GEOPHYSICAL RESEARCH LETTERS, VOL. 27, NO. 19, PAGES 3145-3148, OCTOBER 1, 2000 Accretionof InterplanetaryDustin Polar Ice Edward J. Brook Geology Department, Washington State University, Vancouver, Washington, USA Mark D. Kurz and Joshua Curtice Department ofMarine Chemistry andGeochemistry, Woods HoleOceanographic Institution, Woods Hole, Massachusetts, USA Stuart Cowbum Geology Department, Washington State University, Vancouver, Washington, USA Abstract. Measurements of helium isotopesin particles using 3Heasa proxy[Farley andPatterson, 1995;Patterson and separatedfrom polar ice demonstratethat extraterrestrial3He Farley, 1998;Marcantonioet al., 1996]. Resultsof thesestudies dominates the3HefluxattheGISP2(Greenland) andVostok areequivocal.Marcantonioet al. [1996]arguedfor minimalIDP (Antarctica)ice core sites. Replicatemeasurements of late fluxvariations, whilePatterson andFarley[1998]interpreted Holocene icesamples yield3Hefluxesof 0.62+ 0.27x 10-•2cm3 theirdataasevidence for 100kaperiodicity in IDP accretion. STPcm'2ka'• (GISP2) and0.77ñ 0.25x 10'12cm3STPcm-2ka'l The noblegascontentof particlesin polarice coresmay (Vostok),similarto resultsfrom marinesediments.Theseare the provide another record of IDP flux variations that would shed firstdetailed measurements of 3Hein particles fromice core lightonsomeof theseissues, butpriorto thisstudythispotential samples, andtheydemonstrate the utilityof the ice corerecord has not been investigatedin detail. There are a number of for evaluatingthe temporalhistoryof the extraterrestrial dust possibleadvantages of an ice corerecordof IDP flux. First, flux.Results fromVostok samples from1096-1403 m depth (75- accumulation ratesandtime scalesof deeppolarice coresare 97 ka B.P.) are similarto the late Holocenedata,with the wellknown,allowingaccurate fluxcalculations andcomparison exceptionof two highly anomalous resultsfrom 1307 m. The to otherrecords.Second, ice corerecords containa varietyof latterprobably indicate thepresence of rare,largeor gas-rich climateproxies,includingstableisotopeandtracegasrecords, extraterrestrial particles. allowing examination of postulated linksbetween IDP variability andclimatechange.Finally,thedatapresented heredemonstrate that3Hein ice coreparticulate matteris dominated by Introduction extraterrestrial helium. Accretion of extraterrestrial material to theearthis important to studies of theoriginof theatmosphere andoceans, dynamics Methods of dustin planetary systems, geochemical budgets, astronomical Helium isotopeswere measuredin particles from GISP2 observations, andpossibly terrestrial climatechange [Loveand (centralGreenland)andVostok(centraleastAntarctica)ice core Brownlee, 1993;Muller andMacDonald,1995,1997]. Marine samples.GISP2sampleswerefrom 132-138m depthin the G2 sedimentsand polar ice providean archiveof the influx of extraterrestrial materialin the form of interplanetary dust particles (IDPs)[Parkinand Tilles,1968]. Highlevelsof 3He andothernoblegasisotopes provide a uniquesignature of IDP in pelagic sediments [e.g.,ArnariandOzirna,1988]. Thissignal core and are approximately400-430 years old [Meeseet al., 1997]. Vostoksampleswere from the 110-111m sectionof the BH-5 core and from selecteddepthsbetween1096 and 1404 m fromthe 5G deepcore. The BH-5 samples weredeposited about 3.8 ka B.P andthe 5G samplesweredeposited between75 and97 can,in principle,be usedto constructa recordof extraterrestrial ka B.P.,usingthetime scaleof Sowerset al. [ 1993]. 