Analysis and localization of blue whale vocalizations in the Solomon Sea using waveform amplitude data

A method for tracking blue whales (Balaenoptera musculus) with a widely spaced network of ocean bottom seismometers

Passive acoustic monitoring is an important tool for studying marine mammals. Ocean bottom seismometer networks provide data sets of opportunity for studying blue whales (Balaenoptera musculus) which vocalize extensively at seismic frequencies. We describe methods to localize calls and obtain tracks using the B call of northeast Pacific blue whale recorded by a large network of widely spaced ocean bottom seismometers off the coast of the Pacific Northwest. The first harmonic of the B call at ~15 Hz is detected using spectrogram cross-correlation. The seasonality of calls, inferred from a dataset of calls identified by an analyst, is used to estimate the probability that detections are true positives as a function of the strength of the detection. Because the spacing of seismometers reaches 70 km, faint detections with a significant probability of being false positives must be considered in multi-station localizations. Calls are located by maximizing a likelihood function which considers each strong detection in turn as the earliest arrival time and seeks to fit the times of detections that follow within a feasible time and distance window. An alternative procedure seeks solutions based on the detections that maximize their sum after weighting by detection strength and proximity. Both approaches lead to many spurious solutions that can mix detections from different B calls and include false detections including misidentified A calls. Tracks that are reliable can be obtained iteratively by assigning detections to localizations that are grouped in space and time, and requiring groups of at least 20 locations. Smooth paths are fit to tracks by including constraints that minimize changes in speed and direction while fitting the locations to their uncertainties or applying the double difference relocation method. The reliability of localizations for future experiments might be improved by increasing sampling rates and detecting harmonics of the B call.

Passive stochastic matched filter for Antarctic blue whale call detection

As a first step to Antarctic blue whale (ABW) monitoring using passive acoustics, a method based on the stochastic matched filter (SMF) is proposed. Derived from the matched filter (MF), this filter-based denoising method enhances stochastic signals embedded in an additive colored noise by maximizing its output signal to noise ratio (SNR). These assumptions are well adapted to the passive detection of ABW calls where emitted signals are modified by the unknown impulse response of the propagation channel. A filter bank is computed and stored offline based on a priori knowledge of the signal second order statistics and simulated colored sea-noise. Then, the detection relies on online background noise and SNR estimation, realized using time-frequency analysis. The SMF output is cross-correlated with the signal's reference (SMF + MF). Its performances are assessed on an ccean bottom seismometer-recorded ground truth dataset of 845 ABW calls, where the location of the whale is known. This dataset provides great SNR variations in diverse soundscapes. The SMF + MF performances are compared to the commonly used MF and to the Z-detector (a sub-space detector for ABW calls). Mostly, the benefits of the use of the SMF + MF are revealed on low signal to noise observations: in comparison to the MF with identical detection threshold, the false alarm rate drastically decreases while the detection rate stays high. Compared to the Z-detector, it allows the extension of the detection range of ≃ 30 km in presence of ship noise with equivalent false discovery rate.

Calls reveal population structure of blue whales across the southeast Indian Ocean and the southwest Pacific Ocean

For effective species management, understanding population structure and distribution is critical. However, quantifying population structure is not always straightforward. Within the Southern Hemisphere, the blue whale (Balaenoptera musculus) complex is extremely diverse but difficult to study. Using automated detector methods, we identified “acoustic populations” of whales producing region-specific call types. We examined blue whale call types in passive acoustic data at sites spanning over 7,370 km across the southeast Indian Ocean and southwest Pacific Ocean (SWPO) from 2009 to 2012. In the absence of genetic resolution, these acoustic populations offer unique information about the blue whale population complex. We found that the Australian continent acts as a geographic boundary, separating Australia and New Zealand blue whale acoustic populations at the junction of the Indian and Pacific Ocean basins. We located blue whales in previously undocumented locations, including the far SWPO, in the Tasman Sea off the east coast of Australia, and along the Lau Basin near Tonga. Our understanding of population dynamics across this broad scale has significant implications to recovery and conservation management for this endangered species, at a regional and global scale.

