Professional Documents
Culture Documents
Original Russian Text V.A. Grigorev, B.G. Katsnelson, 2009, published in Akusticheski Zhurnal, 2009, Vol. 55, No. 1, pp. 4755.
OCEAN ACOUSTICS
AND UNDERWATER SOUND
AbstractIntensity variations of high-frequency sound pulses due to the motion of internal solitons in shallow
water are investigated in terms of the ray approximation. It is shown that ray distortions cause intensity fluctuations of about 35 dB. It is found that rays with turning points near the upper boundary of the thermocline play
the dominant role in the formation of these fluctuations. Formulas for estimating the fluctuation frequencies predominantly observed in the spectrum of intensity variations are presented.
PACS numbers: 43.30.Re
DOI: 10.1134/S1063771009010072
INTRODUCTION
x
c2
h2
zu
ht
z1
c1,
n'
rn
vs
h1
H
z
69
cb, b,
S
0
c 0 ( z ) = c 2 ( z z u ), z u < z < z 1 ;
c 1 , z 1 z H,
(1)
m 1 sin n n 1 ( 1 + i ) cos n
-,
V n = ----------------------------------------------------------------------------2
2
m 1 sin n + n 1 ( 1 + i ) cos n
2
(2)
Vol. 55
No. 1
2009
(3)
c ( u, z ) = z 1 + ( u ) z H;
(4)
( u ), z u + ( u ) < z < z 1 + ( u ),
where the function (u) describes the vertical displacement of the thermocline; for this function, we use the
Kortewegde Vries (CdV) soliton
2 u
( u ) = a sech --- .
L
(5)
70
GRIGOREV, KATSNELSON
(u), m
0
2
4
6
8
10
400
200
200
M
V nm . The second subscript m charwhere V =
m=1
acterizes the fact that the angles of incidence on the bottom are different for different reflections.
The cross-sectional area of the nth ray tube near the
receiver is Sn = sin 'n rrn, where 'n is the grazing
angle of the nth ray tube near the receiver (in the general case, 'n differs from n because of the ray refraction in the soliton) and rn is the width of the insonified
region in the radial direction. We assume that the
energy En is uniformly distributed over the tube. Then,
the intensity In of an acoustic pulse1 arriving at the
receiving point along the nth ray tube is
400
u, m
VE cos n
E
I n = ---------n = -------------------------------.
S n
4rr n sin 'n
(8)
I .
(9)
(7)
I ( T ) dT
(10)
G( F) =
I ( T ) exp ( i2FT ) dT .
(11)
RESULTS OF CALCULATIONS
The Field in the Unperturbed Waveguide
Figure 3 shows the angular spectrum of rays
received in the absence of ISs. Note that the grazing
angles of the received rays coincide with the grazing
angles of the corresponding emitted rays, because both
source and receiver are located at the bottom. From
Fig. 3, we can see that the rays fall into two types: bottom rays and bottomsurface rays. The main portion of
the acoustic signal energy is carried by 12 bottom rays
1 Here,
Vol. 55
No. 1
2009
(12)
c
= arccos ----1 = 0.509 rad.
cb
n = 1, 2, ,
2c 1 tan n
2h 1
-,
D n = ------------ + --------------------
tan n
(14)
2h 1
2h 2
2
+ --- ( c 1 tan n c 2 tan n2 ), (15)
- + --------------D n = -----------tan n tan n2
where h1 and h2 are shown in Fig. 1a and tan n2 =
2
(16)
1
D t = --- ( c 1 tan n c 2 tan n2 ).
(17)
ACOUSTICAL PHYSICS
Vol. 55
No. 1
(18)
2009
0.3
0.4
0.5
Grazing angle, rad
(13)
because both source and receiver are located at the bottom. In particular, we have D12 = 833 m for bottom ray
no. 12. On the other hand, for the selected sound velocity profile, ray cycles of the bottom and bottomsurface
rays have respective lengths of
71
2h 1 c 1
2 2 2
- + --- c 2 c 1 = 793 m.
