I Yun Su – r07525010@ntu.edu.tw
Wen-Yang Liu – r06525035@ntu.edu.tw
Chi-Fang Chen – chifang@ntu.edu.tw
Engineering Science and Ocean Engineering,
National Taiwan University,
No. 1 Roosevelt Road Sec.#4
Taipei City, Taiwan

Li-Chang Chuang – eric@ttsb.gov.tw
Kai-Hong Fang – khfang@ttsb.gov.tw
Taiwan Transportation Safety Board
11th Floor, 200, Section 3,
Beixin Road, Xindian District,
New Taipei City, Taiwan

Popular version of paper 2pUWb2
Presented Tuesday afternoon, December 8, 2020
179th ASA Meeting, Acoustics Virtually Everywhere

A flight recorder is installed in every aircraft to record the flight status. When the aviation accident occurs, this recorder can help clarify the cause of the incident. Furthermore, if the plane crashes into the ocean, the underwater locater beacon (ULB) inside the flight recorder will be triggered which would make a sound that could be located by the rescue team.

In 2009, there was a serious accident involving the Air French flight 447. According to the final report, French Civil Aviation Safety Investigation Authority suggested that the ULB should acquire extended transmission time up to 90 days and increased transmission range. In Taiwan, the flight recorder has already been installed with a 37.5 kHz ULB inside the tail section of every vehicle, and now Taiwan Transportation Safety Board considered to put in an additional 8.8 kHz ULB in the flight belly. (Picture 1)

Picture 1: The positions of the 37.5 kHz and 8.8 kHz ULB on the plane.

The main propose of this study is to understand the performance of the newly bought 8.8 kHz ULB – DUKANE SEACOM DK180. Firsts off, I did the simulation on both ULB to compare the detection ranges (DR), and according to the beacon specifications, the source level (SL) of the both is 160 dB re 1μPa.

For the DR to be simulated, the transmission loss (TL) which is affected by a lot of different environmental parameters must be determined first. This study is based on the Taiwan database, and using the Gaussian beam propagation to calculate the TL. After the TL is acquired, the noise level (NL) which also has certain impact on the DR has to be determined. Generally, the lower the frequency, the longer the DR. DR can be determined by passive sonar equation, and can be derive the FOM = SL – NL – DT. The DT is Detection Threshold and the FOM is Figure of Merit, which is the maximum TL that can detect. The intersection of the TL and FOM is DR. In the study, the DT is set to be zero. At the Point A, the NL of the 8.8 kHz is 78 dB re 1μPa and for 37.5 kHz is 65 dB re 1μPa, so the FOM of the 8.8 kHz is 82 dB re 1μPa and for 37.5 kHz is 95 dB re 1μPa. The DR in 8.8 kHz ULB is about twice than 37.5 kHz ULB at Point A. (Picture 2)

Picture 2: Detection Ranges of 8.8 kHz ULB and 37.5 kHz ULB in the Point A.

In the study, I have also done the experiment in Taiwan Miaoli offshore. The results also show that the newly bought 8.8 kHz ULB would have a smaller TL and longer DR. In summary, with an additional 8.8 ULB, the more precise prediction of the beacon location could be obtained.

Jennifer L. Wladichuk – jennifer.wladichuk@jasco.com
David E. Hannay, Zizheng Li, Alexander O. MacGillivray
JASCO Appl. Sci., 2305 – 4464 Markham St.
Victoria, BC V8Z 7X8, Canada

Sheila Thornton
Sci. Branch
Fisheries and Oceans Canada
Vancouver, BC, Canada

Popular version of paper 2pUWb8
Presented Tuesday afternoon, Nov 6, 2018
176th ASA Meeting, Victoria, BC, Canada

