Uwa1 PDF

INSTITUTE OF
W A T E R A C O U S T I C S,
SONAR ENGINEERING AND
SIGNAL THEORY
Underwater Acoustics
Contents
1
2
3
4
5
6
Fundamentals of Ocean Acoustics

Sound Propagation Modeling
Sonar Antenna Design
Sonar Signal Processing
Array Processing
Sonar Data Processing
Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus
INSTITUTE OF
SIGNAL THEORY
1 Fundamentals of Ocean Acoustics

The ocean is an extremely complicated acoustic medium. The
most characteristic feature of the oceanic medium is its inhomogeneous nature. There exist to kinds of inhomogeneities
regular and
random.
Both strongly influence the sound field in the ocean, where

regular variations of the sound velocity versus depth lead to the
formation of an underwater sound channel
random inhomogeneities give rise to scattering of sound waves
and, therefore, to fluctuations in the sound field.
INSTITUTE OF
SIGNAL THEORY
In the following sections we will consider in more detail the

Sound Velocity in the Ocean
typical Sound Velocity Profile
Attenuation of Sound
Ocean surface and bottom interaction
Sound Propagation
Sonar Equation (Performance Prediction)
INSTITUTE OF
SIGNAL THEORY
1.1 Sound Velocity in the Ocean

Variations of the sound velocity c in the ocean are relatively
small. As a rule, c lies between

about 1450 m/s and 1540 m/s.
Even though the changes of c are small, i.e. 3%, the propagation of sound can be significantly effected.
The sound velocity can be directly measured by velocimeters
or calculated by empirical formulae if the
temperature (T)
salinity (S)
hydrostatic pressure (P) or depth (z)
are known.
INSTITUTE OF
SIGNAL THEORY
The measurement error of modern velocimeters is < 0.1 m/s.

The accuracy of calculations using the most complete empirical formulae is about the same.
However, the use of formulae providing such high accuracy is
quit cumbersome.
Therefore, a less but for most applications sufficient accurate
formula is given by
c 1449.2 4.6 T 0.055T 2 0.00029T 3
(1.34 0.01T )(S 35) 0.016 z
with temperature T in (C), salinity S in (ppt), depth z in (m)
and sound velocity c in (m/s). This formula is valid for
0 C T 35 C,
0 ppt S 45 ppt,
0 m z 1000 m.
INSTITUTE OF
SIGNAL THEORY
Assignment 1:
Develop a Matlab program for determining the function
c 1449.2 4.6 T 0.055 T 2 0.00029 T 3
(1.34 0.01T )( S 35) 0.016 z.
Furthermore, to investigate the dependence of c on T, S and z

visualize c versus z for various sets of T and S, i.e.
T = 5 C : 5 C : 30 C and S = 10 ppt : 5 ppt : 35 ppt,
in appropriate diagrams and discuss the results obtained.
INSTITUTE OF
SIGNAL THEORY
1.2 Typical Vertical Sound Velocity Profiles

The form (profile) of the curve
c(z) = c(T(z), S(z), z)
and the derivative (gradient)
d
c(z )
dz
are most important for sound propagation in the ocean.
The profiles c(z)
are different in various ocean regions and
vary with time, e.g. seasonal changes
INSTITUTE OF
SIGNAL THEORY
The variation of T and S at depths below 1 km are usually

fairly weak.
Sound velocity increases almost exclusively due to the increasing
hydrostatic pressure, i.e. sound velocity increases linearly with depth.
1.2.1 Underwater Sound Channel (USC)

Deep water regions, typical profiles possess a velocity
minimum at a certain depth zm, where
zm defines the axis of the Underwater sound channel
above zm the sound velocity increases mainly due to
temperature increases
below zm the increase in hydrostatic pressure is mainly
responsible for increasing the sound velocity
INSTITUTE OF
SIGNAL THEORY
If a sound source is closely located to the axis of the USC

the sound energy is partly trapped within the USC, i.e. some
part of sound does not reach the bottom or surface and therefore does not undergo scattering and absorption at these
boundaries.
Underwater Sound Channel of the First Kind, i.e. c0 < ch
INSTITUTE OF
SIGNAL THEORY
Waveguide propagation can be observed in the domain

0 < z < zc,
where the depths
z = 0 and z = zc
define the boundaries of the USC.
The channel traps all sound rays that leave a source located
on the USC axis at grazing angles
< max with max
2(c0 cm )
,
cm
where cm and c0 are the sound velocities at the axis and the
boundaries of the USC.
10
INSTITUTE OF
SIGNAL THEORY
The greater the sound velocity difference

c = c0 cm
the larger is the angular interval
[0, max]
in which the rays are trapped, i.e. the waveguide is getting
more effective.
The depth zm of the USC axis
lies in a domain of 1000 m to 1200 m at mid-latitudes
tends towards the ocean surface in polar regions
can fall down to 2000 m in tropical areas
11
INSTITUTE OF
SIGNAL THEORY
At moderate latitudes, i.e. 60 S to 60 N, the sound velocity

cm on the USC axis ranges from
1450 m/s to 1485 m/s in the Pacific Ocean
1450 m/s to 1500 m/s in the Atlantic Ocean
Underwater Sound Channel of the Second Kind, i.e. co > ch
12
INSTITUTE OF
SIGNAL THEORY
The USC of the second kind extends from the bottom depth
z = h up to the depth z = zc, where the sound speed equals ch.
USCs of the first kind, i.e. c0 < ch, occur in deep water areas,
whereas USCs of the second kind, i.e. c0 > ch, are more likely
in regions of shallower water.
Typical Zonal Structure of a Sound Field in a USC
For sources near the ocean surface typical so-called

"zonal structures"
of the sound field, i.e. sequences of insonified and shadow
zones, can be observed in the ray diagram
13
INSTITUTE OF
SIGNAL THEORY
where A1, A2, and B1, B2, indicate the shadow zones.
The Shadow zones
decrease as the source depth z1 approaches the USC axis
disappear if the source depth z1 coincides with the depth of the
USC axis zm, i.e. zm = z1.
14
INSTITUTE OF
SIGNAL THEORY
1.2.2 Surface Sound Channel