D cr = -----------------2
2
c2 c1
(19)
n , rad
Dn , m
Dt , m
1
2
3
4
5
6
7
8
9
10
11
12
0.011
0.021
0.032
0.043
0.055
0.068
0.081
0.095
0.112
0.131
0.155
0.192
5.6
5.6
5.6
5.7
5.7
5.8
5.9
6.1
6.4
7.0
8.1
11.7
= 79.2
10000
5000
3333
2500
2000
1667
1429
1250
1111
1000
909
833
16
32
48
65
82
100
120
142
167
196
232
289
72
GRIGOREV, KATSNELSON
0
Depth, m
20
40
12
60
80
5.0
5.4
5.8
6.2
Distance from the source, km
(20)
F* = v/Dcr , cph
F* = v/D12 , cph
Simulation F *madel , cph
3.6
3.45
3.4 0.3
4.7
4.5
4.4 0.7
are nearly sinusoidal, the Fourier spectrum should contain a single main peak at F*.
Figures 58 and Table 2 present the results of
numerical simulation for an IS present in the
waveguide. Ray trajectories were calculated for the
rays with grazing angles at the source within n = (0,
0.5) at a step of = 106104 rad over the time interval T = 1 h at a step of T = 1 min, i.e., for 60 successive positions of the IS along the path. Ray trajectories
before and after the passage of the IS were calculated
by analytical formulas (14)(17). Inside the IS, the trajectories were calculated numerically. From the whole
set of calculated rays, we selected adjacent pairs covering the receiver (boundaries of the ray tubes). The
intensities of these tubes were calculated by Eqs. (8)
(10). The calculations show (Fig. 5) that temporal variations of intensity reach 35 dB, which corresponds to
experimental data [13] on the order of magnitude.
Figure 6 shows the frequency spectra of intensity
fluctuations that were calculated by Eq. (11). In view of
the finite signal length in time and discretization, the
frequency region of the Fourier spectrum is limited to
the band 1/T0.5/T = 130 cph (we considered the
band 229 cps). The spectra are smoothed with the use
of a moving window with a width of 1.2 cph (this is the
resolution of the spectrum for the 1-h observation interval) and correspond to the motion of the IS from the
source to the receiver starting from the initial position
R0 = 5.42 m (position A in Fig. 4). For other values of
R0, the resulting spectra differ from the ones in Fig. 6,
because transform (11) is a window transform, i.e., G =
G(F, R0). Some of the spectra differ significantly. For
example, groups of two to three peaks may nearly
merge into one peak for certain values of R0. However,
the calculations show that, on the whole, Fig. 6 reflects
the typical pattern of intensity fluctuation spectra.
Figure 6 supports the above assumptions concerning
the shape of the spectra. Namely, one can observe a
number of nearly equidistant peaks (eight peaks in Fig. 6a
and six peaks in Fig. 6b), which, in our opinion, characterize the periodic nonsinusoidal behavior of intensity variations. The frequency of these periodic variations F *model can be determined from the simulated
results by averaging the intervals between the adjacent
peaks (Table 2). For inclinations of = 90 and 50, we
obtain F *model = 3.4 0.3 and 4.4 0.7 cph, respectively, where the errors are calculated with a confidence
probability of 0.9. As can be seen, the estimates
obtained with Eq. (20) can be considered satisfactory
(with allowance for the errors). Nevertheless, the mean
values of the confidence intervals are somewhat smaller
than the estimates. If we use the mean frequencies to
calculate the spatial period responsible for intensity
fluctuations, we obtain D = v/ F *model ~ 850 m, which
exceeds the cycle length of ray no. 12 (833 m). This
apparently means that the predominant role in intensity
fluctuations is played by a certain group of critical rays
ACOUSTICAL PHYSICS
Vol. 55
No. 1
2009
73
4
(a)
(b)
Intensity, dB
2
1
0
1
2
10
20
30
40
50
60 0
10
Time, min
20
30
40
50
60
Fig. 5. Temporal variations of intensity for = (a) 90 and (b) 50. The intensity in decibels is calculated with respect to the unperturbed intensity I0.
