Observed and modeled seasonal variation of ... - Wiley Online Library

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 105,NO. D7, PAGES 8891-8900,APRIL 16, 2000

Observedand modeledseasonal variationof 13C,180,and 14C of atmosphericCO at Happo, a remote site in Japan, and a comparisonwith other records ShungoKato, YoshizumiKajii, andHajime Akimoto Research Centerfor AdvancedScienceandTechnology,Universityof Tokyo, Tokyo,Japan

Maya Brfiunlich,ThomasR0ckmann,and Carl A.M. Brenninkmeijer Atmospheric ChemistryDivision,Max PlanckInstitutefor Chemistry,Mainz, Germany

Abstract. Measurements ofthe13C/12C and180/160 ratiosand14Cof atmospheric CO were carriedoutfor 2 years(1997and1998)at Happo(36ø41'N,137ø48'E,1840m abovesealevel) in Japan.Thisis thefirstmeasurement of time seriesfor isotopiccompositions of CO at a remotesitein Asia.Theseasonal cycleofthe•8Ofi60 ratioshows a maximum of 10%0 (Vienna SMOW) in Februaryanda minimumof 2%0in July.This minimumvalueis heavierthanthe valuesreportedfor high latitudein the NorthernHemisphere,andthis indicatesthat CO from

fossilfuelcombustion, whichhasa large•8Ofi60ratio,affects midlatitude in theNorthern Hemisphere considerably. Onthecontrary, the13C/12C ratioshowsa clearseasonal variation withlittlescatter; maximum-24.5%0(ViennaPeedee belemnite) in April andminimum-28.5%0 in July-August. The seasonal variationat Happohasdifferentvaluesandphasescompared to thosein theSouthern Hemisphere andin thenorthernhighlatitude.14COconcentration at Happois similarto that at highlatitudein the NorthernHemisphere. A simpleboxmodelcalculationis presented for theseseasonal variations,andthemodelreproduces theobserved seasonal cyclesof CO concentration, 180/160 ratio,and•4COconcentration withinthelimitationsof the

simplified model,butnot13C/12C ratio.Toreproduce thespringmaximum of COconcentration, an enhanced CO production frombiomass burningor otherCO sources in springareinferredin east Asia.

1.

therearethreestableisotopomers (•3CO,C•sO,andC•70)and oneradioactive variety(•4CO).Stableisotope measurements of

Introduction

Carbonmonoxide(CO) is oneof the importantcomponents in the atmosphere as it is a mainreactionpartnerof OH, which is responsible for scavenging mostof tracespeciesin the atmosphere[Loganet al., 1981].TherearefourmainCO sources: fossil fuel combustion,CH4 oxidation,nonmethanehydrocarbon (NMHC) oxidation,andbiomassburning.Fossilfuel combustion takesplacemostlyin the NorthernHemisphere,whereasbiomassburningis importantin the tropicsandhasa strongimpact on atmospheric chemistry[Crutzenand Andreae, 1990] especially in the SouthernHemisphere.The sink processof atmosphericCO is mainly oxidationby OH radical.SinceCO has a relativelyshort lifetime in the atmosphere(about I month in summerand about 6 monthsin winter) and has a variety of sources,CO concentrationin the atmospherehas considerable spatialandtemporalvariationsas comparedto longerlived species. Presently,a flask samplingnetwork coversmost of the world andrevealsa globalpictureof spatialandtemporalvariationonCO [Novelliet al., 1998].Nevertheless, it is still difficult to quantifythe regionalandglobalbudgetof atmospheric CO in sufficient

detail.

Isotopicmeasurement is a usefulmethodfor better understandingsourceapportionment. In the caseof atmospheric CO,

CO havebeenmadein rural Illinois [Stevenset al., 1972], in Antarctica [Brenninkmeijer,1993], in southernmidlatitude [Brenninkrneijer, 1993], in northernlow latitude[Riickrnann et

al., 1998b],and in the Arctic [Riickrnann and Brenninkrneijer, 1997;Riickmann et al., 1998b].Aircraftmeasurements showthe latitudinalisotopic distribution [MakandBrenninkmeijer, 1998].

Measurements of •4COwereperformed ontheground in thetropics[Mak andSouthon,1998] andby aircraftwhichshoweddifferencesbetweenthe Northernand the SouthernHemisphere [Mak,1992;Mak et al., 1994].Measurements werealsomadein the uppertroposphere [Brenninkmeijer et al., 1995, 1996]. It shouldbenotedthatnomeasurement of isotopic compositions on CO in Asiahasbeenpublished. Thepurpose of thisstudyis to presentthe first observationalresultsof simultaneousmeasure-

ments of C•80,•3CO,and•4COfora complete seasonal cyclein eastAsia at Happo,a midlatituderemotesite in Japan,andto comparewith otherrecords.Box modelcalculations were performedfor the isotoperatiosaswell for the concentration in order to get an insightof sourceapportionment at the midlatitude in east Asia. Model calculations for the CO isotopicratio

(•3C/•2C) havebeencarriedoutfor theSouthern Hemisphere [Manninget al., 1997], andfor biomassburningconditions in Brazil[Connyet al., 1997].Sucha calculation hasnotbeenper-

Copyright 2000bytheAmerican Geophysical Union.

formed for•80/•60ratiodueto thelackof enough observational

Papernumber1999JD901144.

EastAsia is oneof the mostimportantsourceregionsof anthropogenic CO dueto its largepopulation andrapidindustriali-

0148-0227/00/1999JD901144509.00

data for •O/•60 ratios in CO sources.

8891

8892

KATO ET AL.: ISOTOPIC COMPOSITIONS OF CO AT HAPPO, JAPAN

zation. However, there are only limited measurements of CO concentration

at remote sites in east Asia. Seasonal variation

and

air masscategorization of CO concentration havebeenreported at Oki Island,southof theSeaof Japan[Jaffeet al., 1996;Kajii et al., 1997; Pochanart et al., 1998; Pochanart et al., 1999;

Happo

Narita et al., 1999],andat Happo,a mountainsitein the middle of the mainislandof Japan[Kajii et al., 1998].Verticalprofile and seasonal variation

of CO concentration

was also retrieved

fromgrand-based infraredsolarspectraat Rikubetsu [Zhaoet al., 1997].No isotopicmeasurement of CO at a remotesite in Asia hassofar beenreported. 2.

Old,

Method

2.1. IsotopicCompositionsof CO Sources Stableisotopiccomposition is expressed with the6 notation,6

= 1000X(R,ample/Rstandard - 1) %0,where R = •sO/•60 or 13C/12C. Standardsare Vienna SMOW (V-SMOW) -CO2 and Vienna Peedeebelemnite(V-PDB) for oxygenandcarbon,respectively.

