GEOPHYSICAL
RESEARCH
LETTERS,VOL. 23,NO. 5, PAGES523-526,MARCH 1, 1996
Large amplitude solar modulation cycles of 0Be in
Antarctica: implications for atmospheric mixing
processesand interpretation of the ice core record.
E. J.Steig,
1,2P. J.Polissar,
1,3M. Stuiver,
1P M. Grootes,
TMandR. C. Finkel
5
Abstract. 10Be concentrations in an ice core at Taylor
Dome, Antarcticashow greatervariationover the last 75 years
in the low-latitude stratosphere. If
so, then a strong
geomagnetic
signalis expectedin the ice-core10Berecord.
Mazaud et al. [1994] approachedthis questionby optimizing
the fit betweengeomagneticfield variationsestimatedfrom
riodicity which follows expectedchangesin the production ocean sediment cores [Tric et al., 1992], and 10B e
rate of 10Beover the l 1-year solarcycle. Noting that the am- concentrationsin the Vostok ice core [Raisbeck et al., 1987]
plitude of production-ratevariationis both latitude and altitude over the last 150 ka. In this paper,we take a similar approach,
than similar
10Be time-series
from
Greenland.
Like
the
Greenland records, the new Antarctic data exhibit a strongpe-
dependent,
we estimatethe relativecontribution
of 10Befrom
but comparethe Antarctic10Beflux with estimatesof solar
different atmosphericreservoirs.The calculationsyield a relatively small (<35 %) contributionfrom the low-to-mid latitude
stratosphere,suggestinga weak coupling between Antarctic
rather than geomagneticvariability.
10Beand geomagnetic
field intensity.
Relationship
amongsolarvariability,•øBe
concentrationsand atmosphericcirculation
Our goalwasto obtainan estimate
of the amplitude
of the
-l 1-yearsolarmodulationcycle of 10Be in Antarctica.
Introduction
Samples
werecollected
froma firn coreat 2 to 3-yearresolution and in a 4-meterdeep snowpitat -0.5 year resolutionat
Modulation of the cosmicray flux by the geomagneticfield
and the solar wind causesvariation in the productionrate of
cosmogenic isotopes in the atmosphere. Records of
TaylorDome,Antarctica
(77ø50'S, 158040
' E, 2400m).10Be
analysesweremadeat the LawrenceLivermoreCenterfor
cosmogenicløBe in Greenlandice coresexhibit variations AcceleratorMassSpectrometry
[Daviset al., 1990;Southonet
which coincide with known periodicities in solar activity al., 1990]; analyticalprecisionis betterthan5%. Dating was
[Beer et al., 1988; 1990]. The identification of geomagnetic achievedby locatingbomb-radioactivity
horizonsin the firn
field changesin 10Berecordshasprovenmoredifficult,in part [Dibb et al., 1990] andassuming
a constantaccumulation
rate
becausethe expectedlow-frequency variations axe masked by
climatic effects [Yiou et al., 1985; Lal, 1987]. Interpretation
(7.5 g cm'2a'l).Independent
datingcontrolby oxygenisotope
(ti180)stratigraphy
indicatesthat the assumption
of constant
of low-frequency
10Bevariationsis furthercomplicated
by the
accumulationrate is reasonable(Figure 1).
difficulty of isolating the separateeffects of geomagneticand
solar variability [Beer et al., 1987; Raisbeck and Yiou, 1985].
a function of time. Also shown is the neutron counting rate at
In Figure2, TaylorDomeløBeconcentrations
areshownas
Over the 11-yearsolarcycle,10Beproductionvariationsof
about +20% are predicted for the high-latitude stratosphere
[Lingenfelter, 1963; Lal and Peters, 1967; O'Brien, 1979; Lal,
1988], compared with variation of <5% at the geomagnetic
equator. The modulation of cosmic rays by the geomagnetic
field, on the other hand, is greatest at the equator.
Atmosphericmixing thus plays a critical role in determining
DeepRiver,Canada[NRC,1994].Theneutron
counting
rateis
directlyrelatedto the cosmicray flux, and thereforeto the
10Beproductionrate at any point in the earth'satmosphere
[O'Brien and Burke, 1973].
Comparison
of the løBeandneutroncounterdatashowthat
atmosphereto surface.
A questionof particular importanceis whether or not there
there is an --l 1-year periodicitycommonto both records.