3Heflux,andthereby derive a record ofIDPaccretion [Farley Ice sampleswere cut on a bandsaw and weighed.Particles andPatterson, 1995;Farley,1995;Patterson andFarley,1998]. were extracted by melting at room temperaturein a glass MullerandMacDonald [1995,1997]highlighted theimportancefiltration funnel in a class 100 clean hood. Meltwater was of theserecords by suggesting thattheearth's100ka glacial- vacuumfilteredthrough0.45 or 0.2 micronpure silver filters interglacial cyclemightbe drivenby variationsin the tilt of the (Osmonics-Poretics). The filters were rinsed with ethanol, orbital plane.Mostotherexplanations ofthe100kacyclerelyon folded,andplacedin aluminumfoil boats. Helium was extracted astronomically forcedsolarinsolationchanges[Imbrieet al., by meltingthe particles,filter, and foil boat in an ultra-high 1993],andhavedifficulty explaining powerin the100ka cycle. vacuumfurnacewith a tantalumcrucible. Heliumwaspurified Muller and MacDonald[1995]suggested thatvariabilityin and the isotopes measured using static noble gas mass interplanetarydust abundancenear the earth could link orbital spectrometry; the massspectrometerand proceduresused are inclinationand climate. Severalgroupshavesearched marine describedby Kurz et al. [1996]. Blankswere between4 and 6 x sediment records forvariations in IDP fluxon 100katimescales, 10-• cm3 STP4Hewithapproximately atmospheric 3He/4He Copyright2000 by the AmericanGeophysicalUnion. ratios,andrepresent smallcorrections to 4He and3He. The ultimate detection limitof thesystem is approximately 2 x 10'16 cm3 STP3He(5000atoms); thesmallest samples discussed here Papernumber2000GL011813. are approximately20 times largerthan this detectionlimit. All 0094-8276/00/2000GL011813505.00 heliumconcentrations arereported in unitsof cm3 STPg'• ice. 3145 3146 BROOK ET AL.' INTERPLANETARY DUST IN POLAR ICE Table 1. Helium IsotopeData for > 0.45 Micron ParticlesFilteredFromIce Samples. Sample Depth Age• (m) (ka) Wgt.Ice 4He 3He/4He (g) (10'•2cm3g'•ice) (R/R0 3He IceAccum. 2 (10'•8cm3g'•ice) (gcm'2a '•) 3HeFlux (10'•2cm 3cm'2 ka'•) GISP 2 (Greenland) 132.7-133.1 0.40 1092.70 0.77 _+0.02 42.83 +_0.28 45.47 +_0.96 22.4 1.02 +_0.02 133.6-133.7 0.41 338.50 0.99 +_0.02 8.75 +_0.25 12.01 _+0.42 22.4 0.27 _+0.01 134.2-134.3 0.41 800.60 4.26 _+0.09 7.31 _+0.06 43.09 _+0.93 22.4 0.97 _+0.02 137.2-137.7a 5 0.43 751.75 0.59 _+0.01 15.70 +_0.25 12.73 _+0.32 22.4 0.29 +_0.01 137.2-137.7b 5 0.43 938.88 0.82 +_0.02 28.58 _+0.28 32.55 _+0.73 22.4 0.73 +_0.02 137.54-138.18 0.42 1186.23 0.86_+0.02 16.17 _+0.18 19.15 _+0.50 22.4 0.43 _+0.01 137.54-138.183 0.42 1186.23 0.30 _+0.01 0.96 _+0.19 0.39 _+0.08 Mean _+2•: 0.62 _+0.27 0.87 _+0.02 Vostok(Antarctica)4 110.0-110.19a 5 3.80 259.69 1.82 _+0.04 173.51 _+0.92 437.02 _+9.04 2.0 110.0-110.19b 5 3.80 280.35 2.22_+0.04 174.11 _+0.78 535.39_+ 10.97 2.0 1.07 _+0.02 110.19-110.38 3.80 265.96 1.33_+0.03 139.07 _+1.14 255.16_+5.51 2.0 0.51 _+0.01 110.38-110.48a 5 3.80 145.15 1.01 _+0.02 159.05 _+1.72 223.08 _+5.07 2.0 0.45 _+0.01 110.38-110.48b 5 3.80 139.50 2.22_+0.04 152.21 _+1.05 466.61 _+9.87 2.0 0.93_+0.02 1096.0 75.26 154.73 5.16 _+0.01 1145.0 77.99 168.92 1220.0 84.04 152.03 1261.0 87.56 158.99 3.60 _+0.07 116.14 _+0.71 1265.0 87.89 215.93 2.39 _+0.05 90.85 _+0.60 1285.0 89.14 254.01 3.00 _+0.06 1303.3 