Such sound channels are formed when the channel axis coincides with the surface.
b
The sound velocity increases down to the depth z = h and then

begins to decrease.
Rays leaving the source at grazing angles < b propagate
with multiple surface reflection in the surface sound channel.
15
INSTITUTE OF
SIGNAL THEORY
In case of a rough ocean surface, sound is partly scattered into

angles > b at each interaction with the surface, i.e.
rays leave the sound channel

sound levels decay in the surface sound channel and increase below
the surface sound channel.
Surface sound channels frequently occur

in tropical and moderate zones of the ocean, where c increases with
depth due to the positive hydrostatic pressure gradient when T and
S are almost constant due to wind mixing in the upper ocean layer
in Arctic and Antarctic regions, where c monotonically increases
from the surface to the bottom due to the positive temperature and
hydrostatic pressure gradient
in Mediterranean seas, in the tropical zone and in shallow seas,
where c increases with depth if the temperature on the surface
decays due to changes during autumn and winter
16
INSTITUTE OF
SIGNAL THEORY
1.2.3 Underwater Sound Channel with two Axis

This case takes place when surface and deep water sound
channels are present simultaneously.
Such a distribution of the sound velocity with depth is due to

the intrusion of warmer and saltier Mediterranean waters into
the Atlantic Ocean, e.g. off the cost of Portugal.
17
INSTITUTE OF
SIGNAL THEORY
1.2.4 Antiwaveguide Propagation

An antiwaveguide propagation is observed when the sound
velocity monotonically decreases with depth.
Such sound velocity profiles are often a result of intensive
heating by solar radiation of the upper ocean layer.
Formation of a geometrical shadow
18
INSTITUTE OF
SIGNAL THEORY
The shadow zone is not a region of zero sound intensity since

sound waves penetrate into the shadow zone due to
diffraction
reflections at sea floor
scattering by inhomogeneities of the medium
Sound Propagation in Shallow Water
19
INSTITUTE OF
SIGNAL THEORY
1.3
Transmission Loss of Sound
1.3.1 Spreading Loss

Spreading loss is a measure of signal weakening due to the
geometrical spreading of a wave propagating outward from
the source.
Two geometries are of importance in underwater acoustics
spherical spreading , i.e. a point source in an unbounded
homogeneous medium
cylindrical spreading, i.e. a point source in a medium that has
upper and lower boundaries (waveguide case)
If we assume the medium to be lossless the intensity is for

20
INSTITUTE OF
SIGNAL THEORY
spherical spreading inversely proportional to the surface of the

sphere of radius r, i.e.
I
1
( 4 r 2 )
cylindrical spreading inversely proportional to the surface of the

cylinder of radius r and depth d, i.e.
I
1
2 r d
Remark:
For a point source in a homogeneous waveguide one observes
spherical spreading in the near field, i.e. r d
cylindrical spreading at long range, i.e. r >> d
transition region from spherical towards cylindrical spreading in-between
21
INSTITUTE OF
SIGNAL THEORY
1.3.2 Sound Attenuation in Water

The acoustic energy of a sound wave propagating in the
ocean is partly
absorbed, i.e. the energy is transformed into heat
lost due to sound scattering by inhomogeneities
Remark:
It is not possible to distinguish between absorption and scattering effects
in real ocean experiments. Both phenomena contribute simultaneously to
the sound attenuation in sea water.
On the basis of extensive laboratory and field experiments

the following empirical formulae for the calculation of attenuation coefficients in sea water have been derived.
22
INSTITUTE OF
SIGNAL THEORY
Thorp formula (valid for 100 Hz < f < 3 kHz)

0.11 f 2
44 f 2
w
1 f 2 4100 f 2
[dB/km]
with
f in [kHz]
Schulkin and Marsh formula (valid for 3 kHz < f < 0.5 MHz)
S A fT f 2 Bf 2
(1 6.54 10 4 P) [dB/km]
w 8.686 10 2
2
fT
fT f
3
where
A 2.34 10 6 , B 3.38 10 6 , S in [ppt], f in [kHz],
the relaxation frequency

f T 21.9 10 6 1520 / (T 273 )
with
T in [C] for 0 T 30 C
23
INSTITUTE OF
SIGNAL THEORY
and the hydrostatic pressure is determined by

P 1.01(1 z 0.1) in [kg/cm 2 at]
Francois and Garrison Formula (valid for 100 Hz < f < 1 MHz)
A1 P1 f1 f 2
A2 P2 f 2 f 2
w 2
A3 P3 f 2 [dB/km]
2
2
2
f1 f
f2 f
pure water
B(OH)3
boric acid
MgSO 4
magnesium sulphate
viscosity
The coefficients for the contribution of
boric acid, B(OH)3

1245
4
8.686 0.78 ph 5
S
T 273
A1
10
, f1 2.8
10
c
35
P1 1, c 1412 3.21T 1.19 S 0.0167 zmax
24
INSTITUTE OF
SIGNAL THEORY
magnesium sulphate, MgSO4
8.17 1081990 / (T 273)

f2
1 0.0018 ( S 35)
S
A2 21.44 (1 0.025T ),
c
2
P2 1 1.37 104 zmax 6.2 109 zmax
pure water viscosity
4.937 104 2.59 105 T 9.11107 T 2 1.5 108 T 3

for T 20 C
A3
4
5
7 2
10 3
3.964 10 1.146 10 T 1.45 10 T 6.5 10 T for T 20 C
2
P3 1 3.83 105 zmax 4.9 1010 zmax
with f in [kHz], T in [C], S in [ppt]. Furthermore zmax, ph and c denote the

water depth in [m], the ph-value and the sound speed in [m/s] respectively.
25
INSTITUTE OF
SIGNAL THEORY
Comparison of the Thorp, Schulkin-Marsh and FrancoisGarrison attenuation formulae.

