I. N159W

25 Pages·2016·3.09 MB·English

A&A596,A85(2016) Astronomy
DOI:10.1051/0004-6361/201628098
&
(cid:13)c ESO2016 Astrophysics
Radiative and mechanical feedback into the molecular gas
in the Large Magellanic Cloud
I. N159W(cid:63),(cid:63)(cid:63)
M.-Y.Lee1,S.C.Madden1,V.Lebouteiller1,A.Gusdorf2,3,B.Godard4,R.Wu5,M.Galametz6,D.Cormier7,
F.LePetit4,E.Roueff4,E.Bron4,8,L.Carlson9,M.Chevance1,Y.Fukui10,F.Galliano1,S.Hony7,A.Hughes11,
R.Indebetouw12,13,F.P.Israel14,A.Kawamura15,J.LeBourlot4,P.Lesaffre2,3,M.Meixner16,E.Muller15,
O.Nayak17,T.Onishi18,J.Roman-Duval16,andM.Sewiło19
1 LaboratoireAIM,CEA/IRFU/Serviced’Astrophysique,Bât.709,91191Gif-sur-Yvette,France
e-mail:[email protected]
2 LERMA,ObservatoiredeParis,ÉcoleNormaleSupérieure,PSLResearchUniversity,CNRS,UMR8112,75014Paris,France
3 SorbonneUniversités,UPMCUniv.Paris6,UMR8112,LERMA,75005Paris,France
4 LERMA,ObservatoiredeParis,PSLResearchUniversity,CNRS,UMR8112,92190Meudon,France
5 InternationalResearchFellowoftheJapanSocietyforthePromotionofScience(JSPS),DepartmentofAstronomy,
UniversityofTokyo,Bunkyo-ku,113-0033Tokyo,Japan
6 EuropeanSouthernObservatory,Karl-Schwarzschild-Str.2,85748Garching-bei-München,Germany
7 InstitutfürTheoretischeAstrophysik,ZenturmfürAstronomiederUniversitätHeidelberg,Albert-UeberleStr.2,
69120Heidelberg,Germany
8 ICMM,ConsejoSuperiordeInvestigacionesCientificas,28049Madrid,Spain
9 Harvard-SmithsonianCenterforAstrophysics,60GardenSt.,Cambridge,MA02138,USA
10 DepartmentofPhysics,NagoyaUniversity,Chikusa-ku,464-8602Nagoya,Japan
11 CNRS,IRAP,9Av.colonelRoche,BP44346,31028ToulouseCedex4,France
12 DepartmentofAstronomy,UniversityofVirginia,POBox400325,Charlottesville,VA22904-4325,USA
13 NationalRadioAstronomyObservatory,520EdgemontRd.,Charlottesville,VA22903,USA
14 SterrewachtLeiden,LeidenUniversity,POBox9513,2300RALeiden,TheNetherlands
15 NationalAstronomicalObservatoryofJapan,Mitaka,181-8588Tokyo,Japan
16 SpaceTelescopeScienceInstitute,3700SanMartinDr.,Baltimore,MD21218,USA
17 DepartmentofPhysicsandAstronomy,TheJohnsHopkinsUniversity,366BloombergCenter,3400NCharlesSt.,Baltimore,
MD21218,USA
18 Department of Physical Science, Graduate School of Science, Osaka Prefecture University, 1-1 Gakuen-cho, Naka-ku, Sakai,
599-8531Osaka,Japan
19 NASAGoddardSpaceFlightCenter,8800GreenbeltRd.,Greenbelt,MD20771,USA
Received9January2016/Accepted12June2016
ABSTRACT
WepresentHerschelSPIREFourierTransformSpectrometer(FTS)observationsofN159W,anactivestar-formingregionintheLarge
MagellanicCloud(LMC).Inourobservations,anumberoffar-infraredcoolinglines,includingcarbonmonoxide(CO)J = 4 → 3
to J = 12 → 11, [CI] 609 µm and 370 µm, and [NII] 205 µm, are clearly detected. With an aim of investigating the physical
conditions and excitation processes of molecular gas, we first construct CO spectral line energy distributions (SLEDs) on ∼10 pc
scales by combining the FTS CO transitions with ground-based low-J CO data and analyze the observed CO SLEDs using non-
LTE (local thermodynamic equilibrium) radiative transfer models. We find that the CO-traced molecular gas in N159W is warm
(kinetic temperature of 153–754 K) and moderately dense (H number density of (1.1−4.5)×103 cm−3). To assess the impact of
2
theenergeticprocessesintheinterstellarmediumonthephysicalconditionsoftheCO-emittinggas,wethencomparetheobserved
COlineintensitieswiththemodelsofphotodissociationregions(PDRs)andshocks.WefirstconstrainthepropertiesofPDRsby
modelingHerschelobservationsof[OI]145µm,[CII]158µm,and[CI]370µmfine-structurelinesandfindthattheconstrained
PDRcomponentsemitveryweakCOemission.X-raysandcosmic-raysarealsofoundtoprovideanegligiblecontributiontothe
COemission,essentiallyrulingoutionizingsources(ultravioletphotons,X-rays,andcosmic-rays)asthedominantheatingsource
forCOinN159W.Ontheotherhand,mechanicalheatingbylow-velocityC-typeshockswith∼10kms−1appearssufficientenough
toreproducetheobservedwarmCO.
Keywords. ISM:molecules–MagellanicClouds–galaxies:ISM–infrared:ISM
(cid:63) HerschelisanESAspaceobservatorywithscienceinstrumentsprovidedbyEuropean-ledPrincipalInvestigatorconsortiaandwithimportant
participationfromNASA.
