Reflections of an acoustician

8 Pages·2002·2.2 MB·English

by  A C Kibblewhite

RF.Fl ECnONS Op AN ACOUSTICIAN
At KIBB1 EWHITE
Kabblevvluie. AC. 2001 1231: Reflections ofan acoustician. Memoirs oflhz Queewtktnd
Museum 41{2): 491-49$ Brisbane. ISSN 007V-H.S35.
fn tin- early lc>50s a fixed hydrophone airay was .set up offthe coasl ol New Zealand. The
local ocean environment is characterised by a complex OCeaaOgraphic structure, but thcrC
.'.*s. confident expectation that tliis installation would help provide an understandingofthe
phenotftcttu involvedintlu-propagationonovv-frecjuencvsound in localconditions Someof
iik SUTPT1SC8 experienced in this evniuiiiuhoii oi the acoustic properties ol the ocean,
including a report ofthe first probable acoustic encounter with the humpback vrflfljc in the
Southern Hemisphere, are d< crU : 01 thus personal account. G New /.calami marine
envit'Qttitl&M, titoustlcpt'opertteSy hiempbaci whafasang,
I ( ; Kihhlt'wlatt', Department O) PtmtCS, UntwrvUy ol Auckland, Private Btlfr lO(//'>
Auckland, New Zealand 28 MayTOOl
The submnnne almost won World War II for branch ol naval science into the dflft
Germany (Rugc, 1957).Thattheallieseventually Organisations Ol many countries, including
achieved superiority in the Battle ofthe Atlantic Australia and New Zealand. As with main
was due largely to the development ofradar and branches Ol oceanography, (he propagation oi
its joint deployment with sonar (Kemp. 1957). sound in the sea and its noise have been oi im-
Both systems involve the transmission of a portanceto whale.studies Significantcontributions
directional pulse Ofenergy (electromagnetic in have arisen from manyactivities, including those
diecaseofradarandacoustic in thatofsonar)and in Australia (Cato, 1991; Hunter, 10%). Ibis
thedetectionofanyechoreturnedbyan illuminated account deals with someoftheearly experiences
target. ill New Zealand.
On typical anti-submarine warfare vessel PROPAGATION STUDIES
the time, the sonar operated ea. l0-20kl I/., used
transducers of about lm in diameter, and LOTION OF OCEANOGRAPHJC FRON
achieved detection ranges ofaboul 10km li w;i-. I1)n1aIn experiment for evaluating the propagation
well known at the time that the attenuation of propertiesOi theocean,ahydrophoneisdeployed
soundintheoceandecreasedalmost linearly wilh from a receiving ship while another ship (or an
the frequency of the sound used and that aircraft) drops sound signals as it opens range
improved performance should be achieved at Suchaprocedurewasusedtoestablishthe feetors
wlaowserredfurceqeudentcoiaecsh.ieHvoewaevleorng,erifdettheectfiorneqruaenngcey, ctionnetnrtoallliwnagteprrsopaangdatitooninivnesNteigwateZeaalnaynds'esas|onnal
the transducer size had to increase to achieve variation that occurred. On otheroccasions New
comparable directionality; so that engineering Zealand's unique position m the SouthernOcean
problems severely constrained the use of lower allowed examination of longeroceanic paths
frequencies in ship-home sonars. One ofthe firstofthese experiments involved
In the post-war period interest switched from propagation across the Tasman Sea(Kibblevvhite
"active' to "passive' systems, in which the aim & Denham, [967), An aircraft dropped
was to detect and locate the self-noise generated explosives as it llew towards Australia and the
by an underwater source (like a submarine), signals were received on a hydrophone laid off
rather than au echo produced by reflected the southern fiords of west New Zealand
radiation. Such an approach could potentially Amongst other information, the experiment
exploitthe thousand-folddecrease in attenuation dcmonsiiatcd for the first time that oceano
11iiit results when the acoustic frequency of graphic fronts, likethe SubtropicalConvergence,
interest isreduced from 10.0001 1/ to 100Hz* But could beidentifiedandpositionedacousticallyby
any such changecalled fora betterunderstanding the change in transmission characteristics
o\' the factors influencing long-range trans- observed fts the source moved Ifom one water
mission in theoceanand itsambientnoise. These mass to another. The trial also demonstrated, in a
requirements led to the introduction of a new very effective wav. the difference between the
492 MEMOIRS OF THE QUEENSLAND MUSEUM
' FLIGHTTRACK \ \ _^^
FIG. 1. ProjectNeptune, Southern Oceantransmission paths.