89.94 176.94 2.85 _+0.06 1307.0a 5 90.08 67.44 46.34 _+0.93 1307.0b • 90.08 100.98 1307.0c s 90.08 94.16 1316.6a s 90.87 105.46 1.44 _+0.03 163.28 -+ 1.69 1316.6b s 90.87 106.24 1.44 -+ 0.03 173.75 _+1.77 1321.0 91.28 271.70 2.22 _+0.04 1333.0 92.39 221.45 1341.0 92.99 119.85 1346.0 93.36 1403.6 97.30 Mean _+2•: 0.77_+0.25 1.2 0.97 _+0.02 109.98 _+0.56 785.88 _+16.22 3.15 _+0.06 143.79 _+0.80 626.62 _+13.02 1.5 0.92_+ 0.02 2.37 _+0.05 156.14 _+1.04 510.77 _+10.73 1.5 0.75 _+0.02 578.09 _+12.10 1.4 0.80 _+0.02 300.99 _+6.34 1.3 0.40 _+0.01 159.23 _+0.67 660.49 _+13.50 1.4 0.91 _+0.02 128.74 _+0.88 508.51 _+10.75 1.5 0.76 _+0.02 200.05 _+0.86 12829.29 _+262.45 1.4 17.88 _+0.37 1.60 _+0.03 102.38 _+0.98 1.4 37.28 _+0.75 193.13 _+0.92 227.18 _+5.04 9963.83 _+204.85 0.32 _+ 0.01 1.4 13.89 _+0.29 325.71 -+7.33 1.6 0.52 -+0.01 346.75 _+7.78 1.6 0.55 _ 0.01 147.82 _+0.74 453.21 _ 9.34 1.4 0.65 _ 0.01 2.26 _+0.05 136.33 _+0.76 426.69 _ 8.86 1.4 0.62 +_0.01 8.17 _+0.16 149.61 _+0.72 691.59 _+34.79 1.5 2.47 _ 0.05 207.50 5.60 _+0.11 194.93 _+0.77 1512.07 _+30.83 1.5 2.21 _+0.05 106.56 5.80 _+0.12 101.04 _+0.65 811.55 _+17.04 1.6 1.27 +_0.03 •Ages forGISP2fromMeese etal.[1997];ages fromVostok fromSPECMAP timescale ofSowers etal [1993].2GISP2: CuffeyandClow [1997]'Vostok: Sowers etal.(1993).30.2-0.45 m sizefraction. 4Depths forbelow111.0maretopdepths of samples spanning 10-15cm. 5Replicates forHolocene GISP2andVostok samples wereprepared bycutting coresections longitudinally. Replicates of deepVostok samples were madeon unorientedvariablesizedpiecesfrom 10-15 cm intervals. Blank measurementson unusedfilters and proceduralblanks GISP2 siteare - 10 timeshigherthan at Vostok(Table 1), the of the GISP2samples is roughlyhalf of using deionized,filtered water were indistinguishable from mean4Heconcentration furnaceline blanks. As a test of filtration efficiencyand yield the mean for the HoloceneVostok samples. This probably of terrestrial dust onesamplewasfilteredfirstthrougha 0.45micronfilterandthen reflectsthe relativeproximityandabundance through a 0.2micron filter.Thequantity of3Hetrapped onthe sourcesat the GISP2 site, althoughdifferencesin the age, 0.2 micron filter was lessthan 2% of the total trappedby the two composition, or particlesizeof the dustmayalsocontribute. filters(Table1), indicating thatthe0.45 micronfilterseffectively 3He/4Heratiosfor GISP2 samplesare between7.3 and42.8 Ra trapmostof the particulate 3He. Thisis consistent with (Raistheatmospheric 3He/4He ratio, 1.384 x 10'6).Vostok ratios predictionsof the size distributionof He-retainingIDPs [Farley arehigher,between 91 and200 Ra(Table1). Estimates of the etal., 1997],whichsuggest thatthemajority ofIDP-borne 3Heis 3He/4Heratiofor IDPs rangefrom 173 Ra [Nier andSchlutter, carriedin particleswith diametersgreaterthan 1 micron. A 1992]for stratospheric particles to 310 Raforthesolarwind greater quantity of 4Hepassed through thelargerfilter(-26%of [Geiss etal.,1971].Thehigh3He/4He ratios foricecoresamples thetotal;Table1)suggesting thatsome 4Heiscarried in smaller indicatethat virtuallyall of the 3He they containis particles,probablyterrestrialdust. Results extraterrestrial.Using a two componentmixing model [e.g., Patterson andFarley,1998]thefraction 3Heet/3Hetotal isgreater than 99.8% for Greenland samplesand >99.99% for Antarctic samples. This calculationassumesthat 3Heet/3Hetotal =(4He/3Heobs-4He/3Hecrust)/(4He/3Hct-4He/3Hecrust), 3He/4Heet= 290 4.26x 10-•2cm3STPg'•ice(Table1). 