(cid:63)(cid:63) ThefinalreducedHerscheldata(FITSfiles)areonlyavailableattheCDSviaanonymousftptocdsarc.u-strasbg.fr(130.79.128.5)
orviahttp://cdsarc.u-strasbg.fr/viz-bin/qcat?J/A+A/596/A85
ArticlepublishedbyEDPSciences A85,page1of25
A&A596,A85(2016)
1. Introduction LMC(e.g.,Johanssonetal.1994;Fukuietal.1999;Wongetal.
2011),N159Whasbeenfrequentlytargetedforradio,mm,and
Star formation exclusively occurs in molecular clouds, the sub-mmobservationsaswell.TheAustraliaTelescopeCompact
densest component of the interstellar medium (ISM; e.g., Array(ATCA)andtheAtacamaLargeMillimeter/submillimeter
Kennicutt&Evans2012).Themainconstituentofthesemolec- Array (ALMA) have provided the sharpest view of molecu-
ularcloudsismolecularhydrogen(H )whichis,unfortunately, lar gas so far (Sealeetal. 2012; Fukuietal. 2015), revealing
2
not directly observable in the typical conditions of cold molec- the complex filamentary distributions of CO(2–1), 13CO(2–1),
ular gas owing to its symmetric, homonuclear nature. The HCO+(1–0), and HCN(1–0). The presence of high excitation
strong rotational transitions of carbon monoxide (12CO; sim- molecular gas was hinted at by CO(4–3), CO(6–5), and CO(7–
ply CO hereafter) at mm and sub-mm wavelengths have, in- 6) observations by Bolattoetal. (2005), Pinedaetal. (2008),
stead,beenusedascommontracersofmoleculargas.Inparticu- Mizunoetal.(2010),andOkadaetal.(2015).
lar, CO rotational lines have a wide range of critical densities,
Besidestheextensiveobservationsatmultiplewavelengths,
making them accessible probes of the physical conditions of
the presence of various energetic sources makes N159W an
moleculargasindiverseenvironments(e.g.,kinetictemperature
ideal target for our study. As described in the previous para-
T ∼10–1000Kandhydrogendensityn∼103–108cm−3).
k graph, numerous OB-type stars and YSOs exist in the re-
The diagnostic power of CO rotational transitions has been
gion and they can produce UV photons and strong stellar out-
exploited considerably further since the advent of the ESA
flows. In addition, the nearby black hole binary LMC X-1, the
Herschel Space Observatory (Pilbrattetal. 2010). In combi-
most luminous X-ray source in the LMC (X-ray luminosity of
nation with ground-based telescope data, Photodector Array ∼1038 ergs−1; Schlegeletal. 1994), can have a substantial in-
Camera and Spectrometer (PACS; Poglitschetal. 2010), Spec-
fluence on the surrounding ISM. Compared to UV photons,
tral and Photometric Imaging Receiver (SPIRE; Griffinetal.
X-rays penetrate deeper into molecular clouds while dissociat-
2010), and Heterodyne Instrument for the Far Infrared (HIFI;
ingfewermolecules.Asaresult,X-raysproducelargercolumn
deGraauwetal. 2010) spectroscopic observations have en-
densities of warm molecular gas for a given irradiation energy
abled us to construct CO spectral line energy distribu-
(e.g.,Meijerink&Spaans2005).Thesupernovaremnant(SNR)
tions (SLEDs) from the upper level Ju=1 to 50. In recent J0540.0–6944 and its expanding shell are located just ∼75 pc
years, Herschel-based CO SLEDs have been extensively ex-
from N159W (e.g., Chuetal. 1997; Williamsetal. 2000) and
amined for a wide range of Galactic (e.g., photodissocia-
canbeanothersourceofheating.Lastbutnotleast,anumberof
tion regions (PDRs): Habartetal. 2010; Köhleretal. 2014;
studieshaveindicatedthatthe“molecularridge”whereN159W
Ponetal.2014;Stocketal.2015;protostars:Larsonetal.2015)
islocatedmayhavebeenexposedtolarge-scaleenergeticevents
and extragalactic sources (e.g., infrared (IR)-bright galaxies: drivenbymultiplesupernovaexplosions,tidalforce,and/orram
Rangwalaetal. 2011; Kamenetzkyetal. 2012; Meijerinketal.
pressure.Themolecularridgeisthelargestmolecularconcentra-
2013; Pellegrinietal. 2013; Greveetal. 2014; Luetal. 2014; tionintheLMC,comprising∼30%ofthetotalmolecularmass
Papadopoulosetal. 2014; Rosenbergetal. 2014; Schirmetal.
inthegalaxy(e.g.,Mizunoetal.2001).Thedistributionofstar
2014; Mashianetal. 2015; Wuetal. 2015b; Seyfert galax-
formation across the ridge is quite intriguing, increasing from
ies: vanderWerfetal. 2010; Hailey-Dunsheathetal. 2012;
south to north toward the starbursting 30 Doradus region, and
Israeletal. 2014), revealing the ubiquitous presence of warm
thishasledseveralauthorstosuggestsequentialstarformation.
molecular gas (T (cid:38) 100 K). Various heating sources, e.g., ul-
k For example, deBoeretal. (1998) proposed that the motion of
traviolet (UV) photons, X-rays, and cosmic-rays, have been in-
theLMCthroughthehothalogasoftheMilkyWaycreatedbow
vokedtoexplainthepropertiesofthiswarmmoleculargasand
shocksattheleadingedge,consequentlytriggeringthesequen-
the emerging picture is that non-ionizing sources, such as me-
tialstarformation.ThisleadingedgeoftheLMCcorrespondsto
chanicalheating(e.g.,shocksdrivenbymergingactivities,stel-
thesoutheasternHIoverdensityregion,whichappearstomerge
larwinds,andsupernovaexplosions),mustplayacriticalrole.
intotheSmallMagellanicCloud(SMC)throughtheMagellanic
In this paper, we aim at probing the physical conditions Bridge connecting the two Magellanic Clouds (e.g., Kimetal.
