TechTech～テクテク～No.28

9/16

学生中心の研究チームが火星に残る水の量を解明！　東京工業大学大学院、名古屋大学大学院の学生を中心とするメンバーは、臼井助教による火星の水素同位体分析データを用いた理論計算によって、水が失われた時期や量を明らかにした。　それによれば、火星誕生から約4億年の間に火星の水の50%以上が大気を通じて宇宙空間へ流出。一方で、残りの大部分は今も氷などとなって火星の地下に存在する可能性があると結論づけた。　現在火星では、極域で少量の氷が確認されているが、理論計算が導き出した残存量はその何倍にも及ぶ。実際、火星周回機によるレーダー観測でも、今回の研究結果を裏付けるようなデータが得られている。182H.Kurokawaetal./EarthandPlanetaryScienceLetters394(2014)179–185Table2EstimatedwaterlossduringStage-1and-2.Estimatesbasedonoxygenescapecal-culationsarefrom1:Teradaetal.(2009)(18–78mGELintheoriginalpaperwithdifferentconversion)and2:Lammeretal.(2003)(14–34mGELintheoriginalpa-perwithdifferentconversion).Seetextfordetails.MethodStage-1Stage-2BasedonPLD41–99mGEL10–53mGELBasedonCarrandHead(2003)53–280mGEL120–520mGELOxygenescape31–133mGELa24–58mGELbaTeradaetal.(2009).Originally18–78mGELwithadifferentconversioninwhicha1mGELoceancontains8×1042watermolecules(=2.4×1017kg).bLammeretal.(2003).Originally14–34mGELinwithadifferentconversioninwhicha1mGELoceancontains8×1042watermolecules(=2.4×1017kg).Fig.3.Upperdiagram:evolutionofmartianwaterreservoirestimatedfromtheamountofthepresentwaterreservoirinPLD(blue).TheerrorbarsheuncertaintyoftheamountofpresentwaterreservoirRpresent(1200–3000❤)forALH84001.Thecomparisonswithlculationmodelsarealsoshown(orange);etal.(2009)(31–133mGEL).plottedasfunctionB185kmisbegreaterintheolderStage-1thanintheyoungerStage-2duetothehotterexobaseconditioninStage-1thaninStage-2.Ourmodeldoesnottakeintoaccounttheeffectsofwatersupplybymagmaticoutgassingandbylateaccretionofwater-bearingbodiessuchasasteroidsandcomets.Prolongedigneousactivitieswouldhavedeliveredmagmaticwaterfromthemartianinteriortothesurface.Becausesuchmagmaticwaterisassumedtohaveunfractionatedprimordialhydrogenisotopiccompositions(e.g.,≃275❤,Usuietal.,2012),themagmaticoutgassingisex-pectedtosuppressthehydrogenisotopefractionationofthesur-cialandatmosphericwaterreservoirs.Thus,morewatershouldhavebeenlostthanourestimatestoexplaintheD/Hratiosattheboundaryconditionsof4.5Gaand4.1Ga;i.e.,ourmodelagainprovidestheminimumestimatefortheamountsofthesurface-waterloss.Hydrogenisotopiccompositionsofasteroidsandcometsarevariable.Theextremecasesarecarbonaceous-chondriteparentas-teroids(e.g.,from−230to+340❤,Alexanderetal.,2012)andOort-cloudcomets(∼1000❤,Hartoghetal.,2011),respectively.Becausethedeliveryoflow-D/Hwaterbylateaccretionofsuchchondriticbodiestendstomaintainlowerhydrogenisotopecom-positionsofthesurcialwaterreservoirs,itseffectonourcal-culationshouldbesimilartotheeffectofmagmaticoutgassingdiscussedabove.Ontheotherhand,cometaryimpactscanincreasetheD/Hratioofthesurfacewaterreservoirwithoutatmosphericescape.Forexample,thesupplyof1019kgcomets(correspondingto∼100mGEL)withaD/Hratioof1000❤increasestheD/Hra-tioofsurfacewaterreservoirby∼1000❤.However,cometsaretypicallyenrichedinnoblegasesalongwithwater–ice.Lateac-cretionof1019kgcometswithaprobableXe/H2Oratioof∼10−5(Swindle,2012)resultsinasupplyof1014kgXe,whichis106timeslargerthanthemartianatmosphericXe.SinceitisdicultforsuchalargeamountofXetoescapeduring4Gyr,suchasig-nicantsupplyofcometsisunlikely.3.3.ComparisonwithgeologicalrecordsOurmodelrequiresanamountofsurfacewaterataspecicagetocalculateamountsofsurfacewateratotheragesforwhichthemeteoritedataareavailable.Intheprevioussection,wehavereportedtheminimumvolumesforthesurfacewatersat4.5Gaand4.1Ga,respectively,byassumingthepresentwaterinven-tory(Rpresent)of20–30mGEL.However,asnotedearlier,PLDepresentstheminimumestimateforthepresentwaterinventoryistenceof“undetected”currentwater–icereservoirshasthroughutilizinggeologicalestimatesforthesectionexaminesthetransitionofhpresent.spossiblyoc-vi-uringStageTeradaetetetaetaand2:LLaLammmammnversion).SndHead(22000000320l.(2009).OrigiELoceancontaital.(2003).OrigGELoceancontaFig.Fig.3.Upperdiagramamoamountofheunche(120H.Kurokawaetal./EarthandPlanetaryScienceLetters394(2014)179–185181Fig.2.WaterlossduringStage-1(L4.5–4.1Ga,blue)andStage-2(L4.1–0Ga,red)asafunctionofpresentwaterreservoirRpresent.ThewidthoftheblueandredstripesisderivedfromtheδDrange(1200–3000❤)forALH84001.TherangeoftheminimumestimateforthepresentwaterreservoirinPLD(seetextfordetails)isbracketedbydashedlines.(Forinterpretationofthereferencestocolorinthisg-urelegend,thereaderisreferredtothewebversionofthisarticle.)TheD/HratiosofthesurcialwaterreservoirIattheboundaryconditions(4.5Ga,4.1Ga,andthepresent)areobtainedfromhy-drogenisotopeanalysesofmartianmeteorites.TheinitialδDvalueof275❤forthe4.5Gaprimordialmartianwaterisobtainedfromanalysesofolivine-hostedmeltinclusionsfromaprimitivebasalticmeteorite,Yamato980459(shergottite).Thismeteoriterepresentsaprimarymeltfromadepletedmantlesourceformedat∼4.5Ga,anditsmeltinclusionsareinterpretedtopossessundegassedpri-mordialwaterinthemartianmantle(Usuietal.,2012).TheδDrange(1200–3000❤)ofthenear-surcialwaterreservoirat4.1GaisderivedfromanalysesofmagmaticphosphateandsecondarycarbonatemineralsinALH84001(Boctoretal.,2003;Greenwoodetal.,2008).TheδDvalue(5000❤)ofthepresentmartianwaterreservoirisobtainedfromD/Hanalysesofgeochemicallyenrichedshergottites(ShergottyandLAR06319)thatcrystallizednearthesurfaceintherecentpast(0.17–0.18Ga,Greenwoodetal.,2008;Usuietal.,2012).ThishighδDvalueisconsistentwiththoseofwaterinthepresentmartianatmosphere(≃5000❤)determinedthroughbothtelescopic(e.g.,Owenetal.,1988)andtheCuriosityroverobservations(Websteretal.,2013).D/Hratiosofmartianmeteoritesreectcomplexgeologichis-toriesincludingtheterrestrialweatheringafterthemeteoritefalls.ThisstudyemploysD/Hdatasetsobtainedonlyfromrecentinsitumeasurementstominimizetheeffectofterres-4001carbonatesformedbyanciententswhereasthedasresultsshowthatL4.5–4.1GaandL4.1–0GaarepositivelycorrelatedwithRpresent.Ourmodelfurtherindicatesthataratioofwaterlossbetweenthestages(L4.5–4.1Ga/L4.1–0Ga)isindependentofRpresentandthatL4.5–4.1GaisalwaysgreaterthanL4.1–0GaatanyRpresent.DividingEq.