Breakdown of Kennicu0-Schmidt Law at Giant Molecular Cloud ... · mass and star formaon rate (SFR)...
Transcript of Breakdown of Kennicu0-Schmidt Law at Giant Molecular Cloud ... · mass and star formaon rate (SFR)...
BreakdownofKennicu0-SchmidtLawatGiantMolecularCloudScales
SachikoOnodera(MeiseiUniversity,Japan),NROMAGiCprojectteam,FUGINprojectteam
OutlineofToday’sTalk• ResolvedKennicu0-Schmidtlawinnearbygalaxies– NROM33All-disksurveyofGiantMolecularCloudsproject(MAGiC)
– BreakdownoftheK-SlawinM33atGMCscale– Reasonforthebreakdown– StarformaQonrelaQonderivedfromothertracersofstarformaQonrateandgas
• ResolvedstarformaQonrelaQonintheMilkyWay– PreliminaryresultfromNROgalacQc-planeCOsurveyproject(FUGIN)
IntroducQon:MolecularGasandStarFormaQoninNearbyGalaxies
• Kennicu0-SchmidtlawGlobalcorrelaQonbetweengasmassandstarformaQonrate(SFR)ingalaxies(ΣSFR∝ΣgasN)N~1.4(Kennicu01998,Kennicu0&Evans2012)Large-scalestudiesprovideessenQalinformaQonontherelaQonbetweenlarge-scaleproperQesofgalaxiesandstarformaQon
Kennicu0&Evans2012
ResolvedKennicu0-SchmidtLawBigiel et al. 2008
TheprocessofconverQnggasintostarsoccursonasmallerscale,withinmolecularclouds.• HowisthiscorrelaQon
connectedtoeachstar-formingregion?
• TowhichscaledoesthiscorrelaQonhold?
ObservaQonalDifficultyMostgalaxiesaretoofartoresolveeachstar-formingregionbecauseofourresoluQonlimit→closegalaxies logΣHI+H2[Msun/pc2]
logΣ S
FR[M
sun/yr/kpc
2 ]
M33All-disksurveyofGiantMolecularClouds(NROMAGiC)
• Localgroupgalaxy• ClosetoourGalaxy
(D=840kpc) eachGMCcanbe
resolved (NRO45mresoluQon:20"~80pc)
ThebesttargetforstudyingGMCsandstarformaQoninawholegalaxyCO(1-0)surveywithNRO45m+BEARS+OTF
n Moderatelyface-onDistribuQonofGMCsinthewholedisk,inrelaQontoothercomponents(e.g.star-formingregions,arms)
(Arimotoetal.)
n Star-formingregionsexistoverthewholedisk
CO map Spitzer24um(starformaQonindicator)
等強度線: 分子ガス
VarietyofMolecularGasandStarFormaQon
CO(1-0)
Contour:CO
K-SlawisbrokenatsomescalesDatacubeavailableonNROwebsite!
CO map Spitzer24um(starformaQonindicator)
等強度線: 分子ガス
VarietyofMolecularGasandStarFormaQon
mostmassiveGMCbutweakSFacQvity
acQveSF(NGC595)butlessmassive
CO(1-0)
Contour:CO
K-SlawisbrokenatsomescalesDatacubeavailableonNROwebsite!
TowhatscaleistheKennicu0-Schmidtlawvalid?