The Vancouver Fraser Port Authority’s Enhancing Cetacean Habitat and Observation (ECHO) program sponsored deployment of two autonomous marine acoustic recorders (AMAR) in Haro Strait (BC), from July to October 2017, to measure sound levels produced by large merchant vessels transiting the strait. Fisheries and Oceans Canada (DFO), a partner in ECHO, supported an additional study using these same recorders to systematically measure underwater noise emissions (0.01–64 kHz) of whale watch boats and other small vessels that operate near Southern Resident Killer Whales (SRKW) summer feeding habitat. During this period, 20 different small vessels were measured operating at a range of speeds (nominally 5 knots, 9 knots, and cruising speed). The measured vessels were catagorized into six different types based primarily on hull shape: ridged-hull inflatable boats (RHIBs), monohulls, catamarans, sail boats, landing craft, and one small boat (9.9 horsepower outboard). Acoustic data were analyzed using JASCO’s PortListen^® software system, which automatically calculates source levels from calibrated hydrophone data and vessel position logs, according to the ANSI S12.64-2009 standard for ship noise measurements. To examine potential behavioural effects on SRKW, vessel noise emissions were analyzed in two frequency bands (0.5–15 kHz and >15 kHz) corresponding to the whales’ communication and echolocation ranges, respectively (Heise et al. 2015). We found that generally, with increased speed, decibel levels increased across the different vessel types, particularly in the echolocation band (Table 1). However, the speed trends were not as strong as those of large merchant vessels. Of the vessels measured, monohulls commonly had the lowest source levels in both SRKW frequency bands, while catamarans had the highest source levels in the communication band and the landing craft had the highest levels in the echolocation band at all speeds (Figure 1). Another key finding was the amount of noise onboard echosounders produced; a significant peak at approximately 50 kHz was present in some vessels, which is within the most sensitive hearing range of SRKW.

Table 1. Average source level for each vessel type in the SRKW communication and echolocation frequency bands for slow, medium, and fast vessel speeds.

Figure 1. Average one-third octave band source levels for each vessel type for the slow speed passes (≤7 kn, ie. whale-watching speed). Due to non-vessel related noise at frequencies below approximately 200 Hz (grey vertical line), levels at those low frequencies cannot be associated with vessel source levels. The peak observed at around 50 kHz is from onboard echosounders.

Literature cited:
Heise, K.A., L. Barret-Lennard, N.R. Chapman, D.T. Dakin, C. Erbe, D. Hannay, N.D. Merchant, J. Pilkington, S. Thornton, et al. 2017. Proposed metrics for the management of underwater noise for southern resident killer whales. Coastal Ocean Report Series. Volume 2, Vancouver, Canada. 30 pp.

Dag Tollefsen – dag.tollefsen@ffi.no
Norwegian Defence Research Establishment,
Horten, NO-3191, Norway

Stan E. Dosso – sdosso@uvic.ca
School of Earth and Ocean Sciences,
Victoria BC, V8W 3P6, Canada

David P. Knobles – dpknobles@kphysics.org
Knobles Scientific and Analysis, Austin, Texas 78755, USA

Popular version of paper 5aUW1
Presented Friday morning, November 9, 2018
176th ASA Meeting, Victoria, BC, Canada

We infer geoacoustic properties of seabed sediment layers via remote sensing using underwater sound from a container ship.  While several techniques are available for seabed characterization via acoustic remote sensing, this is a first demonstration using noise from a large ship-of-opportunity (i.e., a passing ship with no connection to an experiment).  A benefit of this passive acoustics approach is that such sound is readily available in the ocean.

Our data were collected as part of a large coordinated ocean acoustic experiment conducted at the New England Mud Patch in March, 2017 [1].  To investigate properties of a muddy seabed, the experiment employed multiple techniques including direct measurements by geophysical coring [2] and remote sensing using sound from various controlled acoustic sources.  Our team deployed a 480-m long linear array of hydrophones on the seabed to record noise from passing ships.

The array was located two nautical miles from a commercial shipping lane leading to the Port of New York and New Jersey.  An average of three ships passed the array daily.  We identified ship passages by “bathtub” interference patterns (Fig. 1) in the recordings.  The structure seen in Fig. 1 is due to the interaction between ship-to-hydrophone sound paths, some interacting with the seabed, shifting with time as the ship passes the hydrophone.

Ship locations were obtained from Automatic Identification System data provided by the US Coast Guard Navigation Center.  This enabled us to run a numerical model of underwater sound propagation from source to receivers.  The final step is inverse modeling, where a large number of possible seabed models were sampled probabilistically, with each model used to predict the corresponding sound field that was matched with the recorded data.  We used Bayesian sampling and statistical inference methods based on [3].