speed of sound in water and that of electro- confirmation ofthe remarkable effectiveness of
magnetic radiation in the atmosphere. After the the deep ocean in transmitting low-frequency
pilot's last 'bomb's away' message on the radio sound.
there was adelayofnearly20 minutesbefore the
soundofthedetonationreachedusinNewZealand, CHASE V. In the 1960s the difficulty of dis-
approximately 1,800km away. Other acoustic tinguishingbetweenanatural seismiceventanda
experimentshavesubsequentlydemonstratedthe covert nuclear explosion presented a serious
presence of similar fronts in other oceans impediment to the negotiation of a nuclear
(Kibblewhite & Browning, 1978). test-ban treaty. The Chase V experiment was
PROJECT NEPTUNE. Transmission paths of designed to help resolve these difficulties by
monitoring the seismic and acoustic signals
even greater length were involved in Project
Neptune. In this experimentan aircraftofthe US producedbyalargeunderwaterexplosion.Inthis
projectanoldLibertyship containing ,000 tons
Airforce dropped sound signals along the two of TNT was scuttled at a specific lo1cation off
legs ofa flightpath from Bermuda to Capetown Cape Mendocino and setto explode atthe depth
raensdultCiangpeatcoouwsntictosiPgenratlhs. wReerceormdaidnegsinofNtehwe of the SOFAR channel axis (Fig. 2). Our
participation involved monitoring any signals
Z(Keiablbalnedwhidtuer,ingDetnhheamsec&ondBarlkeegr,of19t6h5e).flTighhet whichmightreachNewZealand(Kibblewhite&
Denham, 1969).
transmission path for the most distant of the
signals received was approximately 10,000km, It transpiredthat, not only was a direct arrival
nearly one quarter ofthe Earth's circumference. observed, but another hour went by before
In spite of the distances involved, the results reverberations from topography throughout the
demonstrated that a propagation path existed Pacific Ocean gradually died away. The first
through the Southern Ocean, provided the great arrival, which took 90 minutes to cross the
circlepathbetweensourceandreceiverwasclear Pacific, confirmed that a direct water path from
of intervening topography (Fig. 1). In all cases Cape Mendocino to New Zealand existed, in
thesignalcharacterwasconsistentwiththetravel spite of the extensive intervening topography.
pathpassingthroughatleastthreedifferentwater The length of the transmission path was again
masses. The results provided, however, further nearly one quarter ofthe Earth's circumference.
REFLECTIONS OF AN ACOUSTICIAN 493
>r.l.njiueyIsland
ftirtls.Mdlld
I1:'»)ll!l.1111I-UK
REFLECTED Silent SUK""
SOUTIIKFRMADECRinijF
vt Fiji ,,Samoa. Rumblein SEAMOUNIS
/
/ Cook.-.."V.
Is."..'Tahiti
; EasterIs
PitcaimIs
FIG. 2. CHASEV, transmission paths.
Six major echoes in the reverberation following
themainarrivalwereparticularly striking. These
signals were subsequently identified with
reflections from Henderson andPitcaim Islands,
and large underwatermountains onthe Tuamotu FIG. 3. Great Barrier Island, the Taupo-White Island-
Raoul Islandvolcanic zoneand the South Kermadec
Ridge which were undiscovered at the time
(Kibblewhite & Denham, 1971 Ridge seamounts.
).
This evidence of a direct acoustic path from purchaseofland,constructionofalaboratoryand
NorthAmericatoNewZealand,togetherwiththe living quarters on an isolated island, a training
evidence from Project Neptune, pointed to the coursetoacclimatisediversforworkinthewater
possibility that a deep-water path might exist depths involved and even the purchase ofModel
from the west to the east coasts of the North 3 of the first television camera employed in
Americanmainland, adistance ofnearlyhalfthe underwater exploration (which was built to
Earth's circumference. support the search for the Comet aircraft that
AMBIENT SEANOISE crashed in the Mediterranean about that time).