FollowingBeer et al. [1990], we attributethis periodicityto
solar modulation of production. An alternate causal
mechanismfor the observedl 1-year periodicityof 10Beis
solarforcingof climate[Lal, 1987]. Thereis indeedevidence
is significantdeposition,at polar latitudes,of 10Be produced
for an approximately-11-year
periodicity
in ti180data
local atmospheric10Beconcentrations
and the 10Beflux from
[Stuiver et al., 1995] from an ice core in central Greenland.
thisperiodicity
is absent
in bothti180andelectrical
1Alsoat Quaternary
Isotope
Laboratory,
University
of Washington,However,
Seattle,WA 98195
2 Now at Instituteof ArcticandAlpineResearch,
University
of
Colorado,Boulder,CO 80309(email:[email protected])
3 Hampshire
College
4 Now at LeibnizC-14 Laboratory
of the ChristianAlbrechts
UniversitatKiel, LeibnizstraBe19, Kiel, Germany
conductivitydata at Taylor Dome. Therefore,althoughlongterm climate changesand high-frequency'noise' may be
superimposed
on production-rate
variations,thereis at present
little evidence for a meteorologicalcontribution to the !l-
yearperiodicity
observed
for 10Beat TaylorDome.
5 Lawrence
Livermore
National
Laboratory
Copyright1996by theAmericanGeophysical
Union.
Papernumber96GL00255
0094-853 4/9 6/96GL- 00255503.00
Amplitudeof the 11-yearcycleat Taylor Dome
To quantifythe relationshipbetweencosmicray flux and
10Bedepositionrates,we applieda 9-13 year bandpass
filter
to the data. Taking autocorrelation
into account,comparison
523
524
0
_
204060
0
•
2
'" 1990
•
150-
1980
1.2
1.0
5
9641•964
61.9 ,
Baaivity
-45
•
(OH / kS)
Figure 1. •80
1875 1900 1925 1950 1975
Figure 3. Lower:bandpass-filtered
(9-13 year)løBe data
-35
a• (•)
and B radiationprofilesat Taylor Dome
Ant•ctica. N•bers
0.8
show datesof countedye• (ass•g
•at
•180 pe•s •e annual)andof •own bomb-radiation
pe•s.
from Taylor Dome, Antarctica (--) and Dye 3, Greenland
( ......... [Beer et al. ,1990]). Dashed line shows filtered snow
pit 10BefromTaylorDome.Dataareplottedrelativeto the
average.For clarity, the Dye 3 data have been shifted forward
by -0.5 years. Upper: filtered Deep River neutron counting
rate (--) and Wolf mean annual sunspotnumbers( ..........).
of the 11-year-periodiccomponents
of the snow-pit10Beand
the neutron data yields a correlation coefficient of r = 0.89 at
lag 1 year (neff = 49). The 1-year lag is expectedfrom the -1-
year residencetime of 10Bein the atmosphere
[Raisbeckand
atmospheric
reservoirthan does 10Bedepositionat Dye 3
(latitude65øN).
We used the calculations of O'Brien et al. [1991] to
Yiou, 1981; Beer et al., 1990]. Before filtering, all data were
numerically re-sampled, using low-degree cubic splines
[Rasmussen, 1991], to obtain average values over identical
one-year intervals. This procedure should underestimate the
11-year amplitudefor the fire-core at Taylor Dome, which was
sampled at relatively low resolution.
integratetotal productionover variouslatitude and altitude
rangesas a functionof the Deep River neutroncountingrate.
Figure3 showsa comparison
of the filtered 10Bedata from
Taylor Dome (A.D. 1920 - 1994) and from Dye 3, Greenland
absoluteproductionrate of 10Be is not importanthere,
tion at each site. Also shown are sunspot.numbers and the
filtered Deep River neutroncounterdata. The amplitudeof the
time) is constant,or at least non-periodic.
O'Brien et al. 's calculationsagree closely with empirical observations of the neutron-flux/altitude relationship [Simpson
et al., 1953], and with the earlier calculationsof Lingenfelter
[1963] and Lal and Peters [1967]. Becausewe use ratios, the
althoughimplicit in our model is that the scavengingratio
(1860 - 1985),plottedrelativeto the average10Beconcentra- (theproportionof atmospheric
10Bethatis deposited
per unit
Figure 4 comparesthe neutron counting rate with the
variation in 10Be at Taylor Dome for the years 1958-1994.
