Underwater Acoustics including - TAU111)r.pdfHistorical Underwater Acoustics Range ~1000 miles polar...

Preview:

Citation preview

Underwater Acoustics including Signal and Array Processing

��  �"3*$��$&".��$/534*$3�����  �"3*$��$&".��$/534*$3�����  �2/0"("4*/.� )&/29�".%��/%&,*.(�

•  !"6&��/%&,3�•  !"6&.5-#&2��.4&(2"4*/.�•  �/2-",��/%&3�•  �"2"#/,*$��15"4*/.�•  �8"-0,&3�

•  �"93��  �/*3&�����.4&2'&2&.$&���  �&.3*4*6*49��&2.&,���  �*(.",��22"9�02/$&33*.(� �  *-&��&6&23",��$/534*$3���  �/*3&������2/#*.(��*(.",��

HIERARCHY(OF(UNDERWATER(ACOUSTIC(MODELS((

Coupled(Modes(

(KRAKEN,CSNAP,COUPLE)(

Coupled(Wavenumber?(

IntegraCon(

(RD?OASES)(

Gridding(Aspects(of(RangeAdependent(PropagaCon(Modeling(

Parabolic(EquaCon(

(RAM)(

Finite(Differences(

Finite(Elements(

Wavenumber(IntegraCon(

Wavenumber(IntegraCon(

�&+&2*3�!"6&(5*%&�!"6&.5-#&2��/,54*/.�

�&+&2*3�!"6&(5*%&�

Growing(ExponenCals(

�&+&2*3�!"6&(5*%&��5-&2*$",,9��4"#,&��/2-�

Decaying(ExponenCals(

z,w!

r,u!

Lame�(constants(

4(Unknowns(

Homogeneous(ElasCc(Layers(Wave(EquaCons(

1((2((3((2((4(

Number(of(Boundary(CondiCons(

Homogeneous(ElasCc(Layers(Boundary(CondiCons(

Modal(EquaCon(

Boundary(CondiCons(Ideal(Waveguide(

Away(from(Source(

0(

Normal(Modes(MathemaCcal(DerivaCon(

F(r)( G(z)(

Classical(SturmALiouville(Eigenvalue(Problem(• (Modal(equaCon(has(infinite(set(of(soluCons(–(modes(of(vibraCng(string(• (Modes(characterized(by((

• (Mode(shape(Ψ (z) (eigenfuncCon) (• (PropagaCon(constant.(k(((.(k((((real((eigenvalue)(• (mAth(mode(has(m(zeros(in([0,D](• (k < ω / c

•  Modes(are(Orthogonal(• (Modes(form(a(Complete(Set(

2(rm rm

rm min

m

From(Mode(EquaCon(

Range(SoluCon(

Modal(Field(SoluCon(

X(

Not(the(Usual(SturmALiouville(Problem!!(

General(StraCfied(Waveguides(

Field(in(boaom(

Field(in(water(

~(

Virtual(Modes(

Normal(Modes(

Normal(and(Virtual(Modes(

Virtual(Modes(Normal(Modes(

Normal(and(Virtual(Modes(

�&&0�!"4&2��2/0"("4*/.�

�"9��/%&��.",/(9�

Modal(EquaCon(

WKB(ApproximaCon(

WKB(ray(theory(ignores(evanescent(field(beyond((turning(points(

Deep(Ocean(PropagaCon(

Finite(Difference(FormulaCon(

Numerical(Approaches((

Modal(EquaCon(

Finite(Difference(FormulaCon((

p!w!=(

VerCcal(Admiaance(

Normalized(Hankel(FuncCons(

Forward(

Backward(

Forward(

Backward(

AsymptoCc(

r(jA1( j(r( r(j+1(

Sector(j(

Normal(Modes(RangeAdependent(Environments(

�Direct(Global(Matrix�(NormalizaCon(

3AD(Modal(Modeling(Framework(

RangeAIndependent(Sectors(3AD(Ocean(Environment(

Full 3-D Mode Coupling Strong Discontinuities 1.  Pre-compute modes for all sectors 2.  Each source-receiver combination