Further, the deployment was not completed
HUMPBACK WHALE SONG. Whatever the without drama. There were ship strikes, cables
became wrapped around ship's screws and near
propagation conditions encountered (and it is
apparent thatthey can be very favourable at low groundingsresulted,diverscollapsedunderwater
frequencies), the successful detection and locat- and the hours at seawere long and exacting.
ion ofan acoustic source by passive techniques However, the time finally came to connect the
willdependultimatelyontherelativestrengthsof underwater array to the shore electronics, albeit
the signal generated by the source and the local in temporary accommodation. A few ofus were
sea noise atthe listening site. Ambient sea noise working late on a coldnight to achieve this. We
is thus as critical as propagationin the successful could well have waited till the morning but
application of underwater sound. Because a curiosity drove us on. We did not know exactly
properunderstandingofanygeophysicalprocess whattoexpectbutweresmuglyconfidentthatwe
requires long-term study, it was decided to base would be listening in to one of the quietest
investigations of this parameter on a fixed, environments in the ocean. Instead we were
semi-permanenthydrophoneinstallation,located 'blasted' by some biological community in full
offGreat Barrier Island (Fig. 3). song. Our expectations of the hoped for 'silent
Preparation for the installation ofthis facility sea'were certainly not fulfilled.
extended overayear. A large sum ofmoneywas
committedtothepurchaseofarmouredcable,the As it turned out we were unknowingly eaves-
commissioning ofa cable ship to lay the cables, dropping on migrating humpback whales. This
construction of large underwater tripods, activity, dubbed by the staff as the 'Barnyard
494 MEMOIRS OF THE QUEENSLAND MUSEUM
20-second Pulse. The '20-second pulse' was so
named by virtue of its pulse length. The
spectrogram (line a) at the top ofFig. 5 and the
higher speed presentation at the bottom of this
figure (line f) show the pulse as a function of
frequency and time. The 1/3 octave analysis
given in lines (b) and (c) show the pulse is an
almost pure tone of about 23Hz. Record (d)
indicates the regularity and spacing of a pulse
sequence. (The 1/3 octave resolution of these
early spectra was the best achievable before
computers gave us access to modem spectral
m m analysis.)
m
fci 5-secondPulse.The '5-secondpulse'wassimilar
in characterbut shorter in length and not so pure
in tone. The dominant frequency was again
around 25Hz. Pulses of similar type were also
observedatothersites around New Zealand, and
MOHNtNOCMORUS a call-answer sequence was often observed. The
signals were attributed to whales but no
particular species was identified. Even so these
observationsoflong, low-frequencypulsesinthe
South Pacific werereassuringtotheacousticians
in the Northern Hemisphere as they confirmed
thatthepulseswereaworldwidephenomenaand
not necessarily a new weapon of the cold war.
Examples are given in figs 4, 5 and 6 ofKibble-
white et al. (1967).
FIG. 4. Humpback whale song recorded off Great Significance ofthe 25Hz Band. The 25Hz band,
Barrier Island 1960.
in which many whale calls operate, coincides
with thatatwhichthe attenuationofsound in the
Chorus', couldpersistfordays at atime (Fig. 4). deep ocean is a minimum (Kibblewhite &
It was most marked between June to December, Hampton, 1980). Given optimum propagation
during the annual migration. Incidence declined conditions, whale calls of this type could
to low levels after 1961, no doubt due to the thereforebe expectedto travel longdistances. In
impact of a local whaling operation (Dawbin, ourearlyexperiments weestimatedsourcelevels
1967). ofthe above pulses were about 50Pa/Hz at one
metre. Usingexpectedspreadingandattenuation
Two features of this early encounter are rates for deep water, we then estimated that
noteworthy. First, the apparent disregard of the measurable signals might prevail at distances of
whales to the harpoon thatwas routinely fired in several thousand kilometres. Speculative as this
their midst during this period (Fig. 4, 2 August was at the time, the US Navy has recently
1960) and second, the activity encountered was reported tracking a particular whale for several
probably the first acoustic experience ofhump- weeks as it moved alongthe European coast and
back whale song in the Southern Hemisphere that they did this from a SOSSUS station on the
(Kibblewhite etal., 1967). other side ofthe Atlantic. Ifwhale hearing is as
good as ours, they have a remarkable com-
OTHER BIOLOGICAL SOURCES. Signals munication system.
