Open Access
Issue
Acta Acust.
Volume 7, 2023
Article Number 46
Number of page(s) 10
Section Underwater Sound
DOI https://doi.org/10.1051/aacus/2023043
Published online 13 October 2023

© The Author(s), Published by EDP Sciences, 2023

Licence Creative CommonsThis is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

1 Introduction

Anthropogenic noise in the ocean is becoming louder, more ubiquitous and more diverse [1]. For example, due to the noise produced by commercial shipping, the ambient noise level (30–50 Hz) in the Northeast Pacific increased by 2.5–3 dB per decade from the 1960s to the 2000s [2]. Sources of anthropogenic noise are categorized as either continuous, such as the operation of offshore wind turbines, or impulsive, such as the noise from air guns used during seismic surveys for subsea deposits of oil [3]. Usually, both noise intensity and directionality within the soundscape change with time at any given location, especially where there is frequent shipping activity. When this occurs, the direction of a source of noise varies through time from the perspective of a listener (receiver). Characterizing noise directionality is needed to understand an organism’s reactions to it, including adverse effects [46]. However, most bioacoustics research reports only noise intensity measured in terms of sound pressure level (SPL). The directionality of the noise is rarely reported. Since the directionality of the noise is what marine animals use to assess where underwater noise comes from [7], it is essential to measure and characterize it in bioacoustics research.

For either omnidirectional or directional sound sources, the incoming sound wave provides marine organisms with information about the direction of the source. If the organism can hear the frequencies of the sound produced by the noise source, determining its direction is necessary to react accordingly, for example, when moving towards or away from relevant stimuli such as conspecifics, predators, or anthropogenic noise.

Sound directionality in water can be measured by using a sonar system in passive mode or an underwater acoustic vector sensor (AVS). Compared to the complexity of a sonar, both in terms of hardware and signal processing, the AVS is a compact and powerful device that measures co-located sound pressure and two (2D) or three (3D) orthogonal components of particle motion, providing data to compute the direction of incoming sound from a listener’s perspective. AVS technology has been used by the US Navy to detect submarines [8], by marine researchers to locate the direction of calls from baleen whales [9] and to assess fish orientation in response to low-frequency sound [10]. An AVS can also detect the directionality of weak sound sources. An AVS deployed along the coast of Belgium revealed the ambient noise (20–40 Hz) coming from the direction of the Thames River [11].

Using a Norwegian fjord in which there is frequent ship traffic and other sources of noise as a case study, we document how an AVS can be used to effectively characterize the directionality of noise from known and unknown sources.

2 Material and methods

2.1 Instrumentation

The AVS instrument (M20-105) was acquired from GeoSpectrum Technologies Inc. (GTI, Halifax, Nova Scotia, Canada). It consisted of an underwater AVS (M20) calibrated by GTI in February 2021, an M209 data logger with 256 GB of internal memory, a deployment frame and a battery pack containing 64 D-Cell alkaline batteries (Fig. 1). The AVS was suspended in the middle of the deployment frame by a pliable rope. This suspension ensured that the AVS kept its vertical position during the deployment, particularly within 15° inclination to avoid touching the frame. The total weight of the instrument package in air was approximately 50 kg. The data logger acquired 4-channel acoustic data: directional x, y and z channels and one omnidirectional acoustic pressure channel the sensitivity of which was −177.6 dB re 1 V/μPa (10–3000 Hz). Data from a magnetic compass sensor located inside the AVS, and an inclinometer located in the data logger, was logged to obtain supporting information about position and movement of the AVS after deployment on the sea bottom. The compass provided the magnetic north to discriminate noise bearing relative to the magnetic north (direction). Hereafter, the bearing is defined as the angle of the incoming acoustic wave when the x-channel was pointing to the north. The angle was measured in degrees and calculated clockwise from the magnetic north. The bearing accuracy of the AVS depends on the receiving signal-to-noise (SNR) ratio; the accuracy can be ± 3° when the SNR is greater than 20 dB and decreases to ± 20° when the SNR is a few dB. The accuracy becomes lower when sound arrives straight at either the X or Y channel and is highest when sound arrives at ±45° to the X or Y channels (GTI Technology, personal communication).

thumbnail Figure 1

M20-105 system (GeoSpectrum Technologies Inc., Halifax, Canada) used to measure underwater noise underwater.