4Heconcentrations ofthe The 4Hecontentof the GISP2 samplesrangesfrom 0.30 to Ra. Thecrustal ratiomaybeashigh Vostoksamples arehigher,reflecting thelowericeaccumulationRa,and3He/4Hec•ust=0.015 ratesat Vostok. AlthoughHoloceneaccumulation ratesat the as 0.33 Ra [Marcantonioet al., 1998]. However,evenusingthis BROOK ET AL.; INTERPLANETARY ratio3HeinVostok samples is> 99.7% extraterrestrial and3He in Greenlandsamplesis > 98.0% extraterrestrial. DUST IN POLAR ICE 3147 The expectedstatisticaldistributionof resultsfor the Holocene Vostoksamples (meanAT=0.01m2a)wasmodeled usingthe Treatingthe 6 GISP2samples as replicates, the mean3He methodof Farley et al. [1997]. (We focuson thesesamples concentration is 28 + 12x 10'•s cm3 STPg'• ice(all means becausethey are most likely to yield statisticallymeaningful reportedas+ 2 x standarderror). For the five VostokHolocene results). We employedthe predictedsize distributionand flux of samples themean 3Hecontent is384+ 123x 10'Iscm IDPsthatwereheatedto lessthan600øCuponatmospheric entry TheGISP2samples spananagerangeof about40 years.If there (andtherefore retained 3He)fromthatstudy.A MonteCarlo were variability in IDP flux to the ice surface on decadal or methodwasusedto calculatethe expecteddistribution of relative shortertime scales[suggested by Kayset et al., 1998] this 3Hecontent withrespect totheglobal mean value, assuming that uncertaintyestimatewould be a maximum. There also could be the3Heconcentrations arecorrelated withparticle surface area scatter in the results due to seasonalvariations in snowfall. The average annual snow accumulation rate at the site is [Farley et al., 1997], generatinga syntheticdistributionof 2000 approximately50% or more of annualsnowfall. Given the small randomlyfrom the 2000 points. points. The distributionhas a meanvalue closeto the global approximately 22 cm/yr,butsummer/fall accumulation is slightly mean,but a standarddeviationof over 100%, reflectingthe greater(-10-20%) thanwinter/spring accumulation [Shuntan et statisticallysmall numbersof IDPs in sampleswith low AT al., 1995]. Althoughthe seasonof deposition is unknown,the values. We then created 1000 model samplesof n=5 for samples areat least10cmthick(Table1) andtherefore represent comparisonwith the Vostok resultsby repeatedlyselecting seasonalvariationsat this site this issue is probablynot Althoughdataare limitedat thispoint,the impliedagreement significant.Theuncertainty for theVostokreplicates is 32 % of of ourreplicatesappearsto be betterthanpredictedby the model themean(Table1), andthesesamples represent deposition over treatment.The probabilityof obtainingan uncertaintyof 32% in - 25years. a sample of n=5forAT=0.01m2aisslightly lessthan15%based Duplicate 3Hemeasurements at1316 matVostok (Table 1)' onthemodel. Although more data arerequired tocharacterize yieldbetteragreement, within6.4 % of themean(Table1). thestatistics of extraterrestrial 3Hein ice coresamples one Given theconsistency ofthisduplicate measurement, results for' explanation forthedifference between thereplicate results and triplicatemeasurements of the 1307 m sample(Table 1) are the model predictionis that the IDP size distributionusedin the difficultto interpret.One replicatehad a concentration of 227 + model is inaccurate. For example,the numberflux