from other biological sources were observed, Further evidence ofthe long distances whale
although less frequently, at several sites around signals can travel is providedbythe comparison
New Zealand. These signals were too low in ofthe spectra ofFig. 6, which were recorded in
frequency to be recorded satisfactorily by the theNorthandSouthPacificOceans(Kibblewhite
tape recorders then available. However typical etal., 1976).Thenorthernspectrumwasrecorded
features could be established (Kibblewhite etah, by a hydrophone moored in deep water in the
1967). middleofthe Pacific, with excellentpropagation
REFLECTIONS OF AN ACOUSTICIAN 495
6 Seconds
B
o PULSE A
* PULSE B
PULSE C
%
1 -30
:>
20 40 60 30 100
Frequency als 13 20 21 22 23 24 25 2G 27 2S 29 30 31
THIRD OCTAVE SPECTRA SPECTRUM OF PULSE B 0.4C/s BANFDreVqUueDnTcHy c/s
^
uuuimiai /f~\\ V^h. *iHfJ\^.^u.rtjA^.__*-^s-Vw..-
>-'
5 Minutes
PULSE SEQUENCE 30 Seconds
THIRDOCTAVEENVELOPES
m
w;:x:wMmm$wMi:^l:$W^ mm
VMNWaW^V
\sf/^
1 Second
V1SIC0RDER RECORD OF PULSE B
FIG.5. '20-second'pulse.A,sonogramsoftwopulses.B,C, 1/3octavespectra.D,pulsesequences.E, 1/3octave
envelopes, pulse B. F, visicorderrecord, pulseB.
occurringinalldirections. The southernspectrum Measurements at numerous sites in the region
was recorded in coastal water. The influence of confirmedasimilardiurnalactivityandshowedit
whales and shipping intheNorth Pacific is clear was influenced by the level of solar intensity,
in the spectral difference. both daily and seasonal. Investigations revealed
thatthe commonseaegg(Evechinuschloroticus),
EVENING CHORUS. The 'Barnyard Chorus' which is endemic around New Zealand, was the
was not the only component ofthe sea noise to most likely source and that the spectral peak
dispelthe conceptthatthe seawas an essentially observed was relatedto the acoustic response of
quiet environment. Experience quickly revealed the animal during feeding. A comprehensive
thata significantincrease innoise level occurred reviewofthiswork,whichincludesacomparison
twice each day. The largerincrease occurredjust with similar biological choruses observed by
around sunset and became known as the Cato and others in Australian waters was
'Evening Chorus', the smaller increase just compiled by Denham in 1994.
before sunrise as the 'Morning Chorus'. The
spectral peak in each case was around 1.2kHz VOLCANISM.Intheearlydaysofoperation,the
(Fig- 7). sea-noise spectrum offGreat Barrier Island was
496 MEMOIRS OF THE QUEENSLAND MUSEUM
AMBIENTNOISFSPECTRA INFLUENCE OF SEA STATE ON AMBIENT
'T
I 1 NOISE
/0*9*00-0-0-0^ O• SNOOURTTHHPPAACCIIFFIICC
80- \ i- SEA NOISE DURING HIGH AND LOW SEA
STATES. Whilemanysourcescontributetonoise
inthesea, the local sea stateusuallyprovidesthe
70
majorinput.AsthespectraofFig. 8demonstrate,
sea surface agitation and wind can significantly
60 modify the local spectrum between 12Hz and
20 110000 220000 1000
1200Hz. The minimum levels shown in these
DIFFERENCE>BETWEENNORT\HANDSOUTHPACIFICSFtCTRA NewZealand spectraare lowbyworld standards
20
r and reflect the country's isolation. While oi~
WHALES1 IFPIN interest in themselves, such levels posed the
\ question as to how low the sea noise would
become, if the surface-generated component
» couldbeeliminatedcompletely. Thecalmwaters
50 100 200 500 1000 underthe ice-sheetatMcMurdo Soundappeared
FREQUENCY.Hi to offer the ideal environment in which to make
&
FIG. 6. Comparison ofambient noise spectra in the this assessment (Kibblewhite Jones, 1976).
North and South Pacific Ocean. Anexploratoryprogrammewasspentadopting
instrumentation to withstand the rigours of the
at times also distorted by an intermittent low- climate and in a second season of operation
frequency component. We came to realise that successful recordings were made. As a further
thisdistortionwastheresultofagenuineacoustic exampleofthemisconceptionscharacterisingthe
source, but in this case one ofgeophysical rather studyofthis subject, weencounteredan acoustic
than biological origin, andultimately traced itto environment completely different from the one
underwater volcanism only 300km from expected. The most immediately apparent
Auckland (Kibblewhite, 1966). The active feature was the high, and almost continuous,
level ofthe biological activity. The chorus ofthe
volcano lies on a linejoiningthe active volcanic local seal community was the dominant feature
centre in the middle of the North Island with but signals produced by cracking ice and
correspondingactivityaround Raoul Island(Fig. humpback whales (perhaps up to 30km away)
3).Abetterpictureofthe levelofactivitybecame were also recognisable. Activity ofthis intensity
apparent when transistor technology allowed an persisted throughout the two weeks ofoperation
improvementofthestation'scapabilityin 1963. and it proved difficult to identify any 'quiet'
06 09
N Z TIME
FIG. 7. Spectral contours ofthe ambientnoise level showing the rise in activity at sunset and sunrise.