(1)forStage-2bythatforStage-1,theratiocanbewrittenasL4.5–4.1GaL4.1–0Ga=(I0GaI4.1Ga)11−f2[(I4.1GaI4.5Ga)11−f1−1](I0GaI4.1Ga)11−f2−1,(4)wheref1andf2arethefractionationfactorsinbothStage-1and-2.AssumingthesamefractionationfactorfforbothStage-1and-2,theratioofwaterlossisgivenbyL4.5–4.1GaL4.1–0Ga=I−11−f4.5Ga−I−11−f4.1GaI−11−f4.1Ga−I−11−f0Ga.(5)ThisequationshowsthattheratioofwaterlossisdeterminedonlyfromtheD/HratiosandthefractionationfactorinStage-1and-2.BecausetheδDvalueof1200–3000❤at4.1Gaisal-readyfractionatedfromtheinitialD/Hratioof275❤,theratioL4.5–4.1Ga/L4.1–0GagivenbyEq.(5)is≃1.2–6.5.ThisindicatesthatthewaterlossismoresignicantinStage-1whencomparedtoStage-2.AstheperiodofStage-1(0.4Gyr)is≃10timesshorterthanStage-2(4.1Gyr),theaverageescaperateinStage-1wouldbemorethan10timeshigherthaninStage-2.3.2.MinimumestimateofwaterlossTheamountofthe“observable”currentsurfacewaterreservoirisdominatedbythepolarlayereddeposits(PLD).AssumingthatthePLDsaremainlycomposedofwater–ice,theyareexpectedtocontainH2Oof1.2–1.6×106km3intheNorthpolarregion(Zuberetal.,1998)and1.6×106km3intheSouthpolarregion(Plautetal.,2007),respectively;theirtotalsum(2.8–3.2×106km3)cor-respondsto20–30mGEL.Weemploythisvalue(20–30m)astheminimumestimatefortheamountofpresentwaterreservoirRpresent,becausetheexistenceof“missing”water–icereservoirshasbeenproposed(e.g.,CarrandHead,2003).Forexample,ice-richmantlesandcoveringsedimentsinthemid-latitudepossiblycontainalargeamountofice(Murrayetal.,2005;Page,2007;Christensen,2006;Pageetal.,2009).Furthermore,thereisincreas-ingevidencethatvastreservoirsofwater–icepotentiallyexistinpartsofthehighlatitudes,asindicatedbygeomorphology(Baker,2001;Kargel,2004;Soareetal.,2007,2011,2012,2013a,2013b;l2009;Lefortetal.,2009;Levyetal.,2009a,2009b,hughthePhoenixLander(Smithetal.,Spectrometer(Boynton00ossdurinesentwamtheimatefodashedheread/Hratins(4.5sotope❤fortesofolrite,Yamarymtsmeltialwae(120erivedbonateal.,200servoirergottiurfaceUsuietwaterithrougroverD/toriehis00This180H.Kurokawaetal./EarthandPlanetaryScienceLetters394(2014)179–185Table1Geologic-mapping-based,time-stratigraphicinformation,includingtheNoachian,Hesperian,andAmazonianPeriods(ScottandCarr,1978).Thegeologicperiodshavebeengivenestimatedabsoluteagerangesbasedonimpactcratermodels(Table1ismodiedfromHartmannandNeukum,2001).Notethatweperformedcomparativeanalysesamongtheestimatedwateramountsofthepre-Noachian,approximatelyreferredtohereasStage-1,andNoachian–Amazonian,approximatelyreferredtoasStage-2,basedonrecentD/Hdatasetfrommartianmeteorites.StagePeriodEstimatedabsoluteagerange[Ga]2Amazonian3.3–2.9topresentHesperian3.7–3.5to3.3–2.9Noachian4.2to3.7–3.5boundaryat4.1Ga1Pre-Noachian4.5to4.2oceans,respectively(Parkeretal.,1993;CliffordandParker,2001;CarrandHead,2003).Althoughthesegeomorphologicstudieshaveprovidedsignicantconstraintsonthehistoryofmartianpaleo-oceans,theylackinformationaboutpre-Noachian(Frey,2