2.2 Analysis method

We studied noise directionality in a large area, where the horizontal distance between the main noise sources and the AVS is much larger than the water depth. In these conditions, the metric of most interest was the azimuth angle ϕ of the arriving noise, which is defined above as bearing. To display noise azimuth, we plotted the bearings of the noise at different frequencies versus time using an azigram [9, 12]), in which ϕ is estimated using the active intensity of an acoustic field [13, 14]. The active intensity is a vector quantity that is used to determine the bearing of a sound source. Essentially, it is the real part of the product of sound pressure and particle velocity, particularly , where p is the sound pressure, v is the particle velocity and Re<> represents the real part operator. This method involves the AVS sound pressure p, orthogonal particle velocities of the horizontal plane, particularly vx and vy. φ(K, T) denotes the azimuth of Fast Fourier Transform (FFT) bin K at sample time T, and it is calculated by

(1)

where FFT() is the FFT operator, and the symbol “*” indicates complex phase conjugation. Equation (1) is analogous to calculating a spectrogram in which the time duration of a processing segment at T can be adjusted to achieve expected frequency resolutions. Overlap between segments can be added to observe time variations in bearing. Following [15], the calculation parameters are kept the same for both the spectrogram and azigram (see Tab. 1), in which the raw data were downsampled to 2000 Hz for calculations.

Table 1

Calculation parameters for spectrogram and azigram.

2.3 Instrument and method validation

The field trials to verify the setup of the AVS system (M20-105) and data processing method were conducted on 14 January 2022 (Fig. 2a), when an active acoustic source (omnidirectional) continuously transmitted 100 Hz with a source level of 175 dB re 1 μPa @ 1 m throughout the trials. The AVS system was placed on the sea bottom at a known location (latitude: 60.0937°, longitude: 5.2670°), which was close to the Institute of Marine Research’s (IMR’s) Austevoll Research Station (https://www.hi.no/en/hi/laboratories/austevoll-research-station) in Austevoll, Norway. The system recorded the source transmission at a depth of 27 m, and its inclination angles (3.6° alonghship and −2.3° athwart) indicate that the system landed on a flat surface. The source (C-Bass M72-110 acquired from GTI) was suspended 5 m below the boat. We performed two trials. During the first trial, the boat drifted freely – away from the AVS – for about 12 min with the boat engine off. During the second trial, the boat was maneuvered for about 12 min to keep the sound source to the west of the AVS deployment location. The Global Positioning System (GPS) positions of the boat and the AVS were recorded to verify the acoustic bearing of the source (given by the boat position) to the AVS.

thumbnail Figure 2

Trials were conducted to verify the instrument (M20-105) using a known source transmitting a signal of 100 Hz. (a) Trial location with the boat tracks. The instrument system (black star) was deployed on the sea bottom. The deployment location (latitude: 60.0937°, longitude: 5.2670°) was about 800 m from the sea cages of the Institute of Marine Research’s Austevoll Research Station (the red circle). The active source was suspended 5 m below the boat Tabben (MMSI: 257033570), and the red (free drift trial) and yellow (maneuvering trial) curves denote the boat tracks. (b) Bearing comparison between AVS and GPS when the boat drifted freely with engine off. (c) Bearing comparison between AVS and GPS when the boat was maneuvered close to, and west/southwest of the deployment location. In (b) and (c), the AVS bearing (blue line) shows the bearing of the 100 Hz signal with local time displayed on the x-axis (HHMMSS). The GPS bearing (orange line) was computed from the GPS positions of the boat (the sound source) to the AVS deployment position.

Using the calculation parameter in Table 1, the bearings of the 100 Hz sound source were extracted from the azigram calculation using equation (1). The inclination of the AVS was not compensated for because it was very small. The AVS location on the bottom was assumed to match the deployment location. The GPS bearings (from the source relative to the AVS) were compared to the 100 Hz acoustic bearings calculated by using the AVS data. During the free drift trial, the difference between AVS and GPS varied from 0 to 50° (Fig. 2b), with a greater difference at the beginning of the drifting. The difference fluctuated more when the distance to the AVS increased. During the trial with the engine switched on, the acoustic bearing was consistent with the variation of the GPS bearing (Fig. 2c). The largest difference of 37° (Fig. 2b) occurred between 13:01:41 and 13:03:21.