Thorp
Schulkin-Marsh
Francois-Garrison
Attenuation [dB/km]
10
10
10
-2
10
-4
10
10
10
10
10
Frequency [Hz]
10
10
26
INSTITUTE OF
SIGNAL THEORY
Frequency dependence of the different attenuation processes

employed in the Francois-Garrison model.
Francois-Garrison Formula
10
Boric Acid
Magnesium Sulphate
Pure Water
All Contributions
Attenuation [dB/km]
10
10
10
-2
10
-4
10
-6
10
10
10
10
10
Frequency [Hz]
10
10
27
INSTITUTE OF
SIGNAL THEORY
Assignment 2:
Develop a Matlab program that calculates the sound
attenuation in seawater by means of the
Thorp formula
Schulkin & Marsh formula
Francois & Garrison formula.
Display and compare the results of the three approaches.

For the Francois and Garrison formula
investigate the dependence on the frequency f, salinity S and

temperature T for a depth of z = 50 m
depict the attenuation versus frequency for a particular S, T
and z and specify the frequency regions where the different
attenuation processes dominate.
28
INSTITUTE OF
SIGNAL THEORY
1.3.3 Sound Attenuation in Sediment

The sound attenuation in the sediment mainly varies with the
bottom type. It can be approximately determined by the empirical formula
n
1
f 1
s
K
,
8.686
1kHz
where K and n denote two bottom type dependent parameters.

The following table provides the values for K and n for four
representative sediment types.
Sediment type
Parameter
very fine silt fine sand medium sand coarse sand
0.17
0.45
0.48
0.53
0.96
1.02
0.98
0.96
29
INSTITUTE OF
SIGNAL THEORY
1.4
Sound Reflection and Transmission at a

fluid-fluid interface
1.4.1 Lossless Media

Reflectivity is the ratio of the amplitudes of a reflected plane
wave to a plane wave incident on an interface separating two
media. It is an important measure for the impact of the sea
surface and bottom on sound propagation.
incident wave
reflected wave
Medium 1
1 1
Medium 2
2, c2
1, c1
transmitted wave
z
30
INSTITUTE OF
SIGNAL THEORY
Assuming the incident wave to have amplitude A and denoting the reflection and transmission coefficients by R and T,
respectively, we can write
pi A exp j ( t k Ti r )
k Ti k1 (sin 1 ,cos 1 )
pr R A exp j ( t k Tr r )
k Tr k1 (sin 1 , cos 1 )
pt T A exp j ( t k Tt r )
k Tt k2 (sin 2 ,cos 2 )
with
k1
c1
, k2
c2
, r ( x , z )T .
The unknown quantities R, T and 2 are determined from the

boundary conditions requiring continuity of pressure and vertical particle velocity across the interface at z = 0.
31
INSTITUTE OF
SIGNAL THEORY
The boundary conditions can be mathematically stated as

p1 pi pr pt p2
and
vz ,1
1 p1
1 p2 vz ,2
,
t
t
1 z
2 z
respectively. After substituting pi, pr and pt in the first boundary condition, we obtain
1 R T exp j ( k1 sin 1 k 2 sin 2 ) x.
Since the left side is independent of x, the right side must also
be independent of x, i.e.
k1 sin 1 k2 sin 2 0.
This leads to 1 + R = T and to the well-known refraction law

sin 1 k 2 c2 c1
n.
sin 2 k1 c1 c2
32
INSTITUTE OF
SIGNAL THEORY
The second boundary condition and k1 sin 1 k2 sin 2 provide

k1
(1 R) cos 1
k2
T cos 2 .
Defining m = 2/1 and using the last equation together with

1 + R = T and n = k2/k1 we find
2
2
m cos 1 n cos 2 m cos 1 n sin 1
R
m cos 1 n cos 2 m cos 1 n 2 sin 21
and
2m cos 1
2m cos 1
,
m cos 1 n cos 2 m cos 1 n 2 sin 21
where n cos 2 n 2 sin 21 for n sin 1 0 has been

exploited.
33
INSTITUTE OF
SIGNAL THEORY
Remarks:
Features of the reflection and transmission coefficient
a) If 1 tends to /2 then R and T tend independently of the
parameters of the media to 1 and 0, respectively.
b) At the angle of incidence 1 that satisfies
sin 1
m2 n2
, i.e. R 0
m2 1
the boundary will be completely transparent.

c) For sin 1 > n and n cos 2 j sin 21 n 2 the reflection coefficient can be expressed by
R
m cos 1 j sin 21 n 2
m cos 1 j sin 21 n 2
finiteness of refracted wave re-

quires negative sign of the root
34
INSTITUTE OF
SIGNAL THEORY
or after some manipulations by

R exp( j ) with
sin 2 n 2
1
R 1 and 2 arctan
m cos 1
i.e., total reflection occurs.

The phase difference between the incident and reflected
waves at the interface is given by .
The angle of incidence satisfying sin 1 n is called
critical angle crit.
For 1 > crit and k2 cos 2 k1n cos 2 j k1 sin 21 n 2
the amplitude of the transmitted sound pressure satisfies
pt exp( z ) with k1 sin 21 n 2 .
(finiteness of refracted wave requires negative sign of the root)

