Low-Frequency Acoustic Attenuation in the South Pacific Ocean

13.2 Received 14 August 1970

Low-Frequency Acoustic Attenuation in the South Pacific Ocean

A. C. KIBBLEWHITE

Physics Department, University of Auckland, Private Bag, Auckland 1, New Zealand

R. N. I)ENHAM

Defence Scientific Establishment, Auckland 9, New Zealand

The first known measurement of the attenuation of sound in deep water at frequencies below 1 kHz in the South Pacific Ocean is described. The receiver was situated off the east coast of the North Island of New

Zealand. It was found that the attenuations are in most cases comparable with the values of attenuation measured in the North Pacific Ocean and the Atlantic Ocean between Bermuda and South America. How-

ever, on one of the paths over which the attenuation was measured, there was an abrupt change in the quality of sound transmission, which is thought to be a result of the path crossing the Subtropical Con- vergence; south of this discontinuity the attenuations were found to be comparable with those obtained previously in the Southern Ocean.

INTRODUCTION

Measurements of the attenuation of sound in the sea

at frequencies below 1 kHz have been carried out in several oceanic areas. •-5 Since the attenuation at low

frequencies is only a small fraction of a decibel per mile, a propagation path of some hundreds of miles in length is required to establish a measurable loss, and the effects of reflection and scattering from the ocean surface and bottom should be minimal to ensure that a reliable

measurement is obtained. To meet these requirements, long-range propagation in deep water in the SOFAR channel is used.

The experiment described here appears to be the first specific measurement of low-frequency attenuation within the SOFAR channel, carried out in the South Pacific Ocean. The bathymetric information available suggested that the paths PA and PB in Fig. 1 would be suitable, and an experiment was designed to test transmission along these paths.

I. EXPERIMENTAL DETAILS

The experiment described here took place in February 1967 when an Orion P3B aircraft of the Royal New Zealand Air Force Maritime Command flew along the paths PA and PB in Fig. 1, dropping explosive sound signals at regular intervals beyond 10 nautical miles (NM) from position "P." These signals were set to detonate at 800 ft.

165 ø E 180' W 165' 150'

o ß øo o

o •

135'

180' W 165ø 150' 135'

Fro. 1. Chart of area showing the paths PA and PB along which the attenuations were measured. The 1000- and 2000-f contours

have been traced from U.S. Naval Oceanographic Chart No. 1262A. P represents the receiving position.

810 Volume 49 Number 3 (Part 2) 1971

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 155.33.16.124 On: Sat, 29 Nov 2014 22:04:05

LOW-FREQUENCY ACOUSTIC ATTENUATION

FIG. 2. Temperature-depth (i) and bathymetric data (ii) obtained during the subsequent survey in March 1967 along (a) path PA and (b) path PB. In diagram (i) the arrows show positions where bathythermo- graph observations were made and the temperature contours are expressed in degrees centigrade.

0-1 [ 16'-'/

[-- •4X. •

•o.,-,% z -(i) I

1000 1200

DISTANCE FROM"P"IN MILES

o ...•,., '"1"' ' ' '"'/" •'/ ..... ,' ' ',•"•",' ' / - 18/ 16'"'

o_ 12

.1• ! _ 13.•

0-2

--Zo. a (i) I

200 400 600 800 1000

DISTANCE FROM""IN MILES

2-0

4-0

6eO --

The Journal of the Acoustical Society of America 811

Redistribution subject to ASA license or copyright; see http://acousticalsociety.org/content/terms. Download to IP: 155.33.16.124 On: Sat, 29 Nov 2014

22:04:05

KIBBLEWHITE AND•DENHAM

VELOCITY IN KM/SEC •.48 1'50 1.52 1'48 t 1.50 1'52

"'/; , , ?q o i

Fro. 3. Sound-velocity-depth profiles obtained during the survey along paths PA [profiles (a) and (b)] and PB [profiles (c)-(f)]. The positions are (a) lat 39ø04'S, long 177ø51'W, (b) lat 38ø48'S, long 169ø55'W, (c) lat 39ø25'S, long 179ø51'E, (d) lat 41ø14'S, long 172ø22'W, (e) lat 42ø04'S, long 168ø06'W, (f) lat 42ø25'S, long 163ø18'W.

For reception of the signals, a hydrophone was laid on the seabed at position P in 600 f (fathoms) of water. Power for the underwater electronics and signal leads were provided by a cable to the shore where the record- ing station was established. The signals from the hydrophone were recorded on an Ampex CP 100 tape recorder and were later analyzed to determine the energy flux received at the hydrophone in octave bands of between 19 Hz and 1.2 kHz. An analog squarer-integrator computer was used to integrate the energy flux over all the arrivals.