2.4 Noise measurement setup

In order to characterize the directionality of underwater noise, we deployed an AVS on the sea bottom in a Norwegian fjord in which there is frequent commercial ship traffic (Fig. 3). Using a drop launch method, the acoustic data recording system was deployed at the sea bottom, about 80 m below the surface (latitude: 60.0871°, longitude: 5.28062°). It was about 820 m to the IMR’s Austevoll Research Station and 380 m to the shore (Fig. 3). The area was carefully selected because of shipping activity. Specifically, the area is in proximity to the route of the fast boat Admiralen (MMSI: 258352500), which cruises from the southeast to north of the location, and there is also noise from electric ferries passing in the Bjørnafjorden. The data from the automatic identification system (AIS) from the vessels with Maritime Mobile Service Identities (MMSIs) contains GPS tracks that were used to verify possible ship noise sources. We collected noise data for about 71 h, from local time 10:24:47 20 May 2022 to 09:22:47 23 May 2022.

thumbnail Figure 3

Location of the field measurements in the Bjørnafjorden, Norway. The measurement instrument system (denoted as AVS; the orange point) was deployed on the sea bottom. The deployment location (latitude: 60.0871°, longitude: 5.28062°) was about 820 m from the Institute of Marine Research’s Austevoll Research Station (the red star), 650 m from the harbor in Kolbeinshamn (the blue star), and 10 km from the ferry stop at Halhjem (the green star). Ship traffic tracks during the measurement period are presented as colored lines, given by the GPS positions of the AIS data. The orange dashed lines assist in visualizing the four quadrants of the directional measurements made by the AVS.

The system (M20-105) was deployed using a hydraulic crane and was lowered to the sea bottom. A rope connected a surface marking float to the top lift ring (see Fig. 1) such that, at high tide, the rope remained slack. The system recorded the noise continuously for 71 h, at a sample rate of 8 kHz, and the data were analyzed after retrieval. The data show that inclination angles were much less than the maximum tilt angle (15° as recommended by GTI) during the deployment, particularly alongship −4.5° and athwart 1.1°, such that there was no inclination problem for the AVS. Therefore, it was not necessary to compensate for the inclination of the AVS.

3. Results

3.1 Directionality of the noise from a fast-moving vessel

An example of a vessel passage in the fjord (fast boat Admiralen) is shown in Figure 4. The vessel was heading to the northwest with a closest point of approach of 690 m (Fig. 4c). The vessel Admiralen was moving at a high speed over ground of about 31 knots, and it generated intense noise over a wide range of frequencies, mostly 100–000 Hz (Fig. 4a). The passage is observable on both the spectrogram and the azigram between local time 13:04:57 and 13:07:51 (Figs. 4a and 4b). The bearing of the vessel relative to the AVS changed from about 110° to 0° (highlighted by the red dot lines in Fig. 4b). The boat noise bearing disappeared when the vessel moved behind the islands surrounding the AVS (close to 0° to the AVS in Fig. 4). The evolution of bearing overtime was validated using the boat’s GPS track.

thumbnail Figure 4

Example of a 30-min measurement covering a vessel passage on 21 May 2022. (a) Spectrogram of power spectral density. (b) Azigram showing only the bearings for the data points in (a) that are >70 dB. The 70 dB is chosen for visualization purposes. The white color denotes the non-visualized bearings. (c) 2.5 min track of the vessel Admiralen (MMSI: 258352500). The vessel headed northwest with a speed over ground of about 31 knots. In (a) and (b), the x-axis is local time (HHMMSS) and calculation parameters are shown in Table 1.

