Basic study on the estimation method of burrows on the seafloor using ultrasound Hajime Tachiki 1 Tokyo University of Science, Graduate School of Science and Technology 2641, Noda, Chiba, Japan Haruki Hirasawa 2 Tokyo University of Science, Graduate School of Science and Technology 2641, Noda, Chiba, Japan Takumi Asakura 3 Tokyo University of Science, Graduate School of Science and Technology 2641, Noda, Chiba, Japan Katsunori Mizuno 4 The University of Tokyo, Graduate School of Frontier Sciences 5-1-5, Kashiwa, Chiba, Japan Koji Seike 5 National Institute of Advanced Industrial Science and Technology 1-1-1 Tsukuba, Ibaraki, Japan The University of Tokyo, Graduate School of Frontier Sciences 5-1-5, Kashiwa, Chiba, Japan

ABSTRACT There have been concerns about the effects of ocean energy facilities on pollution, including chemical, water contamination, and noise. Aquatic environmental measurements have been conducted to assess the effect. However, the effect under the seafloor remains unknown due to the lack of an efficient method for surveying a wide area. The marine environment is affected by the burrows formed by organisms living on the seafloor. Therefore, it is important to understand the morphologies of their burrows from the viewpoint of marine environmental conservation. In addition, the benthic organisms in the burrow can be identified by their burrow because a burrow morphology varies depending on the benthic organisms form it. In this study, we investigated an efficient method for estimating the length and diameter of the burrow using broadband ultrasonic waves. The model experiment and the acoustic simulation using the finite-difference time-domain method (FDTD) were comparatively performed to verify the possibility of estimating the tube length and diameter. It was

1 7522539@ed.tus.ac.jp 2 7521544@ed.tus.ac.jp 3 t_asakura@rs.tus.ac.jp 4 kmizuno@edu.k.u-tokyo.ac.jp 5 seike@ipc.pari.go.jp

Jai. inter noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O ? . GLASGOW

confirmed that the tube length was estimated with a relative error of about 3 to 6% for acrylic and PVC tube models.

1. INTRODUCTION

The environmental pollution caused by marine development may affect some aquatic species such as fish and benthic organisms. For example, noise generated during the construction and operation of some offshore facilities could affect some aquatic species. Furthermore, the rotary movement of tidal wind turbines could degrade temporarily habitat and water quality due to increased turbidity in the water due to disturbances to the seafloor [1]. Various environmental measurements have been conducted to assess the effects of short- and long-term changes in the marine environment caused by such development on marine ecosystems, and much knowledge about the ecosystems in the sea and on the surface of the seafloor [2, 3]. On the other hand, much remains unknown about the ecosystems in seafloor sediments, which require more time and cost to evaluate than those in the sea or on the surface of the seafloor. Some benthic organisms form burrows. These burrows provide shelter for symbiotic organisms and supply oxygen to the seafloor sediments, and they have a considerable impact on the surrounding environment. Techniques to investigate these burrows have been investigated. For example, an evaluation measure of molding the internal morphology of the burrow by pouring resin into the burrow [4] has been investigated. However, there is a need for a costless method suitable for screening surveys to measure many burrows on the seafloor. In this study, we investigate a method to estimate the morphology of the interior space of a burrow by transmitting broadband ultrasonic waves into them. The estimated results of burrow tube length and diameter using both model experiments and acoustic analysis are discussed.

2. ESTIMATION OF TUBE LENGTH

2.1. Experimental setup

The length of the tube is estimated based on the time between the emission of sound waves into the burrow and the observation of echoes, and the speed of sound in the water. This was verified through experiments using tubes of different lengths or with different morphologies. Figure 1 shows a schematic diagram of the model experiments (I-, U-, and Y-shaped tubes). These models were placed in a 30-cm cube water tank and then filled with tap water. After immersing the inside of the model in tap water, an ultrasonic probe (B0.5K10I) made by Japan Probe was placed approximately 1 cm from the model’s tube entrance. This probe radiated a square wave with a center frequency of 500 kHz into the tubes of the model and acquired the echoes. The single probe method was used for the I- and Y-shaped tube models, and the double probe method was used for the U-shaped tube model. I-, Y-tube models and the water tank are made of acrylic, and only the U-shaped tube is made of PVC. Figure 1: Schematic figure of the experimental setup using tube models of (a)I-, (b)U-, (c) Y-shaped tube models, respectively.