•  Horizontal ray tracing, all mode combinations

•  Local single-scattering approximation in plane geometry

•  Approximate accounting for geometric spreading r

COMPUTATIONALLY INTENSIVE

A1/2(

2.5AD(Modal(Modeling(Framework(

RangeAIndependent(Sectors(3AD(Ocean(Environment(

In-Plane Mode Coupling Gradual Range-Dependence 1.  Pre-compute modes for all sectors 2.  Each source-receiver combination

•  In-plane mode propagation between sector boundaries

•  Local single-scattering – No horizontal diffraction

•  Approximate accounting for geometric spreading r

COMPUTATIONALLY EFFICIENT

A1/2(

AdiabaCc(ApproximaCon(

Gulf(Stream(Environment(

WarmACore(Eddy(PropagaCon(RangeAIndependent(

Coupled(Modes(

AdiabaCc(Mode(Theory(

RangeAindependent(cylindrically(symmetric(

RangeAsoluCon(

Slowly(varying(depth(soluCon((envelope)(

Use(Bessel(EquaCon(

Parabolic(EquaCons(

Slowly(varying(envelope:(

NarrowAangle(approximaCon,(valid(for(grazing(angles(less(than(10A15(deg.(

=(0(for(n=n(z),(rangeAindependent(~(0(for(n(r,z)(slowly(varying(in(r((

SoluCon(technique:(Approximate(PseudoAdifferenCal(Operator(Q(

Ignores(backscaaering(

Outgoing( Incoming(

Generalized(DerivaCon(

k

θ#

k r

k z

θ#

0(

Q(=(cos(θ (relates(to(source(angle,(which(–(if(small(–(jusCfies(Taylor(expansion(

0(

RangeAIndependent(Environment(

Local(Plane(Wave(SoluCon(

n = 1

Standard(PE(

Minimizes(phase(errors(0A40(deg(

Standard(and(Wide(Angle(Parabolic(EquaCons(

k 0

θ#

k rm

m

Phase(Errors(and(Angular(LimitaCons(

Exact(Modal(Phase(

Clairbout(PE(Modal(Phase(

PE(Modal(Phases(

“RAM(PE”(

(

Michael(D.(Collins(

Ray methods

•  Long history with contributions by Euclid, Ptolemy, Snell (1626), Fermat (1661), Gauss (1846)

•  Provides an approximate solution. Ray theory is to wave propagation as classical mechanics is to quantum mechanics

•  Provides the language for describing in intuitive terms what occurs

Characteristics of ray models

•  Probably still the most widely used method in operational use

•  Attractive for – High-frequency – Broadband (e.g. tomography, acoustic

communication, active sonar) – Range-dependent problems –  reverberation

The bad news … A robust model is rather difficult to produce

•  Eigenrays are roots of a nonlinear equation •  Ray paths and transmission loss are extremely

sensitive to volume and boundary interpolation •  Tracing must be restarted after each boundary

interaction •  Caustics must be detected even if full caustic

corrections are not used As a result, most ray models do not come close to

the �ray theoretic� result

( )( ) ( ) z-z

,1 s

22

rrrp

zrcp s

ttδδ −

−=+∇

( )( )sxx

x−−=+∇ δω p

cp 2

22

�&,-)/,4:��15"4*/.��

�/6&2.*.(��15"4*/.<�4)&�7"6&�&15"4*/.��

( ) ( ) ( )( )

0

0

/ ck

iA

epj

jji

ω

ωωτ

=

= ∑∞

=

xx x

where,(

)&��"9��.3"4:��&&+�"�3/,54*/.�/'�4)&��&,-)/,4:�&15"4*/.�/'�4)&�'/,,/7*.(�'/2-�

; �",,&%�4)&�����������; �&.&2",,9����������; �2/6*%&3�".���� �����&80".3*/.�

( ) ( )

( )[ ]( ) ( ) ( ) !

"#

$%&

+++−=

()

*+,

-+=

∑∑∑

∑ ∑

=

=

=

=

=

0

,

0

,

0

22

0 0

,

2j

jxxj

jjxj

xj

jj

xxxi

xx

j jjxj

jj

xi

x

iA

iA

iiA

iep

iA

iA

iep

ωωωτ

ωωττω

ωωωτ

ωτ

ωτ

DifferenCate(the(ray(series…(

and(

[ ]( )

( )

( ) !!!!

"

!!!!

#

$

!!!!

%

!!!!