3.2 Overview of the sound pressure

An overview of time-varying noise pressure from all noise sources is displayed as a spectrogram of power spectral density (PSD) (Fig. 5, data from the AVS sound pressure channel), which shows that the fjord is rarely quiet, especially at low frequencies (less than 20 Hz). The spectrogram displays the time-varying noise intensity. The spectrogram is typically used to identify noise sources based on known characteristics of specific sounds. For example, the vertical stripes in the figure are intense broadband noise of short duration, which is typically generated by fast and loud small boats [16]. Activity of small recreational vessels could not be confirmed due to the lack of AIS data; many of these boats are not required to have AIS. The noise highlighted by the white square is typically created by rotational machines [17], as the typical noise spectrum consists of dominant multiple frequency components. Every night at fixed time intervals (highlighted by the red squares), continuous broadband noise is observed. There was other noise of a repetitive pattern, such as that around 300 Hz (examples pointed out by the white arrows) of an intermittent pattern, the noise around 80 Hz (examples pointed out by the red arrows), and the other noises of similar temporal pattern were obviously caused by time scheduled activities, given by the time gap between two occurrences (visual inspection). Using the AVS sound pressure channel (the same as a single hydrophone), the spectral characteristics of the noise over time can be displayed, but the pressure channel alone does not record the directionality of the noise sources.

thumbnail Figure 5

Spectrogram of power spectral density from local time 10:22:47 20 May 2022 to 09:22:47 23 May 2022. Calculation parameters are shown in Table 1. The examples highlighted by the arrows and squares represent different sources of anthropogenic noise.

3.3 Directionality of the noise from a static vessel

An example of 30 min measurement (highlighted by the white square in Fig. 5) is shown in Figure 6. It shows that there was noise of tonal frequencies (pointed by the white arrows in Figure 6a), such as 130 Hz, 180 Hz, 300 Hz, etc., which are typical of rotating machine noise [17]. As shown by the color of the azigram in Figure 6b, multiple frequencies over 180 Hz (above the white dot line) come from a bearing range from 270 to 315°, and frequencies of 50–180 Hz (between the white dot line and dash line) come from around 90° which will be discussed in Section 3.5. After applying a threshold (for visualization purpose) to the data in the azigram, the noise sources from two different bearings are more clearly visible (Fig. 6c). To illustrate the bearing distributions of single frequencies, 440–449 Hz are visually selected from the azigram (Fig. 6c). In Figure 6d, most of those frequencies have approximate bearings of around 280° (the red dashed line), given by the bearing distributions. According to the AIS data, the supply vessel Holmefjord (MMSI: 257586700) was working at the IMR Austevoll Station during this time period. Calculated from the GPS positions, the bearing from the Holmefjord vessel to the AVS was 281°, which matches the focused bearing of the frequencies selected in Figure 6d.

thumbnail Figure 6

Example of a 30 min measurement made on 20 May 2022. (a) Spectrogram of power spectral density. (b) Azigram without a threshold. (c) Azigram showing only the bearings for the data points in (a) that are > 70 dB. The 70 dB is chosen for visualization purposes. The white color denotes the non-visualized bearings. (d) Bearing distributions of the frequencies 440–449 Hz without applying a threshold to the azigram. The boxplots in (d) show minimum, 25th percentile, median, and 75th percentile. In (a) (b) and (c), the x-axis is time (HHMMSS) and the calculation parameters are shown in Table 1.

3.4 Directionality of the noise from power generators

A 30 min measurement of the noise occurring at night (red squares in Fig. 5) was investigated (Fig. 7). At a time of little ship traffic, the noise with PSD of approximately 75 dB re 1 μPa2 Hz−1 at 20–300 Hz is visible in the spectrogram (Fig. 7a). The azigram (Fig. 7b) shows that there were likely two sound sources; one had a bearing of about 225° (pointed by the blue arrows), and the other one had a bearing of about 135° (highlighted by the white square). In order to get an overview of the accumulated acoustic energy versus the bearing from 0 to 360°, the acoustic power of a frequency from a bearing is integrated over the 30 min to obtain the sound exposure level (SEL) (Fig. 7c). It is clear that a lot of acoustic energy (0–1 kHz) came from two dominant bearing ranges, highlighted by the black and red squares, in which more energy (the black square) came from the range with a center bearing of 240° (frequency >150 Hz; Fig. 7c). According to the AIS data during this period, there were no moving or stationary vessels in that bearing range (the black square). The island of Huftarøy is along that bearing (see Fig. 3). After investigation, we learned that several working boats without AIS had switched on power generators at the Kolbeinshamn harbor during those nights, when the engines of the boats were switched off. The bearing from the harbor to the AVS is about 225°, which matches with the bearing information given by the AVS.