and excitation processes of molecular gas traced by CO emis- 1998;Putmanetal.2003).SincetheMagellanicBridgehasbeen
sion in detail on individual molecular cloud scales. To do so, considered to be formed through gravitational interactions be-
we study N159W, an active star-forming region in the Large tween the two Magellanic Clouds (e.g., Bekki&Chiba 2007;
MagellanicCloud(LMC),largelybasedonHerschelPACSand Beslaetal.2012),thishintsthatthetidalforcecouldbeatwork
SPIREobservations.TheLMCisanexcellentlaboratoryforour intheHIoverdensityregion.TheHIoverdensityregionhasalso
studyforthefollowingreasons.Firstofall,theproximityofthe been shown to harbour several supergiant and giant HI shells
LMC(distanceof∼50kpc;e.g.,Pietrzyn´skietal.2013)enables (e.g., Kimetal. 1998, 1999), suggesting that powerful super-
ustoperformhigh-resolutionobservationsofspatially-resolved nova explosions from multiple OB associations have injected a
molecularclouds.Inaddition,theLMCislocatedathighGalac- largeamountofmechanicalenergyintothesurroundingISM.In
ticlatitudeandhasanalmostface-onorientation(inclinationan- Fig.1,wepresentthree-colorcompositeandHIcolumndensity
gleof∼35◦;e.g.,vanderMarel&Cioni2001),providingaview imagesofN159Wanditssurroundingregions.
withlessconfusionandlowinterstellarextinction.
Thispaperisorganizedinthefollowingway.First,wepro-
N159W is one of the three prominent molecular clouds in
vide a description of the multiwavelength datasets used in our
theHIIregioncomplexN159(Fig.1)anditsstellarandgascon-
study (Sect. 2). Next, we discuss spectral line detection in our
tentshavebeenextensivelystudiedatmultiplewavelengths.For
Herschel SPIRE Fourier Transform Spectrometer (FTS) obser-
example, previous optical and near IR studies have identified a
vations(Sect.3)andderivethephysicalpropertiesofmolecular
large number of O- and B-type stars, embedded young stellar
gas by modeling CO lines with the non-LTE radiative transfer
objects(YSOs),andultracompactHIIregions(e.g.,Jonesetal.
codeRADEX(vanderTaketal.2007;Sect.4).Wethenemploy
2005; Fariñaetal. 2009; Chenetal. 2010; Carlsonetal. 2012),
theoreticalmodelsofPDRsandshockstoexaminetheexcitation
suggesting that N159W is one of the most intense star-forming
conditionsofCOinN159W(Sect.5)andfinallysummarizeour
regions in the LMC. As the brightest CO(1–0) peak in the
conclusions(Sect.6).
A85,page2of25
M.-Y.Leeetal.:RadiativeandmechanicalfeedbackintothemoleculargasintheLargeMagellanicCloud.I.
Fig.1. Leftpanel:three-colorcompositeimageoftheHIIregioncomplexN159anditssurroundingenvironment(Spitzer3.6µm/8µm/24µmin
blue/green/red;Meixneretal.2006).TheCO(1–0)integratedintensityfromtheMAGMAsurvey(Wongetal.2011)isshownastheredcontours
withlevelsrangingfrom10%to90%ofthepeak(39.5Kkms−1)in20%steps.Middlepanel:HIcolumndensityimagefromKimetal.(1998)
withthesameMAGMACOcontoursasshownintheleftpanel(nowinblack).SeveralofthesupergiantHIshells,identifiedbyKimetal.(1999)
(clockwisefromleft:SGSs22,17,16,15,13,12,19,and20),areshownasthepurplecircles.Thislarge-scaleHIstructure,wherethemolecular
ridgeisembedded,correspondstothesoutheasternHIoverdensityregionintheLMC.Rightpanel:CO(1–0)integratedintensityimageofN159
withthesamecontoursaspresentedintheleftpanel.ThelocationsoftheSLWandSSWdetectorsareshownastheblueandgreencrosses,while
thecentraldetectorsforthefirstjiggleobservation(SLWC3andSSWD4)areinyellowandorange.AllthesedetectorsarealsoshowninFig.3.
Finally,thethreeprominentsub-regionsinN159arelabeled.
2. Data ThecalibrationwasobtainedfrommeasurementsofUranusand
itsuncertaintywasestimatedtobe∼10%(SPIREManual)1
Inthissection,wedescribethedatainourstudyandsummarize Toderiveintegratedintensityimagesandtheiruncertainties,
theirmainparameters(e.g.,restwavelength,FWHM,1σuncer- weemploythedatareductionscriptbyWuetal.(2015b),which
taintyintheintegratedintensity,luminosity,etc.;Table1). wasrecentlyusedtosuccessfullygenerateFTScubesforM83.
Wefirststartoffbyperforminglinemeasurementofpointsource
calibratedspectraforeachtransition.Asanexample,thespectra
2.1. HerschelSPIREspectroscopicdata
fromthecentralSLWandSSWdetectorsarepresentedinFig.2,
withthelocationsofthespectrallinesobservedwiththeSPIRE
2.1.1. Observations
FTS.Inourlinemeasurement,acombinationofparabola(con-
N159WwasobservedwiththeSPIREFTSinthehighspectral tinuum) and sinc (emission) functions is used to fit a spectral
resolution(∆f ∼1.2GHz),intermediatespatialsamplingmode. lineforthefrequencyrangeofνline±15GHz,whereνline isthe
TheFTShastwospectrometerarrays,SPIRELongWavelength restfrequencyoftheline.Thecontinuumsubtractedspectraare
(SLW)andSPIREShortWavelength(SSW),whichcoverwave- thenprojectedontoacommongridcoveringa5(cid:48) ×5(cid:48) areawith
lengthrangesof303–671µmand194–313µmrespectively.The apixelsizeof15(cid:48)(cid:48) (roughlycorrespondingtothejigglespacing
FTSbeamprofilechangeswithwavelengthandcannotbechar- oftheSSWobservations)toconstructaspectralcube.Thespec-
acterizedbyasimpleGaussianfunctionduetothemulti-moded trumforeachpixeliscalculatedasthe(1/σ2)-weightedsumof
p
natureoffeedhorncoupleddetectors(Makiwaetal.2013).The overlappingspectra,whereσpisthe1σuncertaintyprovidedby
FTSbeamsizevariesfrom17(cid:48)(cid:48)to42(cid:48)(cid:48)(correspondingto4–10pc thepipeline.Theoverlappingspectraarescaledinproportionto
atthedistanceoftheLMC;Wuetal.2015b)andispresentedin their covering areas in the pixel before the summation. Finally,
Table1.TheSLWandSSWarraysconsistof19and37detectors theintegratedintensity(ICO, ICI,or INII)isderivedbyperform-
respectively,whicharearrangedinahexagonalpatterncovering ing line measurement of the constructed spectral cube and its
a ∼3(cid:48) × 3(cid:48) area. In the intermediate spatial sampling mode, the uncerta(cid:112)inty(σf)isobtainedbyaddingtwoerrorsinquadrature,
SLW and SSW are moved to four jiggling positions with ∼28(cid:48)(cid:48) σ = σ2+σ2, where σ is the statistical error based on the
f s c s
and ∼16(cid:48)(cid:48) spacings respectively. The locations of all detectors residual from line measurement and σ is the calibration error
c
are shown in Figs. 1 and 3. Note that our final maps are sub- of10%.