35
INSTITUTE OF
SIGNAL THEORY
Example: Water-Air / Air-Water Interface

With the densities and sound velocities of water and air
w 1
g
kg
m
1000
,
1500
w
cm 3
m3
s
a 1.3 10 3
g
kg
m
1.3 3 , ca 333 ,
3
cm
m
s
we obtain for a sound wave impinging perpendicular from

water into air and vice versa the reflection and transmission
coefficients
R 1, T 0
and
R 1, T 2,
respectively.
36
INSTITUTE OF
SIGNAL THEORY
The intensities (power flux densities) of the incident, reflected

and transmitted waves are defined by
pi
pr
A2
Ii
,
2 1c1 2 1c1
R 2 pi
R 2 A2
Ir
2 1c1
2 1c1
2 1c1
and
2
T 2 pi
T 2 A2
.
It
2 2 c2
2 2 c2
2 2 c2
pt
Moreover, to derive the power reflection and transmission

coefficients RP and TP, one has to take into account that the
refraction at the interface changes the intensity because of a
change in the cross-sectional area, cf. following figure.
37
INSTITUTE OF
SIGNAL THEORY
bundle of incident rays
bundle of reflected rays
Si S r
1 1
Medium 1
1, c1
St
Medium 2
2, c2
x
bundle of transmitted rays
The cross-sections of the incident, reflected and transmitted

bundle of rays are given by
Si S r l y x cos 1 and St l y x cos 2 .
38
INSTITUTE OF
SIGNAL THEORY
Using these expressions, we obtain

m cos 1 n cos 2
P SI
I
RP r r r r R 2
Pi Si I i I i
m
cos
n
cos
1
2
and
TP
Pt St I t cos 2 I t cos 2 1c1 2

T
Pi Si I i cos 1 I i cos 1 2c2
cos 2 n
4m 2 cos21
4nm cos 1 cos 2
.
2
2
cos 1 m m cos 1 n cos 2
m cos 1 n cos 2
One can now prove that the law of energy conservation, i.e.
Pi = Pr + Pt implying RP + TP = 1 is satisfied. Furthermore, it
can be shown that RP and TP remain unchanged if we change
the ordering of the media, i.e.
1 2 , c1 c2 , 1 2 .
39
INSTITUTE OF
SIGNAL THEORY
1.4.2 Lossy Media

In the former section the reflection and transmission properties of sound at plan boundaries have been deduced if the absorption of the media can be neglected.
In contrast to sound propagation at the water-air-boundarylayer experimental investigations at the water-sediment-boundary-layer show that the theory agrees sufficiently exact with
the results of measurements only if the absorption in the sediment is taken into account.
Therefore, the results of section 1.4.1 are now extended for the
case of a boundary layer between a absorption-free medium 1
(water) and an absorption-afflicted medium 2 (sediment).
40
INSTITUTE OF
SIGNAL THEORY
The absorption is introduced by means of the complex wave

number
k2 k2,R j k2, I
with
k2,R
c2
and k2, I 2 ,
where c2 and 2 denote the velocity of sound and attenuation

of the medium 2 (sediment) respectively. The attenuation of
medium 1 (water) can be neglected.
Thus, the refraction law can be written as
k
sin 1 k2 k2,R

j 2, I nR j nI n.
sin 2 k1
k1
k1
41
INSTITUTE OF
SIGNAL THEORY
The wave number k1 and the incidence angle 1 are always

real. Thus, the product k1sin1 is also real. Since, the wave
number k2 is complex and k2sin2 has to be real due to the
refraction law,
k sin 1
sin 2 1
k2
has to be complex. Hence,

k2 cos 2 k2 1 sin 2 2 k2 1 k12 sin 21 k22
k1 n 2 sin 21 k2,R j k2, I
and
k1 sin 1 k2 sin 2 .
The transmitted sound pressure can now be expressed by

42
INSTITUTE OF
SIGNAL THEORY
r ,
pt T A exp 0, k2, I r exp j t k1 sin 1 , k2,R r
T A exp k TA r exp j t k Tp
where
kA
with
0, k
k P k1 sin 1 , k2,R
T
2, I
k P sin 2,P ,cos 2,P
k A sin 2, A ,cos 2, A
k P ( k1 sin 1 ) ( k2,R ) 2 k1 sin 21 Re
and
k A k2, I k1 Im
n sin 1
n 2 sin 21 .
43
INSTITUTE OF
SIGNAL THEORY
The angles of refraction for the wave fronts of constant phase

and constant amplitude are given by
2,P
and
k1 sin 1
arg k P arctan
arctan
k
Re
2,R
n 2 sin 21
sin 1
2, A arg k A 0,
respectively. Furthermore, the phase velocity of the wave in

the sediment can be written as
cP
kP
c1
sin 1 Re
n sin 1
44
INSTITUTE OF
SIGNAL THEORY
planes of constant amplitude and phase

(homogeneous plane wave)
angle of
incidence
planes of constant amplitude
water
planes of constant phase

(inhomogeneous plane wave)
2,P
sediment
angle of
refraction
45
INSTITUTE OF
SIGNAL THEORY
1.4.3 Forward Reflection Loss

A rough sea surface or seafloor causes attenuation of the
acoustic field propagating in the ocean waveguide. The
attenuation increases with increasing frequency. The field is
scattered away from the specular direction.
Scattering characteristic for

A: smooth bottoms
B: rough bottoms
46
INSTITUTE OF
SIGNAL THEORY
The forward reflection loss due to a rough boundary is often

simply modeled by incorporating an additional loss factor into
the calculation of the specular reflection coefficient. A formula
often used to describe reflectivity from a boundary is
R ( ) R( ) e p 2 ,
where
2
p( ) 2 k cos
denotes the so-called Rayleigh parameter,

2
c
k
with
f
the wavenumber, the RMS (root mean square) roughness

and the angle of incidence.
47
INSTITUTE OF
SIGNAL THEORY
The roughness of the ocean surface caused by wind induced

waves is often modeled by the Pierson-Moskowitz or the
Pierson-Neumann spectrum. The RMS roughness (wave
height) of a fully developed wind wavefield can be approximately determined by
PM 1.4 10 5 v w4
(Pierson-Moskowitz)
PN 0.341 10 5 v w5
(Pierson-Neumann),
where vw denotes the wind speed in [kn] (1 kn = 0.514 m/s).

