Ying-Tsong Lin, James F. Lynch, Timothy F. Duda, Arthur E ...acoustics.whoi.edu/sw06/sw06_austin_workshop/lin_3-D_sound... · Woods Hole Oceanographic Instiution. 3-D oceanographic

3-D sound propagation in a shallow ocean

3-D acoustic effects from shelfbreak fronts and

submarine canyons

Ying-Tsong Lin, James F. Lynch, Timothy F. Duda, Arthur E. Newhall

Woods Hole Oceanographic Instiution

3-D oceanographic featureShelfbreak front

Vertical StructureScanfish sound speed dataGlen Gawarkiewicz, WHOI

Sea Surface Temperature Satellite Image

2o 2o block centered at Hudson CanyonThe Coastal Ocean Observation Lab (COOL) at

Rutgers University, NJ

10:22 Sep. 9, 2006 (GMT)10:32 to 13:57, Sep. 9, 2006 (GMT)

Field observations during the SW06 experiment

• Submarine canyons are commonly seen in the continental shelf and

shelfbreak areas, e.g., Hudson Canyon in the Mid Atlantic Bight.

3-D oceanographic featureSubmarine Canyon

Hunson Canyon (continental shelf portion) Hunson Canyon (shelfbreak portion)

3-D acoustic effects from shelfbreak fronts

(Oceanic whispering gallery effect)

lower sound speed higher sound speed

sloping bottom

Slope angle 1/10º (~1.75/1000 slope)

Water sound speed

inside 1,500m/s, outside 1,530m/s

Normalized Frequency 1,500/ 0

• Continuous wave signal propagation

Source moves from the front to the wedge apex on the same

angle, following the red line shown below.

frontal interface


Idealized model study:

3-D Rigid-Bottom Wedge with a Frontal Interface

pressure-release surface

Penetrable Bottom

Case

Slope angle 1

Water sound speed

1,500m/s, 1,530m/s

Bottom sound speed

1,650m/s with atten.

coef. 0.1db/

Source frequency

200Hz

PE starter angle ~20

Not seeing the

second mode

because source is too

far from the front.

X

YZ


• A Cartesian 3-D PE program by

T. Duda is employed.


3-D Soft-Bottom Wedge with a Wavy Frontal Interface

Y

X


• 4-D ocean fields from a data assimilation model (HOPS), provided by P. F. J.

Lermusiaux, are employed.

Realistic model study: 4-D Ocean Acoustic Field Prediction

• Acoustic normal mode wavenumbes are calculated, and the indices of modal refraction are presented in

the lower-right two panels. The modal phase speeds in red areas are faster, which cause acoustic

modes to refract away and propagate toward lower phase speed areas (blue and white areas).

Half million

points are

calculated in

each time frame

TL CalculationsCase 1: “Cold Water Pool”

Sep 04 18:00

Refractive Index of Mode 1

n1 = k1/k1,ref = Cp1/ C

p1,ref

Attenuation Coefficient of Mode 1

dB/ = 20 log10 e Im(k)

CASE 1: “Cold Water Pool” (Source Frequency 300 Hz)


• An approach of vertical modes and horizontal PE is employed to investigate acoustic modal propagation, currently without considering mode coupling. Kraken and RAM programs are used.


Mode 1 intensities (contours), plotted along with modal refractive index and attenuation coefficient images



• A PE program (RAM) is modified and employed to calculate the modal intensities in

the field. The calculation grid sizes is 6m 1m.






• To identify the 3-D acoustic effects appearing in the modal intensity field, another

solution from only taking into account geometrical and attenuation losses, hence

no horizontal refraction, is obtained and compared to the vertical modes and

horizontal PE solution.

TL CalculationsCase 2: Mesoscale Eddy

Sep 04 18:00


n1 = k1/k1,ref = Cp1/ C

p1,ref



CASE 2: Mesoscale Eddy (Source Frequency 300 Hz)


• Vertical modes and horizontal PE





• Modal intensities predicted by the vertical modes and horizontal PE approach






• Comparison of the vertical modes and horizontal PE solution to the geometrical plus

attenuation losses

TL CalculationsCase 3: SW06 Acoustic Site

Sep 04 18:00


n1 = k1/k1,ref = Cp1/ C

p1,ref




CASE 3: SW06 Acoustic Site (Source Frequency 300 Hz)

• Vertical modes and horizontal PE




• Modal intensities predicted by the vertical modes and horizontal PE approach





• Comparison of the vertical modes and horizontal PE solution to the geometrical plus

attenuation losses



3-D acoustic effects from submarine canyons

modeled domain in this talk


Idealized model study: Gaussian Canyon Model

• The canyon shape across the Y axis is a Gaussian function. A Munk’s deep water sound speed profile is used, and the SOFAR channel axis is at depth 1,100 m.