Acoustically invisible feeding blue whales in Northern Icelandic waters

Fixed passive acoustic monitoring can be used for long-term recording of vocalizing cetaceans. Both presence monitoring and animal density estimation requires the call rates and sound source levels of vocalizations produced by single animals. In this study, blue whale calls were recorded using acoustic bio-logging systems in Skjálfandi Bay off Húsavík, Northeast Iceland, in June 2012. An accelerometer was attached to individual whales to monitor diving behavior. During 21 h recording two individuals, 8 h 45 min and 13 h 2 min, respectively, 105 and 104 lunge feeding events and four calls were recorded. All recorded calls were down-sweep calls ranging from 105 to 48 Hz. The sound duration was 1-2 s. The source level was estimated to be between 158 and 169 dB re 1μPa rms, assuming spherical sound propagation from the possible sound source location to the tag. The observed sound production rates and source levels of individual blue whales during feeding were extremely small compared with those observed previously in breeding grounds. The feeding whales were nearly acoustically invisible. The function of calls during feeding remains unknown.

Blue whale vocalizations recorded around New Zealand: 1964-2013

Previous underwater recordings made in New Zealand have identified a complex sequence of low frequency sounds that have been attributed to blue whales based on similarity to blue whale songs in other areas. Recordings of sounds with these characteristics were made opportunistically during the Southern Ocean Research Partnership's recent Antarctic Blue Whale Voyage. Detections of these sounds occurred all around the South Island of New Zealand during the voyage transits from Nelson, New Zealand to the Antarctic and return. By following acoustic bearings from directional sonobuoys, blue whales were visually detected and confirmed as the source of these sounds. These recordings, together with the historical recordings made northeast of New Zealand, indicate song types that persist over several decades and are indicative of the year-round presence of a population of blue whales that inhabits the waters around New Zealand. Measurements of the four-part vocalizations reveal that blue whale song in this region has changed slowly, but consistently over the past 50 years. The most intense units of these calls were detected as far south as 53°S, which represents a considerable range extension compared to the limited prior data on the spatial distribution of this population.

Comparative analysis of localization algorithms with application to passive acoustic monitoring

The problem of sound source localization using sparse arrays of bottom-mounted synchronized hydrophones is addressed. The closed-form representations for several time-differences of arrival based localization algorithms are given, and their accuracies are compared using both statistical simulations and in situ measurements. In most of the tests, the lowest localization error was provided by the maximum likelihood algorithm.

Applying distance sampling to fin whale calls recorded by single seismic instruments in the northeast Atlantic

Automated methods were developed to detect fin whale calls recorded by an array of ocean bottom seismometers (OBSs) deployed off the Portuguese coast between 2007 and 2008. Using recordings collected on a single day in January 2008, a standard seismological method for estimating earthquake location from single instruments, the three-component analysis, was used to estimate the relative azimuth, incidence angle, and horizontal range between each OBS and detected calls. A validation study using airgun shots, performed prior to the call analysis, indicated that the accuracy of the three-component analysis was satisfactory for this preliminary study. Point transect sampling using cue counts, a form of distance sampling, was then used to estimate the average probability of detecting a call via the array during the chosen day. This is a key step to estimating density or abundance of animals using passive acoustic data. The average probability of detection was estimated to be 0.313 (standard error: 0.033). However, fin whale density could not be estimated due to a lack of an appropriate estimate of cue (i.e., vocalization) rate. This study demonstrates the potential for using a sparse array of widely spaced, independently operating acoustic sensors, such as OBSs, for estimating cetacean density.

Tracking fin whales in the northeast Pacific Ocean with a seafloor seismic network

Ocean bottom seismometer (OBS) networks represent a tool of opportunity to study fin and blue whales. A small OBS network on the Juan de Fuca Ridge in the northeast Pacific Ocean in ~2.3 km of water recorded an extensive data set of 20-Hz fin whale calls. An automated method has been developed to identify arrival times based on instantaneous frequency and amplitude and to locate calls using a grid search even in the presence of a few bad arrival times. When only one whale is calling near the network, tracks can generally be obtained up to distances of ~15 km from the network. When the calls from multiple whales overlap, user supervision is required to identify tracks. The absolute and relative amplitudes of arrivals and their three-component particle motions provide additional constraints on call location but are not useful for extending the distance to which calls can be located. The double-difference method inverts for changes in relative call locations using differences in residuals for pairs of nearby calls recorded on a common station. The method significantly reduces the unsystematic component of the location error, especially when inconsistencies in arrival time observations are minimized by cross-correlation.