Nyquist sampled because of the detector spacing that roughly In this paper, we combine the FTS data with other tracers
corresponds to the FTS beam size. The observations were per- of gas and dust (Sects. 2.2–2.4). To compare them at a com-
formedonJanuary8,2013withatotalintegrationtimeof5707s mon resolution, we smooth the FTS maps to 42(cid:48)(cid:48) resolution
(Obs.ID:1342259066;PI:S.Hony). (∼10pcattheLMCdistance),whichcorrespondstotheFWHM
for the CO(4–3) transition, by convolving with proper kernels.
These kernels were generated by Wuetal. (2015b) based on
2.1.2. Dataprocessingandmap-making the fitting of a two-dimensional Hermite-Gaussian function to
the FTS beam profiles, essentially following the method by
WeprocesstheFTSdatausingtheHerschelInteractiveProcess-
ing Environment (HIPE) version 11.0.2825 and the SPIRE cal- 1 http://herschel.esac.esa.int/Docs/SPIRE/html/spire_
ibrationversion11.0(Fultonetal.2010;Swinyardetal.2014). om.html
A85,page3of25
A&A596,A85(2016)
Table1. Spectrallinesanddustcontinuumemissioninourstudy.
Species Transition Restwavelengtha Ea nb FWHMc σd,g,h σe,g,h Luminosityf,g
u crit s,med f,med
(µm) (K) (cm−3) ((cid:48)(cid:48)) (10−11Wm−2sr−1) (10−11Wm−2sr−1) (L )
(cid:12)
12CO J=1–0 2600.8 6 4.7×102 45 0.1 0.8 0.9±0.1
12CO J=3–2 867.0 33 1.5×104 22 1.4 25.1 33.5±1.9
12CO J=4–3 650.3 55 3.7×104 42 15.3 29.4 67.8±2.2
12CO J=5–4 520.2 83 7.2×104 34 11.6 34.6 88.5±2.6
12CO J=6–5 433.6 116 1.3×105 29 6.6 36.9 101.5±3.0
12CO J=7–6 371.7 155 2.0×105 33 6.3 32.5 86.5±2.6
12CO J=8–7 325.2 199 2.9×105 33 17.2 29.4 71.5±2.6
12CO J=9–8 289.1 249 4.0×105 19 14.7 22.6 48.7±1.7
12CO J=10–9 260.2 304 5.3×105 18 18.3 22.7 33.4±1.5
12CO J=11–10 236.6 365 7.0×105 17 20.6 23.2 24.3±1.3
12CO J=12–11 216.9 431 9.0×105 17 17.9 23.2 18.9±1.1
12CO J=13–12 200.3 503 1.1×106 17 27.1 27.9 –
[CI] 3P –3P 609.1 24 4.9×102 38 17.9 20.0 18.8±1.2
1 0
[CI] 3P –3P 370.4 62 9.3×102 33 6.3 17.4 48.8±1.5
2 1
[CII] 2P –2P 157.7 91 2.7×103 12 368.8 9439.1 3907.2±399.9
3/2 1/2
[OI] 3P –3P 145.5 327 1.5×105 12 88.1 622.7 250.3±25.4
0 1
[OI] 3P –3P 63.2 228 9.7×105 10 354.5 6444.5 2435.1±248.9
1 2
[OIII] 3P –3P 88.4 163 5.0×102 10 544.8 1.1×104 5460.5±600.8
1 0
[NII] 3P –3P 205.2 70 4.5×101 17 27.5 42.5 99.2±3.5
1 0
L – 60–200 – – 42 – 3.9×104 (3.3±0.1)×105
FIR
L – 3–1000 – – 42 – 8.1×104 (7.5±0.1)×105
TIR
Notes.(a)DatafromLeidenAtomicandMolecularDatabase(exceptfor[NII]and[OIII],whosedatacomefromCarilli&Walter2013).(b)Critical
density(COdatafromWalkeretal.2015andtherestfromTielens2005).FortheCO,[CI],[CII],and[OI]lines,thecriticaldensitiesareevaluated
atthekinetictemperatureof100K.(c) Angularresolutionoftheoriginaldata.(d) Medianσ intheintegratedintensityat42(cid:48)(cid:48) resolution(σ =
s s
statistical1σuncertainty).(e) Medianσ intheintegratedintensityat42(cid:48)(cid:48) resolution(σ =final1σuncertainty;statisticalandcalibrationerrors
f f
addedinquadrature).(f)LuminosityderivedbyintegratingoverallpixelswithS/N >5at42(cid:48)(cid:48)resolution.(g)ExceptforCO(1–0)at45(cid:48)(cid:48)resolution.
s
(h)ForCO(1–0)andCO(3–2),theconversionfactorsof1.6×10−12and4.2×10−11areusedtoconvertKkms−1intoWm−2sr−1.(d,e,f)Different
spatialareasareconsideredfortheestimates:∼1.0(cid:48)×1.5(cid:48)forthe[CII],[OI],and[OIII]transitionsand∼2.5(cid:48)×2.5(cid:48)fortherest.