006)oceansbecausenogeologicrecordsareavailable.Furthermore,theshoreline-demarcationapproacheswouldnotbeapplicabletotheyoungestAmazonian(3.1Gatopresent)era,duringwhichthesur-facewaterwouldhaveoccurredmostlyasice(CliffordandParker,2001;CarrandHead,2010).Thisstudyendeavorstotracetheglobalinventoryofsur-cialwaterthroughtimebeginningwiththeembryonicstagesofdevelopmentofMars(i.e.,4.5Ga)topresentdaybasedonageochemicalapproachofhydrogenisotopes(D/H:deuterium/hy-drogen).Hydrogenisamajorcomponentofwater(H2O)anditsisotopesfractionatesignicantlyduringhydrologicalcyclingbetweentheatmosphere,surfacewater,andgroundandpolarcapices.TelescopicstudieshavereportedthatthehemisphericmeanofthemartianatmospherehasaD/Hratioof∼6times(δD≃5000❤)theterrestrialvalues(Owenetal.,1988);δD=[(D/H)sample/(D/H)reference−1]×1000,wherethereferenceisStan-dardMeanOceanWater(SMOW).BecausethehighatmosphericD/Hratioisinterpretedtoresultfromthepreferentiallossofhy-drogenrelativetotheheavierdeuteriumfromthemartianatmo-spherethroughouttheplanet’shistory(Lammeretal.,2008),thedeuteriumenrichmentcanbeusedtoestimatetheamountofwa-terlossduetotheatmosphericescape.Comparedtoanumberofgeomorphologicstudies(e.g.,Scottetal.,1995;Headetal.,1999;CliffordandParker,2001;CarrandHead,2003;DiAchilleandHynek,2010),onlyafewgeochemi-calinvestigationshavebeenconducted(ChasseèreandLeblanc,l2003)Thisispartlybecausetherehavertianme-Fig.1.Schematicillustrationofthetwo-stagemodelfortheevolutionoftheglobalsurfacewaterreservoironMars.Rpresentisthesizeofthepresentwaterreservoir,L4.5–4.1GaandL4.1–0GaarethewaterlossduringStage-1and-2,andfisthefrac-tionationfactor(seetext).ouranalyses.Combiningourresultswithgeologicalestimatesforthevolumeofmartianpaleo-oceans,weproposethatunidentiedsurcialwater–icereservoirsshouldcurrentlyexistandthevolume(≃100–1000mGEL)shouldexceedtheestimatedpresentwaterinventory(20–30mGEL,Christensen,2006)onMars.2.CalculationTheamountofwaterlossduetotheatmosphericescapebe-tweentimet1andt2canbecalculatedfromanassumedamountofpresentwaterreservoirusingthefollowingequations:Lt1−t2=Rt1−Rt2=Rt2×[(It2It1)11−f−1],(1)andf=d[D]/[D]d[H]/[H].(2)HereLt1−t2istheamountofwaterlossduetotheatmospherices-capeduringthetimefromt1tot2,RandIareanamountofwaterreservoirandaD/Hratioateachtime,respectively,fisthefrac-tionationfactor,and[H]and[D]aretheabundancesofHandDinthecombinedreservoirsinatomscm−2(Lammeretal.,2003).Boththevolumesofwaterreservoirandwaterlossareexpressedinoceandepth[m]asaglobalequivalentlayer(GEL).Usingtheden-sityofwaterof103kgm−3andthesurfaceareaof1.4×1014m2,1mGELcorrespondsto1.4×1017kgofwater.Eq.