Theamount of •4COis expressed in unitsof molecule cm-3at the standard condition(1013.25hPa,273.15 K). The carbonisotopicratio of CO from automobileexhaustis Figure 1. Locationof samplingsite, Happo (36ø41'N, closeto the isotopicratioof thefuel, andStevens et al. [1972]es- 137ø48'Eand1840m abovesealevel).The locationof Oki Island is also shown. timated-27.4%0for the world average.CH4oxidationis believed to produceverylightcarbon(-52.6%o),whichwasdeducedfrom combining the CH4 isotopicratio (-47.2%0[Loweet al., 1994]) thatthe sampledair at thissite is basically and the kineticisotopeeffect(KIE, 5.4%0)of CH4 oxidationby ness,it is expected OH [Cantrellet al., 1990].CO fromNMHC is estimated to have free from localpollution.Possiblelocal pollutionsourcesare of snowmobiles, but pollution 6•3C= -32.2%0 [Stevens andWagner, 1989].Thecarbon isotopic bonfireof garbageandapproach ratio of CO emittedfrom biomassburningwas measuredand it by theseCO sources wasavoidedduringsampling. At thissamconwascloseto originalbiomass[Stevens andEngelkemeir, 1988], plingsite we alsohavebeenmeasuringCO concentration from1996with a nondispersive infrared(NDIR) CO whereasKato et al. [1999b]observeda largeKIE duringsmol- tinuously analyzer(KimotoModel540S), andozonefrom 1994by a UV dering. Concerning the oxygenisotopicratio of CO sources,automo- absorption analyzer (TECOmodel49) [Kajiiet al., 1998].CO

bileexhaust COhasalmostthesame6•80asatmospheric oxygen (23.5%o),which is clearlyheavierthan in the atmospheric CO [Stevens et al., 1972;Sakugawaand Kaplan, 1997;Kato et al., 1999a].AlthoughCO from CH4 andNMHC oxidationwas estimatedearlierincorrectly by Stevensand Wagner[1989]to have

6•80= 14.9%o, Stevens [1996]revisedthevalueto about-3%0. Brenninkmeijer and R6ckmann[1997] also proposeda light valueof about6•80= 0%0.Theoxygenisotopic ratioof CO from

biomass burningwasmeasured as 6xsO= 18%o[Stevens and Wagner, 1989],whereas a largeKIE wasalsoobserved for6•80

concentration has small diurnal variation which maximizes in the

afternoonup to 10%, but ozonehas almostno diurnalvariation

exceptfor spring-summer, when photochemical productionof ozonegiveshigherconcentration at daytimeby abouta fewppb. Thustheeffectof theupslope/downslope exchange of air along mountain slopeis notthoughtto affecttheisotopicdataof CO at Happosignificantly. Air samples weretakenevery2 weeksfor 2 yearsfromFebruaryin 1997to January1999.Sampling at Happowasmadein daytime,but the collected time wasdifferentfor eachsample.

Air wassampled witha modified 3 stageRIX compressor [Mak andBrenninkmeijer, 1994] into 5.9 L aluminumair cylinders (Scott-Marrin Inc.)at about100bar.At theinletof thecompressor,sampledair was driedby meansof Drierite(Hammond, Thebackground dependenton pressure:1.006 at 1 atm anddecreases as pressure Ohio).It takesabout1 hourto fill onecylinder. in thissampling methodis lessthan2 decreases.On the contrary, oxygen shows a negative K1E levelof CO concentration ppbvandtheCO canbe keptstablein thecylinders [Makand (kc•6o+oH / kc•So+oH = 0.990), nearlyindependent onpressure. duringsmoldering [Katoet al., 1999b]. The oxidationof CO by OH radicalsinducesa KIE andthus affectsisotopiccompositions of atmospheric CO significantly [Stevens and Wagner,1989]. For carbon,k•-cO+OH / k'•cO+OH is

2.2.

Measurement Site and SamplingProcedure

Happois locatedat latitude36ø41N,longitude137ø48•E, and at 25 km distancefrom the Sea of Japanas shownin Figure 1. For the purposeof later discussion, the locationof Oki is also shownin Figure 1. The nearestcity to Happois Shinano-omachi (population- 31,000) at 25 km southandthe largercity of Nagano(population - 360,000)is locatedat 50 km to the east.The samplingsite is on a mountainslope(1840 m abovesealevel), near a terminal

of ski lifts. Because of its elevation

and remote-

Brenninkmeijer, 1994].The filled air cylindersweresentto the Max PlanckInstitutein Germanyand processed therefor isotopicanalysis. 2.3. AnalyticalProcedure

Detailedanalyticalprocedure for CO isotopiccompositions hasbeendescribed byBrenninkmeijer [1993]anda briefdescriptionis givenhere.The air sampledin thecylinderis introduced into the CO extraction line via a mass flow controller at about 5

L min4 (STP).At first,twoveryeffective liquidnitrogen cold

KATO ET AL.: ISOTOPIC COMPOSITIONS OF CO AT HAPPO, JAPAN

trapsnamedRussian doll trap[Brenninkmeijer, 1991]remove H20,CO2,N2¸, andhydrocarbons (except for CH4).Next,CO whichpasses through thetrapis oxidized to CO2by means of Schiitze reagent (I205supported onsilicagel).Theproduced CO2 is condensed in anotherliquid nitrogencold trap downstream. Afterthe air processing is finished,theCO2derivedfromCO is driedwith P205andits amountis measured with a piezoresistive

300

i

i

i

i

8893

i

i

i

i

i

i

-

100

i

o

Ie• I• ßß•lß OO O

200 rj

o

ß

-

1

(a):

ß

e• _

I 1

pressure sensor in a calibrated smallvolume.FromtheCO2volumethusmeasuredand the total volumeof sampledair measuredby a calibratedgasmeterat the airflowexhaust,the CO concentration of the sampledair canbe deduced.Stableisotope ratiosare measuredby an isotopemassspectrometer (Finnigan MAT model252). Theoxygenisotopic composition of CO canbe calculatedfrom the oxidizedCO2 isotopiccomposition,since oxidationprocess by SchQtze reagentis just addingoneoxygen atomontoCO, withoutcausingany isotopicexchange[Stevens andKrout, 1972;Brenninkmeijer andR6ckmann,1997].For the

0 12

8

-I

I I

i ß

difference between the results of the extraction method, which

O1

I t

I

I

I ß

I

0

t I --

I

I

-

I

-24

I o

-:28

1998are alsoplottedby opentriangleswith 25-75%range.The

0 0 I

CO concentration, 0.2%0for 6x3C,0.4%0for 6xgO,and2% of x4CO concentration [Brenninkmeijer, 1993].

In Figure2a the CO concentration as measuredby the extraction methodin 1997 and 1998 is plottedby circlesandsquares, respectively. The medianof the monthlyaveragedCO concentration measuredcontinuously by the NDIR methodfrom 1996 to

•

4

-26 -II • II

Results

I

l

measurements of x4Cin CO, theextracted CO2samples derived fromCO areanalyzedby an accelerator massspectrometer after dilutedwithx4C-free CO2.Theerrorsof theseanalysis are2% of

3.

el

l

(c)_

I

I

I

0

n

-27 I

I

20 16

lie

I

i

ß

I

_

I I

ß

I 1 _

$1

12

-

.•

I

_

-29

I

I

o

rn I

I I

o

I _

8

takesabout1 hourfor air sampling,andthecorresponding hourly F M A M J J A S O N D meanNDIR data is lessthan20% (typicallywithin 10%). As seenin Figure2a, someCO concentration datameasuredby the Figure 2. Concentration and isotopiccompositions of CO extractionmethoddeviatedfrom the monthlyaveragedvalue of measured at Happo.(a) CO, (b) OgO,(c) x3CO, and(d) x4CO. thecontinuous measurement. This canbe ascribedto the tempo- Circlesandsquares represents the measurements in 1997 andin ral fluctuationof CO concentration; the corresponding hourly 1998,respectively. Opensymbols represent dataaffectedby polmeanNDIR data alsodeviatedfrom the monthlymeanvalues. lution. CO concentrations measuredby the NDIR methodare Particularlyin summer,Japanis underthe Pacificmarine air alsoplottedin Figure2a by opentriangles.They are given as massandair parcelspassingoverthe mainlandof Japansome- medianwith errorbarsindicating the25 and75% spread. timespickup air pollutantsof largecities,andhighCO concentrationat Happois observed.The data ascribedto this caseare denotedby open symbolsin Figure 2. When ignoringthese canbe explainedqualitatively asfollows.Durhigherconcentration data,the seasonalcyclehasthe maximum variationof fixgO heavycarexhaust (fixgO is about CO concentration of about230 ppbvin April andthe lowestCO ingwinter,COfromisotopically because of low OH concentration. Toward concentration of about120 ppbvin July.This seasonalvariation 23.5%0)accumulates increases andenhances production of is similarto the measurements at Oki Islandin the Seaof Japan summer,OH concentration [Pochanart et al., 1999]bothin seasonal cycleandconcentration lighter CO from methane and NMHC oxidation. At the same (maximum248 ppbvin April andminimum105ppbvin August). time, the reactionof CO + OH whichis accompanied by a negaTheseCO concentrations are higherthan the recordat Oregon tive KIE further reduces fixgO. These two factors would contrib(45øN,maximumabout160 ppbvin April andminimumabout ute to the low fixgO in summer. It should also be noted that the