Probe

Pulser / Receiver

W

Probe

Radiation

t

Water

Wave

Water

(b)

PC

Probe

Echo

Tube model

(a)

Water

(c)

2.2. Experimental result

The time waveform data for each tube obtained from the model experiments are shown in Fig. 2. Furthermore, the echo arrival times and calculated tube lengths are shown in Table 1. The speed of sound in water is given by Eq. (1)[5].

c = 1492.9 + 3 × (𝑇−10) −6 × 10 ିଷ × (𝑇−10) ଶ −4 × 10 ିଶ × (𝑇−18) ଶ

+ 1.2 × (𝑆−35) −10 ିଶ × (𝑇−18) × (𝑆−35) + 𝐷/61 (1)

where the c is the speed of sound in water and T are the water temperature and S are the salinity of water and D is the depth.

The tube length calculated from echo arrival times and the speed of sound in the water is close to the actual dimensions for all tube geometries. The relative error is approximately 3.2 % for the I- shaped tube, 3.3 % for the U-shaped tube, and 6.4 % for the Y-shaped tube. The more complex the tube geometry, the larger the difference between the actual and estimated values. This is because the sound waves do not propagate along the axis of the tube. Since the echoes in the shorter Y-tube overlapped with multiple reflections from the sloping section immediately after the entrance, only the longer Y-tube is described in Table 1.

(a)

Normalized Voltage

1

0.5

0

-0.5

-1

0 50 100 150 200 250 300

Time [μs]

(b)

Normalized Voltage

0.2

0.1

0

-0.1

-0.2

0 100 200 300 400 500 600

Time [μs]

(c)

Normalized Voltage

0.4

0.2

0

-0.2

-0.4

0 200 400 600 800

Time [μs]

Figure 2: Waveforms of the (a)I-, (b)U-, and (c)Y-shaped tube models, respectively.

Table 1: Comparison of actual and calculated propagation times of reflected waves.

Tube length [mm]

Tube Shape Time [µs]

Water temperature [℃] Estimated

(Actual)

I 155.1 ± 0.7 104.8 ± 0.5

(100) 19.7 ~ 20.3

I (long) 290.7 ± 0.8 205.9 ± 0.6

(200) 20.7 ~ 21.8

U 326.5 ± 0.2 463.5 ± 0.3

(447) 19.9 ~ 20.0

Y (long) 301.7 ± 0.5 214.0 ± 0.4

(228) 18.6 ~ 21.9

3. ESTIMATION OF TUBE DIAMETER

3.1. Experimental and numerical setups

To estimate the diameter of the tube, the following acoustic mechanism was considered. The intervals between each of these multiple reflections should widen as the diameter of the tube increases. To verify this, an I-shaped model made of oil clay was placed in a water tank with 10-cm side and 20-cm height, immersed in water, and then a square wave with a center frequency of 500 kHz was radiated from the probe into the tube. The probe was placed approximately 1 cm from the model’s tube entrance as before. In addition, to measure differences in the time patterns of the echoes due to changes in tube diameter, the diameter of a 10 cm long I-tube was varied to 26 mm, 32 mm, and 38 mm, respectively. Furthermore, a two-dimensional acoustic simulation using FDTD was performed to visualize the propagation of sound waves in the tube. Figure 3 shows the numerical results of the time-series contour maps obtained by FDTD. The dashed lines in Fig. 3 indicate the acrylic model. A radiated sound wave travels in the I-shaped tube while forming multiple reflections can be seen.