of small 5 x 10-]acm3 STPg-•,similar, though slightly lowerthan, particles(10 micronand less)couldbe underestimated; this is a relativelylow valuesfoundin thetwo samples belowthisdepth plausiblehypothesisbecausethe IDP size distributionis not well (Table1). The3Hecontent of theothertworeplicates of this knownfor smallparticles.Fragmentation of particlesduring sampleareextremelyhigh,12,829+ 263 and9,964+ 205 x 10'•a atmospheric entryor in the ice is anotherpossible explanation cm3STPg'],- 10times higher thanthemean forallother Vostok [Peucker-Ehrenbrink and Ravizza,2000]. Futureanalysisof samples. Thesesamples alsohad5-20 timesmore4Hethanthe size-fractionated samples mayshedlightonthisissue. others(Table 1). The origin of thesehigh concentrations is Theaccumulation rateattheGISP2sitewas22.4g cm'2yr'• in unclear butmayresultfromsampling relatively raregas-rich or thedepthrangeof thesamples analyzed [CuffeyandClow,1997]. largeIDPs[Patterson andFarley,1998]. 3He/4He ratiosfor 3Hefluxeswerecalculated assuming that•He•ux=•Heconc * b, thesetwo replicates are200 and193Ra,thehighestin thedata where b is accumulationrate. At Vostok accumulationrates were set,supporting thisconclusion. Thesehigh3He/4He ratiosare basedon the accumulation-temperature relationship[Sowerset fluxesare0.62+ 0.27x 10'•2cm3 similarto thoseobtained for individual IDPs[NierandSchlutter, al., 1993](Table1). Resulting 1992], illustrating that extraterrestrialhelium dominatesthe STPcm'2ka-• for GISP2and0.77+ 0.25x 10-•2cm3 STPcm'2ka' particulatematterheliumbudget. The occurrence of thesetwo • forVostok.These results aresimilar toprevious estimates of highest concentrations in ourdatasetat thesamedepthlevelis theextraterrestrial 3Hefluxtomarine sediments overthepast200 curious, andunlikelyif particles arerandomlydistributed.This ka (Figure 1) [Pattersonand Farley, 1998; Marcantonioet al., may indicatethat there were temporalvariationsin the size 1996,1999].3Hefluxes atVostok inthe1096.0-1403.6 mdepth distribution of3He-bearing extraterrestrial particles. interval(Figure2) arebetween0.32-2.47x 10'•2 cm3 STPcm'2 Discussion measurements of the 1307 m sample). The mean flux in this intervalis 0.94+ 0.32x 10']2 cm3 STPcm'2ka-]. Thelattervalue ka-• (excluding theanomalous datafortwoof threereplicate Thenumberof IDPs in a sampleis proportional to samplesize and inverselyproportionalto accumulationrate. Following Farley et al. [ 1997] the total accumulationof ice or sedimentin a 2.0 samplecan be describedby the ratio of the samplemassto accumulation rate,or "area-timeproduct"(AT). Largervaluesof =, AT are preferableto reducestatisticaleffects relatedto the small numberof IDPs in typical samples(Farley et al., 1997). The meanAT forGISP2samples discussed hereis 0.004m2a,andfor Vostok samples themeanAT is0.01m2a.Forcomparison, most marinesediment3He datacomesfrom sediments with AT values •-n 1. 0 0.5 - GISP21 Pacific 2 • Pacific4 • Atlantic 3J[ indian• VostokI of 0.02 to 0.5 m2a[Farleyand Patterson,1995; 1997; Marcantonio et al., 1996; Patterson and Farley, 1998]. Uncertaintyin multiplereplicatesof marinesediments from ODP Figure 1. Comparison of ice coredatapresented herewith recent estimates of theextraterrestrial 3Heflux.1 = thispaper.2 = Site806 withAT of 0.125m2aranges from15-185%(one Averagefor last 200 ka from Marcantonioet al. [1996]. 