75 ppbv in July-September) [Khalil and Rasmussen, 1984], outliers of CO shownby opensymbols havealsoheavierfixgO whichis locatedat slighthigherlatituderepresenting mostlyma- valuesin Figure2a. This fact agreeswith the interpretationthat rine air mass.This differencewouldindicatethe regionalan- theseair parcelspassedoverlargecitiesin Japanandwere polthropogenic pollutionof eastAsia andincreaseof hemispheric lutedby automobileexhaust. CO source between 1980-1982 and 1997-1998 to some extent. Theresultsof themeasurements of x3COareshownin Figure Figure 2b showsthe observedseasonalvariationof 2c. Thereis a very clearseasonalcyclewith a maximumof about whichnearlyparallelsthat of the CO concentration: a winter- -24.5%0aroundApril-May, and minimum of-28.5%0 in Julyspringmaximumand a summerminimum.However,the maxi- August.In contrastto the largeanomalyfor thepollutedsamples mumof fixgO appears earlier,in February ascompared to April foundin Figures2a and2b, the deviationsof thesedatafrom the for the CO concentration. Thus 6xgOhas a maximum of 10%o in

seasonal variation curve of 6x3C are rather small. This could be

Februaryand a minimumof about2%0in July. The seasonal explained by anassumption thatthedifference between the6x3C

8894

KATOETAL.:ISOTOPIC COMPOSITIONS OFCOAT HAPPO, JAPAN

value of CO in the background air and that of pollutedCO fore14CO is almost independent fromCOconcentration change sources is small.Theseasonal variationof b•3Cis explained as due to pollutionby fossilfuel combustion. When significant follows.During winter, the OH concentration is low and CO amounts of CO from localbiomassburningcontributes, $4CO is affectedby the small amountof 0.38 molecules emissions fromcombustion sources (carexhaust, domestic heat- concentration ing, andindustry;about-27%o)are dominantsources. However, cm -3per10ppbCO[Bergamaschi etal., 1998].Using backward the observed maximum of-24.5%o is much heavier than the iso- trajectorycalculation,the clearly lower data of •4COdeviated toperatioin suchsources. Thereforeanother heavyCO sourceis fromtheaverage seasonal cycleindicated byarrows in Figure2d neededin thisperiod.No suchheavyCO sourcehasbeenknown are attributedto air from lower latitudes,where $4COconcentraexceptfor C4plantsburning,whichtakesplacemainlyin thesa- tionis lowerthanat midlatitude. Theseasonal variation in Figcycleof OH radical vannaregions (C4plantshavea b•3Cvalueof about-13%o). ure2d wouldbe explainedby the seasonal Asidefromthecontribution of C4plantsburning,thepositive•3C concentration, whichis lowestin Januaryandhighestin July KIE of CO + OH reactionneedsto be considered. In spring,OH [Spivakovsky et al., 1990]. The seasonaldependence of trobeginsto increaseand the reactionof CO + OH makesthe at- posphere-stratosphere exchange needsto be considered for more

mospheric CO moreenriched in •3C.Duringsummer productiondetailed discussionon $4CO.

ofverylightCOin •3CfromCH44-OHreaction (about -52%•)is enhanced andthe atmospheric CO becomes lighterin b•3Cand 4. givesa minimumof b•3C.The enriched valuesof •3COin winter ascompared to thatof combustion sources maybe partiallydue to a KIE effect. These two factors would determine the basic fea-

Discussion

4.1. Comparisonwith other remote sites

tureof a seasonal cycleof b•3Cwitha maximum in latespring

Thereis onlya limitednumberof published dataof seasonal (April-May)anda minimumin summer. cycles forisotopes in atmospheric COatremote sites. In Figures Figure2d showstheresultsof •4Cmeasurements. A clearsea- 3a-3d,reported seasonal changes in COconcentration, b•sO,b•3C, sonal variation can be seen with a maximum of about 23 mole-

and•4CO areplotted together. Details ofthesampling sites and

culescm-3in February-March anda minimum of about9 mole- measurements are summarized in Table 1. Alert (81øN)

cules cm-3inJuly-August. Themaximum of14CO appears earlier [R6ckmann etal., 1998b]andSpitsbergen (79øN)[R6ckmann et than that of CO concentrationand b•3C.The values of $4COhave al., 1998b]represent northern highlatitude,lzafia(28øN)

smalldeviations fromtheseasonal cycle,evenfor pollutedair [R6ckmann et aL, 1998b]represents northern low latitude,Barmeasurements. $4CO is produced by cosmic raysandthe•4CO ingHead(41øS)[Manning etal., 1997]represents southern mid-

sink is solelya reactionwith OH. CO fromfossilfuel combus- latitude, andScottBase(78øS)[Brenninkrneijer, 1993]repre-

tioncontains no$4C since itslifetime isabout 5730years. There- sents southern highlatitude. For•4CO,results atJi. ilich(51øN) SH months

SH months

J A

300

S

0

ND

J FM

AM

i

i

J

i

i

J

-2

A i

S i

O i

ND i

i

FM i

AM

i

i

J

i

i

(c)-

4-

4'

+

:.H.• .H.++

_

200

J

i

_+

+

+%.,+q'+'•

+ + o

+

+

•ß •).

+ +++'

+ +4- ++-k -+

ø 100

-

0

{ 0 16

i

>

8

i

}

i

i

i

i

o o

i

(d)4-

_

+? +

.+*;• 4-+;,+

n++++4-ff +ß

4-

Tm+

+

-++4-+ +++11=1:• ß 'bøo'm, red,

+4+

+

4

ß

o

i

o

(b•

• 12 •

i

o

ß ++

+

ß

+

0

-

I

J

I

F

I

MA

I

I

M

oo I o {

I

{

J A

NH months

I

S

I

O

n+

+o +o

.. J

_

o '•'•-,e •d• +++++ + o •oo • o + E!El+ 0•

.

-

•

•

J

AS

O

v v

I

N

D

J

F

M

AM

J

O ND

NH months

Figure3. Comparison ofseasonal variations ofreported COconcentrations andisotopic compositions overthe world: (a)CO,(b)C•SO, (c)•3CO, and(d)•4CO. Pluses, Happo (37øN); SolidSquares, Alert(81øN)andSpitsbergen(79øN);Solidtriangles, lzafia(28øN);Opentriangles, Barbados (13øN);Opencircles,BaringHead (41øS); Solidcircles, ScottBase(78øS); Opensquares, Ji. ilich(51øN).Theresults in theSouthern Hemisphere areshifted byhalfa yearforeasier comparison ofseasonal variation (upper scale).