thumbnail Figure 7

Example of 30 min noise measurements on 21 May 2022 at night. (a) Spectrogram of power spectral density. (c) Azigram showing only the bearings for the data points in (a) that are >65 dB. The 65 dB is chosen for visualization purposes. The white color denotes the non-visualized bearings. (c) Accumulated energy distribution over 30 min versus bearing and frequency. In (a) and (b), the x-axis is time (HHMMSS) and calculation parameters are shown in Table 1. In (c), the frequency and bearing resolutions are 1 Hz and 1°.

3.5 Directionality of the noise from electric ferries and small boats

By observing the SEL versus bearing throughout the 71-hour deployment of the AVS (Fig. 8a), the two bearing ranges, highlighted by the black and red squares, are identified as directions where the most noise energy was coming from. Since the AIS data only confirms the activities of big vessels, it is assumed that most of the noise to the left side of the AVS (bearing 180–360°) was generated by small boat traffic. This is supported by the fact that numerous small vessels transit in that area regularly, from/to the Kolbeinshamn harbor, the residential area, and IMR Austevoll Research Station. From the bearing distributions of example frequencies (Fig. 8b), there are two bearing ranges, denoted as B1 and B2; where B2 (200–290°) is consistent with the assumption that noise originated from small boat traffic. The arrival time of the noise of a repetitive pattern, as highlighted by the red and white arrows in Figure 5, is consistent with the official ferry time at Halhjem (Fig. 8c) which is confirmed by the AIS data. For instance, the noise at 300 Hz (highlighted by the white arrows in Fig. 5) has a clear time pattern, becoming less frequent when there were less frequent ferry crossings between midnight and early morning. This noise pattern at 300 Hz from the electric ferry was probably not vessel dependent. This was confirmed by the ferry operator – all four ferries cruising in the area were built in 2019 and are outfitted with similar mechanical equipment. From the ferry tracks, it is visible that the distance between the ferries and the AVS (Fig. 8c) ranged from 3.7 to 9.5 km, and the bearing from 58.7° to 117.2°.

thumbnail Figure 8

Measurement covering 71 h. (a) Accumulated energy distribution of different frequency versus bearing. (b) Bearing distributions of selected frequencies. B1 and B2 denote two bearing ranges. (c) GPS tracks of the ferries. The azigram calculation parameters are shown in Table 1. In (a), the frequency and bearing resolutions are 1 Hz and 1°. The ferries are Lysoey (MMSI: 257056190), Flatoy (MMSI: 257055990), Faeroes (MMSI: 257055910) and Huftaroy (MMSI: 257055930), and they were built in 2019 according to the registration.

The ferry noise is likely to be responsible for the energy accumulated in the bearing range B1. Analogous observations are shown in Figures 6c and 7b when looking at the bearings at 50–150 Hz versus time. From the spectrogram overview (Fig. 5), the ferry noise (around 300 Hz) is relatively weak which was due to the transmission loss over long distance, but its directional energy accumulated in those 3 days is detected by the AVS.

4. Discussion

From the analysis of the data collected with the AVS in the Bjørnafjorden, an area with frequent ship traffic, we determined the directionality of anthropogenic noise from various sources. Using the AVS, we could resolve the direction of noise sources, such as the electric ferries crossing the fjord. The analysis shows that the bearings were often frequency dependent, and multiple noise sources were often present at the same time. This case study illustrates that the soundscape is complex not only with respect to time-varying sound pressure but also in terms of noise directionality, which varies in the frequency domain as well. Thus, measurements of acoustic particle motion obtained from an AVS provide valuable spatial information, particularly the directionality of anthropogenic noise.