3 = standard deviation) with the more extreme values driven Averagefor last 20 ka from Marcantonio et al. [1998]. 4= primarily by one or two anomalouslyhigh concentrations averagefor the past114ka fromPattersonand Farley [ 1998]. 5 [Pattersonand Farley, 1998]. = averagefor thelast200 ka fromMarcantonioet al. [ 1999]. 3148 BROOK ET AL.' INTERPLANETARY DUST IN POLAR ICE References I i i &l & 250.• 15o 50 ::!3 Amari, S. and M. Ozima, Extra-terrestrial noble gases in deep sea sediments,Geochim. Cosmochim.Acta, 52, 1087-1095, 1988. Cuffey, K.M. and G. D. Clow, Temperature,accumulation,and ice sheet elevationin centralGreenlandthroughthe last deglacialtransition,J. Geophys.Res. 102, 26,383-26, 396, 1997. Farley, K.A., Cenozoic variations in the flux of interplanetarydust recorded by3Heindeep-sea sediment, Nature,376,153-156, 1995. Farley, K.A., S.G. Love, and D.B. Patterson,Atmosphericentry heating and helium retentivity of interplanetary dust particles, Geochim. 10 Range at3.8ka Cosmochim.Acta, 61, 2309-2316, 1997. Farley, K.A. and D. B. Patterson,A 100-kyr periodicity in the flux of extraterrestrial 3Heto theseafloor,Nature,378,600-603,1995. Geiss, J., F. BQehler, H. Cerutti, P. Ebderhardt,and C.H. Filleaux, Solar m 0.1 T' 75 I 80 I 85 I 90 I 95 100 wind composition experiments, Apollo 16 Preliminary Science Report,NASA SP315, 14.1-14.10,1972. Imbrie, J. et al., 1993, On the structureand origin of major glaciation cycles; 2, The 100,000-yearcycle, Paleoceanography, 8, 698-735, 1993. Kayser, R., J. Wu., E. A. Boyle, and R. Sherrell, A seasonalcycle in cosmicIridium depositionin central Greenland:Does it result from meteorologicalfocussing?,Trans.AGU, 45, F47, 1998. 3Hefluxes and3He/4He ratios fromVostok, 1096-1404Kurz, Mark D., T. Kenna, J. Lassiter,and D. DePaolo, Helium isotopic Age (ka) Figure2. m. Gray bar showsrangeof fluxesfrom replicatemeasurements evolution of Mauna Kea Volcano; first results from the 1-km drill of late HoloceneVostoksamples.High valuesat - 90 ka are for core,J. Geophys.Res.,101, 11,781-11,791,1996 two of threereplicatesfor the 1307 m sampleandmay be dueto Love, S.G. and D.E Brownlee, A direct measure of the terrestrial mass accretionrate of cosmicdust,Science,262, 550-553, 1993. rare,largeor gasrich IDPs. Marcantonio,F. R.F. Anderson,M. Stute, N. Kumar, P. Schlosser,and A. Mix, Extraterrestrial 3Heasa tracerof marinesediment transport and accumulation,Nature, 383, 705-707, 1996. is alsosimilarto recentestimates spanning the last200 ka (Figure Marcantonio,F., S. Higgins, R.F. Anderson,M. Stute,P. Schlosser,and E.T. Rasbury,Terrigenoushelium in deep-seasediments,Geochim. 1). However,thereis no significantdifferencebetweenthe mean Cosmochim.Acta, 62, 1535-1543, 1998. of all of the Holoceneresults(GISP2 and Vostok, n=l 1) and Marcantonio,F., K.K. Turekian,S. Higgins,R.F. Anderson,M. Stute,and glacial results (Vostok, n=15) at the 95% confidenceinterval (two-tailedstudents t-test,t=1.240, outliersat 1307m neglected). The overallmeanflux for the dataset,(neglectingthe outliersat 1307m)is0.83+ 0.20x 10'12cm3STPcm'2ka'l. Thisvaluemay P. Schlosser, The accretionrate of extraterrestrial 3He basedon oceanic 23øTh flux andtherelationto Os isotope variation overthe past 200,000 years in an Indian Ocean core, Earth and Planet. Sci. Lett, 170, 157-168, 1999. Meese,D. A. et al., The GreenlandIce SheetProject2 depth-agescale; be more appropriatefor comparisonto marine sedimentdata