(1)canberewrittenas:1(3)ngSwaterδDforthdlinederistioso5Ga,eanathe4livineYamatomeltfroltincluaterin00–3000dfromeminer08).Therisobtatites(Shinthertal.,20200101111intheprghbothtobservati/HratiosesincludinsstudylculatlculatsstudyembeengivenestimodiedfromHartmannandNeukum,2001).Noanalysesamongtheestimatedwateramountsofthepre-Noachian,approximatyreferredtohereasStage-1,andNoachian–Amazonian,approximatelyreferredtoasStage-2,basedonrecentD/Hdatasetfrommartianmeteorites.StagePeriodEstimatedabsoluteagerange[Ga]2Amazonian3.3–2.9topresentHesperian3.7–3.5to3.3–2.9Noachian4.2to3.7–3.5boundaryat4.1Ga1Pre-Noachian4.5to4.2oceans,respectively(PPParerettal.,199kkkeerarkrkPaPkea3;CliffordandParker,2001;CarrandHead,20033).Althoughthesegeomorphologicstudieshaveprovidedsignicantconstraintsonthehistoryofmartianpaleo-oceans,theylackinformationaboutpre-Noachian(Frey,2006)oceansbecausenogeologicrecordsareavailable.Furthermore,theshoreline-demarcationapproacheswouldnotbeapplicabletotheyoungestAmazonian(3.1Gatopresent)era,duringwhichthesur-facewaterwouldhaveoccurredmostlyasice(CliffordandParker,2001;CarrandaeHHeeaaeedd,220010d).Thisstudyendeavorstotracetheglobalinventoryofsur-cialwaterthroughtimebeginningwiththeembryonicstagesofdevelopmentofMars(i.e.,4.5Ga)topresentdaybasedonageochemicalapproachofhydrogenisotopes(D/H:deuterium/hy-drogen).Hydrogenisamajorcomponentofwater(H2O)anditsisotopesfractionatesignicantlyduringhydrologicalcyclingbetweentheatmosphere,surfacewater,andgroundandpolarcapices.TelescopicstudieshavereportedthatthehemisphericmeanofthemartianatmospherehasaD/Hratioof∼6times(δD≃5000❤)theterrestrialvalues(Owenetal.,1988);δD=[(D/H)sample/(D/H)reference−1]×1000,wherethereferenceisStan-dardMeanOceanWater(SMOW).BecausethehighatmosphericD/Hratioisinterpretedtoresultfromthepreferentiallossofhy-drogenrelativetotheheavierdeuteriumfromthemartianatmo-spherethroughouttheplanet’shistory(Lammeretal.,2008),thedeuteriumenrichmentcanbeusedtoestimatetheamountofwa-terlossduetotheatmosphericescape.Comparedtoanumberofgeomorphologicstudies(e.g.,Scottett1191999995;Headala1tal..,aetal.,1999;CliffordandParker,2001;CarrandHH220003;DiAchaadd,,HHeeaadd,,illeandHynek,2010),onlyafewgeochemi-calinvestigationshavebeenconducted(ChasseèreandLeblanc,l2003)Thisispartlybecausetherehavertianme-Fig.1.Schematicillustrationofthetwo-stagemodelfortheevolutionoftheglobalsurfacewaterreservoironMars.Rpresentisthesizeofthepresentwaterreservoir,L4.5–4.1GaandL4.1–0GaarethewaterlossduringStage-1and-2,andfisthefrac-tionationfactor(seetext).ouranalyses.Combiningourresultswithgeologicalestimatesforthevolumeofmartianpaleo-oceans,weproposethatunidentiedsurcialwater–icereservoirsshouldcurrentlyexistandthevolume(≃100–1000mGEL)shouldexceedtheestimatedpresentwaterinventory(20–30mGEL,Christensen,2006)onMars.2.CalculationTheamountofwaterlossduetotheatmosphericescapebe-tweentimet1andt2canbecalculatedfromanassumedamountofpresentwaterreservoirusingthefollowingequations:Lt1−t2=Rt1−Rt2=Rt2×[(It2It1)11−f−1],(1)andf=d[D]/[D]d[H]/[H].