thesurfacedensityofmoleculargasmassvsSFRinahighresoluQon(~80pc)• CorrelaQonsareevidentin
1kpc,500pc• Becomelooserwithhigher
resoluQon,andfinallybreaksdown(dispersionover~fourordersofmagnigude)
Schruba+(2010),Boquien+(2015)whichusedUV,Hα,MIR,&FIRhaveshownthesimilarresults⇒GMCscale(~80pc)iswhere
theK-Slawbecomesinvalid
Σ(SFR)
Hα,corr[M
oyr-1pc
-2]
Onoderaetal.2010Σ(H2)[Msunpc-2]
ResoluQon~80pc ~250pc
~500pc ~1kpc
2σ
ReasonsforbreakdownoftheK-SlawatGMCscales
• Drisofclustersfromtheirparentclouds• StochasQcityeffectsontheesQmaQonofSFR• DifferentevoluQonarystagesofGMCs
ReasonforbreakdownoftheK-SlawatGMCscales
• Drisofclustersfromtheirparentclouds– InMilkyWay,youngclustersrecedefromtheirparentGMCswithv>10km/s(Leisawitz+1989)
– theseparaQonbetweenthemisexpectedtobe~100pcataclusterageof10Myr
⇒comparabletoourresoluQonof80pc ⇒ However,itcannotexplainsomestarformingregionswithoutassociatedGMCsnorGMCswithoutassociatedstarformingregions
ReasonforbreakdownoftheK-SlawatGMCscales
• StochasQcityeffectsontheesQmaQonofSFR– duetosmallsamplingathighspaQalresoluQons– IMFisnotfullypopulatedwhensmallerregionsthatcontainonlyafewstarsinclustersaresampled.
– Thus,inregionswithweakerexQncQon-correctedHαemission,thiseffectmayleadtosignificantsca0ersintheesQmatedΣ(SFR).
– AssumingT〜104Kandne〜100cm−3inHIIregions,theionizingphotonfluxisQ(H0)=7.3×1011L(Hα)corrs−1(Kennicu01988;Brocklehurst1971).
– Σ(SFR)=10−9Msunyr−1pc−2 ⇒L(Hα)corr=6.6×1035ergs−1withinthe80pcbeam ⇒Totalionizingphotonflux: Q(H0)=4.9×1048s−1.
StochasQcityeffectsontheesQmaQonofSFR
– TheraQooftheerrorinQ(H0)tothemeanvalue,σ[Q(H0)]/Q(H0),is∼10M−1/2(M:clustermass)(Cervinoetal.2002)
⇒theclustermassrequiredtoproducetheionizingfluxis50〜5×104Msun,whichresultsinσ[Q(H0)]/Q(H0)≈1to0.04.
• Insummary,theerrorduetotheeffectofstochasQcityatΣ(SFR)=10−9Msunyr−1pc−2is0.04–1×10−9Msunyr−1pc−2inthe80pcscale.
• TheerrordecreaseswithhigherΣ(SFR)andlowerspaQalresoluQons.
AlthoughtheerrorissignificantatsmallerΣ(SFR)atthe80pcresoluQon,itcannotexplainthesca0erofoverfourordersofmagnitudeinΣ(SFR).
EvoluQonsofGMCsinM33DifferentevoluQonarystagesinM33usingCO(3-2)• TypeA:nosignofmassive
SF• TypeB:onlywithrelaQvely
smallHIIregions• TypeC:withbothHII
regions&young(<10Myr)stellargroups
• TypeD:withbothHIIregions&relaQvelyold(10–30Myr)stellargroups
thedispersionintheK-SlawislargelyduetovariaQonsintheevoluQonarystageoftheGMCs
TypeA(1%)
TypeB(20%)
TypeC(45%)
TypeD(34%)
Miuraetal.2012
K-SlawatmatchedageinTaffyI
• ThedispersionσofcorrelaQon:14%(constantconversionfactor)25%(differentconversionfactor)excludingregionD• atthesameresoluQon(700pc)σ=50%(M51),27%(NGC3521)(Liuetal.2011)43%(M33:500pcresoluQon)(Onoderaetal.2010)thedispersionintheK-SlawislargelyduetovariaQonsintheevoluQonarystageofGMCs