The RMS roughness of an ocean seafloor is related to the
mean grain size of the sediment. The following table provides
the values of mean grain size and the RMS roughness for
various sediment types.
48
INSTITUTE OF
SIGNAL THEORY
Mean Grain Size

[ = log2(a)]
RMS Roughness
[cm]
2.5
0.5
2.25
coarse sand
0.5
1.85
medium sand
1.5
1.45
fine sand
2.5
1.15
very fine sand
3.5
0.85
coarse silt
4.5
0.7
medium silt
5.5
0.65
fine silt
6.5
0.6
very fine silt
7.5
0.55
slity clay
8.0
0.5
clay
9.0
0.5
Sediment type
sandy gravel
very coarse sand
49
INSTITUTE OF
SIGNAL THEORY
1.5
Sound Scattering
The sea contains, within itself and on its boundaries, inhomogeneities of many different kinds. These inhomogeneities reradiate a portion of the acoustic energy incident upon them. This
reradiation of sound is called scattering. The total sum of all
scattering contributions is called reverberation.
The reverberation basically produced by scatterers
in the ocean volume (marine life, inanimate matter),
on or near the ocean surface (roughness, air bubbles),
on the ocean bottom (roughness)
is called volume reverberation, surface reverberation and

bottom reverberation, respectively.
50
INSTITUTE OF
SIGNAL THEORY
Back scattering from ocean surface
T/R
Back scattering from ocean volume
Target
Back scattering from ocean bottom
Surface Backscattering
Because of its roughness and the occurrence of air bubbles
beneath it, the sea surface is a significant scatterer of sound.
Experiments indicate that the backscattering strength of the
sea surface varies with the
grazing angle ( = / 2 with = angle of incidence),
sound frequency and
wind speed induced roughness,
51
INSTITUTE OF
SIGNAL THEORY
and that the collected measurements can be fitted by the following empirical expression
Ss 10 log10 105.05 (1 vW )2 ( f 0.1)vW
150
tan ( )
with
vW 2
1/ 3
1/8
2.5 ( f 0.1) 4 cos ( ),
vW 1
where SS represents the surface backscattering coefficient

in [dB/m2]. The parameters f, vw and denote the sound
frequency in kHz, the wind speed in knots and the grazing
angle, respectively.
52
INSTITUTE OF
SIGNAL THEORY
frequency = 50 kHz
frequency = 100 kHz

10
surface reverberation [dB/m2]
10
0
-10
-20
-30
-40
ws
ws
ws
ws
-50
-60
-70
20
=
=
=
=
5 kn
10 kn
20 kn
40 kn
40
60
grazing angle [deg]
0
-10
-20
-30
-40
-60
-70
80
ws
ws
ws
ws
-50
20
frequency = 200 kHz
40
60
grazing angle [deg]
80
10
5 kn
10 kn
20 kn
40 kn
frequency = 400 kHz
10
0
-10
-20
-30
ws
ws
ws
ws
-40
-50
-60
-70
=
=
=
=
20
=
=
=
=
5 kn
10 kn
20 kn
40 kn
40
60
grazing angle [deg]
0
-10
-20
-30
-50
-60
-70
80
ws
ws
ws
ws
-40
20
=
=
=
=
5 kn
10 kn
20 kn
40 kn
40
60
grazing angle [deg]
80
53
INSTITUTE OF
SIGNAL THEORY
wind speed = 5 kn
wind speed = 10 kn
10
f
f
f
f
0
-10
-20
10
= 50 kHz
=100 kHz
= 200 kHz
= 400 kHz
-30
-40
-50
-60
-70
20
40
60
grazing angle [deg]
0
-10
-20
-30
-40
-60
-70
80
f
f
f
f
-50
20
wind speed = 20 kn
80
10
40
60
grazing angle [deg]
wind speed = 40 kn
10
0
-10
-20
-30
f
f
f
f
-40
-50
-60
-70
= 50 kHz
=100 kHz
= 200 kHz
= 400 kHz
20
= 50 kHz
=100 kHz
= 200 kHz
= 400 kHz
40
60
grazing angle [deg]
80
0
-10
-20
-30
f
f
f
f
-40
-50
-60
-70
20
= 50 kHz
=100 kHz
= 200 kHz
= 400 kHz
40
60
grazing angle [deg]
80
54
INSTITUTE OF
SIGNAL THEORY
Bottom Backscattering
The bottom acts, like the sea surface, due to its roughness as
an reflector and scatterer of sound, cf. figure on p. 46.
In section 1.4.3 the specular direction has been considered as
part of the sound propagation via the forward reflection loss.
Now, we are going to model the backscattering behavior of
the seabed. Experimental investigations have shown that the
backscattering strength of the bottom varies with the
grazing angle ( = / 2 with = angle of incidence),
sound frequency and
bottom type induced roughness.
Furthermore, it could be observed that a Lamberts law relationship between the backscattering strength and the grazing
angle fits to many experimental data satisfactorily accurate for
angles below 60.
55
INSTITUTE OF
SIGNAL THEORY
Consequently, the backscattering strength can be described by

Lamberts law and an empirically specified scattering coefficient, i.e.
SB K ( f , bt ) 10 log10 sin2 ( )
0 60,
where K( f,bt) denotes the scattering coefficient depending on

the frequency of sound f and bottom type bt.
Due to the empirical definition of K( f,bt) it is evident that one
will have considerable difficulty in determining an appropriate
value for the backscattering strength in practice.
Therefore, easier applicable and over the entire grazing angle
domain sufficient accurate bottom backscattering models are
of interest.
56
INSTITUTE OF
SIGNAL THEORY
More accurate bottom scattering curves have been derived

from measurements
SEARAY Model (20 kHz f 500 kHz)
APL-UW Model (1 kHz f 500 kHz)
For the SEARAY model the so-called reverberation coefficient is defined by

S B 10 log10 3.03 f 3.20.8bt 102.8bt 12 10 4.42
with
sin( ) 0,19
bt cos16 ( )
and
1 125 exp 2.64 (bt 1.75) 2
50
cot 2 ( ) ,
bt
57
INSTITUTE OF
SIGNAL THEORY
where f, bt and denote the sound frequency in kHz, the

bottom type and the grazing angle respectively.
The bottom type parameter is defined as follows
bt = 1 mud
bt = 2 sand
bt = 3 gravel
bt = 4 rock.
In principle bt can be any real number satisfying
1 bt 4.
This allows an improved bottom type specification.