(2)HereLt1−t2istheamountofwaterlossduetotheatmospherices-capeduringthetimefromt1tot2,RandIareanamountofwaterreservoirandaD/Hratioateachtime,respectively,fisthefrac-tionationfactor,and[H]and[D]aretheabundancesofHandDinthecombinedreservoirsinatomscm−2(Lammeretal.,2003).Boththevolumesofwaterreservoirandwaterlossareexpressedinoceandepth[m]asaglobalequivalentlayer(GEL).Usingtheden-sityofwaterof103kgm−3andthesurfaceareaof1.4×1014m2,1mGELcorrespondsto1.4×1017kgofwater.Eq.(1)canberewrittenas:1(3)EarthandPlanetaryScienceLetters394(2014)179–185ContentslistsavailableatScienceDirectEarthandPlanetaryScienceLetterswww.elsevier.com/locate/epslEvolutionofwaterreservoirsonMars:ConstraintsfromhydrogenisotopesinmartianmeteoritesH.Kurokawaa,b,∗,M.Satoc,b,M.Ushiodab,T.Matsuyamab,R.Moriwakib,J.M.Dohmd,T.UsuibaDepartmentofPhysics,NagoyaUniversity,Furo-cho,Chikusa-ku,Nagoya,Aichi464-8602,JapanbDepartmentofEarthandPlanetarySciences,TokyoInstituteofTechnology,2-12-1Ookayama,Meguro,Tokyo152-8551,JapancDepartmentofEnvironmentalChanges,KyushuUniversity,744Motooka,Nishi-ku,Fukuoka819-0395,JapandEarth-Life-ScienceInstitute,TokyoInstituteofTechnology,2-12-1-1E-1Ookayama,Meguro-ku,Tokyo,152-8550,JapanarticleinfoabstractArticlehistory:Received11December2013Receivedinrevisedform12March2014Accepted13March2014AvailableonlinexxxxEditor:T.M.HarrisonKeywords:MarsmeteoriteswaterreservoirisotopeatmosphericescapeMartiansurfacemorphologyimpliesthatMarswasoncewarmenoughtomaintainpersistentliquidwateronitssurface.WhilethehighD/Hratios(∼6timestheEarth’soceanwater)ofthecurrentmartianatmospheresuggestthatsignicantwaterhasbeenlostfromthesurfaceduringmartianhistory,thetiming,processes,andtheamountofthewaterlosshavebeenpoorlyconstrained.Recenttdevelopmentsofion-microprobeanalysisofmartianmeteoriteshaveprovidedhydrogenisotopecompositions(D/H)ofmartianwaterreseformed.BasedontheD/Hdatafromthemeteithepre-Noachian(>41–99mlmartianhistory(f　大学、大学院時代は、地球の地質学を専門にしていた臼井助教。もともとこの分野を選んだのは、「地層や岩石の調査で、世界中いろんなところに行けそうだと思ったから（笑）」という。それが博士号を取得し研究員となった頃、米国のある専門紙を見たのをきっかけに、地球から宇宙へと研究の対象を変えることとなる。　「今後も研究者としてやっていくにあたり、別の分野の可能性も考えていたときに目にしたのが、米国の大学が火星探査の研究員を募集する広告。未知の領域の魅力に惹かれ、迷わず応募しました。今から10年くらい前のことです。採用後、すぐ米国に渡り、その後NASAにも勤めました。地層や岩石の研究という基本は変わりませんが、惑星科学の醍醐味はその変化の激しさ。新しい情報や研究成果がどんどん発表され、5年前と現在とでは研究の前提や条件がまったく違う。研究者間の競争は厳しいですが、とても刺激的です」　臼井助教は、自身の研究活動のやりがいについてはどう考えているのだろう。　「僕にとって、それは難しい質問。研究者には、例えば『とにかく実験が楽しい』『新たな事実が見つかったときがたまらない』といった人もいます。でも自分の場合、特別にこの時間、この瞬間が─というのはあまりないんです。実験中も、論文をまとめているときも、人の論文を読んでいるときも、例えるなら子どもが夢中でブロック遊びをしているような感じ。とにかく研究自体に没頭していたいタイプなんだと思います。ただ一歩引いて、研究全体の面白味ということで言えば、自ら問題を設定し、それを解く手法やルールも自身で決められるというのが大きい。自分だけで問題が解決できなければ、必要な技術を持っている人と共同で取り組むのも自由です。特に惑星科学は、研究分野そのものをデザインしていけるまだ新しい領域。志のある学生にはお勧めです」　最後に、火星における生命の源「水」の現在を突き止めた臼井助教に、火星を含めた地球外生命の可能性について聞いてみた。　「広い宇宙で、生命がいるのは地球だけというのはやっぱり考えにくい。その発見には関心があります。もちろんそれが、DNAを持つなど地球上の生物と同じ構造かどうかもわからないので、決着をつけるには生命かその痕跡を見つけなければなりません。その意味で 僕は、地質学的な研究に留まらず、火星探査にも強い興味を持っています。先頃、宇宙航空研究開発機構（JAXA）は火星衛星からのサンプル・リターン計画（2020年代初頭打ち上げ予定）を発表しました。火星の衛星から岩石や砂を地球に持ち帰るプランですが、実は火星圏からのサンプル・リターンは、まだ世界のどの国も成功していません。こうしたプロジェクトに携わることができ、もし、火星の生命を発見することができたら、さすがに研究者として“最高の瞬間”と感じるでしょうね」 Photo by：NASA求人広告がきっかけで火星研究へ2015 Autumn9