&

'

+∇

⋅∇+

∇+∇−

=∇

=

=

=

0

2

0

0

222

2 2

jjj

jjj

jj

j

i

iA

iA

i

iA

i

ep

ω

ωτω

ωτωτω

ωτ

SubsCtute(back(in(the(Helmholtz(equaCon…(

Equate(terms(of(like(order(in((

( ) ( )

( ) ( )( ) ( ) ,...2,12:

Transport 02:

Eikonal:

,1221

02

0

222

=−∇=∇+∇⋅∇

=∇+∇⋅∇

=∇

−−

jAAAo

AA

c

jjjj ττω

ττωο

τωο x

Rays and wavefronts

Ray coordinates

τ∇= cdsdx

Define(rays(as(curves(perpendicular(to(the(wavefronts(of((

But,(phase(is(sCll(unknown.(

Lots(of(work…(

Rays(are(now(defined(in(terms(of(

ccds

dcds

d∇−=#

$

%&'

(211 x

�/,6*.(�4)&��*+/.",���&4)/%�/'��)"2"$4&2*34*$3��

( )xτ

( )xc

⋅=)(

1ds

sc

(eikonal(equaCon(in(ray(coordinates).((Original(nonlinear(PDE(is(now((linear(ODE.(

SoluCon:(

Phase(of(the(wave(is(delayed(by(its(travel(Cme.(

Eikonal(equaCon:(

�/,6*.(�4)&��*+/.",�

211 cds

dxc

=⋅∇τ

21c

=∇⋅∇ ττ

( ) ( )( )∫+=

sds

scs

0'.

'10ττ

Therefore,(

or,(

( )

( )

( )[ ( ) ]

( )

( )0sin

, z

, 0

cos ,

,

1 ,

,1

,

s

2

2

cz

crr

szsr

dzdc

cdsdsc

dsdz

drdc

cdsdsc

dsdr

s

θζ

θξ

ζζ

ξξ

==

==

−==

−==

is(the(trajectory(of(the(ray.(

IniCal(condiCons(

�"9��15"4*/.3�*.��9,*.%2*$",��//2%*."4&3�

Building a Matlab ray code We are solving an IVP (initial value problem) of the form

Where s is arclength Step 1: define a subroutine that calculates the sound speed and its

gradient Step 2: Solve the ray equations for a fan of take-off angles (initial

conditions) (e.g. ODE45)

][]//)()([),(

),(

22

ζρ

ξρ

zrccccscscsf

sfdsd

zr

=

−−=

=

xx

xx

Sound speed profile and ray trace in the Balearic Sea

Ray artifacts in shadow zones

Ray artifacts (caustics)

Effect of profile

interpolation

�*&$&7*3&�,*.&"2�

�5#*$�30,*.&�

Review(of(the(2AD(Model(((((((Cerveny,((Popov,(Psencik…)(

• Trace(a(fan(of(the(beams(from(source((“space(filling”)(

• At(any(given(receiver(locaCon(AA>Eliminates(the(need(for(source(to(receiver(eigenrays(

Sum(up(the(influences(of(all(the(beams(

Beam(has(a(Gaussian(crossAsecCon.(

Gaussian beam tracing

Bellhop •  “Beam(Tracing”(model(for(predicCng(acousCc(pressure(field(in(the(ocean(

environment.(•  Beam(structurOCEAN(ACOUSTICS(LIBRARY:(hTp://oalib.hlsresearch.com(•  es(include(Gaussian(and(hatAshaped(beams(with(both(geometric(and(

physicsAbased(spreading.(•  Can(produce(a(variety(of(outputs(including(basic(raytracing(plots,(

transmission(loss,(eigenrays,(arrivals,(and(received(Cme(series.((•  Allows(for(range(dependence(in(top(and(boaom(boundaries(as(well(as(

sound(speed(profile((with(userAdefined(environmental(files).(•  Top(and(boaom(reflecCon(coefficients(can(be(provided.(•  Bellhop(is(implemented(in(Fortran,(Matlab,(and(Python(and(works(on(Mac,(

Windows,(and(Linux.((

•  hap://hlsresearch.com/personnel/porter/papers/JASA/JASA%20gbt%20bw%20with%20errata.pdf(