58
INSTITUTE OF
SIGNAL THEORY
bt
bt
bt
bt
-10
=
=
=
=
frequency = 100 kHz

bottom reverberation [dB/m2]
frequency = 50 kHz
1
2
3
4
-20
-30
-40
0
20
40
60
grazing angle [deg]
bt
bt
bt
bt
-10
-30
-40
80
20
-20
bt
bt
bt
bt
20
40
60
grazing angle [deg]
=
=
=
=
1
2
3
4
-10
40
60
grazing angle [deg]
80
frequency = 400 kHz
-40
1
2
3
4
-20
frequency = 200 kHz
-30
=
=
=
=
0
-10
-20
bt
bt
bt
bt
-30
-40
80
20
40
60
grazing angle [deg]
=
=
=
=
1
2
3
4
80
59
INSTITUTE OF
SIGNAL THEORY
f
f
f
f
-10
bottom type = 2
bottom type = 1
= 50 kHz
=100 kHz
= 200 kHz
= 400 kHz
-20
-30
-40
0
20
40
60
grazing angle [deg]
f
f
f
f
-10
-20
-30
-40
80
20
0
-10
-20
-30
-40
0
20
= 50 kHz
=100 kHz
= 200 kHz
= 400 kHz
40
60
grazing angle [deg]
40
60
grazing angle [deg]
80
bottom type = 4
bottom type = 3
f
f
f
f
= 50 kHz
=100 kHz
= 200 kHz
= 400 kHz
80
0
-10
-20
f
f
f
f
-30
-40
0
20
= 50 kHz
=100 kHz
= 200 kHz
= 400 kHz
40
60
grazing angle [deg]
80
60
INSTITUTE OF
SIGNAL THEORY
Volume Backscattering
In section 1.3.2 we considered the frequency dependence of
the sound attenuation which is partly caused by scattering in
the water volume.
This also produces a backscattered sound field. However
most volume reverberation is thought to arise from biological
organisms and turbidity. The volume reverberation can be
modeled by the so-called volume reverberation coefficient.
SV Sp 7 log10 f ,
where f and Sp indicate the frequency in kHz and the

particle contribution in [dB/m3]. The particle contribution
parameter Sp is defined as follows
61
INSTITUTE OF
SIGNAL THEORY
Sp 50 dB High
Sp 70 dB Moderate Particle density
Sp 90 dB Low
Assignment 3:
Develop a Matlab program for computing the surface, bottom
and volume reverberation coefficient.
Plot the coefficients SS and SB versus the grazing angle for
various sets of (f, vw) and (f, bt), respectively.
Plot the volume reverberation versus frequency for high,
moderate and low particle densities.
Explain the results.
62
INSTITUTE OF
SIGNAL THEORY
1.6 Ambient Noise

The isotropic noise level consists of the following components
turbulence noise (1Hz to 10 Hz)
NLturb ( f )=30 30 log10 ( f ) ,
f in [kHz]
far shipping (traffic) noise (10 Hz to 300 Hz)
3 108
,
NLtraffic ( f )=10 log10
4
4
1
+
10
f
f in [kHz]
sea state noise (300 Hz to 100 kHz)
vw2
,
NLss ( f,vw ) 40 10 log10
5/ 3
1
+f
f in [kHz]
where vw denotes the wind speed in [kn].

63
INSTITUTE OF
SIGNAL THEORY
thermal noise (molecular agitation) (100 kHz to 1 MHz)
NLtherm( f ) 15 20 log10 ( f ) , f in [kHz]
rainfall noise (1 kHz to 5 kHz)
NLrain ( f,r ) in [dB],

where f and r denote the frequency and rainfall rate, respectively.
biological noise (fishes, scrimps etc.)
NLbio ( f,s ) in [dB],

where f and s denote the frequency and season, respectively.
self (vessel) noise of sonar platform
NLvessel ( f,vv ) in [dB],

where f and vv denote the frequency and vessel speed, respectively.
64
INSTITUTE OF
SIGNAL THEORY
Thus the isotropic noise level can be determined by

NL( f , vw , r, s, vv ) 10 log10 (100.1NLturb 100.1NLtraffic 100.1NLss
100.1NLtherm 100.1NLrain 100.1NLbio 100.1NLvessel )
65
INSTITUTE OF
SIGNAL THEORY
Ambient Noise
Ambient Noise
130
130
Rate of Rainfall = 10cm/h
wind speed = 5 knots (SS1)
120
110
110
100
90
Level [dB]
Level [dB]
100
80
90
80
70
70
60
60
50
50
40
-3
10
-2
10
-1
10
10
Frequency [KHz]
10
10
Rate of Rainfall = 2.5cm/h

120
40
-3
10
10
-2
10
-1
10
Ambient Noise
10
Frequency [KHz]
10
10
Ambient Noise
130
140
120
110
100

120
100
Level [dB]
90
Level [dB]
10
80
70
80
60
60
50
40
40
30
-3
10
-2
10
-1
10
10
Frequency [KHz]
10
10
10
20
-3
10
-2
10
-1
10
10
Frequency [KHz]
10
10
10
66
INSTITUTE OF
SIGNAL THEORY
Ambient Noise
Ambient Noise
130
130
shrimp noise = 65dB

120
110
100
110
100
90
Level [dB]
Level [dB]
90
80
70
80
70
60
60
50
50
40
40
30
-3
10
shrimp noise = 55dB

120
-2
10
-1
10
10
Frequency [KHz]
10
10
30
-3
10
10
-2
10
-1
10
10
10
140
shrimp noise = 45dB
120
100
80
80
60
60
40
40
-2
10
-1
10
10
Frequency [KHz]
10
10
shrimp noise = 35dB