KATO ET AL.: ISOTOPIC COMPOSITIONS OF CO AT HAPPO, JAPAN

8895

Table1. Remote Sampling SitesforSeasonal Variationof CO Isotopic Measurements Site ScottBase BaringHead

Latitude 78øS 41øS

Location Antarctica New Zealand

SamplingPeriod July 1991 - Feb. 1992 June1989 - June1995

Barbados

13 øN

Barbados

Jul 1996-

Izafia Happo Jfilich Spitsbergen

28øN 37øN 51øN 79øN

Spain Japan Germany Norway

Nov. 1996 Feb. 1997 - Jan. 1999 May 1977 - Sep.1978 April 1995- June1995 Nov. 1996 -

Alert

8 iøN

Canada

Nov. 1996 -

Reference Brenninkmeijer [ 1993] Manning et al. [1997] Brenninkmeijer[ 1993] Brenninkmet.'jer et al. [ 1992] Mak and Southon[1998] R6ckmannet al. [ 1998b] this work

Volz et al. [ 1981] R6ckmannand Brenninkmeijer[1997] R6ckmannet al. [1998b] R6ckmannet al. [ 1998a] R6ckmannet al. [1998b]

R6ckmannet al. [1998a]

[Volzet al., 1981] and at Barbados(13øN) [Mak and Southon, 1998] are alsoshown.In thesefiguresthe date in the Southern Hemisphere (ScottBaseandBaringHead)are shiftedby half a yearfor easiercomparison of the seasonalvariationbetweenthe hemispheres. The data of Alert and Spitsbergen, BaringHead, andBarbadosare monthlyaveragevalues.As the resultsmeasuredat Spitsbergen andAlert are similar,the data are grouped togetherand plots are given by the samesymbol.Since these measurementsare based on the same standard gas

[Brenninkmeijer andR6ckmann, 1997]exceptfor14CO, theserecordscan be comparedwith eachotherdirectly.Stevenset al. [1972]reportedsimilarseasonal variationsof CO concentration and isotopiccompositions at rural Illinois;their data are not includedin Figure3 becauseof difficultyof readingvaluesfrom theirfigures.Thereis no published reportfor stableisotopes in CO in Asia,andat a siteasremoteasHappoin thenorthernmidlatitude,where regionalanthropogenic CO emissionsare the

heavierCO emissions from combustion (613C= -27 %0)are smallerin the SH, whereasthe input of isotopicallyvery light

CO fromCH4 (613C= -52 %0)is aboutthesamein bothhemispheres. The6•3Cmaximum at Happoappears in April,whereas in the northernhigh latitudes(Spitsbergen) thereseemto be a

steep6•3Cincrease duringpolarsunriseperiod.R6ckmann and Brenninkmeijer [1997]explained thisrapid6•3Cincrease asa result of the KIE of CO oxidationby OH. However,CO formation from CH4 oxidationby OH, whichgivesCO muchdepletedin

6a3C, wasnotconsidered in theirexplanation. In theSH,the613C maximized in December (earlysummerin theSH). Thisshiftof peak period is thought to be causedby biomassburning [Brenninkmeijer, 1993;Manninget al., 1997].After maximum,

613C decreases rapidlyandgivesminimumin Marchin theSH. In Figure3d, •4COmeasurements overtheworldare shown. Although the•4COproduction is knownto be affected bythe11

yearsolarcycle,thisfactoris notconsidered hereandall dataare plottedtogether. The datafor BaringHeadarethe averaged valFromFigure3a, onecanfind that CO concentrations aremuch uesfromBrenninkmeijer [1993].In spiteof latitudinaldifference, lowerin theSouthern Hemisphere (SH) andshowmuchsmaller measurements in themiddleandhighNorthernHemisphere have

most active.

seasonal variations thanin the NorthernHemisphere (NH), as similarvaluesandseasonal cycles.Notethattheperiodof thesohasbeenwell known[Novelliet al., 1998]. Spitsbergen and lar cycle is almost same for the measurementsat Jfilich, Alert measurements are about30 ppbvlowerthanthe Happo Spitzbergen, andHappo.Themaximumappears in February, and measurements. Izafia and Barbados are located at low latitude theminimumappears in July.The measurements in thetropics, andhavea nearlyconstant CO concentration of around100ppbv Barbados, showthelowest14CO. Mak andSouthon [1998]exwith a small seasonal variation with a minimum in summer. The

plained thattheselow•4COwerestillhigherduringwinterthan

difference betweenthe NH andthe SH is largestin February- expected by modelingstudies,and the transportof •4COfrom March(winterin theNH) andminimumin July-August (summer higher latitudewas suggested.In the SH the maximum and in theNH). minimumperiodis almostthe sameas in the NH (half year As shownin Figure3b, thereis largedifference of 6•80 be- shifted), butthe•4COconcentration islower,especially inwinter. tweenthe NH and SH. The interhemispheric differenceis dis- Because the•4COproduction bycosmic rayis nearlysymmetric tinct in winter and is muchsmallerin summer,when 6•80 at in bothhemispheres [J6ckelet al., 1999],Brenninkmeijer et al. northernhighlatitudeapproaches to thatof the SH. About90% [1992]initiallypointed outa possible interhemispheric asymmeoftheheavy COfromcombustion (6•O isabout23.5%t>) isemit- try of averageOH concentration: higherin the SH than in the tedin theNH, andthiscauses theinterhemispheric difference of NH.

6•O. In thenorthern highlatitude, 6•O iscomparable to northern midlatitudeduringwinterandspring.SinceOH concentrationsare low in winter andonly a smallfractionof CO is consumed, polluted air emitted in midlatitude accumulatesand

4.2.

Box Model

Calculations

Box modelcalculationsare performedfor the CO concentra-

wouldmix with the high latitudein the NH. Duringsummer, tion, 6•80, 6•3C,and •4CO.No modelcalculationhas been re6•O atmidlatitude does notbecome aslightasathighlatitude in ported for stable isotopesin the NH, and there is a twothe NH. This canbe explainedby the fact that the sources of dimensional (2-D) calculation onlyfor6•3Cin theSH [Manning havebeenreheavyCO aremostlyin themidlatitude andthefreshlyemitted et al., 1997].For •4CO,somemodelcalculations ported[Volzet al., 1981;Mak, 1992;Maket al., 1994;Mak and CO affectspredominantly in themidlatitude. Seasonal variations of 613Careshownin Figure3c.The6•3C Southon,1998;J6ckelet al., 1999]. Thereis a limitationfor a is heavierin the NH thanin the SH. Thiscanbe explainedas simplebox model;it can not representthe entire hemisphere

8896

KATO ET AL.: ISOTOPIC COMPOSITIONS OF CO AT HAPPO, JAPAN

Table 2. Production RatesandIsotopicCompositions of CO Sources UsedfortheBoxModel FossilFuel 640x0.94 23.5 -27

production rate•) {5•SO, %0 {5•3C, %o

Biomass Burning 520x0.63 18 -27

NMHC 320x0.72

CH4Oxidation

5 -30

0 -52

') Unitis(TgCOyr-t).Multiplied factors arerepresenting theproportion intheNorthern Hemisphere.