(

3D sound speeds and bathymetry

3D(bathymetry(causes(horizontal(refracCon(

Horizontal(waveguide(formed(by(two((syntheCc)(soliCons(limits(spreading(and(provides(beaer(detecCon(opportuniCes(

Focus(range(varies(with(elevaCon(angle(

Horizontal(fan(of(beams(at(0(degrees(

Horizontal(fan(of(beams(at(2.5(degrees(

Paul Hursky, Martin Siderius and Michael B. Porter (HLS)

Ocean(AcousCc(Models(Summary(

Ray(Tracing((

IntuiCve,(computaConally(efficient(RI(and(RD(environments(

HighAfrequency(approximaCon(Does(not(incorporate(diffracCon(Does(not(handle(causCcs(properly(

Wavenumber(IntegraCon((

Exact,(efficient(soluCon(for(RI(waveguides(Full(seismoAacousCcs,(including(poroAelasCcity(Coupled,(twoAway(propagaCon(for(stepwise(RD(

ComputaConally(intensive(for(RD(

Normal(Modes((

Highly(efficient(for(RI(environments(Efficient(coupled(and(adiabaCc(modes(for(stepwise(RD(

Backscaaer(by(twoAway(coupled(modes(HighAfrequency(approximaCon(

PreAcomputaCon(of(modes(and(coupling(coefficients(for(general(3D(environments((

Parabolic(EquaCons((

Highly(efficient,(Inherently(rangeAdependent(ConCnuous,(gradual(rangeAdependence(AdAhoc(approximaCons(for(backAscaaer(

No(preAcomputaCon(gain(for(general(3D(environments(and(sourceAreceiver(configuraCons.((

Underwater Acoustics including Signal and Array Processing

��  �"3*$��$&".��$/534*$3�����  �"3*$��$&".��$/534*$3�����  �2/0"("4*/.� )&/29�".%��/%&,*.(��  �/*3&�����.4&2'&2&.$&�

•  �/*3&��.�4)&��$&".�•  �/%&,*.(��$&".��!"6&(5*%&���/*3&�

•  �0&$42",��/%&��/%&,�•  ����/*3&��/%&,�•  �"9��

��  �*(.",��22"9�02/$&33*.(���  �&.3*4*6*49��&2.&,� �  *-&��&6&23",��$/534*$3���  �/*3&������2/#*.(��*(.",��

MOTIVATION:ONE(MAN�S(NOISE(IS(ANOTHER(MAN�S(SIGNAL(

• WHITE(NOISE:(�UNCORRELATED�(sensor(to(sensor(ANONAPROPAGATING(• ISOTROPIC(NOISE:PROPAGATING(IN(ALL(DIRECTIONS(WITH(EQUAL(AMPLITUDE(• CONFUSED(WITH(WHITE(NOISE(BECAUSE(WHEN(SENSED(AT(HALF(WAVE(LENGTHAAsame(a(white(noise(• SEA(SURFACE(NOISE:(PROPAGATING(PARTIALLY(CORRELATEDAHAS(DIRECTIONALITY:((• SHIPPING(NOISE:(PROPAGATING(WHEN(FAR(AWAYAAPPEARS(TO(HAVE(SAME(PROPERTIES(AS(SURFACE(NOISE(• BIOLOGICAL(NOISE(• SEISMIC(NOISE(• THERMAL(NOISE(

___________________________________________________________(• USUALLY(TRY(TO(DETECT(SIGNAL(IN(NOISE(• USE(DIRECTIONALITY/CORRELATION(OF(NOISE(FOR(INVERSION(• ACOUSTIC(DAYLIGHT,INVERSION(• THIS(TALK((WILL(END(WITH(UTILIZING(NOISE:(((((((((((((((((((((((((((((((((((((((((TREATING(NOISE(AS(THE(SIGNAL(((

58

HISTORY (2)

FOG ALERT SYSTEM: 100 HZ TRAVELS FURTHER --BUT 300-400 HZ MORE EASILY HEARD BECAUSE OF “ENVIRONMENTAL NOISE” : FIRST USE OF SIGNAL TO NOISE TERMINOLOGY LASKY, JASA 1977.