120
Level [dB]
100
Level [dB]
10
Ambient Noise
Ambient Noise
140
20
-3
10
10
Frequency [KHz]
10
20
-3
10
-2
10
-1
10
10
Frequency [KHz]
10
10
10
67
INSTITUTE OF
SIGNAL THEORY
Assignment 4:
Develop a Matlab program for calculating the isotropic
ambient noise level. Plot the ambient noise level versus
frequency for wind speeds of 5:5:30 kn and where NLvessel,
NLrain and NLbio are set to 999 dB. Indicate the frequency
domains where either NLtraffic, NLturb, NLss or NLthermal
dominate.
68
INSTITUTE OF
SIGNAL THEORY
1.7 Sonar Performance Prediction

1.7.1 Performance parameters
To assess the capabilities of a sonar system, parameters that
measure the performance have to be defined, e.g.
EL:
EE:
SN:
SE:
Echo Level
Echo Excess
Signal to Noise Ratio
Signal Excess
1.7.2 Sound Propagation Related Parameters

Transmission Loss (TL)
The Transmission loss is given by
TL = spreading loss + attenuation, [dB]
69
INSTITUTE OF
SIGNAL THEORY
where TL is defined to be 0 dB on a sphere around the source

of radius r = 1 m.
For a constant sound velocity profile and therefore spherical
spreading the transmission loss can be determined by
TL(r , z ) 20 log10 ( R) ( R 1 m)
with
R ( r r0 )2 ( z z0 )2 ,
where r0, z0 denote the horizontal and vertical coordinates of

the source locations and the receiver position, respectively.
In case of depth and range dependent cylinder symmetric
sound velocity profiles the TL can be calculated by
70
INSTITUTE OF
SIGNAL THEORY
TL( r, z ) 20 log10 ( R ) 10 log10 F ( r, z ) ( R 1m),
where F(r,z) denotes the so called focusing factor given by

F ( r, z )
actual spreading at r, z
.
spherical spreading at r, z
Isotropic Noise Level (NL)

The isotropic noise level NL(f,vw,r,s,vv) describes for particular
vw, r, s and vv the noise power within a 1Hz band around frequency f.
Thus, assuming NL approximately white over the frequency
band B of interest, the noise level is given by
NLB NL( f , vw , r, s, vv ) 10 log10 ( B ).
71
INSTITUTE OF
SIGNAL THEORY
Bottom Reverberation Strength (RSB)

The bottom reverberation coefficient SB(f,bt, ) describes the
reverberation strength of an insonified area of 1 m2.
With c
= sound speed
=
2h =
2h,T =
2h,R =
r0, z0 =
r,z =
pulse length
min(2h,T , 2h,R)
horizontal 3 dB beam width of transmitter
horizontal 3 dB beam width of receiver
coordinates of transmitter / receiver configuration
coordinates of a particular point on the sea floor
the bottom reverberation strength can be determined for given

f and bt as function of by
72
INSTITUTE OF
SIGNAL THEORY
RS B S B ( f , bt , ) 10 log10 ( AB ),
where AB denotes the insonified bottom area.

AB 2 h R
with
2 v
c
2 cos
R (r r0 ) 2 ( z z0 ) 2
c
2 h
c 2
c (2 cos )
AB
73
INSTITUTE OF
SIGNAL THEORY
Surface Reverberation Strength (RSS)

Analog to the bottom reverberation strength, the surface
reverberation strength is provided by
RS S S S ( f , vw , ) 10 log10 ( AS ) ,
where AS denotes the insonified sea surface area
c
AS 2 h (r r0 ) 2 ( z z0 ) 2
2 cos
with
=
2h =
2h,T =
2h,R =
r0, z0 =
r, z =
sound speed
pulse length
min(2h,T, 2h,R)
horizontal 3 dB beam width of transmitter
horizontal 3 dB beam width of receiver
coordinates of transmitter / receiver configuration
coordinates of a particular point on the sea surface
74
INSTITUTE OF
SIGNAL THEORY
Volume Reverberation Strength (RSV)

The volume reverberation coefficient SV (Sp,f ) describes the
reverberation strength of an insonified volume of 1m3. Thus,
the volume reverberation strength can be calculated by
RSV SV ( Sp, f ) 10 log10 (V ),
where V denotes the insonified volume (isovelocity)

c
V
2 h 2 v R 2 and R (r r0 ) 2 ( z z0 ) 2
2
with
= sound speed
= pulse length
= min(2h,T , 2h,R) / min(2v,T , 2v,R)
2h / 2v
2h,T / 2v,T = horizontal / vertical 3 dB beam width of transmitter
2h,R / 2v,R = horizontal / vertical 3 dB beam width of receiver
75
INSTITUTE OF
SIGNAL THEORY
r0, z0 = coordinates of transmitter / receiver configuration

r, z = coordinates of a particular point on the water volume
2 v
R
c 2
2 h
76
INSTITUTE OF
SIGNAL THEORY
1.7.3 Sonar Equation

To determine the aforementioned reverberation levels the
following environmental parameters have to be specified.
bt:
vw:
S:
T:
c:
Bottom type
Wind speed
Salinity
Water temperature
Sound-speed-profile
For given sonar parameters, i.e.

SL:
f:
B:
:
Source Level in dB at 1m
Center frequency of the sound signal
Bandwidth of the sound signal
Pulse length of the sound signal
77
INSTITUTE OF
SIGNAL THEORY
DI:
BPT:
BPR:
2h :
2v:
Directivity Index of Receiver Array

Transmitter Beam pattern (vertical)
Receiver Beam pattern (vertical)
Horizontal 3 dB Beamwidth min{2h,T , 2h,R}
Vertical 3 dB Beamwidth min{2v,T , 2v,R}
and Target parameters i.e.