Fig.2. PointsourcecalibratedFTSspectraobtainedwiththetwocentraldetectors,SLWC3(red)andSSWD4(blue),forthefirstjiggleobservation.
ThepositionsofthetwodetectorsareshownastheyellowandorangecrossesinFigs.1and3andthespectrallinesobservedwiththeSPIREFTS
areindicatedastheblackdashedlines.
Gordonetal.(2008).Inaddition,werebinthesmoothedmapsto Finally,tocross-checkourmap-makingprocedure,wecom-
haveafinalpixelsizeof30(cid:48)(cid:48),whichroughlycorrespondstothe pare the FTS CO(4–3) integrated intensity image with NAN-
jigglespacingoftheSLWobservations.Wepresenttheresulting TEN2 CO(4–3) observations at 38(cid:48)(cid:48) resolution (Mizunoetal.
integrated intensity images in Fig. 3 and Appendix A and refer 2010). For the comparison, we convolve the NANTEN2 data
toWuetal.(2015b)fordetailsonthemap-makingprocedure. with a Gaussian kernel to have a final resolution of 42(cid:48)(cid:48) and
A85,page4of25
M.-Y.Leeetal.:RadiativeandmechanicalfeedbackintothemoleculargasintheLargeMagellanicCloud.I.
provided by PACSman) and the calibration uncertainty of 22%
(σ ; 10% for spaxel-to-spaxel variations and 12% for absolute
c
calibration)areaddedinquadrature.Wepresentfinalintegrated
intensity maps and a sample of PACS spectra in Figs. 4 and 5
and refer to Lebouteilleretal. (2012) and Cormieretal. (2015)
fordetailsonthedatareductionandmap-makingprocedures.
ThelimitedspatialcoverageofthePACSdata,unfortunately,
results in only several pixels to work with when the maps are
smoothedandregriddedtomatchtheFTSresolution(42(cid:48)(cid:48))and
pixelsize(30(cid:48)(cid:48)).ThecommonregionbetweenthePACSandFTS
data at 42(cid:48)(cid:48) resolution has a size of ∼1.0(cid:48) ×1.5(cid:48) (e.g., Fig. 13)
and all fine-structure transitions are clearly detected with the
statisticalsignal-to-noiseratioS/N (integratedintensitydivided
s
byσ )>5(ourthresholdforlinedetection;Sect.3.1).
s
2.3. Ground-basedCOdata
Fig. 3. CO(7–6) integrated intensity image at 42(cid:48)(cid:48) resolution (pixel
size=30(cid:48)(cid:48)).InourFTSobservations,CO(7–6)isoneofthemostsen- 2.3.1. MopraCO(1–0)data
sitivetransitionswiththelowestmedianstatisticaluncertainty(σ ;
s,med
Table 1). The SLW and SSW arrays are shown as the blue and green We use the CO(1–0) data from the MAGellanic Mopra As-
crosses, except the central detectors for the first jiggle observation sessment (MAGMA) survey (Wongetal. 2011). This survey
(SLWC3andSSWD4)inyellowandorange.Thespectrawiththestatis- targeted bright CO complexes that were previously identified
ticalsignal-to-noiseratioS/Ns(integratedintensitydividedbyσs)>5 from the NANTEN survey (Mizunoetal. 2001) and observed
are presented in red (our threshold for detection; Sect. 3.1) and their them with the 22 m Mopra telescope on 45(cid:48)(cid:48) scales. To esti-
x-axis(inGHz)andy-axis(in10−18Wm−2Hz−1sr−1)rangesareshown mate the CO(1–0) integrated intensity, each data cube was first
inthebottomrightcornerwiththespectrumofthepixelobservedwith smoothedto90(cid:48)(cid:48) resolutionandamaskwasgeneratedbasedon
SLWC3 and SSWD4. The black dashed circle delineates the 2(cid:48) unvi-
the 3σ level. The generated mask was then applied to the orig-
gnetted field-of-view for FTS observations and almost all pixels are
inal cube and the CO(1–0) emission within the mask was in-
within this field-of-view. Note that we do have a spectrum for every
tegrated from3 v = +200 kms−1 to +305 kms−1. The un-
detector and the blank pixels within the SLW coverage in the current LSR
imagesimplyresultfromtherebinningprocess. certainty in the integrated intensity was derived by multiply-
ingtheroot-mean-square(rms)noiseperchannelbythesquare
root of the number of channels that contribute to the intensity
findthattheFTSandNANTEN2dataareconsistentwithin1σ
map at that position. To take a systematic error into account,
uncertainties: the ratio of the FTS to NANTEN2 data ranges
we combine this uncertainty with the calibration uncertainty of
from∼0.7to∼1.2,suggestingthatourmap-makingprocedureis
25% (T.Wong, priv. comm.) and add them in quadrature. For
accurate.
theareathatoverlapswiththeFTScoverage(2.5(cid:48)×2.5(cid:48)insize;
Fig. 7), the final uncertainty at 45(cid:48)(cid:48) resolution has a median of
2.2. HerschelPACSspectroscopicdata ∼5.3 Kkms−1 (Table 1) and the CO(1–0) transition is detected
everywherewithS/N >5.
N159WwasobservedwiththePACSspectrometeronMay24, s
2011,aspartoftheHerschelguaranteedtimekeyprojectSHIN-
ING(PI:E.Sturm).Thefourfine-structurelines,[CII]158µm, 2.3.2. ASTECO(3–2)data
[OI]63µm,[OI]145µm,and[OIII]88µm,weremappedinthe
unchoped scan mode (Obs. IDs: 1342222075 to 1342222084). WeusetheCO(3–2)dataobtainedbyMinamidanietal.(2008).