USEFUL(Noise(+(TR:Track(of(storm(from(microseisms((0.2(Hz)(

Gerstoft et al., 2008; Zhang et al 2009, see also Wilson,Makris, JASA 2006 (UWA frequencies)

Webb, Rev Geo. 1998

Another(Example(of(Using(SHIP(NOISE(AA(FOR(INVERSION(

Battle, JASA 2004

Range ~1000 miles

polar latitudes

Mid latitudes

array

Typical mid-latitude sound speed profile

Typical northern

sound speed profile

Radiated noise

Sea mountain or continental

shelf

Ray trapped in the Deep Sound Channel

(DSC)

Depth ~10000 ft

Layers of constant sound speed

C (m/s)!

Historical Underwater Acoustics

Range ~1000 miles

polar latitudes

Mid latitudes

array

Typical mid-latitude sound speed profile

Typical northern

sound speed profile

Radiated signal

Sea mountain or continental

shelf

Ray trapped in the Deep Sound Channel

(DSC)

Depth ~10000 ft

Layers of constant sound speed

C (m/s)!

Historical Underwater Acoustics

Noise at Array

AMBIENT NOISE SPECTRA (WENZ)

AMBIENT NOISE SPECTRA (WENZ)

Thermal Noise “FLOOR”

BIOLOGICS

66

NOISE LEVELS AND

SOURCE LEVELS

BRADLEY,STERN NRC 2008

Ships Underway Broadband Source Level (dB re 1 Pa at 1 m)

Tug and Barge (18 km/hour) 171 Supply Ship (example: Kigoriak) 181

Large Tanker 186 Icebreaking 193

Seismic Survey Broadband Source Level (dB re 1 Pa at 1 m )

Air gun array (32 guns) 259 (peak) Military Sonars Broadband Source Level

(dB re 1 Pa at 1 m ) AN/SQS-53C

(U. S. Navy tactical mid-frequency sonar, center frequencies 2.6 and 3.3 kHz)

235

AN/SQS-56 (U. S. Navy tactical mid-frequency sonar, center

frequencies 6.8 to 8.2 kHz)

223

SURTASS-LFA (100-500 Hz) 215 dB per projector, with up to 18 projectors in a vertical array operating

simultaneously Ocean Acoustic Studies Broadband Source Level

(dB re 1 Pa at 1 m ) Heard Island Feasibility Test (HIFT)

(Center frequency 57 Hz 206 dB for a single projector, with up to 5

projectors in a vertical array operating simultaneously

Acoustic Thermometry of Ocean Climate (ATOC)/North Pacific Acoustic Laboratory

(NPAL) (Center frequency 75 Hz)

195

Source Broadband Source Level (dB re 1 Pa at 1 m )

Sperm Whale Clicks 163-223 Beluga Whale Echolocation Click 206-225 (peak-to-peak) White-beaked Dolphin Echolocation Clicks 194-219 (peak-to-peak) Spinner Dolphin Pulse Bursts 108-115 Bottlenose Dolphin Whistles 125-173 Fin Whale Moans 155-186 Blue Whate Moans 155-188 Gray Whale Moans 142-185 Bowhead Whale Tonals, Moans and Song 128-189 Humpback Whale Song 144-174 Humpback Whale Fluke and Flipper Slap 183-192 Southern Right Whale Pulsive Call 172-187 Snapping Shrimp 183-189 (peak-to peak)

Man Made Sounds

Animal Sounds

68

NOISE LEVELS +PASSING SHIP

BRADLEY,STERN, NRC 2008

Shipping and Wind Noise

Carey,Evans,CTCA 2010

TYPICAL SONAR VIEW OF NOISE: NUISANCE

GOAL OF ARRAYS OR ANTENNAS: 1.  ADD UP MORE SIGNAL THAN NOISE 2.  LOOK IN A CERTAIN DIRECTION

1.  TOWARD A SIGNAL OF INTEREST 2.  LOW SIDELOBES

3.  ADAPTIVE PROCESSING: USE DATA FOR HIGH RESOLUTION AND MINIMUM

SIDELOBE

ARRAY(GAIN:(Signal(adds(up(faster(than(noise(

Isotropic(noise( is(uncorrelated(at(( λ / 2

AG(=(10(log(m(

Incoherent:(no(correlaCon((between(sensors:(no((XAterms(in(sumA>m(terms(Vs(m2(for(coherent(

GEOMETRY(FOR(SURFACE(DISTRIBUTED(NOISE(MODEL(

TheoreCcal(approach(((discrete,(modal(part)(

B(

A(

t(t(

surface(noise(

The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.