properlyfor a relativelyshort-livedspecieslike CO, andthereis of the CO production ratesaregivenin Figure4. Sinceforestfire a difficultyto choosep?operparameters. Evenso, a simplebox is enhanced duringspringin the NH [e.g.,Cahoonet al., 1994], modelis usefulto get insightintothebasiccharacters of sources 6/10 of total BB productionis presumedto occurevenlyover a and sinks. yearand4/10 of themis superimposed on it asa GaussiandistriIn the presentbox modelcalculationthe intervalof the calcu- butionwith a peak in April with the width of about3 months. lationstepis 1 day,andthe resultsin the thirdyearareusedfor The seasonal variation of NHMC emission is derived from fluxes the discussion. The box is assumedto presentthe Northern of isoprene andterpenes at 10ø-90øNlatitude[Guentheret al., Hemisphere. Sincethe CO observations at Happowere doneat 1995] (The data are availableon the homepageof The Global 1840m altitude,theparameters (temperature, pressure, OH con- Emissions Inventory Activity (GEIA; centration, andKIEvaluefor613C) arededuced fromtheaverage http://blueskies.sprl.umich.edu/geia/)). The CHqconcentration is of under0-3 km altitudeweightedby molecular density.The at- assumed to be constantat 1780 ppb and its seasonalvariationis mospheric scaleheightis assumed to 7.5 kin,thatis,thepressure ignored.The rateconstantof CHq+ OH usedin thecalculationis is assumed to reduce 1/e of the surface at this altitude. The CO

2.45x10-nx exp(-1775/T)cm3 molecule -1s-1 [DeMoteet al.,

sourcesconsidered are fossil fuel combustion (FF), biomass burning(BB), oxidation of nonmethane hydrocarbons (NMHC), andoxidationof CHq.Only the reactionwith OH radicalis considered asa sink.Directemission fromoceanandplants,andup-

1997].Sincethisrateconstant hhstemperature dependence,

weightedaverageof the temperatureby the moleculardensity under3 km (279 K) is used.The seasonal variationof the temperature(279 _ 10 K) is assumed to follow a sinecurvewith a takeby soil are not considered here.The valuesusedfor consid- maximumin July and a minimumin Januarywith an amplitude eredCO sources are listedin Table2. The CO production rates of 10 K. The conversion factorof CO productionfrom oxidized for FF, BB, and NMHC are calculated from the recent estimate

CHqis assumed to be 0.8 [Loganet al., 1981]. OH seasonal of globalemissions (640,520,and320 Tg CO yr-1,respectivelyvariationis deducedfrom the estimationby Spivakovskyet al. [Novelliet al., 1998])andareportioned to theNH (94, 63, and [1990] at midlatitude(28øN-52øN).The weightedaverageOH 72%, respectively), referringto Loganet al. [1981].The CO concentration by moleculardensityunder3 km for four seasons emittedfromthesesurfaceCO sources (FF, BB, andNMHC) is is smoothedand its seasonalvariationis shownin Figure 4. is 17.1x10 s molecules assumed to mix intothewholetroposphere (under10 kin), and Daily averageof the OH concentration the CO production ratesin the box are deduced.However,since cm-3forthemaximum valuein Julyand2.4x10 smolecules cm-3 thereis generally a verticalconcentration gradient with higher for the minimumvaluein January.The rate constantof the sink valuesnearsurface[Zhaoet al., 1997],CO concentration in the reactionCO + OH is 1.5x10-13x(l+0.6xP) cm3 molecule -1s4 lower altitude would be underestimated when we use the CO [DeMoreet al., 1997], whereP indicatesthe pressure in units production ratesdeducedfrom the averageof the wholetro- atm.The weightedaverageby moleculardensityunder3 km deposphere. Thereforethe production ratesof FF, BB, andNMHC ducestheemployed P (0.84 atm). The 61sO and 613Cvalues for CO sources from FF are estideduced abovearemultipliedby 1.3. Thefactorof 1.3 is selected to give a reasonablecalculationresulton CO concentrationcom- matedto be 23.5%0(atmospheric oxygenisotopiccomposition) paredto ourobservation at Happo.Assumed seasonal variations and-27.4%•[Stevens et al., 1972],respectively. ForBB,61sO =

]

[

[

[

[

]

[

]

[

]

[

I2

[OHI/-•.

1.5

]

BB

1.5.•?• FF •

o.s

'•

o.s

o

o

j

F

M

A

M

J

J

A

S

O

N

D

Figure 4. Theseasonal variationsof CO production ratesfor theboxmodelcalculation (rightscale).The seasonalvariationof OH concentration is alsoshown(left scale).

KATO ET AL.' ISOTOPIC COMPOSITIONS OF CO AT HAPPO, JAPAN

25O

200

=-- 0 -- ..._.

o

,

0

O0

(D

8897

der 3 km altitude.The 14COproduction rate usedhere(1.18 molecules cm-2s-1)is selected to givethebestfittingof theob-

0

served data as discussed later.

¸

The calculatedresultsare plottedby solidlinesin Figures5a-

5d forCOconcentration, 61sOand613C and14CO, respectively. In

150 100 E

oo

o

-

spite of usinga simple one-boxmodel, the calculationsreproduce the observeddata reasonablywell for CO concentration, 61sOand 14CO,but not for 613C.The calculatedCO maximum andminimumin concentration appearsin March-AprilandJulyAugust,respectively, reproducing the observedseasonalcycleas

0

50 0

•

12

•

10

6 4

-

ooo

derived from the NDIR

-

o 'xX,%

_

¸ ¸•

-

m

2

o

culatedspringpeakhasresultedfrom the enhancedCO production of biomassburningin spring.When no seasonalvariationof the BB sourceis assumed,the maximumappearsin February,as discussed later. The agreement of the calculated61•Owith the observation is satisfactory; the maximumappearsin February,

oo o

o

o

0

earlierthanCO concentration and613C,andtheminimumappearsin July-August. For613C thedelayed phaseof theseasonal

0 -24

(c)

•

-25

amplitudeof the seasonalvariationis slightlysmallerthan the observedvalues.When the KIE value of 1.006 for 613Cis used, whichcorresponds to the value at 1 atm, the calculatedresults wouldbe shiftedto theheavysideasshownby thedottedline in

• -27

•,• -:28 -29

Figure5c. However,sincethe pressure dependence of the KIE

-30

•,

variationis reproduced: higherin late springand lower in summer. However, the calculated value is about 1%, lower and the

13 -26

•

method. It should be noted that this cal-

valuefor613C iswellestablished [Stevens etal., 1980;Smitetal., (d)

25

•

20

g

•0

•

5

1982;R6ckmann et al., 1998b],usingthe KIE valueof 1.006for thepresentboxmodelwouldnotbejustified.Whenthe613Cvaluesof CO from CH4 or NMHC are assumedto be heavierthan employed, thecalculated 613Cwouldbecomeheavier.However, the seasonal variation becomes even smaller since oxidation of

CHnandNMHC is enhanced duringsummer. Whenthe613C valuesof FF andBB are parameterized to fit the observedresults,

0

J

F

M

A

M

J

J

A

S

O

N

D

theyhaveto be -24 and-23%,,respectively. A valueof 613C= 24%,forFF is unacceptably high,anda contribution of CO from

theburning of C4plants(613C '- -13%,[Deines, 1980])needs to Figure 5. Resultsof the box modelcalculation: (a) CO, (b)

be about30% to deducethis heavyBB value.In the present ClSO,(c) 13CO, and(d) 14CO. Opencirclesaremeasurements at study,the reasonof the discrepancy for 613Ccouldnotbe eluci-

Happoexcludingthe dataaffectedby pollution.Opentriangles dated. Asfor14CO, agreement ofthephase oftheseasonal variain Figure5a are the monthlyaveragedCO concentration meastion is satisfactory. The production rate of 14CO (1.18 molecules uredby the NDIR method.Dottedline in Figure5c is the result when the KIE value for carbon is 1.006.

cm-2s-1)wasselected togivethebestresultasshown in Figure 5d.The14C production isminimum inthesolarmaximum period.