As described in Poglitschetal. (2010), the PACS spectrometer Minamidanietal. (2008) observed N159 with the 10 m Ata-
is an integral field spectrometer that consists of 25 (spatial) × camaSubmillimeterTelescopeExperiment(ASTE)telescopeat
16(spectral)pixels.Thespectrometercoversawavelengthrange 22(cid:48)(cid:48) resolution. In order to derive the integrated intensity, we
of 51–220 µm with a projected footprint of 5 × 5 spatial pix- integrate the CO(3–2) emission from v = +220 kms−1 to
LSR
els (“spaxels”) on the sky (corresponding to a ∼47(cid:48)(cid:48)×47(cid:48)(cid:48) +250 kms−1. This velocity range is slightly different from that
field-of-view). The FWHM depends on wavelengths, ranging used to estimate the CO(1–0) integrated intensity, but the dis-
from∼10(cid:48)(cid:48)at60µmto∼12(cid:48)(cid:48)at160µm(PACSManual)2. crepancy would not make a significant impact on the CO(3–2)
ThePACSspectroscopicdataarefirstreducedwiththeHIPE integrated intensity considering that the spectra contain essen-
version12.0.0(Ott2010)fromLevel0toLevel1.TheLevel1 tially noise beyond the velocity range of 220–250 kms−1 (e.g.,
cubes(calibratedinbothfluxandwavelength)arethenexported Fig. 2 of Minamidanietal. 2008). The final uncertainty in the
and processed with PACSman (Lebouteilleretal. 2012) to cre- integratedintensityisthenestimatedinthesamewayaswedo
ate integrated intensity images. Each spectrum is fitted with a forCO(1–0):addthestatisticalerrorderivedfromthermsnoise
combinationofpolynomial(baseline)andGaussian(line)func- perchannelandthecalibrationerrorof20%(Minamidanietal.
tionsandthelinefluxesofallspaxelsareprojectedontoagrid 2008) in quadrature. When smoothed to 42(cid:48)(cid:48) resolution and
with a size of ∼1(cid:48) × 2(cid:48). The pixel size of ∼3(cid:48)(cid:48) (corresponding regriddedtomatchtheFTSdata,theCO(3–2)observationshave
to ∼1/3 of the spaxel size) is chosen to recover the best spa- amedianuncertaintyof∼5.9Kkms−1 (Table1)andS/N > 5
s
tialresolutionpossible.Forthe1σerrorintheintegratedinten- isachievedeverywhere.
sity,theuncertaintyfrommapprojection/linemeasurement(σ ;
s
2 http://herschel.esac.esa.int/Docs/PACS/html/pacs_om. 3 Inthispaper,allvelocitiesarequotedintheLocalStandardofRest
html (LSR)frame.
A85,page5of25
A&A596,A85(2016)
Fig.4. IntegratedintensityimagesofPACSfine-structurelines.Theangularresolutionsare∼10(cid:48)(cid:48)for[OI]63µmand[OIII]88µmand∼12(cid:48)(cid:48)for
[OI]145µmand[CII]158µm.TheFTScoverage(e.g.,Fig.3)isindicatedastheblacksolidline.NotethatthePACSobservationsarespatially
limitedandhenceonlypartoftheFTScoverageisshownhere.Ineachimage,thelocationofthespaxelwhereanexamplespectrumisextracted
isalsooverlaidastheredcross.ThisspecificspaxelischosenastheclosestonetotheFTSSLWC3detector(yellowcrossinFig.3).Theextracted
spectraarepresentedinFig.5.
2.4. DeriveddustandIRcontinuumproperties 36(cid:48)(cid:48) resolution and estimated their uncertainties by performing
MC simulations where the measured IR fluxes were perturbed
based on 1σ errors and the SED fitting was repeated 20 times.
We use the dust and IR continuum properties of N159W esti-
In our study, we use the parameters estimated on 42(cid:48)(cid:48) scales to
mated following Galametzetal. (2013). To derive these prop-
matchtheFTSresolution.WerefertoGalametzetal.(2013)and
erties, Galametzetal. (2013) applied the dust spectral energy
Gallianoetal.(2011)fordetailsondustSEDmodeling.
distribution (SED) model by Gallianoetal. (2011) to Spitzer
(3.6, 4.5, 5.8, 8, 24, and 70 µm) and Herschel (100, 160, 250,
350,and500µm)photometricdata(Meixneretal.2006,2013).
3. Results
Theamorphouscarbon(AC)compositionwasusedforthispur-
pose,asitwasdesignedtoconsistentlyreproducetheHerschel
3.1. FTSlinedetection
broadband emission of the LMC. It is more emissive than the
standard Draine&Li (2007) model. In essence, Gallianoetal. WeshowtheFTSspectrainFig.3andAppendixA.1andfocus
(2011)’s approach is twofold: (1) modeling of dust SED for on CO and [CI] line detection in this section. A detailed anal-
a single mass element of the ISM with uniform illumination ysis on the physical properties and excitation conditions of the
and (2) synthesizing several mass elements to account for the CO-emitting gas will be presented in Sects. 4 and 5. Through-
variations in illumination conditions. In the SED fitting proce- out this paper, we apply a threshold of S/N > 5 for line
s
dure,independentfreeparametersarethetotaldustmass(M ), detection and categorize the CO transitions into three groups
dust
PAH-to-dust mass ratio (f ), index for the power-law distri- (e.g., Köhleretal. 2014): low-J for J ≤ 5, intermediate-J for
PAH u
butionofstarlight intensities(α ),lowercut-off forthepower- 6≤ J ≤9,andhigh-JforJ ≥10whereJ istheupperlevelJ.
U u u u
lawdistributionofstarlightintensities(U ),rangeofstarlight In N159W, FTS CO transitions with J = 4–12 are de-
min u
intensities (∆U), and mass of old stars (M ). Based on these tected. These rotational lines have upper level energies of
star
free parameters, the following properties can also be estimated E ∼55–431 K and critical densities of n ∼ 104–106 cm−3
u crit
(Fig. 6 and Table 1): far IR luminosity (60–200 µm; L ), (Table 1; Yangetal. 2010), suggesting that they are valuable
FIR
total IR luminosity (3–1000 µm; L ), and dust temperature probesofmoleculargasoverarangeofdensityandtemperature.