r
REFLECTIONS OF AN ACOUSTICIAN 497
10ij i i i | ' 11 i r i | M i 1 | I1 1- M
„£^*7800"--^** ^ +" C*B 111544000000DDDEEECCCEEEMMMBBBEEERRR897,,,WKWIIINNNDDD0ZZ,EE5RRr,OO,ML "
High Sea Stale
80 S Slfc. "*^..
X -^>jL '""- SE*STATE-T
-\
Low Sea Stat,.e-' j*50-. ^C*S^ , <><^^"s*"—^£!N^>s--BI^OLOGICALCOMPONENT
S;40 ^^*^$^>^--
60 SeLF^NOISE -_
30, 1 1 1 1 1 1 1 .1*W«—*Mi-J-J
FREQUENCY-<H.
,1,1 FIG. 9. Sea-statespectraatlowwindspeeds underthe
I'll 1 1 .1 t ( ' l1 I I McMurdo Sound ice sheet.
FREQUENCr-kHi
FIG. 8. Typical sea-noise spectra in high and low Investigations by acousticians, in many parts of
sea-state conditions. theworld,havenow shownthatthehigh spectral
levels below 10Hz arise from a particular
periods at all. An uncontaminated sea-noise interaction ofocean surface-waves and are thus
spectrum at low wind speed was however ob- highly wind dependent (Kibblewhite & Ewans,
tained. Spectrallevelsdidindeedliesubstantially 1985; Kibblewhite & Wu, 1996).
below those normally associated with sea-state Typical sea-noise spectra between 0.1-3kHz
zero (Fig. 9). andwindspeeds5-30ms_1 incorporatingthislow
,
INFRASONIC SEA NOISE. Our initial frequency component, are shown in Fig. 10.
These infrasonic acoustic components are
experience of sea-noise at frequencies below
m1OaHdzeiotcpcousrsribeldewthoeinncotrrpaonsriastteorantaemcphlnioflioegryinfitrhset hroebosswpceouvnrseeisrb,rleecaofrroedrinmlgiescaorrfonssieeniigssmmotloaocgtieisxvtpislt.yoAiwctohuistcthhiecoimfatnestn,o
fhiorustsienxgpeorfimtehentunadceurtw-aotfefrfrheyqdureonpchyonoef.1FHozrwtahse advantage. We have recently used them to
measure the spectral properties of an offshore
selected.
wave climate, using seismic sensors on land
Deploymentofthis systemwentverywell and
aslelvewreerestaodrmmirbilnegwitusp.quTaloitoyurwhdeinsmaaymosdeear-antoeilsye rwaatvhee-rritdhaenrsthaetmsoera.e eItxpheanssiavlesoanbdeveunlnserhaobwlne
theoretically, that certain properties of the
levels rose al—armingly (in spite ofthe depth of
deployment approx. 250m) and eventually sea-noiseinthispartofthespectrumareuniquely
overloaded the underwater electronics. Tests suitable for the location of oil and gas bearing
showed that nearly all ofthis
unexpected noise occurred in Etf'Av'ti:.--
thenewbandbelow 1OHz.The
hydrophone system was
recovered and the low
frequency response of the
underwater amplifieradjusted,
justintimetorecordthesignal
from theCHASE Vexplosion g
|
described earlier. The ex- ?
periencewasagaincompletely s
unexpected. I
Some ofthe features ofthis
infrasonic component of the
sea-noise spectrum were
established at the time but a
full understanding had to
await the technological
developments which were to
appear some 20 years later. FIG. 10. Typical sea-noise spectra(0.1-3kHz) forwind speeds 5-30ms"
498 MEMOIRS OF THE QUEENSLAND MUSEUM
structuresontheworld's continentalshelves. This by long-range acoustic propagation experiments.
component may therefore become critical in the Deep SeaResearch 25: 1107-1118.
search forfutureenergyreserves (Kibblewhite& KIBBLEWHITE,A.C.&DENHAM,R.N. 1967.Long
Wu, 1996). range propagation in the South Tasman Sea.