3.2. Experimental and numerical results

Time-frequency analysis by the wavelet transform with the Gabor function as the mother wavelet was used to obtain the frequency characteristics of time waveforms of the echo time pattern acquired from the model experiment and the acoustic simulation by FDTD. And the results were compared for each tube diameter. Figures 4 and 5 show the spectrogram of echoes obtained from the simulation and the model experiment, respectively. The interval of multiple reflections after 160 µs (black arrow in Fig. 4) becomes wider as the tube diameter increases. In Fig. 5, the larger the tube diameter, the faster the relative energy level in the 500 kHz band of multiple reflections decay can be seen. However, there is no change in the interval of multiple reflections as seen in Fig. 4. The reason for this difference may be due to the difference in the source characteristics between the spherical sound source of the simulation and the plane-wave sound source of the experiment. So, the interval of experimentally-obtained multiple reflections does not differ as much as in the simulation. Here, the peaks of multiple reflections in the simulation were extracted and the interval of multiple reflections was linearly approximated. The results are shown in Figure 6 and Table 2. Figure 7 also shows the relationship between the coefficients of the regression line in Table 2 and the tube diameter. Eq. (2) is the regression line in Figure 7. These suggest that spherical waves are effective in estimating the tube diameter. On the other hand, Plane waves have less distance attenuation than spherical waves. This makes plane waves more suitable for measuring the length of a tube. That indicates the directive characteristics of the sound source considerably affect the estimated results, and the optimal selection

of the sound source characteristics not only in numerical analysis but also in actual measurement is future work.

d = 7.98 × 10 ଺ 𝑎+ 16.7 (2)

where the d is the estimated tube diameter and a is the coefficients of the regression line.

16.0

40.0

76.0

20

40

, wlll | Ae! —_— A

60

y

80

100

120

40

50

10

20

30

40

50

10

20

30

40

50

10

20

30

x

Unit : mm

Figure 3: Time-series contour maps obtained from the acoustic simulation of FDTD on the I- shaped tube model.

(a)

(b)

(c)

Figure 4: Spectrogram of the echo time pattern from the acoustic simulation of FDTD in the I- shaped tube model with diameters of (a)26 mm, (b)32 mm, (c)38 mm, respectively.

Figure 5: Spectrogram of the echo time pattern measured in the I-shaped tube model with diameters of (a)26 mm, (b)32 mm, (c)38 mm, respectively.

2.E-05

Interval of multiple reflection [µs]

2.E-05

2.E-05

1.E-05

1.E-05

9.E-06

26mm

7.E-06

32mm

38mm

5.E-06

0 1 2 3 4 5 6 7

Cycle

Figure 6: Transition of multiple reflection intervals for each tube diameter.

Table 2: Comparison of regression lines for the interval of multiple reflections.

Tube Diameter [mm] Regression line [µs] 𝑹 𝟐

26 𝑦= 1.19𝑥+ 0.578 0.956 32 𝑦= 1.88𝑥+ 0.587 0.991 38 𝑦= 2.69𝑥+ 0.730 0.980

40

38

36

Tube diameter [mm]

34

32

30

28

26

24

1.E-06 2.E-06 2.E-06 3.E-06 3.E-06

Coefficients of the regression line [µs]

Figure 7: Relationship between tube diameter and coefficient of the regression line. 4. CONCLUSIONS

We conducted a basic investigation of a method to measure a burrow's length and diameter using broadband ultrasonic waves and verified whether it is possible to estimate the length and diameter of a tube from echoes in tubes with various morphologies. It has been confirmed that the lengths of I-, U-, and Y-shaped acrylic and PVC tubes can be experimentally estimated with an error of less than 7% at most. The simulated results have suggested that using a spherical wave as a source of sound effectively estimates tube diameter. Further measurements under the conditions closer to more practical cases will be conducted as future works.

6. REFERENCES

1. Gasparatos, A., Doll, N. H. C., Esteban, M., Ahmed, A., & Olang, A. T., Renewable energy and

biodiversity: Implications for transitioning to a Green Economy, Renewable and Sustainable Energy Reviews, 70 , 161–184 (2017) 2. Katsunori, M., Kei T., Seiichiro, H., Shigeru, T., Shingo, S., Toshihiro, O., Kenichi, S. &

Hironobu, F., An efficient coral survey method based on a large-scale 3-D structure model obtained by Speedy Sea Scanner and U-Net segmentation. Scientific Reports, 10 , 12416 (2020). 3. Hiroyuki, Y., Survey, monitoring and assessment of marine communities conducted by new

technology. BUTSURI-TANSA(Geophysical Exploration) , 73 , 53–63 (2020) 4. Koji, S., Robert G. J., Hiromi W., Hidetaka, N., & Kei, S., Novel use of burrow casting as a

research tool in deep-sea ecology. Biology Letters, 8 , 648–651 (2012). 5. Robert, J. U. , Principles of Underwater Sound. In Japanese: Suichu Onkyou no Genri, edited by

Akira T., Kyoritsu Shuppan, 1978.