After the experiment, the laboratory's former research ship RNZFA TUI was sent along the paths PA and PB to gather hydrological and bathymetric data. The results of this survey are shown in Figs. 2 and 3, which show, respectively, the temperature-depth and bathymetric data along the paths PA and PB and examples of velocity-depth profiles obtained by in situ measurements with a velocimeter to a depth of about 1.5 km. (At depths below 1.5 km other available information shows that the velocity increases monotonically with depth from about 1.493 km/sec at 2 km to 1.522 km/sec at 4 kin.)

II. RESULTS

Figures 4 and 5 present the received energy, corrected for cylindrical spreading, as a function of range in

T^nL•z I. Attenuation coefficients (dB/Myd).

Frequency band (Hz) PA

,..

19-37.5 1.74-0.7 0.84-0.2 37.5-75 0.34-0.8 ....

75-150 0.64-0.9 0.24-0.4 150-300 2.74-1.1 2.94-0.5 300-600 7.64-1.2 7.1 4-0.8

PB1 PB2

(20-950 NM) (1000-1350 NM)

2.04-1.3

3.74-O.9 5,74-O.9 7.24-!.6

11.24-3.0

,.

a A negative value of attenuation is obtained in this case; for this reason, no value is quoted here and the corresponding regression line has been omitted in Fig. 5.

nautical miles for the paths PA and PB, respectively. Where possible the results were fitted, using the method of least squares, to a law of the form

E(r) = Eo- 10 logr--ar, (1)

where E(r) is the energy in decibels, r is the range, and a is the attenuation coefficient. The regression lines so obtained are shown in Figs. 4 and 5, and the resulting values of the attenuation coefficient, together with the standard errors involved, are presented in Table I. No attenuation values have been presented for the 600- to 1200-Hz band as the results in this case were rather

poor. The most reliable value of attenuation in this band is 25:1:4 dB/Myd, which was derived by fitting the energy values observed between 20 and 360 NM along the path PB to Eq. 1.

In the case of the path PA, the above analysis could be carried out sensibly only for signals dropped at

-4O

-80

O ß

'-J eeel e ß - + -

,,, -}20 (E)

_

-}6ol I I I I 0 400 800

RANGE IN MILES

Fro. 4. Energy levels, corrected for cylindrical spreading, obtained for the path PA. Conditions: (A) 19-37.5 Hz, (B) 37.5-75 Hz (--20 dB), (C) 75-150 Hz (--40 dB), (D) 150-300 Hz (--60 dB), (E) 300-600 Hz (--80 dB).




ranges of less than 620 NM from position P. Although ,0o hydrological factors cannot be ruled out, the most probable explanation for the erratic signal level beyond 620 NM is given by the bathymetric profile obtained during the subsequent survey. It may be seen in Fig. 2(a) that between 620 and 670 NM from P there is a large underwater feature which rises from a depth of • over 4.5 km to within 1.2 km of the surface. This ß

feature could be associated with the recently discovered • Eltanin fracture zone. • (See Fig. 1.) = lO

,I

Figure 5 shows that along the path PB there is a • sudden abrupt drop in signal level at a range of 950 NM •z from the hydrophone. At ranges beyond 1000 NM, • however, it is clear that the results could again be fitted [ to Eq. 1. However, as may be seen by comparing the • second and third columns of Table I, the values of o attenuation obtained are greater than those observed z for the interval 20-950 NM. Unfortunately the TUœ o survey could not be extended beyond 950 NM along ,• 1• the PB track, and it is thus not certain whether the • discontinuity in transmission is caused by another • topographic feature associated with the Eltanin ,• fracture zone or a hydrological feature. In this case, however, it is quite probable that at this range the sound path crosses the hydrological feature known as the Subtropical Convergence. 7 In an earlier propagation experiment which was known to cross the convergence, a similar drop in level, accompanied by other changes in the quality of transmission, was observed at the intersection of the propagation path and this feature. 8

When the results given in Ref. 8 were analyzed, the drop in level was puzzling as no topographic feature existed at this range, and yet, a thorough check of the

0.1

• PA

x PB1

+ PB2

ß Ref. 4

10 100 lO0O

FREQUENCY in Hz

FzG. 6. Attenuations as a function of frequency. The curves $, T, and U65 represent Eqs. 2-4; the Ref. 4 results represent the Southern Ocean and the bars labeled U66, the attenuations given by Ref. $.

--40

--80

+ --120

--160

ß e•e ß

i i i i I 1•o0 0 400 8oo

RANGE IN MILES

FIG. 5. Energy levels, corrected for cylindrical spreading, obtained for the path PB. Conditions: as in Fig. 4.

instrumentation indicated that this was a genuine effect. However, it now appears that, whenever a propagation path crosses a hydrological boundary such as the Subtropical Convergence, a marked drop in level can occur.