TIR
(T ). Galametzetal.(2013) derived all these parameters at To estimate the total CO luminosity, we add up the integrated
dust
A85,page6of25
M.-Y.Leeetal.:RadiativeandmechanicalfeedbackintothemoleculargasintheLargeMagellanicCloud.I.
Fig.5. ExamplePACSspectra.Thelocationofthespaxelwhereeachspectrumisextractedisindicatedinthetopleftcornerofeachplot,aswell
asinFig.4astheredcross.
Fig.6. DustandIRcontinuumproperties:farIRluminosityL (left),totalIRluminosityL (middle),anddusttemperatureT (right).All
FIR TIR dust
imagesareat42(cid:48)(cid:48) resolutionwithapixelsizeof30(cid:48)(cid:48) andtheSpitzer24µmemissionat6(cid:48)(cid:48) resolutionisoverlaidastheredcontourswithlevels
rangingfrom10%to90%ofthepeak(2149MJysr−1)in10%steps.
intensities at 42(cid:48)(cid:48) resolution over all pixels with S/N > 5 and non-LTE radiative transfer modeling (Sect. 4.2) suggests that
s
provide the results in Table 1. We find the total CO luminos- high-J transitions with J ≥ 13 could contribute up to 8% of
u
ity of L = (575.5 ± 6.8) L (including J = 1,3,4,...,12), thetotalCOluminosity4.
CO (cid:12) u
whichis∼8×10−4 ofthetotalIRluminosity.ThistotalCOlu- Both atomic carbon fine-structure lines at 609 µm and
minosity is dominated by the rotational lines with 4 ≤ Ju ≤ 8 370 µm are detected in N159W. Their upper levels lie ∼24 K
(low-J andintermediate-J)andthelargestcontribution(∼20%) and∼62Kabovethegroundstateandtheircriticaldensitiesare
comes from the CO(6–5) transition. Note, however, that our lowwithn ∼102–103cm−3(Table1;Launay&Roueff1977).
crit
estimate is limited up to Ju = 12. Considering that the ob- UnderthetypicalconditionsofPDRs(densityn∼0.5–107cm−3
served CO SLEDs are relatively flat up to J = 12 for some
u
pixels (Sect. 4.1), hinting at more CO emission at Ju ≥ 13, the 4 Thisestimateisbasedontheassumptionthatthereisnoadditional
actual total CO luminosity is likely higher. To be specific, our warmcomponentemittingatJ ≥13.
u
A85,page7of25
A&A596,A85(2016)
Fig.7. Topleft:CO(1–0)integratedintensityimageat45(cid:48)(cid:48) resolution.TheSpitzer24µmemissionatitsoriginalresolutionof6(cid:48)(cid:48) isoverlaidin
redwithcontourlevelsrangingfrom10%to90%ofthepeak(2149MJysr−1)in10%steps.Topright:sameasthetopleftpanel,butforCO(6–5).
Bottomleft:normalizedCO(1–0)integratedintensityasafunctionofangulardistancefromitspeakat(RA,Dec)=(05h39m37s,−69◦45(cid:48)43(cid:48)(cid:48)).
Theindividualpixelvaluesareshownasthegraycirclesandthe(1/σ2)-weightedmeanvaluesareoverplottedastheredcircleswhenthereismore
s
thanonedatapointineachangulardistancebin.Bottomright:sameasthebottomleftpanel,butforCO(6–5).
and temperature T ∼ 10–8000 K; e.g., Hollenbach&Tielens pixelvaluesasgraycircles,whileoverplotting(1/σ2)-weighted
s
1997; Tielens 2005), the two [CI] lines therefore can be easily values as red circles when there is more than one data point in
excited and thermalized. In N159W, the ratio of the 370 µm to eachangulardistancebin.Ingeneral,wefindthatintermediate-J
609µmlinesis∼1.5–2.3fortheregionswherebothlinesarede- andhigh-J transitionsaremorecompactthanlow-J transitions.
tected,whichindicatesopticallythinemission(e.g.,Pinedaetal. Tobespecific,thewidthathalfmaximumis∼42(cid:48)(cid:48)–60(cid:48)(cid:48) forthe
2008; Okadaetal. 2015). As we do for the CO lines, we then low-J CO lines (J ≤ 5), while not resolved (<42(cid:48)(cid:48)) for the
u
estimate the total [CI] luminosity of L = (67.6±1.8) L by intermediate-J and high-J CO lines (J ≥ 6). Similar results
CI (cid:12) u
summing up the measured integrated intensities at 42(cid:48)(cid:48) resolu- werefoundintheFTSobservationsofGalacticPDRs,OrionBar
tionforthepixelswithS/N > 5(Table1fortheluminosityof andNGC7023(Habartetal.2010;Köhleretal.2014).Wealso
s
eachtransition).Thiscorrespondsto∼9×10−5ofthetotalIRlu- findthatthe[CI]609µmand370µmtransitionsareascompact
minosity.IncomparisonwithCO,wefindthatthe[CI]emission astheCOlineswith J ≥ 6.Note,however,thatourradialpro-
u
ismuchweaker.Tobespecific,thetotalCO-to-[CI]luminosity fileanalysisissomewhatlimitedduetotheincompletecoverage
ratiovariesfrom∼5to∼18withamedianof∼9. oftheFTSmaps(e.g.,blankpixelsattheedge)andlarge-scale
observationsarehencerequiredtostudythespatialdistribution
oftheCOand[CI]emissionmoreaccurately.