The solarcycleis 11 yearsandthe periodof measurements at Happo(1997-1998)is closeto the solarmaximumperiod.The 18%,[Stevens and Wagner,1989] and613C.m-27%,(typical globalaverage of 14CO production rateis 1.57molecules cm-2s-1 value for C3 plants)are used,respectively.There is no actual in solarmaximum periodandroughly 50% of all 14COis promeasurement of isotopiccompositions of CO from NMHC oxida- duced in thestratosphere [J6ckel et al., 1999].Alsotheproduction. Brenninkmeijer and R6ckmann[1997] proposed0%, de- tionof14CO haslatitudinal gradient andit is 80%oftheglobal rived from the observation in the SH. Here 61sO is assumed to be

average at Happo(37øN)[J6ckelet al., 1999].If thereis no

5%, consideringa contributionfrom the oxidationby ozone, stratosphere-troposphere exchange, the calculated 14COproducwhichis knownto produce veryenrichedoxygen[R6ckmann et tionratein theboxshould be0.63molecules cm'2s-1(1.57x 0.5

al., 1998a].The613C is assumed to be -30%,,whichis slightly x 0.8).Thisvalueis muchsmallerthantheemployed valueof lighterthan613Cof isoprene (-27.9%o [ConyandCurtie,1996] 1.18,whichgivesthe bestresultfor simulating the observed and-29.4%,[Shatkey et al., 1991]).For CH4oxidation, 6180is 14CO. Thedifference wouldbe explained by theintrusion of assumedto be 0%, as estimatedby Brenninkmeijerand R6ck- stratospheric air, andalsomixingof air fromthe highlatitude mann[1997],and613C = -52%*asdeduced fromisotopic compo- since theproduction of14CO islargerinthehigher latitude twice sitionof CH4(-47.2%,)[Loweet aL, 1994]andtheKIE of CH4+ asmuchastheaverage production rate[J•ckeletal., 1999].Mak OH (5.4%o)[Canttellet al., 1990].The employedvaluesof the andSouthon [1998]alsosuggested fastmixingof higherlatitude KIE for CO + OH are0.990 for oxygenand 1.0045 for carbonin

tropospheric air fromthe comparison betweenthe observation thisboxmodel[Stevens andWagner,1989].The lattervaluehas and2-D modelcalculation in thetropics. pressuredependence, and it decreases as the pressuredecrease The estimateof the production rateof CO sources is associ-

(KIE= 1.006at1 atmand1.001at0.5atm[Stevens etal., 1980]). atedwithconsiderable uncertainty. Although theproduction rate Theselected valueisweighted average bymolecule density un- reported byNovellietal. [1998],whichweemployed forthebox

8898

KATO ET AL.: ISOTOPICCOMPOSITIONSOF CO AT HAPPO,JAPAN

300

suggestsan enhancedCO productionfrom BB or other CO sources in springin eastAsia. As for the oxygenisotopiccompositions of CO from CH4 oxidationand NMHC oxidation,only limited informationis avail-

250

2OO

able.Stevens and Wagner[1989] originallyestimated fixso=

150

revisedinterpretation of the Illinois1971data[Stevens, 1996].

15%o,but now it is correctedto be much lower at -3%0basedon a

The other estimationwas proposedby Brenninkmeijerand

R6ckmann [1997],fixsOof about0%0from CH4 oxidationis neededto balancetheobserved resultsin the SH. Theyapplied 0%0and10%ofor isotopiccomposition of NMHC oxidationand statedthat the formergivesbetter agreement with the observa-

50

tion.According to theirstudy,fixsO - 0%0forCO fromCH4oxidationis usedin thepresentboxmodelcalculation. ForCO from J F M A M J J A S O N D NMHC, 5%0is usedsincetheremaybe a contribution fromthe Figure 6. Model resultswith and withoutenhancedbiomass oxidationby ozone,whichis knownto producevery enriched et al., 1998a].In Table3 the maximumand burning in spring. The solid line representsthe basic run oxygen[R6ckmann (biomass burningin springis 4/10 of totalamount),andthedot- theminimum fixsO valuesin theboxmodelresults usingdiffertedline represents no enhancedCO production of biomassburn- entfixsO values forCH4 andNMHCoxidation sources arepreing in spring. sented. Whentheheavier fixsO (15%o) is assumed forCH4oxidation,the resultis tooheavybothfor the maximumandthe mini-

mum.Fromthisresult,it is alsosupported thattheCO produced fromCH4oxidation wouldbe aroundzeroasproposed byBrenninkmeijer andR6ckmann [1997].For NMHC oxidationthere-

sultusingtheheavier fixsO (15%o) is alsoconsistent withtheobservedvalueswithin the fluctuationof the data. However,it is model,did notgivethe uncertainties, thefollowinguncertainties notenough to conclude thattheCO fromNMHC oxidation prowerequotedby otherresearchers [Volz et al., 1981;Sellerand duces heavier 6•sO. Conrad,1987;Pacynaand Graedel,1995]:FF (31-34%), BB (29-76%),andNMHC (50-80%).WhenFF areincreased or de-

creased by50% in theboxmodel,CO concentration, fixsO, and 5. 6x3C become about50 ppb,2%0and1%ohigheror lower,respec-

Conclusions

Isotopicratiosof atmospheric CO are measuredat a remote site in eastAsia, in Japan.The seasonal variationsof fixsOand

tively.In the caseof BB, shift of about30 ppb,0.4%0and0.7%0 are broughtby the 50% changeof the sourcestrength,respectively. Similarly,about15 ppb, -1%oand0.3%0changeare seen for the 50% changeof NMHC productionrate, respectively. Thus, althoughFF is recognizedto be moresensitiveto the concentrationand isotopiccompositions, estimationof FF production ratehaslessuncertaintythanBB andNMHC.

6•3Caredifferent fromtheSouthern Hemisphere andfromthe highnorthernlatitudeandrepresent an enhanced contribution of fossil fuel combustion in northern midlatitude. Box model calcu-

lationsfor northernmidlatitudereproduce the observed results for CO concentration, lSO/X60 ratioandx4COconcentration, but

notforx3C/X2C. Theresults suggest thatthereis anenhanced CO production by biomass burningor otherCO source in springin biomassburning(BB) in the midlatitudein the NH is not availthisregion.ThefixsO valueof CO fromCH4 oxidation will be able. In the lower latitudein the NH, BB takesplacein March lightaround zero.However, thereissomelimitation to a simple (for example,in Thailand (P. Pochanart,personalcommunicaboxmodel.Thereis a difficulty in choosing suitable representation,1999)).BB in Siberiahasalsobeenrecognized asan importivevaluesforparameters. Formoreaccurate descriptions, 2-D tantCO source[Cahoonet al., 1994],whichis enhanced in April or 3-D models will be needed. At this moment there are not Quantitative data of seasonal variation of CO emission from

andMay. In Figure6 the response of the modelto changingthe BB in springis shownfor the CO concentration. The solidline is the resultof the basiccondition(4/10 of totalBB is in spring) andthe dottedline is for no enhancedBB in spring(constant overa year).WithoutenhancedBB in spring,the CO concentration maximumappearsin Februaryand this is not consistent with the observation. From theseresultsthe observed"spring peak,"thatis, thedelayedCO concentration peakin March-April,

enoughmeasurements for detailedmodels,because isotopic compositions of CO seemtobe quitevariablewith latitude.Even

withlargeuncertainty ofmodel inputs, it isagoodfirststepthat thisboxmodelcanreproduce someof theobservations. Acknowledgments.We would like to thank W. Hanewackerfor CO

analysis operation andtheair compressor construction. We wouldalsolike to thankS. WakamatsuandNaganoResearch Institutefor HealthandPol-

Table3. MaximumandMinimumfixsO Valuesof COin theBoxModelUsingDifferent fixsO ValuesforCH4 and NMHC

Oxidation

Sources

Estimated b•So of CO Sources

CH4Oxidation

0 %oa) btSomaximum,%o btSominimum,%o

10.1 0.2

NMHC Oxidation

15%o 11.8 4.1

0 %o 9.5 -1.0

•)Standard condition ofthepresent boxmodel calculation.