TS:
Ll:
Lr:
Target Strength
Target extent in lateral direction
Target extent in radial direction
the performance parameters can be determined by
EL( r, z ) 10 log10 el ( r, z ) 10 log10 sl bpT ,E bpR ,E ts tlE2
10 log10 100.1SL 10
0.1BPT , E
10
0.1BPR , E
100.2TLE 100.1TS
SL BPT ,E BPR ,E 2TLE TS

78
INSTITUTE OF
SIGNAL THEORY
EE ( r, z ) 10 log10 ee( r, z ) 10 log10 el di nlB
10 log10 100.1EL 100.1( NLB DI ) EL ( NLB DI )

SL BPT ,E BPR ,E 2TLE TS ( NLB DI )
SN ( r, z ) 10 log10 sn( r, z ) 10 log10 el til
10 log10 100.1EL 100.1TIL ) EL TIL

SL BPT ,E BPR ,E 2TLE TS TIL
and
SE ( r, z ) 10 log10 se( r, z ) 10 log10 sn dt
10 log10 100.1SN 100.1DT ) SN DT

SL BPT ,E BPR ,E 2TLE TS TIL DT ,
79
INSTITUTE OF
SIGNAL THEORY
where DT denotes the detection threshold, TIL the total

inference level
TIL( r, z ) 10 log10 til ( r, z ) 10 log10 nlB di rlB rlS rlV
10 log10 100.1( NLB DI ) 100.1RLB 100.1RLS 100.1RLV
and RLB, RLS, and RLV the reverberation level of the bottom,
surface and volume, respectively, i.e.
RLB SL BPT ,B BPR ,B 2TLB RS B
RLS SL BPT ,S BPR ,S 2TLS RS S
RLV SL BPT ,V BPR ,V 2TLV RSV .
For c = const., we can write

TLE TLB TLS TLV .
80
INSTITUTE OF
SIGNAL THEORY
Furthermore, the following abbreviations have been used.

BPT ,E : Transmitter Beampattern value for the ray directed
BPR ,E : Receiver toward the target position

BPT ,B : Transmitter Beampattern value for the ray directed
BPR ,B : Receiver toward the insonified bottom area

BPT ,S : Transmitter Beampattern value for the ray directed
BPR ,S : Receiver toward the insonified surface area

TLE
TLB Tranmission loss for the echo, bottom,
TLS surface and volume reverberation, respectively

TLV
81
INSTITUTE OF
SIGNAL THEORY
Assignment 5:
Develop a Matlab program for determining the SN(r,z) and
carry out calculations for the following parameters.
z / r: up to 50 m / 600 m
S:
33 ppt
bt:
T:
mud, sand, gravel

15
vw:
c:
5, 15, 25 knots
1480 m/s
SL:
B:
DI:
2v,R:
z s:
220 dB re1Pa@1m
10 kHz
30 dB
180
5m
f:
BPT:
2h,R:
2v,T:
100 kHz
0 dB (90)
0.5
180
:
BPR:
2h,T:
r s:
100 s
0 dB (90)
90
0m
TS:
15 dB
Discuss the observations, i.e. the impact of the bottom type

and wind speed on the signal to noise ratio.
82
INSTITUTE OF
SIGNAL THEORY
Literature
[1] Brekhovskikh, L.M.; Lysanov, Y.P.: Fundamentals of Ocean
Acoustics, Springer, 2003
[2] Etter, P.C.: Underwater Acoustic Modeling, Spon Press, 2003
[3] Jensen, F.B: Computational Ocean Acoustics, Springer, 2000
[4] Lurton, X.: An Introduction to Underwater Acoustics, Springer,
2004
[5] Medwin, H.;Clay C.S: Acoustical Oceanography, Academic Press,
1998
[6] Tolstoy, I.; Clay C.S: Ocean Acoustics, AIP-Press, 1987
[7] Urick, R.I: Principles of Underwater sound, McGraw Hill, 1983
83

Uwa1 PDF

Diunggah oleh

Informasi Dokumen

Deskripsi Asli:

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Uwa1 PDF

Diunggah oleh

Hak Cipta:

Format Tersedia

INSTITUTE OF

Fundamentals of Ocean Acoustics

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

1 Fundamentals of Ocean Acoustics

Both strongly influence the sound field in the ocean, where

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

In the following sections we will consider in more detail the

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

1.1 Sound Velocity in the Ocean

small. As a rule, c lies between

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

The measurement error of modern velocimeters is < 0.1 m/s.

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

Furthermore, to investigate the dependence of c on T, S and z

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

1.2 Typical Vertical Sound Velocity Profiles

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

The variation of T and S at depths below 1 km are usually

hydrostatic pressure, i.e. sound velocity increases linearly with depth.

1.2.1 Underwater Sound Channel (USC)

If a sound source is closely located to the axis of the USC

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

Waveguide propagation can be observed in the domain

< max with max

The greater the sound velocity difference

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

At moderate latitudes, i.e. 60 S to 60 N, the sound velocity

Underwater Sound Channel of the Second Kind, i.e. co > ch

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

For sources near the ocean surface typical so-called

1.2.2 Surface Sound Channel

The sound velocity increases down to the depth z = h and then

In case of a rough ocean surface, sound is partly scattered into

rays leave the sound channel

Surface sound channels frequently occur

1.2.3 Underwater Sound Channel with two Axis

Such a distribution of the sound velocity with depth is due to

1.2.4 Antiwaveguide Propagation

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

The shadow zone is not a region of zero sound intensity since

Sound Propagation in Shallow Water

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

Transmission Loss of Sound

1.3.1 Spreading Loss

If we assume the medium to be lossless the intensity is for

spherical spreading inversely proportional to the surface of the

cylindrical spreading inversely proportional to the surface of the

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

1.3.2 Sound Attenuation in Water

On the basis of extensive laboratory and field experiments

Thorp formula (valid for 100 Hz < f < 3 kHz)

the relaxation frequency

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

and the hydrostatic pressure is determined by

The coefficients for the contribution of

boric acid, B(OH)3

Chapter 1 / Underwater Acoustics / Prof. Dr.-Ing. Dieter Kraus

magnesium sulphate, MgSO4

8.17 1081990 / (T 273)

pure water viscosity

4.937 104 2.59 105 T 9.11107 T 2 1.5 108 T 3

with f in [kHz], T in [C], S in [ppt]. Furthermore zmax, ph and c denote the