Acoustic particle motion, which provides directional information about underwater sound, is rarely considered in research on the effects of noise on marine organisms [18, 19]. That is at least partly because of the challenges associating, measuring and interpreting it. This study shows that an AVS can be used to quantify the directionality of the noise propagating from vessels, and that the directional information is preserved even in noisy environments. Specifically, trials with an AVS could characterize the directionality, from a listener’s perspective, of the noise from a specific vessel through time and across frequencies. Information such as this would support the interpretation of fish behavior in response to such noise.

Since both sound pressure and directionality characterize an acoustic field (a soundscape), it is recommended that acoustic particle motion be measured in research on the effects of sound on marine organisms [5]. The results reported in this study show that an AVS can be used to characterize noise directionality from known and unknown sources of sound. This has potential for applications in bioacoustics research to aid in the interpretation of behavioral responses of marine organisms to underwater noise. Specifically, an AVS can be used to characterize the directional features of the noise that marine organisms would detect. The AVS can also be used to describe directional features in sound exposure experiments in which sound projectors are often used to expose animals to specific sounds.


a

These authors contributed equally to the work.

Acknowledgments

Thanks to the reviewers and the Norwegian Coastal Administration for the AIS data and IMR engineer Glenn Sandtorv for assistance with the field work.

Conflict of interest

The Authors declare no conflict of interest.

Funding

This work was funded by the Norwegian Institute of Marine Research’s project “Assessing the effects of offshore wind power generating facilities on the early life stages of fish” (Project 15655) to Howard I. Browman.

Data availability statement

Data are available on request from the authors.