5 %oa)

Observed Values

15%o

atHappo

10.1

11.2

10

0.2

2.7

2

KATO ET AL.: ISOTOPIC COMPOSITIONS OF CO AT HAPPO, JAPAN lutionfor supporting air samplingat Happo.This researchis supported by CREST(CoreResearch for Environmental ScienceandTechnology) of the JapanScienceandTechnology Corporation.

8899

Kato, S., H. Akimoto, M. Brfiunlich, T. R6ckmann, and C.A.M. Brennink-

meijer,Measurements of stablecarbonandoxygenisotopiccompositions of CO in automobile exhausts and ambient air from semi-urban Mainz,

Germany,Geochern.J., 33, 73-77, 1999a. Kato, S., H. Akimoto, T. R6ckmann, M. Brfiunlich, and C.A.M. Brennink-

References Bergamaschi, P., C.A.M. Brenninkmeijer,M. Hahn, T. R6ckmann,N.F. Elansky,I.B. Belikov,N. Trivett,andD. Worthy,Isotopeanalysisbased sourceidentificationfor atmospheric CH4 and CO sampledacrossRussiausingtheTrans-Siberian railroad,J. Geophys.Res.,103, 8227-8235, 1998.

Brenninkmeijer,C.A.M., Robust,high-efficiency,high-capacitycryogenic trap,Anal. Chem.,63, 1182-1184, 1991.

Brenninkmeijer, C.A.M.,Measurement of theabundance of 14COin theatmosphere andthe13C/12C and180/160 ratioofatmospheric COwithapplicationsin New Zealand and Antarctica,J. Geophys.Res., 98 , 10,595-10,614, 1993.

Brenninkmeijer, C.A.M., and T. R6ckmann,Principalfactorsdetermining

the180/160ratioof atmospheric CO asderived fromobservations in the SouthernHemispherictroposphere and lowermoststratosphere, J. Geophys.Res.,102, 25,477-25,485, 1997. Brenninkmeijer,C.A.M., M.R. Manning, D.C. Lowe, G. Wallace, R.J. Sparks,andA. Volz-Thomas,Interhemispheric asymmetryin OH abun-

danceinferred frommeasurements of atmospheric 14CO,Nature,356, 50-52, 1992.

Brenninkmeijer, C.A.M., D.C. Lowe,M.R. Manning,R.J. Sparks,andP.F.J.

Velthoven, The13C,14C,and180isotopic composition of CO,CH4,and CO2 in the highersouthernlatitudeslower stratosphere, J. Geophys. Res., 100, 26,163-26,172,

1995.

Brenninkmeijer,C.A.M., R. Miiller, P.J. Crutzen,D.C. Lowe, M.R. Man-

ning,R.J. Sparks,and P.F.J.Velthoven,A large 13COdeficitin the lowerAntarcticstratosphere dueto "ozonehole"chemistry:Part I, observations, Geophys.Res.Lett.,23,2125-2128, 1996. Cahoon,D.R., Jr., B.J. Stocks,J.S. Levine, W.R. Cofer III, and P.J.M., Sat-

ellite analysisof the severe1987 forestfires in northernChina and southern Siberia,J. Geophys.Res.,99, 18,627-18,638, 1994. Cantrell, C.A., R.E. Shetter,A.H. McDaniel, J.G. Calvert, J.A. Davidson,

D.C. Lowe,S.C. Tyler, R.J. Cicerone,andJ.P. Greenberg,Carbonkineticisotope effectin theoxidation of methane by thehydroxylradical,J. Geophys.Res.,95, 22,455-22,462, 1990. Conny,J.M., and L.A. Currie, The isotopiccharacterization of methane, non-methane hydrocarbons andformaldehyde in the troposphere, Atrnos.

meijer, Stable isotopiccompositionsof carbonmonoxidefrom biomass burningexperiments, Atrnos.Environ., 33, 4357-4362, 1999b. Khalil, M.A.K., and R.A. Rasmusssen, Carbon monoxide in the Earth's at-

mosphere: Increasingtrend,Science,224, 54-56, 1984. Logan,J.A., M.J. Prather,S.C. Wofsy. and M.B. McElroy, Tropospheric chemistry:A global perspective,J. Geophys.Res., 86, 7210-7254, 1981.

Lowe, D.C., C.A.M. Brenninkmeijer,G.W. Brailsford,K.R. Lassey,and

A.J.Gomez,Concentration and13Crecords of atmospheric methane in New ZealandandAntarctica:Evidencefor changesin methanesource,J. Geophys.Res.,99, 16,913-16,925, 1994. Mak, J.E., Evidencefor a missingcarbon monoxidesink based on tro-

pospheric measurements of 14CO,Geophys. Res.Lett.,19, 1467-1470, 1992.

Mak, J.E., and C.A.M. Brenninkmeijer,Compressed air sampletechnology for isotopicanalysisof atmospheric carbonmonoxide,J. Atrnos.Oceanic Technol., 11,425-431,

1994.

Mak,J.E.,andC.A.M.Brenninkmeijer, Measurement of •3COandC180in thefreetroposphere, J. Geophys.Res., 103, 19,347-19,358, 1998. Mak, J.E., and J.R. Southon,Assessment of tropicalOH seasonalityusing

atmospheric 14COmeasurements fromBarbados, Geophys. Res.Lett., 25, 2801-2804, 1998.

Mak, J.E.,C.A.M. Brenninkmeijer, andJ. Tamaresis, At•nospheric 14CO observations andtheirusefor estimating carbonmonoxideremovalrates, J. Geophys.Res., 99, 22,915-22,922, 1994. Manning,M.R., C.A.M. Brenninkmeijer, andW. Allen, Atmosphericcarbon monoxidebudget of the southernhemisphere:Implicationsof

•3C/12C measurements, J. Geophys. Res.,102,10,673-10,682, 1997. Narita, D., et al., Seasonal variation of carbon monoxide at remote sites in

Japan,Chernosphere: Global ChangeSci., 1,137-144, 1999. Novelli, P.C., K.A. Masarie, and P.M. Lang, Distributionsand recent changes of carbonmonoxide in thelowertroposphere, J. Geophys.Res., 103, 19,015-19,033, 1998.