In Fig. 6, our values of attenuation are compared with other values quoted in the literature. The line S represents the empirical formula of Sheehy and Halley • for the North Pacific. This formula, with f expressed in kilohertz, is

•=33fi '• dB/Myd. (2)

The curve T represents an empirical formula due to ThorpO:

•=0.1f2(l+f•)-•+40f•(4100+f•) -• dB/kyd, (3)

where the frequency f is expressed in kilohertz, and the curve U63 represents the formula which Urick fitted to his 1963 resultsS:

o:= 1.5+8.2fkI•,. dB/Myd. (4)

The bars labeled U66 show the results that Urick ob-

tained in the North Atlantic during Part I of the experi-



KIBBLEWHITE AND DENHAM

men t known as Project Neptune. 5 In addition, we have plotted the results we obtained during Part III of Project Neptune using a receiving position on the southwest coast of New Zealand: These results may be said to represent SOFAR attenuation in the southern ocean.

An examination of Fig. 6 shows that the values of attenuation obtained along the path PA, and from 20 to 950 NM along the path PB, are comparable with those obtained by Urick in the North Atlantic in Project Neptune. Although the values obtained in the 35- to 75- and 75- to 150-Hz bands appear to be somewhat lower, Table I shows that the standard errors are such that the attenuations are probably not significantly different from those observed by Urick in Project Neptune at these frequencies. We have therefore com- bined our results obtained for the 19- to 37.5-, 150- to 300-, and 300- to 600-Hz bands with the attenuations obtained by Urick in Project Neptune, to produce the empirical formula

a=44f•F1--l.4+O.O20fm dB/Myd. (5)

It should be noted, furthermore, that the attenuations in the first and second columns of Table I are also comparable with the values obtained by Sheehy and Halley in the North Pacific.

On the other hand, as mentioned earlier, the attenuations obtained along the path PB from 1000 to 1350 NM are all somewhat greater than those applying between 20 and 950 NM along this path. In fact, it may be seen from Fig. 6 that these attenuations are comparable with those obtained in the Southern Ocean. 4 This may therefore be taken as further evidence that the portion of path PB beyond 950 NM from P does lie south of the Subtropical Convergence and within the Southern Ocean.

III. DISCUSSION

Several -- ' ' o 4 9--11 mechanisms •', have been invoked to explain the origin of the deep-water attenuation observed at frequencies below 1 kHz. The values obtained are always an order of magnitude higher than the attenuations predicted from the relaxation absorption caused by the dissociation of magnesium sulphate ions in seawater, as given by the second term of Eq. 3. One mechanism that has been discussed previously by Urick 2 and by the authors 4,8 is scattering caused by temperature and salinity inhomogeneities existing below the thermocline. There is now much evidence for the existence of turbulent fluctuations in temperature in and below the permanent thermocline. 12-15 It is found that the turbulence is highly intermittent, with the turbulent patches being flattened by the stable density gradient into horizontal disks or ribbons (b/ini). 16 The nature of the low-frequency attenuation curve thus probably arises from two mechanisms.

At very low frequencies the attenuation is likely to become high because the SOFAR modes become leaky: The cutoff frequency for the first mode is typically about 2 Hz, and at 10 Hz there will usually be only two or three modes trapped in the SOFAR channel. As the frequency increases, more modes will become trapped and the attenuation will tend to decrease. On the other hand, the attenuation resulting from the turbulent inhomogeneities will tend to increase with frequency. There should thus be a minimum of attenuation at a frequency that is governed by the character of the turbulence and the geometry of the situation.

In the case of the South Pacific Ocean, such a minimum certainly appears to exist and occurs at around 50 Hz. (See Table I and Fig. 6.) It may also be seen from Fig. 6 that Urick's Neptune attenuations exhibit a minimum at the same frequency, so that this effect appears to be well established in at least two oceanic areas.

If now the values of attenuation are examined closely, it is clear that those obtained by Urick in Project Neptune are comparable to those recorded in the South Pacific Ocean. It would appear, therefore, that the magnitude and spectrum of the turbulence in the two cases, that is, in the South Pacific Ocean and along the North Atlantic path between Bermuda and South America, are similar. It is apparent from Fig. 6, however, that no minimum exists in Urick's 1963 results, although the data were obtained from an experiment also carried out in the North Atlantic. In this case, however, the transmission path lay to the northeast of Bermuda, and the difference in the attenuation-versus- frequency curve presumably reflects the fact that water masses were encountered which were different in

character from those existing along the path south- eastward from Bermuda.

These results also indicate that the characteristics of the turbulence in the South Pacific Ocean are different from those which appeared to be present in the Southern Ocean during Project Neptune. 4 It has been stated previously 4,6 that this contrast can be associated with the different characteristics of the water masses north and south of the Subtropical Convergence. This difference in water masses appears to cause an abrupt change in the quality of sound transmission similar to that observed at 950 NM along the path PB.