Journal oftheAcoustical Society ofAmerica41:
DISCUSSION 401-411.
1969. Hydroacoustic signals from the CHASE V
It is clear that the 'Silent Sea' is anything but explosion. Journal ofthe Acoustical Society of
quiet, and instead displays the output of many America45: 944-956.
contributing acoustic sources. Some of these 1971. The CHASE V explosion - submarine
result from environmental factors; others from topographic reflections from the vicinity of
biological and geophysical activity, while others PitcaimIsland. DeepSeaResearch 18:905-911.
are man-made. Perhaps the most intriguing, KIBBLEWHITE,A.C,DENHAM,R.N.&BARKER.
however,arethoseproducedbycetaceans. These P.H. 1965. Long-range sound propagation in the
signals are ofparticular interest given that some Southern Ocean. Journal of the Acoustical
display characteristics of language (Noad et al., SocietyofAmerica38: 629-643.
2000),thedevelopmentofwhichhasbeenoneof KIBBLEWHITE,A.C,DENHAM, R.N.&BARNES,
the major achievements ofthe human race. The D.J. 1967. Unusual low-frequency signals
sounds that we recorded in the 1950s and early observed in New Zealand waters. Journal ofthe
1960s are now an important feature ofmodern KIBBALcoEuWstHiIcaTlES,ociAe.tCy.of&AmeErWiAcaNS41,: 6K44.-C655.1985.
research following the recognition of Northern
Wave-wave interactions, microseisms and
Hemisphere humpback whale song in the early
1970s (Payne & McVay, 1971) andagaininNew Aincroruasstoinciacl SnoociiseetyinoftAhemeorciecaan.78:Jo9u8r1n-a9l94o.f the
Zealand waters by Helweg et al. (1998). KIBBLEWHITE, A.C. & HAMPTON, L.D. 1980. A
LITERATURE CITED review ofdeep-ocean sound attenuation data at
very low frequencies. Journal ofthe Acoustical
CATO, D.H. 1991. Songs of humpback whales: the SocietyofAmerica 67: 147-157.
Australian perspective. Memoirs ofthe Queens- KIBBLEWHITE,A.C.&JONES,D.A. 1976.Ambient
landMuseum 30(2): 277-290. noise under Antarctic sea ice. Journal of the
DAWBIN,W.H. 1967.WhalinginNewZealandwaters Acoustical Society ofAmerica59: 790-798.
1791-1963.In,AnencyclopaediaofNewZealand. KIBBLEWHITE, A.C, SHOOTER, J.A. &
(Government Printer: Wellington). WATKINS, S.L. 1976. Examination of
DENHaAroMu,ndR.NN.ew199Z4e.alMaanrdi.neDebifoelnocgiecalScicehnotriufsiecs attenuation at very low frequencies using the
deep-water ambient noise field. Journal of the
EstablishmentReportGTP-9,VI.
HELWGEAGR,RIDG.UAE.,, CC.AT&O,McD.CHA.U,LEJEYN,KIRN.DS., 1P9.9F8.., KIBBALcEouWsHtiIcTalE,SocAie.tCyo&fAmWeUr,icaCY60.: 11094906-.10W47a.ve
Geographic variation in south Pacific humpback interactions as aseismo-acoustic source. Lecture
whale songs. Behaviour 135: 1-27. Notes in Earth Science 59. (Springer-Verlag:
HUNTER, W.F. 1996. The development ofthe RAN Berlin, Heidelberg).
ResearchLaboratory, DSTO-GD-0078. NOAD, M.J., CATO, D.H., BRYDEN, M.M.,
KEMP, P.K. 1957. Victory at sea 1939-1945. JENNER, M.N. & JENNER, K.CS. 2000.
(ShakespeareHead: London). Cultural revolution in whale songs. Nature 408:
KIBBLEWHITE, A.C. 1966. The acoustic detection 537.
and location of an underwater volcano. New PAYNE,R.S. &McVAY, S. 1971. Songs ofhumpback
ZealandJournal ofScience 9: 178-199. whales. Science 173: 585-597.
KIBBLEWHITE, A.C. & BROWNING D.G. 1978. RUGE, F. 1957. Sea warfare 1939-1945. A German
The identification ofmajoroceanographic fronts viewpoint. (Cassell & Co.: London).