(a)
160
120
80
40
3.0
3.8
6.8
3.7
10.5
3.1
13.6
3.1
3.0
3.6
4.1
16.7 19.7
23.3
27.4
(b)
200
1.0
0.8
0.6
0.4
0.2
0
160
120
80
40
3.5
3.7
7.2
5.2
12.4
10
5.7
18.1
14
18
F, cph
4.1
22.2
22
4.4
26.6
26
Vol. 55
No. 1
2009
12
(a)
b
c
d
12
(b)
b
c
d
0.1
0.2
0.3
Grazing angle, rad
0.4
0.5
Fig. 7. Angular spectrum of the rays received in the presence of an IS (L = 95 m, the IS is in position B, see Fig. 4).
The angular spectrum is given for (a) the radiation angle n
and (b) the reception angle n' . The dashed lines show the
unperturbed spectrum. Letters a, b, c, and d denote identical
rays in both patterns.
74
GRIGOREV, KATSNELSON
Intensity, arb. units
1.0
0.8
11
10
9
8
0.6
0.4
0.2
0
1.0
0.8
0.6
0.4
0.2
0
12
(a)
12
(b)
11
10
8 9
CONCLUSIONS
11
0.1
0.2
0.3
Grazing angle, rad
0.4
0.5
Fig. 8. Angular spectrum of the rays received in the presence of an IS (L = 95 m) versus the average grazing angle
a = (n + 'n )/2. The soliton is located in positions (a) A
and (b) B (see Fig. 4).
Thus, the motion of an IS (a train of IS) approximately along the acoustic path causes the intensity of
the field at the receiver to undergo quasi-periodic variations. The predominant frequency of these variations
is proportional to the velocity of the IS motion along
the path and inversely proportional to the cycle length
of the critical ray. The critical ray (or, to be more accurate, the narrow group of critical rays) is characterized
by the following features:
they have a turning point near the upper boundary
of the thermocline,
they travel the longest distance in the thermocline,
they carry a major portion of signal energy.
These features determine the dominant contribution
of critical rays to intensity variations. As the IS moves
along the path, the spatial distribution of intensity varies mainly due to refraction of the critical rays with the
period of the IS passage through the turning points.
The simulation showed that the amplitude of intensity variations can be as high as 5 dB. The frequency
spectrum of variations is represented in the form of a
number of approximately equidistant peaks. The peakto-peak intervals are determined by the predominant
frequency, whose typical value under realistic conditions may be about 35 cycles per hour. A change in the
IS velocity along the path (this may take place because
of a change in the direction of the path) results in a proportional change in the peak-to-peak distance in the
spectrum.
We note an interesting consequence. Let us consider
two acoustic paths making an angle and possessing a
common source and different receivers and suppose
that a moving IS train covers both paths. Then, signal
variations at the receivers will have different spectra
and this difference will be determined by the angle
between the paths. This effect is experimentally measurable and can serve to verify the theory described
above.
ACOUSTICAL PHYSICS
Vol. 55
No. 1
2009
ACKNOWLEDGMENTS
This work was supported by the Russian Foundation
for Basic Research, project no. 06-05-64853.
REFERENCES
1. K. D. Sabinin and A. N. Serebryany, Akust. Zh. 53, 410
(2007) [Acoust. Phys. 53, 357 (2007)].
2. R. Oba and S. Finette, J. Acoust. Soc. Am. 111, 769
(2002).
3. M. Badiey, B. G. Katsnelson, J. F. Lynch, and S. Pereselkov, J. Acoust. Soc. Am. 122, 747 (2007).
4. J. Zhou and X. Zhang, J. Acoust. Soc. Am. 90, 2042
(1991).
5. K. V. Konyaev, D. E. Lekin, K. D. Sabinin, et al., Akust.
Zh. 44, 476 (1998) [Acoust. Phys. 44, 407 (1998)].
6. T. F. Duda, J. Acoust. Soc. Am. 119, 3717 (2006).
ACOUSTICAL PHYSICS
Vol. 55
No. 1
2009
75
Translated by A. Vinogradov