Inferred geoacoustic profiles (Fig. 2) indicate fine-grained sediment (mud) in the upper seabed and coarse-grained (higher sound speed and density) sediment (sand) in the lower seabed.  Results are in overall good agreement with sediment core data.  This work establishes that noise from large commercial ships can contain usable information for seabed characterization.

[Work supported by the Office of Naval Research, Ocean Acoustics. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of ONR].

Figure 1. Spectrogram of noise measured on a hydrophone on the seabed (left) due to a passing container ship (right).

Figure 2. Inversion results in terms of probability densities for sediment layer interface depths (left), and geoacoustic properties of sound speed (middle), and density (right) as a function of depth into the seabed. Warm colors (e.g., red) indicate high probabilities.

[1]            P.S. Wilson and D.P. Knobles, “Overview of the seabed characterization experiment 2017,” in Proc. 4th Underwater Acoustics Conference and Exhibition, 743-748, ISSN 2408-0195 (2017).

[2]       J.D. Chaytor, M.S. Ballard, B. Buczkowski, J.A. Goff, K. Lee, A. Reed, “Measurements of Geologic Characteristics, Geophysical Properties, and Geoacoustic Response of Sediments from the New England Mud Patch,” submitted to IEEE J. Ocean Eng. (2018).

[3]            S.E. Dosso, J. Dettmer, G. Steininger, and C.W. Holland, “Efficient trans-dimensional Bayesian inversion for seabed geoacoustic profile estimation,” Inverse Problems, 30, 29pp (2014).

Tsuwei Tan – ttan1@nps.edu
Oleg A. Godin – oagodin@nps.edu
Physics Dept., Naval Postgraduate School
1 University Cir.
Monterey CA, 93943, USA

Popular version of paper 5aUW7
Presented Friday morning, November 9, 2018, 10:15-10:30 AM
176th ASA Meeting, Victoria, BC Canada

Introduction
Scientists have long used sound to probe the ocean and its bottom. Breaking waves, roaring earthquakes, speeding supertankers, snapping shrimp, and vocalizing whales make the ocean a very noisy place. Rather than “shouting” above this ambient noise with powerful dedicated sound sources, we are now learning how to measure ocean currents and seafloor properties using the noise itself. In this paper, we combine long recordings of ambient noise with a signal processing skill called time warping to quantify seafloor properties. Time warping changes the signal rate so we can extract individual modes, which carry information about the ocean’s properties.

Experiment & Data
We pulled our data from Michael Brown and colleagues [1].  They recorded ambient noise in the Straits of Florida with several underwater microphones (hydrophones) continuously over six days (see Figure 1). We applied time warping to this data. By measuring (cross-correlating) noise recordings made at points A and B several kilometers apart, one obtains a signal that approximates the signal received at A when a sound source is placed at B. With this approach, a hydrophone becomes a virtual sound source. The sound of the virtual source (the noise cross-correlation function) can be played in Figure 2. There are two nearly symmetric peaks in the cross-correlation function shown in Figure 1 because A also serves as a virtual source of sound at B. Having two virtual sources allowed Oleg Godin and colleagues to measure current velocity in the Straits of Florida [2].

Retrieving Environmental Information
Sound travels faster or slower underwater depending on how soft or hard the seafloor is. We employ time warping to analyze the signal produced by the virtual sound source. Time warping is akin to using a whimsical clock that makes the original signal run at a decreasing pace rather than steadily (Figure 3a  3b). The changing pace is designed to split the complicated signal into simple, predictable components called normal modes (Figure 3c  3d). Travel times from B to A of normal modes at different acoustic frequencies prove to be very sensitive to sound speed and density in the ocean’s bottom layers. Depth-dependence of these geo-acoustic parameters at the experimental site as well as precise distance from B to A can be determined by trying various sets of the parameters and finding the one that best fits the acoustic normal modes revealed by the ambient noise measurements. The method is illustrated in Figure 4. The sound of the virtual source (Figure 2), which emerges from ambient noise, reveals that the ocean bottom at the experimental site is an 11 m-thick layer of sand overlying a much thicker layer of limestone (Figure 5).