Also shown are predicted variations integratedover l) the
(1875-1985) as large as the largest observedat Taylor Dome
whole atmosphere,2) stratosphereonly, and 3) the polar
over the last 75 years. This suggestsa fundamentaldifference stratosphere.We estimate the uncertainty in the average
between the northern and southern hemisphere sites. A
slopesof the theoreticalcurvesto be +10%, basedon comparpossibleexplanation
is that 10Bedeposition
at TaylorDome ison amongdifferent calculationsof cosmogenicnuclidepro(77ø50'S) reflects 10Be production from a higher-latitude duction rates [Lal, 1988; Lingenfelter, 1963; O'Brien et al.,
1991]. A best-fit straightor parabolicline to the Taylor Dome
data is statisticallydistinct (>95% confidence)from both the
whole atmosphereand whole stratosphere
predictedvalues.
løBe variationsat Taylor Dome is on averagelarger than at
Dye 3. There are no variations in the entire Dye 3 data set
Discussion
'•30
]ß, 25
•o
I !\i:i ß
•'
lo /i
1.8•
.
20
The data suggestthat there is some contribution,possibly
significant,of polar stratospheric
10Beto the Antarctic10Be
flux. This conflicts with the traditional view that at high
latitudes there is little input of cosmogenicisotopesproduced
in the stratosphere[Lal and Peters, 1967], but is consistent
with
1925
1945
1965
1985
observations
of ozone
and aerosol
concentrations
in the
Antarctic [Fox et al., 1995; Maenhaut et al., 1979; Manney et
al., 1994; Santee et al., 1995; Toon et al., 1986; Crutzen and
Arnold, 1986; Deshler et al., 1995; Gobbi et al., 1991].
To quantifythe relativecontributionof non-polar10Beto
Figure 2. 10Beconcentrations
at TaylorDomefroma firn polar 10Be,we calculatea sensitivityfactor,
core (lower) and a snow pit (upper). Dashed line showsmean
annualneutroncount(106/hr)at DeepRiver,Canadafor
comparison.
•)(10Be/10Bemean
)••
525
1.4
'
5.0
'......
''I'l'ay!•.r.
'''IDome
ß''I'''I'''
1.•
+
O.
8
1.8
•p
4.0
-
3.5
0.2
0.4
0.6
0.8
low-latitudestratosphere
fraction
1.9
2
2.1
2.2
Riverneurons
(1• •un•)
Figure 5. Sensitivity
of thepolar10Beflux to changes
in
the cosmic ray flux, as characterizedby the Ehmert potential,
•, asa functionof thefractionof løBederivedfromthelowFigure4. Comparison
of predicted
andmeasured
10Bepro- latitude stratosphere. Curved lines show theoretical
duction as a function of countingrates at the Deep River neutron monitor, 1958 - 1994. Crosses show 1-year mean data
from bandpass-filteredTaylor Dome data shown in Figure 3.
Lines show predicted values for entire atmosphere ( .... ),
stratosphereonly (---), and polar stratosphereonly (--).
where • is the Ehmert potential,a measureof solarmodulation
of the cosmicray flux [Ehmert, 1960]. For a typical solar
cycle,• variesfrom 300 to 600 MeV, with an averagevalueof
sensitivities near a mean value of • = 450 MeV; bold line is
from calculation of O'Brien et al. [1991]. Dotted lines denote
estimated2(• uncertainty.Arrows show empirically determined
values
forTaylorDomeandDye3. Leastsquares
2ndorder
polynomial fit gives 95% confidencelimits of -I- 0.4 keV for
the Taylor Dome value.
eta/., 1993] to achieve their best-fit match introduces errors,
3) the fraction of 'modulated'10Be in the ice core record is
variable, or 4) the theoreticalproductionrate variationsare in
450 MeV [La/, 1988];• = 450 MeV corresponds
to a counting error. The first three possibilities draw into question the
rateatDeepRiverof 1.97X 106neutrons/hour
[O'Brien
and validity of using 10Be ice core recordsas proxiesfor the
Burke, 1973].
geomagneticmodulationof cosmogenicisotopeproduction.
We divide the atmospheric
10Beproductioninto geomagnetically 'modulated' and 'unmodulated'components,correspondingto the low latitude (0ø to 60ø) and high-latitude(60ø
to 90ø) atmosphere.The boundary at 60ø correspondsto a
break in slope in theoretical cosmogenicisotope production
curves and is meaningful also from the perspective of
atmosphericdynamics,since 60ø approximatesthe location of
the polar vortex [e.g. Kakegawa et al., 1986]. Although
geomagnetic and geographic latitudes are different, this
difference is negligible becauseproductionrates are nearly
independentof latitude poleward of 60ø [e.g. Lingenfelter,
1963].