References

  1. C. Duarte, L. Chapuis, S. Collin, D. Costa, R. Devassy, V. Eguíluz, C. Erbe, T. Lamont, B. Halpern, H. Harding, M. Havlik, M. Meekan, N. Merchant, J. Miksis-Olds, M. Parsons, M. Predragovic, A. Radford, C. Radford, S. Simpson, F. Juanes: The soundscape of the Anthropocene ocean. Science 371 (2021) eaba4658. [CrossRef] [PubMed] [Google Scholar]
  2. M.A. McDonald, J.A. Hildebrand, S.M. Wiggins: Increases in deep ocean ambient noise in the Northeast Pacific west of San Nicolas Island, California. Journal of the Acoustical Society of America 120 (2006) 711–718. [CrossRef] [PubMed] [Google Scholar]
  3. B.L. Southall, A.E. Bowles, W.T. Ellison, J.J. Finneran, R.L. Gentry, C.R. Greene, D. Kastak, D.R. Ketten, J.H. Miller, P.E. Nachtigall, W.J. Richardson, J.A. Thomas, P.L. Tyack: Marine mammal noise-exposure criteria: initial scientific recommendations. Bioacoustics 17 (2008) 273–275. [CrossRef] [Google Scholar]
  4. A. Chahouri, N. Elouahmani, H. Ouchene: Recent progress in marine noise pollution: A thorough review. Chemosphere 291 (2022) 132983. [CrossRef] [PubMed] [Google Scholar]
  5. A.D. Hawkins, A.N. Popper: A sound approach to assessing the impact of underwater noise on marine fishes and invertebrates. ICES Journal of Marine Science 74 (2016) 635–651. [Google Scholar]
  6. R. Williams, A.J. Wright, E. Ashe, L.K. Blight, R. Bruintjes, R. Canessa, C.W. Clark, S. Cullis-Suzuki, D.T. Dakin, C. Erbe, P.S. Hammond, N.D. Merchant, P.D. O’Hara, J. Purser, A.N. Radford, S.D. Simpson, L. Thomas, M.A. Wale: Impacts of anthropogenic noise on marine life: Publication patterns, new discoveries, and future directions in research and management. Ocean & Coastal Management 115 (2015) 17–24. [CrossRef] [Google Scholar]
  7. R. Fay: Soundscapes and the sense of hearing of fishes. Integrative Zoology 4 (2009) 26–32. [CrossRef] [PubMed] [Google Scholar]
  8. R.A. Holler: The evolution of the sonobuoy from World War II to the Cold War. U.S. Navy Journal of Underwater Acoustics 62 (2014) 322–347. [Google Scholar]
  9. A.M. Thode, T. Sakai, J. Michalec, S. Rankin, M.S. Soldevilla, B. Martin, K.H. Kim: Displaying bioacoustic directional information from sonobuoys using “azigrams”. The Journal of the Acoustical Society of America 146 (2019) 95–102. [CrossRef] [PubMed] [Google Scholar]
  10. A. Cresci, G. Zhang, C.M.F. Durif, et al.: Atlantic cod (Gadus morhua) larvae are attracted by low-frequency noise simulating that of operating offshore wind farms. Communications Biology 6 (2023) 353. doi: 10.1038/s42003-023-04728-y. [CrossRef] [PubMed] [Google Scholar]
  11. P. Rogers, E. Debusschere, D.D. Haan, B. Martin, H. Slabbekoorn: North Sea soundscapes from a fish perspective: Directional patterns in particle motion and masking potential from anthropogenic noise. Journal of the Acoustical Society of America 150 (2021) 2174–2188. [CrossRef] [PubMed] [Google Scholar]
  12. P. Stinco, A. Tesei, G. Ferri, S. Biagini, M. Micheli, B. Garau, K.D. LePage, L. Troiano, A. Grati, P. Guerrini: Passive acoustic signal processing at low frequency with a 3-D acoustic vector sensor hosted on a buoyancy glider. IEEE Journal of Oceanic Engineering 46 (2021) 283–293. [CrossRef] [Google Scholar]
  13. G.L. D’Spain, W.S. Hodgkiss, G.L. Edmonds: Energetics of the deep ocean’s infrasonic sound field. Journal of the Acoustical Society of America 89 (1991) 1134–1158. [CrossRef] [Google Scholar]
  14. S. Nedelec, M. Ainslie, M. Andersson, C. Sei-Him, M. Halvorsen, M. Linné, B. Martin, A. Nöjd, S. Robinson, S. Simpson, L. Wang, J. Ward: Best Practice Guide for Underwater Particle Motion Measurement for Biological Applications. Technical report by the University of Exeter for the IOGP Marine Sound and Life Joint Industry Programme, 2021. [Google Scholar]
  15. S.P. Robinson, P.A. Lepper, R.A. Hazelwood, Good practice guide for underwater noise measurement, National Measurement Office, Marine Scotland, The Crown Estate, Teddington, England. 2014, p. 95. (NPL Good Practice Guide No. 133). http://dx.doi.org/10.25607/OBP-21 [Google Scholar]
  16. L. Hermannsen, L. Mikkelsen, J. Tougaard, K. Beedholm, M. Johnson, P.T. Madsen: Recreational vessels without Automatic Identification System (AIS) dominate anthropogenic noise contributions to a shallow water soundscape. Scientific Reports 9 (2019) 15477. [CrossRef] [PubMed] [Google Scholar]
  17. W.K. Blake, Chapter 6 – Noise from rotating machinery, in: W.K. Blake (Ed.), Mechanics of Flow-Induced Sound and Vibration, Volume 2 (Second Edition), Academic Press, 2017, pp. 505–658. [CrossRef] [Google Scholar]
  18. S.L. Nedelec, J. Campbell, A.N. Radford, S.D. Simpson, N.D. Merchant: Particle motion: the missing link in underwater acoustic ecology. Methods in Ecology and Evolution 7 (2016) 836–842. [CrossRef] [Google Scholar]
  19. A.N. Popper, A.D. Hawkins: The importance of particle motion to fishes and invertebrates. Journal of the Acoustical Society of America 143 (2018) 470–488. [CrossRef] [PubMed] [Google Scholar]

Cite this article as: Zhang G. Cresci A. & Browman HI. 2023. Determining the directionality of anthropogenic noise using an underwater acoustic vector sensor: a case study in a Norwegian fjord. Acta Acustica, 7, 46.

All Tables

Table 1

Calculation parameters for spectrogram and azigram.

All Figures

thumbnail Figure 1

M20-105 system (GeoSpectrum Technologies Inc., Halifax, Canada) used to measure underwater noise underwater.