• TaffyI:interacQnggalaxythatexperiencedadirectcollisiononly~20Myrago
• AgesoftheSFblobsesQmatedfromPaαEquivalentwidth• AgesofallblobsexceptforregionBandregionD:∼7Myr
RegionD:<3.5Myr
Komugietal.2012
K-SlawusingYSOasSFRtracerinLargeMagellanicCloud
• therelaQonbetweenΣ(SFR)&Σ(HI+H2)inLMCforsixGMCs
• largesca0erinΣ(SFR)derivedfromHα+24um〜twoordersofmagnitude
• theΣ(HI+H2)lieswithinafactoroftwo
• sca0erinΣ(SFR)derivedfromYSO〜anorderofmagnitude
⇒YSO:abe0ertracerofSFR
Chenetal.2010
PreviousStudies:CO(3-2)-SFRcorrelaQonsinNearbyGalaxies(SeealsoMichiyama’sposter)
M83 disk¢er
(Muraoka+ 2009)
In 500pc~kpc resolution CO(3-2) intensity: slightly better correlation with SFR than CO(1-0)
(Komugi+ 2007) log I(CO)[K km/s]
Nearby galactic centers
I(CO(1-0)) I(CO(3-2))
CO(1-0) vs. SFR
CO(3-2) vs. SFR
CO(3-2)&CO(1-0)map
Color:CO(3-2)Grey:CO(1-0)
Miuraetal.2014
SimilardistribuQon
SFRvs.CO(1-0),CO(3-2)inGMCs
• NocorrelaQonisfoundbetweenI(CO1-0)&Σ(SFR)inGMCscales• I(CO3-2)-Σ(SFR)correlaQon(alreadyknowninkpc-scale)isvalid
downtoGMCscales⇒StarformaQoniscloselyassociatedwiththewarmanddense
regionstracedbyCO(3-2)line,alsoineachGMC
Onodera+2012
corr.coeff.=0.22 corr.coeff.=0.68
StarformaQonrelaQonsintheMilkyWay• Spitzerc2d&Gould’sBeltsurvey20low-massstar-formingmolecularclouds
• Avmap→Σgas• YSOnumberdensity→ΣSFR • resoluQon:sizeofmolecularclouds(〜3pconaverage)
• indexissuperlinear:largelydifferentfromthatoftheextragalacQcK-Slaw(N=1〜1.4)
→star-formingthresholdingassurfacedensityof〜100Msunpc-2?
Heidermanetal.2010
StarformaQonrelaQonsintheMilkyWay
Sixhigh-massstar-formingmolecularclouds• classIYSOnumber→ΣSFR• HerschelFIR→Av→Σgas• SinglepowerlawineachcloudbutNisdifferentamongclouds
• NisalsosuperlineareveninΣgas>100Msunpc-2
5pc Willisetal.2015
FORESTUnbiasedGalacQcplaneImagingsurveywiththeNobeyama45mtelescope
(FUGINproject)
• unbiasedCOlinemappingoftheGalacQcPlane(2014~2017)• 10°<l<50°,|b|<1°(80deg2)and198°<l<236°,|b|<1°(76deg2)• spiralarms(Perseus,Sagi0arius,Scutum,andNormaarms),bar• 12CO,13CO,andC18O(J=1–0)lines,simultaneously• newmulQ-beamreceiver,FOREST,ontheNobeyama45m
telescope• angularresoluQonofthefinalmap:20″(12CO),21″(13CO&C18O)
Redsquare:observaQonalregions,colormap:CfAsurveyCOmapbyDame+(2001)
FORESTUnbiasedGalacQcplaneImagingsurveywiththeNobeyama45mtelescope
(FUGINproject)
Umemotoetal.2017
12COtotalmoleculargas diffuseregions〜102cm-3
13CO&C18OdenserandopQcallythinnermoleculargas〜103to104cm-3
PreliminaryResultsinM17Region
confidenQal
Comparisonwithpreviousresults
confidenQal
Summary• The“classical”K-SrelaQonbetweenΣ(SFR)derivedfromHα,IRandUVandΣ(H2)derivedfromCObreaksdownattheGMCscale– DrisofnewbornstarsfromtheparentcloudsandstochasQcityeffectsontheesQmaQonofSFRarenotlargeenoughtoexplainthedispersion
– DifferentevoluQonarystagesofGMCsarethemainreason