Figure 5 depictsthe sensitivityof 10Be production to
changes in Ehmert potential for different mixing ratios
between10Bederivedfrom high latitudesand from the lowlatitude stratosphere.Comparison between the calculated and
observed
sensitivities
shows that at most 35% of the Antarctic
10Beflux can be derived from low-latitude sources.In contrast,
the amplitudeof variationof 10Beobserved
in theDye 3 core
[Beer
et al.,
1990]
is consistent
with
a well-mixed
stratosphere [Monaghan, 1987]. This conclusionis at odds
with recent work of Mazaud et. al., [1994], who were able to
produce a reasonablematch between the Vostok, Antarctica,
Conclusions
Evidently, a relatively small proportion of the Antarctic
løBe flux is derivedfrom latitudesat whichvariationsin geomagneticfield strengthare important.If our 75-year recordis
applicableto the distant past, then the 10Be record from
Antarctic ice cores should contain only a small geomagnetic
signal. The global average variation over the last 10 ka is on
the order of -I-25%; assuming a maximum 35% contribution
from the low-latitude stratosphere,geomagneticmodulation
accountsfor variation of at most +10% at Taylor Dome.
This conclusionis satisfying from the point of view of
glaciology,becauseit suggeststhat 10Be concentrations
can
be used to determine accumulationrates in ice cores [Lorius et
al., 1989; Raisbeck et al., 1987; 1992; Steig eta/., 1995;
Yiou et al., 1985]. The rather minimal effect of geomagnetic
modulation on the Antarctic 10Be flux makes reasonable the
assumption of a "constant" 10Be flux over millennial
timescales.The effect of solar variability, of course,must still
be considered,
but our resultssuggestthatcomparison
of 10Be
recordsamongAntarcticand Greenlandice coresmay allow us
to separate the geomagnetic and solar contributions to
10Berecordandocean-sediment
proxyrecordsof geomagnetic
changesin cosmogenicisotope productionrates.
field strength. As has been done here, Mazaud et al. [1994]
dividedthe atmospheric
10Besourcesinto 'unmodulated'
and
'modulated'components;they determined a value of 75% for
the modulated component.The discrepancywith our value of
<35% suggeststhat 1) ocean-sedimentgeomagneticrecords
approximate geomagnetic field variations rather poorly
[Raisbeck et al., 1994] 2) the slight adjustmentmade by
Mazaud•t al. [1994]to theaccepted
Vostoktimescale
[Jouzel
Acknowledgments. We are gratefulto J. Dibb for B-decaymeasuremerits,to K. Taylor for help with the electricalconductivity
measurements,
to E. Waddington,D. Morse and manyothersfor field
support.,K. O'Brienfor tabulatedcosmogenic
isotopeproductionrate
data,and to J. Beer, M. Pavichand an anonymousreviewerfor
constructivecriticsim. This studywas supportedby the U.S. Antarctic
ProgramundergrantsNSF-OPP 9316162 and 89-15924,and by a U.S.
DOE GlobalChangeFellowshipto EJS.
526
References
Mazaud, A., C. Laj and M. Bender,A geomagneticchronologyfor
Antarctic
iceaccumulation,
Geoph,
y•s.
Res.Lett.,21,337-340,1994.
Monaghan,
M. C., Greenland
ice løBeconcentrations
andaverage
Beer, J., A. Blinov, G. Bonani,R. C. Finkel, H. J. Hofmann,B. Lehmann,
H. Oeschger,A. Sigg,J. Schwander,
T. Staffelback,B. Stauffer,M.
precipitation
ratesnorthof 40øNto 45øN,EarthPlanet.Sci.Lett.,84,
197-203, 1987.
SuterandW. W/51fli,Useof løBein polariceto tracethe l 1-year NRC, CosmicRay NM-64 NeutronMonitorData, CosmicRay NM-64
NeutronMonitor Data,, HerzbergInstituteof Astrophysics,
National
ResearchCouncil of Canada(NRC), Ottawa, 1994.
Oeschger,
U. Siegenthaler
andR.C.Finkel,
løBemeasurements
on O'Brien,K., Secularvariationsin the productionof cosmogenicisotopes
in the Earth'satmosphere.,
J. Geophys.
Res.,84, 423-431, 1979.
polar ice: comparisonof arctic and antarcticrecords,Nucl. Instr.