• Two different source positions are considered. One is on the canyon axis, and another is not.

Munk profile

Frequency = 200 Hz

Source depth 70 m

• Numerical TL calculation: The Cartesian 3-D PE program by T. Duda is

employed.


Idealized model study: Gaussian Canyon Model

Source on the canyon axis Source off the canyon axis

211 211 Y-Z grids (total ~4 millions points) marching in

the X direction every 7.5 m

grid intervals

(dy, dz) = (2.68 m, 2.68 m)


Realistic model study: Realistic Hudson Canyon Model

• NGDC 3 Arc-Second Coastal Relief Model (~70 m resolution) is used to construct the Hudson Canyon topography on the continental shelfbreak.

• Water column property profiles are obtained from P. F. J. Lermusiaux’s HOPS model results in the Hudson Canyon area. SOFAR channel axis is found at depth 720 m on Sep. 4th.

• TL from 4 difference source positions are calculated.

Sound speed profile

from the HOPS ocean

model

1st Source

Position

Frequency = 100 Hz

Source depth 70 m

A Cartesian 3-D PE program

is used, and the angle

coverage of the PE starter is

40 degrees.

211 211 Y-Z grids (total ~4

million points) marching in the

X direction every 15 m

grid intervals

(dy, dz) ~ (9.76 m, 3.25 m)

2nd Source

Position

Frequency = 100 Hz

Source depth 70 m




40 degrees.

3rd Source

Position

Frequency = 100 Hz

Source depth 70 m




40 degrees.

4th Source

Position

Frequency = 100 Hz

Source depth 70 m




40 degrees.

4th Source

PositionCalculation

Domain Rotated

Frequency = 100 Hz

Source depth 70 m

A Cartesian 3-D PE program is

used, and the angle coverage

of the PE starter is 40 degrees.

Summary• 3-D acoustic effects from shelfbreak fronts

– Two idealized cases have been studied using analytic and numerical approaches.

– Realistic ocean model from the HOPS has been employed, and an approach of vertical modes and horizontal PE enables us to investigate normal mode propagation in the field.

• 3-D acoustic effects from submarine canyons– Sound propagation in a idealized Gaussian canyon model is first

analyzed using a Cartesian 3-D PE program.

– Realistic Hudson canyon mode is also studied. Complex sound propagation situations are found.

• Consequences of the 3-D effects– Shelfbreak fronts: noise pool, longer propagation distance

– Submarine canyons: sound ducting, noise in the continental shelf radiating from canyons

Backup Slides

wedge

apex

frontal interface

y

rpressure-release surface

•The frontal interface separates different water sound speeds.

•A cylindrical coordinate (r, , y) is used. The symmetric axis y is lying on the wedge apex. is the clockwise angle to the apex. r is the radial distance.

•Angular modal solution can be found.


3-D Rigid-Bottom Wedge with a Frontal Interface

Effects of a shelfbreak frontal system on 3-D sound propagation

(Oceanic whispering gallery effect)

Source Frequency 200Hz

Receiver/Source Depth 20m

Slope 100m/40km

without foot of front with foot of front

Sound energy is trapped in the water

column and reflects/refracts from the

shelfbreak front.

Effects of a shelfbreak frontal system on 3-D sound propagation

SW06 climatology data model study:

Oceanic Whispering Gallery Effect in the Present of the Foot of Front

• The 3-D propagation program in the Acoustic Toolbox by M. Porter is

employed. (An approach of vertical modes and horizontal Gaussian beams)

4th Source

PositionCalculation

Domain Rotated

Frequency = 100 Hz

Source depth 70 m

Cartesian 3-D PE program


coverage of the PE starter

is 40 degrees.

Documents

Ying-Tsong Lin, James F. Lynch, Timothy F. Duda, Arthur E ...acoustics.whoi.edu/sw06/sw06_austin_workshop/lin_3-D_sound... · Woods Hole Oceanographic Instiution. 3-D oceanographic