R(

at(frequency(ω,#

( ) ( ) ( )( )RkHRkH21RkJ n0n0n0 −+=with#

Z�(

Z1( Z2(

W.A.(Kuperman(&(F.(Ingenito,(JASA,(1980(

NOISE STRUCTURE IN SHALLOW WATER (mode or spectral model)

(((

GEOMETRY FOR PE NOISE MODEL (March toward array adding random surface sources)

A PE NOISE MODEL RUN

80 C. G. Anderson 1972 NOSC Tech Note 800

Deep Water Noise Notch (Atlantic)

up

81

DOWNSLOPE CONVERSION

WAGSTAFF, JASA 1981

82

Ambient Noise Notch: Model/Data Comparison

-12

-8

-4

0

-30 -20 -10 0 10 20 30

data model

look direction (deg)

-12

-8

-4

0

outp

ut (d

B r

e: m

ax) -12

-8

-4

02 kHz

3 kHz

4 kHz

Model/Data Comparison

Frequency (Hz)

Beam

form

er

look d

irection (

deg)

noise

notch

Result: Good model/data agreement for both width and depth of noise notch.

Compare ECS data and model for

beamformer output at 2, 3, and 4 kHz.

Data

Applied Physics Laboratory • University of Washington

SHALLOW WATER NOISE NOTCH- INT WAVES

ROUSEFF, TANG JASA 2006

QUESTIONS?(

Underwater Acoustics including Signal and Array Processing

��  �"3*$��$&".��$/534*$3�����  �"3*$��$&".��$/534*$3�����  �2/0"("4*/.� )&/29�".%��/%&,*.(��  �/*3&�����.4&2'&2&.$&���  �&.3*4*6*49��&2.&,���  �*(.",��22"9�02/$&33*.(� �  *-&��&6&23",��$/534*$3���  �/*3&������2/#*.(��*(.",��

WHAT(PART(OF(THE(WATER(COLUMN(MOST(AFFECT(THE(PROPAGATION?(

•  FAF(experiments(–  Data(

(•  Environmental(perturbaCons(

–  RelaCng(∆p(to(∆c((

•  AcousCc(SimulaCons(–  SensiCvity(maps(

(•  Inversion(esCmates(

–  Simulated(((

•  Future(work(

Focused(AcousCc(Fields((FAF)(

•  Pianosa,(near(Elba(Island((Italy)(

•  NATO(Undersea(Research(Center((NURC)(

–  Formerly,(SACLANT(URC(

•  NRV(Alliance(

Map(from(hap://commons.wikimedia.org/wiki/File:Tuscan_archipelago.png(

29(x(32(Sets(of(Greens(funcCons,(recorded(every(20s(Each(represents(the(arrival(structure(due(to(mulCpath(propagaCon(

R = 4 km

SA(29(Elements(~3.5(kHz(

RA(32(Elements(

D =

100

m

Experimental(setup(

RA(SA(

SA element 4

SA element 15

5.75( 5.76( 5.77( 5.78( 5.79( 5.8(

30(

40(

50(

60(

70(

80(

90( -30(

-25(

-20(

-15(

-10(

-5(

0(

5.75( 5.76( 5.77( 5.78( 5.79( 5.8(

30(

40(

50(

60(

70(

80(

90( -30(

-25(

-20(

-15(

-10(

-5(

0(

Time (s)

Dep

th (m

) D

epth

(m)

Signals received on RA

Source(signal:(3(kHz(A(4(kHz(LFM(chirp(

Data(example(

Data(A(single(receiver(

23

( )( | ) 2 ( | ) ( | ) ( )( )r s r sV

c rG r r G r r G r r dV rc r

ω"Δ

" " "Δ = − ⋅ ⋅"∫∫

Born(approximaCon(to(a(perturbaCon(

sr

rr

r!

r!

r!

Temporal(sensiCvity(of(received(pressure(

23

( | ; ) ( )1 2 ( | ) ( | )( ) 2 ( )

i tr s sr s

p r r t PG r r G r r e dc r c r

ωωω ω

π

∞−

−∞

∂& &= ⋅ −

& &∂ ∫

r!

sr

rr

r!