O'Brien, K. and G. Burke, Calculated cosmic ray neutron monitor
Method.Phys.Res.,B29, 203-206, 1987.
response
to solarmodulationof galacticcosmicrays.,J. Geophys•
Beer, J., F. Joos, C. Lukasczyk, W: Mende, J. Rodriguez, U.
cycleof solaractivity,Nature,347, 164-166,1990.
Beer, J., G. Bonani, H. F. Hofmann,M. Suter,A. Synal, W. W51fli, H.
Siegenthaler
andR. Stellmacher,
løBeas an indicator
of solar
variabilityandclimate,in Nesme-Ribes,
E., ed. TheSolarEngineand
Its Influence on Terrestrial Atmosphereand Climate, SpringerVerlag, Berlin,221-233, 1994.
Beer, J., U. Siegenthaler,G. Bonani, R. C. Finkel, O. H., M. Suter and
W. W61fli, Informationon pastsolaractivityandgeomagnetism
from
løBeintheCampCentury
icecore,
Natura331,675-679,
1988.
Crutzen, P. J. and F. Arnold, Nitric acid cloud formation in the cold
Antarcticstratosphere:
a major causefor the springtimeozonehole.,
Nature, 324, 651-654, 1986.
Davis, J. C., I.D.
Proctor, J. R. Southon, M. W. Caffee, D. W.
Heikkinen, M. L. Roberts,T. L. Moore, K. W. Turteltaub,D.e.
Nelsøn,D. H. Loyd and J. S. Vogel, LLNL/UC AMS facility and
researchprogram,Nucl. Instr. Method. Phys. Res., B52, 269-272,
1990.
ß
Deshler, T., A. Adriani, G. P. Gobbi, D. J. Holmann, G. Di
Donfrancesco and B. J. Johnsen, Volcanic aerosol and ozone
depletionwithin the antarcticpolar vortexduringthe australspringof
1991,Geophys.
Res.Lett.,19, 1819-1822,1995.
Dibb, J. E., P. A. Mayewski, C. S. Buck and S. M. Drumracy,Beta
radiation from snow, Nature, 345, 25, 1990.
Res., 78, 1973.
O'Brien,K., A. Lerner, M. A. Sheaand D. F. Smart,The productionof
cosmogenic
isotopesin the Earth'satmosphereandtheir inventories,
in Sonnet,C. P., M. S. GiampapaandM. S. Matthews,eds.TheSunin
Time,Universityof ArizonaPress,Tucson,990, 1991.
Raisbeck,
G. M. andF. Yiou, Cosmogenic
1øBe/7Be
as a probeof
atmospheric
transportprocesses,
GeOphys.
Res.Lett.,8, 101501018,
1981.
Raisbeck,
G.M. andF.Yiou,løBeinpolariceandatmospheres,
•Annals
Glac. 7, 138-140, 1985.
Raisebeck,G. M., F. Yiou, D. Bourles, C. Lorius, J. Jouzel and N. I.
Barkov,Evidence
for twointervals
of enhanced
løBedeposition
in
Antarcticice duringthe last glacialperiod,Nature, 326, 273 - 277,
1987.
Raisbeck,
G.M., F. YiouandS.Z. Zhou,Palaeointensity
puz•e,Nature,
371, 207-208, 1994.
Rasmussen,A., Piecewise integral splines of low degree, Comp.
Geosci., 17, 1255-1263, 1991.
Santee,M. L., W. G. Read,J. W. Waters,L. Froidevaux,G. L. Manney,
D. A. Flower, R. F. Jarnot, R. S. Harword and G. E. Peckham,
Interhemispheric
differencesin polar stratosphericHNO3, H20,
C10,and0 3, Science,
267, 849-852,1995.
Ehmert,A., On the modulationof the primary spectrumof cosmicrays
Shaw, G. E., Aerosoltransportfrom sourcesto ice sheets,in Oeschger,
from solaractivity,Proc.!nt. CosmicRay Conf.,4, 140-152,1960.
H. andC. C. Langway,eds.The EnvironmentalRecordin Glaciers
Fox, L. E., D. R. Worsnop,M.S. ZahniserandS.C. Wofsy, Metastable
and Ice Sheets,JohnWiley & Sons,New York, 13-27, 1989a.
phases
in polarstratospheric
aerosols,
Science,
267, 351-355,1995.
Simpson,J. A., W. FongerandS. B. Treiman,CosmicradiationintensityGobbi, G. P., T. Deshler, A. Adriani and D. J. Hofmann, Evidence for
time variationsand their origin. 1. Neutronintensityvariationmethod
denitrificationin the 1990 Antarcticspringstratosphere:
I Lidar and
andmeteorological
factors,Phys.Rev.,90, 934-950, 1953.
temperature measurements,Geophys.Res. Lett., 18, 1995-1998,
1991.