Figure 5. Ocean bottom properties retrieved from ambient noise. Blue and red lines show sound speed in water and bottom, respectively, at different depths below the ocean surface. The ratios ρ_s and ρ_b of the bottom density to seawater density are also shown in two bottom layers.  

Conclusion Ambient noise does not have to be an obstacle to acoustic remote sensing of the ocean.  We are learning how to use it to quantify ocean properties. In this research, we used ambient noise to probe the ocean bottom. Time warping has been applied to ambient noise records to successfully measure sound speeds and densities at different depths below the seafloor in the Straits of Florida. Our passive acoustic approach is inexpensive, non-invasive, and environmentally friendly. We are currently working on applying the same approach to the extensive underwater ambient noise recordings obtained at several sites off New Jersey during the Shallow Water 2006 experiment.  

Reference

[1] M. G. Brown, O. A. Godin, N. J. Williams, N. A. Zabotin, L. Zabotina, and G. J. Banker, “Acoustic Green’s function extraction from ambient noise in a coastal ocean environment,” Geophys. Res. Lett. 41, 5555–5562 (2014). [

2] O. A. Godin, M. Brown, N. A.  Zabotin, L. Y. Zabotina, and N. J. Williams, “Passive acoustic measurement of flow velocity in the Straits of Florida.” Geoscience Lett. 1, 16 (2014).

J. Daniel Park (ARL/PSU)
Daniel A. Cook (GTRI)

Lay-language paper for abstract 3aUW8 “Representation trade-offs for the quality assessment of acoustic color signatures”
presented at the 175^th Meeting of Acoustical Society of America in Minneapolis.

We use sound to learn about the underwater environment because sound waves travel much better in water than light waves do. Similar to using a flashlight to find your lost car keys in the woods, sound wave pulses are used to ‘light up’ the sea floor. When carefully organized, sound echoes from the surroundings can be shown as sonar imagery such as Figure 1.

Figure 1. A sonar image is generated from a collection of sound recordings by carefully organizing them into a spatial representation, and we can see various features of the sea floor and even shadows cast by sea floor textures and objects, similar to when using a flash light.

Images are easy for us to understand, but not all of the useful information embedded in the sound recordings is represented well by images. For example, a plastic and a metallic trash bin may have the same cylindrical shape, but the sounds they make when you knock on them are easy to distinguish. This idea leads to a different method of organizing a sound recording, and the resulting representation is called acoustic color, Figure 2. This shows how different frequencies emerge and fade as you ‘knock’ on the object with sound from different directions.

Figure 2. Acoustic color of a solid aluminum cylinder. It shows the strength of frequency components when seen from different viewing angle. Source, University of Washington PONDEX 09/10

This representation has the potential to be useful for distinguishing objects that have similar shapes in visual imagery, but have noticeably different acoustic spectral responses. However, it is not easy to extract relevant information that can help discriminate different objects as seen in Figure 2. One of the reasons is the weak and dispersed nature of the object signatures, which makes it difficult to mentally organize them and draw conclusions. We want to explore other ways of organizing the acoustic data in order to make it intuitive for us to ‘see’ what is yet to be uncovered from the environment. Certain animals, such as dolphins and bats are able to take advantage of complicated acoustic echoes to hunt for prey and understand their environment.

One representation under consideration is time-varying acoustic color, Video 1, which provides the ability to observe the time-evolving characteristics of the acoustic color, with some loss in the ability to precisely distinguish frequencies. This helps one understand how different spectral signatures appear and change, and eventually fade out. This short timescale evolution is important information not easily extractable in the typical acoustic color representation.

<Video 1 missing>

Another representation under consideration is an approach called symbolic time series analysis. By representing a short segment of the raw time series as a symbol and assigning the same symbol for segments that are similar, the time series is transformed into a sequence of symbols as illustrated in Figure 3. This allows us to use tools developed for sequence analysis such as the ones used in DNA sequencing for comparing and recognizing patterns in the sequence. It may prove to be an effective approach to extracting underlying patterns from acoustic data that are not as easily accessible in other more common modes of visualizing the data.

Figure 3. Each time series is transformed into a sequence of symbols, then can be further analyzed to characterize temporal patterns. We can use tools developed for other applications such as DNA sequencing, and extract information that are not as easily accessible from more common modes of visualizing the data.