methodsandresults,J. Geophys.Res.,102, 26,411-26,423,1997. whichare averagedover longtime periodsby stripsampling[e. Muller, R.A. and G.J. MacDonald, Glacial cyclesand orbital inclination, Nature, 377, 107, 1995. g., Pattersonand Farley, 1998]. Thereis alsoapparenttemporal Muller, R.A. and G.J. MacDonald, Glacial cycles and astronomical variability in the Vostok results for the samplesdeposited forcing,Science,277, 215-218, 1997. between 75 and 97 ka (Figure 2). Apart from the highly Nier, A.O. and D.J. Schlutter, Extraction of helium from individual anomalousresultsfrom the sampleat 1307 m, the Vostok results interplanetaryparticles by step-heating,Meteoritics, 27, 166-173, indicatean apparent variabilityin the3Heflux of a factorof almost 4 on time scalesof a few thousandyears (Figure 2). 1992. Parkin, D.W. and D. Tilles, Influx measurementsof extraterrestrial material, Science,159, 936-946, 1968. Peucker-Ehrenbrink, B. andG. Ravizza,The effectsof samplingartifacts Apparent fluxesbetween 97 and93 kaare1.27- 2.47x 10'22cm3 on cosmic dust flux estimates: A re-evaluation of non-volatile tracers STPcm-2ka'•. In thesubsequent periodbetween 92 and84 ka (Os,Ir), Geochmim.Cosmochim.Acta, 64, 1965-1970. fluxesrangedfrom0.32 - 0.91 cm3 STP cm'2ka'•, andthen 3He in seafloor returned to highervaluesbetween 0.92and0.97x 10'•2cm3 STP PattersoreD.B. and K.A. Farley, Extraterrestrial sediments,Evidencefor correlated100 kyr periodicityin the accretion cm'2ka'• by78 ka.). Suchrapidchanges notappear in marine rateof interplanetarydust,orbitalparameters,andQuaternaryclimate, sedimentdata,but bioturbationand samplingtechniquessmooth Geochim. Cosmochim.Acta, 62, 3,669-3,682, 1998. the marine record to varying degrees,while ice cores should Sowers,T. et al., A 135,000 year Vostok-SPECMAP commontemporal framwork,Paleoceanography, 8, 737-766, 1993. preservea higherresolutionrecordof flux variations. There are a number of possiblereasonsfor the apparent Shuman,C.A., R.B. Alley, S. Anandakrishnan,J.C. White, P.M. Grootes, temporal 3Hefluxvariations intheVostokcore.First,theymay reflectthe statisticsof smallIDP populations.Second,theremay be terrestrialfactorsthat affectthe IDP contentof polarice. For example, variations in wind velocity might changethe size distribution of deposited IDPs. Finally,the3Hefluxvariability may reflect changesin the dynamics of interplanetarydust production,transport, or accretion. This latter possibility is intriguing,as it implies that ice core samplescould be usedto examineIDP flux variationsoverthe past-• 400 ka. Acknowledgments. We thank the National Science Foundation (OPP-9909384, OPP-9909663, and EAR-9807951) and NASA (NAG59345) for support,Ken Farley andKarl Turekianfor helpfulreviews,and DempseyLott for importantlaboratoryassistance. and C.R. Stearns,Temperatureand accumulationat the Greenland summit:Comparisonof high-resolutionisotopeprofilesand satellite passivemicrowavebrightnesstemperaturetrends,J. Geophys.Res., 100, 9,165-9177, 1995. E. Brook and S. Cowburn, Geology Department,WashingtonState University, 14204 NE SalmonCreek Ave, Vancouver,WA 98686 (email: brookor cowburn•vancouver.wsu.edu) M. Kurz and J. Curtice, Department of Marine Chemistry and Geochemistry,WoodsHole Oceanographic Institution,WoodsHole, MA, 02543 (e-mail:mkurzorjcurtice•whoi.edu). (Received April2, 2000;revised June27,2000; accepted August8, 2000.)