Kerri D. Seger

A decade of marine mammal acoustical presence and habitat preference in the Bering Sea

As Arctic seas rapidly change with increased ocean temperatures and decreased sea ice extent, traditional marine mammal distributions may be altered, and non-traditional Arctic species may shift poleward. Extant and seasonal continental shelf odontocete species in the Arctic Ocean include sperm whales, killer whales, beluga whales, delphinids, harbour porpoises, and Dall’s porpoises. Until recently, recording constraints limited higher sampling rates, preventing detection of many of these high frequency-producing species in the Arctic seas. Using one of the first long-term data sets to record clicks, buzzes, and whistles, multiple species have been detected and classified to explore shifting distributions in the Arctic Corridor. The data were then paired with environmental variables for generalized linear and additive modeling. Pacific white-sided dolphins were detected for a few years during minimal ice coverage. Dall’s porpoises and Risso’s dolphins may have been acoustically detected in areas farther n...

3D source location using overlapping multipath propagation

In passive acoustic monitoring, source localization using multipath propagation can be challenging whenever the source-receiver configuration leads to overlapping multipath arrival patterns. We propose a single-hydrophone model-based method with matched-impulse response (IR) processing that exploits unresolved multipath arrival patterns to locate the source in 3D, making the best use of the bathymetric information as well as the time differences of groups of arrivals. When only groups of paths are resolved, the identity of the paths constituting the groups remains unknown, preventing the use of classical localization methods. To overcome this critical limitation, we apply an approach based on the characterization of groups of paths to estimate the source location.

Automatic classification of humpback whale social calls

Acoustic methods are an established technique to monitor marine mammal populations and behavior, but developments in computer science can expand the current capabilities. A central aim of these methods is the automated detection and classification of marine mammal vocalizations. While many studies have applied bioacoustic methods to cetacean calls, there has been limited success with humpback whale (Megaptera novaeangliae) social call classification, which has largely remained a manual task in the bioacoustics community. In this project, we automated this process by analyzing spectrograms of calls using PCA-based and connected-component-based methods, and derived features from relative power in the frequency bins of these spectrograms. These features were used to train and test a supervised Hidden Markov Model (HMM) algorithm to investigate classification feasibility. We varied the number of features used in this analysis by varying the sizes of frequency bins. Generally, we saw an increase in precision, ...

Marine mammal range and depth estimation using non-separated multipath propagation

In passive acoustic monitoring, source localization using multipath propagation can be challenging whenever the source-receiver configuration leads to unresolved paths. Here, we propose a single-hydrophone method to estimate the range of a source based on two steps. First, we define a time deformation operator (e.g., Bonnel et al. 2014) to estimate the signal’s impulse response (IR). Each group of paths, when warped at its time of arrival, is transformed into a tone that corresponds to a peak in the IR, which can be isolated in the warped time-frequency domain, permitting a significant improvement of the time resolution of the multipath, or groups of multipath. Next, relative arrival times are used to estimate the source range and depth. When only groups of paths are resolved, the identity of the paths constituting the group remains unknown, preventing the use of classical localization methods. To overcome this critical limitation, we apply an approach based on the characterization of groups of paths to estimate source location. We evaluate the method’s performance on synthesized data, and examine its potential use on humpback whale low-frequency modulated calls using data collected off Cabo San Lucas in 2013.

Preliminary analysis of social calls used by tagged humpback whales in the Los Cabos, México, breeding ground

Humpback whales produce a large variety of diverse sounds beyond their well-known songs. Mothers, calves, and non-breeding whales may use these “social sounds” to maintain group cohesion, facilitate feeding, and/or increase a calf’s safety. To date, social sounds have been studied off Hawaii, Alaska, and Australia. During the 2014-2015 breeding seasons, acoustic tag data were collected from humpback whales off Los Cabos, Mexico. Twenty-one tags successfully recorded data (nine in 2014; 13 in 2015) from three mother/calf pairs, ten mother/calf/escort groups, and eight competitive pods varying in size from four to fourteen individuals. A subsequent manual analysis found 1587 social sounds in 2258 total minutes of data. Currently, there are 42 identified distinct social sounds used by humpback whales in the Los Cabos breeding ground, six of which seem unique to the Cabo region as compared to those published from Australia, Hawaii, and Alaska. Call type usage, call rates, and repertoire diversity (measured us...