This analysis emphasizes the need to specify the oceanic areas in which measurements are made before attempting to compare deep ocean attenuation values.

Comment has already been made on the fact that our attenuations above 150 Hz are comparable with the values reported by Sheehy and Halley for the North Pacific, rather than those predicted by Thorp's empirical formula • (Eq. 3). Inadequate as the data are at around 1 kHz (it will be remembered that our value in the 600- to 1200-Hz band is 25 dB/Myd), they do not seem to support Thorp's proposal of a low-frequency relaxation mechanism. On the other hand, the Southern Ocean




data suggest a relaxation mechanism at around 100-200 Hz, but additional evidence is required to confirm this. It was unfortunate that no Neptune data existed above 100 Hz.

IV. CONCLUSIONS

The values of attenuation obtained over the path PA and from 20 to 950 NM along the path PB are comparable with the attenuations measured in the North Pacific and the Atlantic Ocean between Bermuda and

South America, and indicate the existence of a minimum around 50 Hz. The implication is that the magnitude and spectrum of the turbulence of the water masses of the South Pacific, North Pacific, and North Atlantic between Bermuda and South America are similar if the

attenuation at low frequencies is caused by scattering from temperature inhomogeneities. The attenuations measured in the South Pacific, on the other hand, are much lower than those obtained in the southern ocean, a consequence no doubt of the different water masses existing north and south of the Subtropical Con- vergence. Finally, it appears that the path PB crosses the Subtropical Convergence at about 950 NM from "P," causing the abrupt change in the quality of transmission observed at this point; the values of attenuation observed between 1000 and 1350 NM

along the path PB provide additional evidence that this segment of the path lies within the Southern Ocean, and emphasize the dependence of low-frequency attenuation on the oceanic area. The change in transmission characteristics observed here implies that a suitably designed propagation experiment can be used as a sensitive method of identification of hydrological boundaries such as the Subtropical Convergence.

ACKNOWLEDGMENTS

This paper is published by permission of the Ministry of Defence (Navy). The authors wish to acknowledge the assistance of J. W. F. Mason, F. G. Crook, A. E. Dunn, A. S. Gannaway, A. Howard-Taylor, and M.D. Wilson of the laboratory staff with the experimental measurements and analysis, and the authors are grateful to the Royal New Zealand Air Force Maritime Com- mand, whose willing cooperation contributed greatly to the success of the experiment.

1 M. J. Sheehy and R. Halley, J. Acoust. Soc. Amer. 29, 464-469 (1957).

'"R. J. Urick, J. Acoust. Soc. Amer. 35, 1413-1422 (1963). a W. H. Thorp, J. Acoust. Soc. Amer. 38, 648-654 (1965). 4 A. C. Kibblewhite, R. N. Denham, and P. H. Barker, J.

Acoust. Soc. Amer. 38, 629-643 (1965). 5 R. J. Urick, J. Acoust. Soc. Arner. 39, 904-906 (1966). 0 B.C. Heezen and M. Tharp, "Pacific Ocean Floor," map

published by the Nat. Geograph. Soc., Washington, D.C. (1969); appeared in Nat. Geograph. 136, No. 4 (Oct. 1969).

7 G. E. R. Deacon, in The Sea, M. N. Hill, Ed. (Wiley, New York, 1963), Vol. 2, Chap. 12.

8 A. C. Kibblewhite and R. N. Denham, J. Acoust. Soc. Amer. 41, 401-411 (1967).

9 W. H. Thorp, J. Acoust. Soc. Arner. 42, 270(L) (1967). l0 M. Schulkin, J. Acoust. Soc. Amer. 35, 253-254(L) (1963). n C. B. Brown and S. J. Raft, J. Acoust. Soc. Arner. 35, 2007-

2009 (1963). • C. F. Black and P.M. Gluckman, "Large Scale Structure of

Turbulence beneath the Mixed Layer," in Ocean Science and Ocean Engineering (Marine Technology Society/American Society of Limnology and Oceanography, Washington, D.C., 1965), Vol. 2, pp. 687-699.

•a J. R. Lovett, Limnol. Oceanog. 13, 127-142 (1968). •4 H. L. Grant, A. Moilliet, and W. M. Vogel, J. Fluid Mech.

34, 443-448 (1968). x5 I. Orlanski and K. Bryan, J. Geophys. Res. 74, 6975-6983

(1969). x00. M. Phillips, The Dynamics of the Upper Ocean (Cambridge

U. P., New York, 1966), Chap. 6, p. 199.



Documents

Low-Frequency Acoustic Attenuation in the South Pacific Ocean