SimulaCons(

•  PE acoustic propagation model

•  Pekeris waveguide –  100m deep x 1km wide

•  Broadband source signal –  3kHz to 4kHz

SimulaCons(A(pressure(record(

1(2(

3(4(

5(

SimulaCons(A(kernels(1(

3(2(

∂ P(t)∂c(r' )

=Pr (t) ∂P( t)

∂c(r ')( )r +Pi(t) ∂P( t)∂c(r ')( )i

P(t)

∂θ(t)∂c(r' )

=Pr (t) ∂P (t )

∂c(r ')( )i −Pi (t)∂P (t )∂c(r ')( )r

P(t) 2

Phase(and(amplitude(kernels(

•  Amplitude/envelope sensitivity

•  Phase sensitivity €

dA per ms−1( )

dθ(radians) per ms−1( )

1(

SpaCal(spectrum(of(sensiCvity(

3(

SpaCal(spectrum(of(sensiCvity(

EvoluCon(of(kernels(with(Cme(

Arrays(of(kernels(

1(

Arrays(of(kernels(

2(

Arrays(of(kernels(

3(

Arrays(of(kernels(

4(

Inversion(for(SoundASpeed(

1 2

1 2 1

1 2 2

1 2

1 2

1 1 11

2 2 22

3 3 33

4 4 4 4

r r R

r r R r

r r R r

r r R

RN N N N

r r R

h h hph h h cph h h cp

p h h hc

p h h h

! ! !

! ! ! !

! ! ! !

! ! !

!

! ! !

" #" #Δ % &% & Δ" #% &Δ% & % &% &% & ΔΔ % &% &=% & % &% &Δ% & % &% &% & Δ% &% & ' (% & % &Δ% &' ( % &' (

RA(SA(

ˆ Inv( )

P H C

C H P

Δ = Δ

⇒ Δ = Δ

( | ; )( )r s

rp r r thc r!

∂=

!∂

Δˆ p (t) = ∂p( t)∂c(r ') Δc(r' )dV

V∫

RealisCc(example(

•  Mean profiles from FAF�05 CTD measurements –  Temperature –  Salinity –  Sound-speed

Internal(waves(

•  Provides(realisCc(perturbaCons(

Internal(waves(

•  Helps(constrain(inversion(problem(

Inversion(results(

•  Input(perturbaCon(

•  Estimated perturbation

LAB(MEASUREMENT(((((((Direct(Path(((((((Surface(Reflected(Path(

RA(SA(

Range ~ 4 km (a) (b)

~120(m(

Mean sound speed (m/s)

Dep

th (m

)

1500( 1520( 1540(

0(

20(

40(

60(

80(

100(

SA(

RA(

0( 1( 2( 3(Rms sound speed (m/s)

Eigenray/SensiCvity(Kernel(“Micro?Surgery”(at(Sea( Roux(et(al,(FluctuaCons…,(JASA(2013(

Refer/further/to/Time/Reversal//analysis/in/subsequent/lecture/

QUESTIONS(

Wave(Models(

•  Wavenumber(IntegraCon(

•  Normal(Modes(

•  Parabolic(EquaCon(•  Shallow(water(examples(

(

Gradient(Fluid(Layers(

Airy(FuncCons(

Ai Bi

n2(linear(profile(

ElasCc(Ice(Cover(

Fluid(Water(Column(

Fluid(((Sediment(Layer(

ElasCc(Sediment(Layer(

ElasCc(Halfspace(

Boundary(CondiCons(

Wavefield(Unknowns(

Vacuum(0(

4(

2(

2(

4(

2(

2(3(

2(

3(

4(

14(unknowns( 14(equaCons(

Global(EquaCons(and(Unknowns(

Outgoing(waves(only(

Wavenumber(IntegraCon(

• Slow(evanescent(decay(if(receiver(depth(close(to(source(depth(• Homogeneous(soluCon(decay(slowly(if(source(close(to(interfaces(• Short(range(–(slow(exp()(oscillaCon(A>(Evanescent(contribuCon(significant(• Long(range(–(fast(exp()(oscillaCon(A>(Evanescent(contribuCon(insignificant(

