error standard deviation, all with estimated uncertainties. A simulation study shows that accurate prior environmental knowledge is not required for accurate localization. Joint inversion of multiple recorded calls is shown to substantially reduce localization, source IF, and relative clock drift uncertainties. Whale location uncertainties are estimated between 30 and 160 m and clock drift uncertainty is estimated between 3 and 26 ms. The clock synchronization provided by the inversion is sufficient for localizing other types of marine mammal calls using simpler time-difference-of-arrival methods.

Waveguide mode amplitude estimation using warping and phase compensation

“…This procedure is robust to mismatches in t r , as long as t start is correctly chosen. As a result, isovelocity warping has been successfully applied in many scenarios where t r was completely unknown (e.g., passive source localization with unknown r [19][20][21] ). Figure 4 plots the Artic time-domain signal using two different t start values, with the time axis defined so that t start ¼ 0 to facilitate reading.…”

“…The warping function that is (nearly) unanimously used in the ocean acoustics community derives from an ideal-isovelocity waveguide approximation. [14][15][16][17][18][19][20][21][22][23][24][25] The utility of such a model is that it provides a simple analytical warping solution that is applicable to every mode present in the signal. Other approximations exist that also permit warping every mode at once, e.g., approximated Pekeris waveguide, 13 waveguide invariant 26,27 or beam-displacement ray-mode theory.…”

“…Warping based on the ideal-isovelocity approximation, called "isovelocity warping" here, has been found to be robust to environmental mismatch, allowing successful modal filtering for real data in relatively complex shallow-water environments. It has been used by various researchers for numerous applications, including geoacoustic inversion, [15][16][17] water column tomography, 18 marine mammal localization, [19][20][21] etc. Nonetheless, modal filtering using isovelocity warping presents some limitations.…”

“…Modal filtering using isovelocity warping provides a reliable estimate of the mode phase; however, it seems to be less reliable for mode amplitude estimation, in particular, the first arriving mode. Indeed, for all the successful applications of warping previously cited, [14][15][16][17][18][19][20][21] the inversion algorithm was based on the modal phase (or, equivalently, on the modal TF dispersion curves that directly derives from the modal phase). On the other hand, very few at-sea applications have benefited from mode amplitude estimated using warping.…”