Southon, J. R., M. W. Ca_free,J.C.Davis, T. L. Moore, I.D. Proctor, B.
Schumacherand J. S. Vogel, The new LLNL AMS spectrometer,
Jouzel, J., N. I. Barkov, J. M. Barnola,M. Bender,J. Chappellaz,C.
Nucl. Instr.Method.Phys.Res.,B52, 301-305, 1990.
Genthon, V. M. Kotlyakov, V. Lipenkov, L. C, J. R. Petit, D.
Raynaud,G. Raidbeck,C. Ritz, T. Sowers,M. Stievenard,F. Yiou Steig, E. J., P. J. Polissar and M. Stuiver, Cosmogenicisotope
concentrations
at Taylor Dome, Antarctica,Ant. Jour. U.S., in press,
and P. Yiou, Extendingthe Vostok ice-corerecordof palaeoclimate
1995.
to the penulitmateglacialperiod,Nature, 364, 407-412, 1993.
Kakegawa, H., T. Yasunari and T. Kawamura, Seasonaland intra- Stuiver,M., T. Braziunasand P. Grootes,The GISP2 tSlSOclimate
recordof the past 16,500 yearsand the role of the sun, ocean,and
seasonalfluctuationsof polar anticycloneand circumpolarvortex
volcanoes.,Quat. Res., 44, 341-354.
over Antarctica,Merrc Nat. Inst. Polar Res., 45, 19-29, 1986.
of HNO3
Lal, D., løBein polarice:datareflectchanges
in cosmic
ray flux or Toon,O. B., P. Hamill, R. P. TurcoandJ. Pinto,Condensation
and HC1 in the winter polar stratospheres,
Geophys.Res. Lett., 13,
polarmeteorology,
Geophys.
Res.Lett.,14, 785-788,1987.
1284-1287, 1986.
Lal, D., Theoreticallyexpectedvariationsin the terrestrialcosmic-ray
productionratesof isotopes.,in Castagnoli,G. C., ed. International Tric, E., J.-P. Valet, P. Tucholka,M. Pateme,L. Labeyrie,F. Guichard,
L. Tauxe and M. Fontugne,Palcointensityof the geomagneticfield
Schoolof Physics"EnricoFermi", North-Holland,Varennasul Lago
duringthelast80,000years,J. Geophys.
Res.,97, 9337-9351,!992.
di Como, 216-233, 1988.
D. Bourles,
C. LoriusandN. I. Barkov,løBe
Lal, D. and B. Peters,Cosmicray producedradioactivityon the earth, Yiou,F., G. M. Raisbeck,
in ice at VostokAntarcticaduringthe last climaticcycle,Nature, 316,
Handbuchder Physik;46, 551-612, 1967.
616-617, 1985.
Lingenfelter,R. E., Productionof carbon14 by cosmic-rayneutrons,
Rev.Geophys.,
I, 35-55, 1963.
Lorius, C., G. Raisbeck, J. Jouzel and D. Raynaud, Long-term
environmentalrecordsfrom Antarctic ice cores.,in Oeschger,H.
E.J. Steig,M. StuiverandP.M. Grootes,Dept.of Geological
andC. C. Langway,eds.TheEnvironmental
Recordin Glaciersand Sciences,351310,Univ. of Washington,Seattle,WA 98195
Ice Sheets,JohnWiley, New York, 1989.
P.J.Polissar,Schoolof NaturalSciences,HampshireCollege,
Maenhaut,W., W. H. Zoller and D. G. Coles, Radionuclidesin the South
Amherst, MA 01002
Poleatmosphere,
J. Geophys.Res.,84, 3131-3137,1979.
R. C. Finkel, Centerfor AcceleratorMass Spectrometry,Lawrence
Manney,G. L., L. Froidevaux,J. W. Waters,R. W. Zurek,W. G. Read, LivermoreNationalLaboratory,L-231, Livermore,CA 94550
L. S. Elson,J. B. Kumer,J. L. Mergenthaler,A. E. Roche,A. O'Neill,
R. S. Harwood,I. MacKenzie and R. Swinbank,Chemicaldepletion
of ozone in the Arctic lower stratosphereduring winter 1992-93,
Nature, 370, 429-433, 1994.
(ReceivedSeptember
29, 1995;acceptedDecember19, 1995.)
,