Pacyna,J.M., andT.E. Graedel,Atmospheric emissions inventories: Status andprospects, Annu.Rev.EnergyEnviron.,20, 265-300, 1995. Pochanart,P., J. Hirokawa, Y. Kajii, and H. Akimoto, Surfaceozoneand carbonmonoxidemeasurementat Oki islands,Japan, during March 1994 to February 1996, in AtmosphericOzone, Proceedingsof the Environ., 30, 621-638, 1996. XVII Quadrennial Ozone Symposium,editedby R.D. Bojkov and G. Conny,J.M., R.M. Verkouteren, andL.A. Currie,Carbon13 composition Visconti,pp. 571-574, InternationalOzonecommission, Gen•ve, 1998. of tropospheric CO in Brazil:A modelscenario duringthebiomass season,J. Geophys.Res.,102, 10,683-10,693, 1997. Pochanart, P., J. Hirokawa,Y. Kajii, H. Akimoto,andM. Nakao, Influence of regional-scale anthropogenic activity in northeastAsia on seasonal Crutzen,P.J.,andM.O. Andreae,Biomass burningin thetropicsimpacton atmospheric chemistryandbiogeochemical cycles,Science,250, 1669variationsof surfaceozoneandcarbonmonoxideobservedat Oki, Japan, 1678, 1990. J. Geophys.Res., 104, 3621-3631, 1999. R6ckmann,T., andC.A.M. Brenninkmeijer,CO andCO2 isotopiccomposiDeines,P., in Handbookof Environmental IsotopeGeochemistry, Vol.i, tion in Spitsbergen duringthe 1995 ARCTOC campaign,Tellus,49B, editedby P. FritzandJ.C.Fontes,pp.329-434,ElsevierSci.,New York, 1980.

DeMore,W.B., S.P.Sander,D.M. Golden,R.F. Hampson,M.J. Kurylo,C.J. Howard, C.J. Ravishankara, C.E. Kolb, and M.J. Molina, Chemical ki-

neticsand photochemical data for usein stratospheric modeling,JPL Publ., 97-4, Eval. 12, 1997.

455-465, 1997.

R6ckmann,T., C.A.M. Brenninkmeijer,P. Neeb, and P.J. Crutzen, Ozonolysisof nonmethanehydrocarbonsas a sourceof the observedmass independent oxygenisotopeenrichment in tropospheric CO, J. Geophys. Res., 103. 1463-1470, 1998a.

G. Saueressig, P. Bergamaschi, J.N. Guenther, A., et al., A globalmodelof naturalvolatileorganiccompound R6ckmann,T., C.A.M. Brenninkmeijer, emissions, J. Geophys. Res.,100, 8873-8892,1995. Crowley,H. Fischer, andP.J.Crutzen,Mass-independent oxygenisotope Jaffe,D.A., R.E. Honrath,L. Zhang,H. Akimoto,A. Shimizu,H. Mukai,K. fractionation in atmospheric CO as a resultof the reactionCO + OH, Science, 281,544-546, 1998b. Murano,S. Hatakeyama, andJ. Merrill,Measurements of NO, NOy,CO, and03 andestimation of theozoneproduction rateat Oki Island,Japan, Sakugawa, H., andI.R. Kaplan,Radio-andstable-isotope measurements of duringPEM-West,J. Geophys.Res.,101, 2037-2048, 1996. atmospheric carbonmonoxide in LosAngeles, Geochern. J., 31, 75-83, J6ckel, P., M.G. Lawrence,and C.A.M. Brenninkmeijer,Simulationsof

cosmogenic 14COusing the three-dimensional atmospheric model MATCH: Effectsof 14Cproduction distribution andthesolarcycle,J. Geophys.Res.,104, 11,733-11,743 1999.

Kajii, Y., H. Akimoto,Y. Komazaki,S. Tanak,H. Mukai, K. Murano,and H.T. Merrill, Long-rangetransportof ozone,carbonmonoxide,and acidic trace gasesat Oki Island, Japan,during PEM-WEST B / PEACAMPOT B campaign,J. Geophys.Res., 102, 28,637-28,649, 1997.

Kajii, Y., I42Someno,H. Tanimoto,J. Hirokawa,and H. Akimoto,Evidence for the seasonalvariation of photochemicalactivity of tropospheric ozone:Continuousobservation of ozoneand CO at Happo, Japan,Geophys. Res.Lett.,25, 3505-3508,1998.

1997.

Seiler, W., and R. Conrad, Contributionsof tropical ecosystemsto the

globalbudgetsof tracegasesespecially CH4,H2, CO andN20, in Geophysiologyof Arnazonia,editedby R. Dickinson,pp. 133-162, John Wiley, New York, 1987.

Sharkey,T.D., F. Loreto,C.F. Delwiche,andI.W. Treichel,Fractionation of carbonisotopesduringbiogenesis of atmospheric isoprene,Plant Physiol.,97, 436-466, 1991. Smit,H.G.J.,A. Volz, D.H. Ehhalt,andH. Knappe,The isotopefractionationduringtheoxidationof carbonmonoxide by hydroxyl-radical andits implicationfor the atmospheric CO-cycle,in StableIsotopes,editedby H.L. Schmidt,H. F6rstel,andK. Heinzinger,pp. 147-152, ElsevierSci., New York, 1982.

8900

KATO ET AL.: ISOTOPICCOMPOSITIONSOF CO AT HAPPO,JAPAN

Spivakovsky, C.M., R. Yevich,J.A. Logan,S.C.Wofsy,andM.B. McElroy,. Volz, A., D.H. Ehha!t,andR.D. Derwent,Seasonal andlatitudinalvariation Tropospheric OH in a three-dimensional chemicaltracermodel:An asof14CO andthetropospheric concentration ofOHradicals, J. Geophys. sessment basedon observations of CH3CCI3,J. Geophys.Res., 95, Res.,86,5163-5171, 1981. 18,441-18,471,

1990.

Stevens,C.M., CO fluxesin the NorthernHemispheretemperatezone in 1971 (abstract),EOS Trans.AGU, 77(46), Fall Meet. Suppl.,F109, 1996.

Stevens,C.M., and L. Krout, Method for the determination of the concentra-

tionof the carbonandoxygenisotopiccomposition of atmospheric carbonmonoxide, Int. J. MassSpectrorn. Ion Phys.,8,265-275, 1972. Stevens,C.M., and A. Engelkemeir,Stable carbonisotopiccompositionof Methanefrom somenatural and anthropogenic sources,J. Geophys. Res., 93, 725-733, 1988.

Stevens, C.M., andA.F. Wagner,The roleof isotopefractionation effectsin atmospheric chemistry, Z. Naturforsch.,A,44, 376-384, 1989. Stevens,C.M., L. Krout, D. Walling, A. Venters,A. Engelkemeir,and L.E. Ross,The isotopiccomposition of atmospheric carbonmonoxide,Earth

Zhao,Y., Y. Kondo, F.J.Murcray, X. Liu,M. Koike,K. Kita,H. Nakajima, I. Murata, and K. Suzuki, Carbon monoxide column abundancesand

tropospheric concentrations retrieved fromhigh-resolution ground-based infraredsolarspectra at 43.5øNoverJapan,J. Geophys. Res.,102, 23,403-23,411, 1997.

H. Akimoto,Y. Kajii, andS. Kato, ResearchCenterfor AdvancedScienceand Technology, Universityof Tokyo, 4-6-1 Komaba,Meguro-ku,

Tokyo153-8904,Japan. (shungo@atmchem. rcast.u-tokyo.ac.jp) M. Brfiunlich, C.A.M. Brenninkmeijer, andT. R6ckmann, Atmospheric Chemistry Division,Max PlanckInstitutefor Chemistry, Johann-BecgerWeg29, Mainz 55128, Germany.

Planet. Sci. Lett., 16, 147-165, 1972.

Stevens, C.M., L. Kaplan,R. Gorse,S. Durkee,M. Compton,S. Cohen,and K. Bielling,The kineticsisotopeeffectfor carbonandoxygenin the reaction CO + OH, Int. J. Chern.Kinet., 12,935-948, 1980.

(Received June9, 1999;revised November 18, 1999; accepted November 22, 1999.)