+( A(

TruncaCon(of(IntegraCon(Interval(

Integral(TruncaCon(

FFP(IntegraCon(

Example(Pekeris(Waveguide(

FFP(IntegraCon(

AaenuaCon(in(Seawater((Jkps(Eq.(1.34)((α(=(3.3eA3(+(0.11*f^2/(1+f^2)((α:( (AaenuaCon((dB/km)(f:( (Frequency((kHz)((

Pekeris(Waveguide(DepthAdependence(clear(

D=100;(z0=D;(c1=1500;(c2=1800;(rho1=1000;(rho2=1800;(f=20;(omega=2*pi*f;(alpha2=0.0;(delta2=alpha2/54.58;(k1=omega/c1;(k2=omega/c2*(1+i*delta2)(((((((zs=36;(nk=2^10;(km=2*real(k1);(dkr=km/(nkA1)(kr=[0:dkr:km];(nkr=length(kr);(eps=3/(2*pi*log10(exp(1)))(*dkr;(kr=kr(A(i*eps;((nzw=41;(dz=D/(nzwA1);(z=[0:dz:D*1.2];(nz=length(z);((pk=pek_kernel(kr,z,zs,D,z0,k1,k2,rho1,rho2);((figure(1)(subplot(2,1,1);(hold(off(wavei(dba(pk),real(kr),DAz,A60,40);(h=Ctle('Wavenumber(Kernel');(set(h,'Fontsize',16);(h=xlabel('Horizontal(Wavenumber((m^{A1})');(set(h,'Fontsize',14);(h=ylabel('Depth((m)');(set(h,'Fontsize',14);(hold(on;(h=plot([0(real(kr(nkr))],([0(0],'m');(set(h,'Linewidth',2);((%(Transmission(loss([pr,r]=ffp(pk,kr,z);((subplot(2,1,2);(hold(off;(wavei(dba(pr(:,1:nkr/2)),1eA3*r(1:nkr/2),DAz,A120,A20)(h=Ctle('Transmission(Loss');(set(h,'Fontsize',16);(h=xlabel('Range((m)');(set(h,'Fontsize',14);(h=ylabel('Depth((m)');(set(h,'Fontsize',14);(hold(on(h=plot([0(1eA3*r(nkr/2)],([0(0],'m');(set(h,'Linewidth',2);(axis([0(3(DAz(nz)(DAz(1)]);((

pekeris_c.m(

r(jA1( j(r( r(j+1(

Sector(j(

Sector(Interface(Coupling(ConCnuity(of(Pressure(

r(jA1( j(r( r(j+1(

Sector(j(

Sector(Interface(Coupling(ConCnuity(of(ParCcle(Velocity(

r(jA1( j(r( r(j+1(

Sector(j(

Interface(Coupling(EquaCons(

�ReflecCon�(

r(jA1( j(r( r(j+1(

Sector(j( Source(

RadiaCon(CondiCon(b(((=(0(N(

IniCal(and(RadiaCon(CondiCons(

r(jA1( j(r( r(j+1(

Sector(j(

RadiaCon(CondiCon(b(((=(0(N(

Mean(of(pressure(and(velocity(coupling((Other:(p/(ρc)(((conCnuous(1/2(

OneAway(Coupled(Modes(

Split(Step(Parabolic(EquaCons(

Environment( PropagaCon(

θ#

0(

n = 1

=(k rm#

Range-Independent Environment

2# ψ#=(A(k rm#

2# (Φ� � 2ik Φ) #

Phase(Errors(and(Angular(LimitaCons(

0#

PE(Propagates(Normal(Modes(Undistorted(

PE(Modal(Phase(

Phase(Errors(and(Angular(LimitaCons(

Exact(Modal((Phase(

PE(Modal(Pressure(Field(

Phase(Errors(and(Angular(LimitaCons(

c(((=(1590(m/s(ρ((((=(1200(kg/m( 3(2(2(

PE(Workshop(Case(3B(

OUTLINE •  OVERVIEW OF NOISE

– OCEAN ENVIRONMENT – PROPAGATION – DIFFERENT TYPES OF NOISE – SIGNAL PROCESSING: From NOISE AS A

NUISANCE •  EXTRACTING COHERENT

INFORMATION FROM NOISE – THERMAL NOISE – SHIPPING/BIOLOGICAL NOISE – NATURAL: SEISMIC AND SURFACE

131

Recommended