In the text
thumbnail Figure 2

Trials were conducted to verify the instrument (M20-105) using a known source transmitting a signal of 100 Hz. (a) Trial location with the boat tracks. The instrument system (black star) was deployed on the sea bottom. The deployment location (latitude: 60.0937°, longitude: 5.2670°) was about 800 m from the sea cages of the Institute of Marine Research’s Austevoll Research Station (the red circle). The active source was suspended 5 m below the boat Tabben (MMSI: 257033570), and the red (free drift trial) and yellow (maneuvering trial) curves denote the boat tracks. (b) Bearing comparison between AVS and GPS when the boat drifted freely with engine off. (c) Bearing comparison between AVS and GPS when the boat was maneuvered close to, and west/southwest of the deployment location. In (b) and (c), the AVS bearing (blue line) shows the bearing of the 100 Hz signal with local time displayed on the x-axis (HHMMSS). The GPS bearing (orange line) was computed from the GPS positions of the boat (the sound source) to the AVS deployment position.

In the text
thumbnail Figure 3

Location of the field measurements in the Bjørnafjorden, Norway. The measurement instrument system (denoted as AVS; the orange point) was deployed on the sea bottom. The deployment location (latitude: 60.0871°, longitude: 5.28062°) was about 820 m from the Institute of Marine Research’s Austevoll Research Station (the red star), 650 m from the harbor in Kolbeinshamn (the blue star), and 10 km from the ferry stop at Halhjem (the green star). Ship traffic tracks during the measurement period are presented as colored lines, given by the GPS positions of the AIS data. The orange dashed lines assist in visualizing the four quadrants of the directional measurements made by the AVS.

In the text
thumbnail Figure 4

Example of a 30-min measurement covering a vessel passage on 21 May 2022. (a) Spectrogram of power spectral density. (b) Azigram showing only the bearings for the data points in (a) that are >70 dB. The 70 dB is chosen for visualization purposes. The white color denotes the non-visualized bearings. (c) 2.5 min track of the vessel Admiralen (MMSI: 258352500). The vessel headed northwest with a speed over ground of about 31 knots. In (a) and (b), the x-axis is local time (HHMMSS) and calculation parameters are shown in Table 1.

In the text
thumbnail Figure 5

Spectrogram of power spectral density from local time 10:22:47 20 May 2022 to 09:22:47 23 May 2022. Calculation parameters are shown in Table 1. The examples highlighted by the arrows and squares represent different sources of anthropogenic noise.

In the text
thumbnail Figure 6

Example of a 30 min measurement made on 20 May 2022. (a) Spectrogram of power spectral density. (b) Azigram without a threshold. (c) Azigram showing only the bearings for the data points in (a) that are > 70 dB. The 70 dB is chosen for visualization purposes. The white color denotes the non-visualized bearings. (d) Bearing distributions of the frequencies 440–449 Hz without applying a threshold to the azigram. The boxplots in (d) show minimum, 25th percentile, median, and 75th percentile. In (a) (b) and (c), the x-axis is time (HHMMSS) and the calculation parameters are shown in Table 1.

In the text
thumbnail Figure 7

Example of 30 min noise measurements on 21 May 2022 at night. (a) Spectrogram of power spectral density. (c) Azigram showing only the bearings for the data points in (a) that are >65 dB. The 65 dB is chosen for visualization purposes. The white color denotes the non-visualized bearings. (c) Accumulated energy distribution over 30 min versus bearing and frequency. In (a) and (b), the x-axis is time (HHMMSS) and calculation parameters are shown in Table 1. In (c), the frequency and bearing resolutions are 1 Hz and 1°.

In the text
thumbnail Figure 8

Measurement covering 71 h. (a) Accumulated energy distribution of different frequency versus bearing. (b) Bearing distributions of selected frequencies. B1 and B2 denote two bearing ranges. (c) GPS tracks of the ferries. The azigram calculation parameters are shown in Table 1. In (a), the frequency and bearing resolutions are 1 Hz and 1°. The ferries are Lysoey (MMSI: 257056190), Flatoy (MMSI: 257055990), Faeroes (MMSI: 257055910) and Huftaroy (MMSI: 257055930), and they were built in 2019 according to the registration.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.