A preliminary acoustical survey of echolocating marine mammals in the Bering Sea

As the Arctic seas rapidly change with increased ocean temperatures and decreased sea ice extent, traditional Arctic marine mammal distributions will be altered, and non-traditionally Arctic species may shift poleward. Arctic species typically include sperm, bowhead, humpback, right, gray, fin, and blue whales; odontocetes, specifically killer and beluga whales; and several pinnipeds species. Their acoustic presence has been documented because they produce relatively low-frequency sounds that are detectable by many common, remotely-deployed recording platforms. Until recently, however, recording constraints of power and storage limited higher sampling rates and prevented the detection of many high-frequency-producing species in the Arctic seas. Such species likely include Baird’s, Cuvier’s, and Stejneger’s beaked whales, as well as Northern right whale and Pacific white-sided dolphins. Using one of the first long-term data sets to record relatively high frequencies in the Bering Sea, signal types similar ...

A seven-year review of ambient acoustic environments in the Beaufort Sea

Over seven seasons (2007 to 2014) Greeneridge Sciences, Inc., deployed passive acoustic recorders (DASARs) between August and October at five sites in the Beaufort Sea off the Alaskan North Slope to collect acoustic data during the fall bowhead whale migration. Each site consisted of 7–11 DASARs, arranged in triangular grids with 7 km spacing between each unit. The shallowest DASAR unit in each set was deployed 15–33 km due north of the coast in 22–39 m of water. The acoustic environments between sites differ due to varying ocean depths and sound source contributions (whales, winds, and human activities). Here, we conduct a multi-year bulk analysis of the summer Beaufort Sea acoustic environment, comparing noise properties as a function of both distance offshore and longitude within a given year, and across sites across all eight seasons. Comparisons were conducted by investigating the percentile distributions of ambient noise intensity over several bandwidths that are representative of the different soun...

Long-term measurements of the directionality and active intensity of the underwater noise field in the shallow Beaufort Sea

In each of the past seven years, at least 35 Directional Autonomous Seafloor Acoustic Recorders (DASARs) have been deployed over a 280 km swath of the Beaufort Sea continental shelf (20–55 m depth) during the open-water season to monitor the westward bowhead whale migration. DASARs have one omnidirectional pressure sensor and two orthogonal particle velocity sensors that permit measurements of the azimuths of both transient and continuous sounds, including diffuse ocean noise. Here, we map the azimuthal directionality of the Beaufort ambient noise field as a function of frequency and location across all seven seasons. Dominant directionalities exist in the diffuse ambient noise field, which change with frequency, time, and location. We examine how localized storms, heavy whale calling activity, seismic exploration, and other industrial activities influence the noise directionality. We also examine how both the active and reactive intensity of the noise evolves with frequency and time, by comparing the pha...

The ambient acoustic environment in Laguna San Ignacio, Baja California Sur, Mexico

Each winter gray whales (Eschrichtius robustus) breed and calve in Laguna San Ignacio, Mexico, where a robust, yet regulated, whale-watching industry exists. Baseline acoustic environments in LSIs three zones were monitored between 2008 and 2013, in anticipation of a new road being paved that will potentially increase tourist activity to this relatively isolated location. These zones differ in levels of both gray whale usage and tourist activity. Ambient sound level distributions were computed in terms of percentiles of power spectral densities. While these distributions are consistent across years within each zone, inter-zone differences are substantial. The acoustic environment in the upper zone is dominated by snapping shrimp that display a crepuscular cycle. Snapping shrimp also affect the middle zone, but tourist boat transits contribute to noise distributions during daylight hours. The lower zone has three source contributors to its acoustic environment: snapping shrimp, boats, and croaker fish. As suggested from earlier studies, a 300 Hz noise minimum exists in both the middle and lower zones of the lagoon, but not in the upper zone.

Estimating relative abundance of singing humpback whales in Los Cabos, México, using diffuse ambient noise

Previous research has speculated that diffuse ambient noise levels can be used to estimate relative cetacean abundance in certain locations when baleen whale vocal activity dominates the soundscape (Au et al., 2000; Mellinger et al., 2009). During the 2013 and 2014 humpback whale breeding seasons off Los Cabos, Mexico, visual point and line transects were conducted alongside two bottom-mounted acoustic deployments. As theorized, preliminary analysis of ambient noise between 100 and 1,000 Hz is dominated by humpback whale song. It also displays a diel cycle similar to that found in the West Indies, Australia, and Hawai’i, whereby peak levels occur near midnight and troughs occur soon after sunrise (Au et al., 2000; McCauley et al., 1996). Depending upon site and year, the median band-integrated levels fluctuated between 7 and 16 dB re 1 uPa when sampled in one hour increments. This presentation uses analytical models of wind-generated noise in an ocean waveguide to analyze potential relationships between s...