3.2. SpatialdistributionoftheneutralgastracedbytheCO
and[CI]emission
4. PhysicalpropertiesoftheCO-emittinggas
Inthissection,weexaminethespatialdistributionoftheneutral
gasprobedbytheCOand[CI]emissionon42(cid:48)(cid:48) scales(∼10pc
4.1. ObservedCOSLEDs
at the LMC distance). First of all, we find that the CO and
[CI] integrated intensities peak at comparable locations, (RA, WepresenttheobservedCOSLEDsfrom J = 1to13(J = 2
u u
Dec) ∼ (05h39m40s, −69◦45(cid:48)30(cid:48)(cid:48)). These peaks are adjacent to notincluded)inFig.8.ToconstructtheCOSLEDonapixel-by-
the peak of the Spitzer 24 µm emission (Fig. 7), which traces pixel basis, we use the integrated intensity images smoothed to
thewarmdustheatedbyyoungstars.Toquantifythespatialdis- thecommonresolutionof42(cid:48)(cid:48)andapplyathresholdofS/N >5
s
tribution of neutral gas, we derive a radial profile for each CO forlinedetection.OurCOSLEDimageclearlyshowsregional
and [CI] transition and measure the radius at which the profile variations in the shape of the CO SLEDs, indicating different
reaches half its maximum. As examples, CO(1–0) and CO(6– physical conditions of the CO-emitting gas. For example, the
5) radial profiles are presented in Fig. 7. We show individual locationofthepeakandtheslopebeyondthepeakaresensitive
A85,page8of25
M.-Y.Leeetal.:RadiativeandmechanicalfeedbackintothemoleculargasintheLargeMagellanicCloud.I.
Fig.9. High-Jslopeimageoverlaidwiththe24µmemissioncontours.
NotethatthetwopixelswhoseCOSLEDsdonothaveclearpeaks(red
inFig.8)aremaskedhere.
Fig.8. COSLEDsforJ =1,...,13(J =2notincluded).Inthisimage,
thepixelscorrespondtoutheindividuauldatapoints(30(cid:48)(cid:48) insize)inour CO SLEDs found for many Galactic and extragalactic sources
FTS maps at 42(cid:48)(cid:48) resolution and the pair of numbers in the top right (e.g., Habartetal. 2010; vanderWerfetal. 2010; Köhleretal.
cornerofeachpixelshowsalocationofthepixelrelativetothecenter 2014;Kamenetzkyetal.2014;Mashianetal.2015).
oftheFTScoverage([0,0]is[RA,Dec]=[05h39m37s,−69◦46(cid:48)13(cid:48)(cid:48)]).
ThecirclesrepresentdetectedCOtransitionsandthedownwardarrows
showupperlimitsbasedon5σ.MostoftheobservedCOSLEDsshow 4.2. Non-LTEmodeling
s
peakat J ≥ 6(gray),whilesomepixelshavepeaksthatareuncertain
u WeanalyzetheobservedCOSLEDsusingthenon-LTEradiative
(red)oroccuratthetransitionslowerthan J = 6(green).Finally,the
u transfercodeRADEX(vanderTaketal.2007).Tocomputethe
best-fitRADEXmodelsdeterminedinSect.4.2areoverlaidinblue.
intensitiesofatomicandmolecularlines,RADEXsolvesthera-
diativetransferequationbasedontheescapeprobabilitymethod.
to the density and temperature, while the CO column density Byassumingthatthespectrallinesareproducedinanisothermal
determinesthelineintensitymagnitude.InN159W,mostofthe andhomogeneousmedium,thissimplemodelcanthenbeused
observedCOSLEDspeakatJ ≥6(grayinFig.8;tenpixelsat toconstrainthephysicalconditionsofthemedium,e.g.,kinetic
u
J =6andonepixelat J =7),whilesomeshowpeaksthatare temperature,density,andcolumndensityofeachspecies.
u u
uncertainduetonon-detections(red;twopixels)oroccuratthe To model the intensity of each CO transition, we consider
transitionslowerthanJu =6(green;onepixelatJu =5andtwo a grid of the following input parameters: kinetic temperature
pixelsatJu =4). Tk = 10–103 K, H2 density n(H2) = 102–105 cm−3, CO col-
It is interesting to notice that the CO SLEDs with peaks at umn density N(CO) = 1015–1020 cm−2, and beam filling factor
Ju ≤5arealllocatedattheedgeofourFTSmaps,somedistance Ω = 10−3–1. These input parameters are sampled uniformly in
awayfromtheactivestar-formingregions.Inaddition,theslope log space with 50 points, except for N(CO) where 100 points
ofeachCOSLEDbeyonditspeakspatiallychanges.Toquantify areused.Inaddition,weusethecosmicmicrowavebackground
howdifferenttheslopesare,wecomputethe“high-J slope”by radiation temperature of 2.73 K and the FWHM linewidth of
∆ICO,norm =[ICO(Jp+3)−ICO(Jp)]/ICO(Jp),whereICO(Jp)isthe 10kms−1 (basedonPinedaetal.2008;andMizunoetal.2010,
integrated intensity of the peak transition Jp and ICO(Jp +3) is where CO transitions up to J = 7–6 were spectrally resolved
theintegratedintensityofthe Jp+3transition(Fig.9).Bydef- for N159W). In our modeling, we compare the observed inte-
inition, the high-J slope is always negative and a smaller value grated intensities (I ) with RADEX predictions scaled by the
obs
representsasteeperdecrease.TheJp+3transitionischosenhere beam filling factor (ΩImod) and find the best-fit model with the
to calculate the high-J slope for all pixels except for those two minimumχ2whereχ2isdefinedas
whosepeaksareuncertain.Wefindthatthehigh-Jslopechanges
byafactorofthreeacrossN159W.Inparticular,thetwopixels (cid:88)13 (cid:34)I −(ΩI )(cid:35)2
correspondingtotheactivestar-formingregionsat(RA,Dec)∼ χ2 = i,obs i,mod · (1)
(05h39m40s, −69◦46(cid:48)10(cid:48)(cid:48)) (traced by the Spitzer 24 µm emis- i=1 σi,obs
sioninFig.9)havetheflattestCOSLEDswith∆I ∼−0.3
CO,norm
while the pixels at the northwest edge of our FTS maps show Hereσ istheuncertaintyintheobservedintegratedintensity
obs
thesteepestdecreasewith∆I ∼−0.7.Asimilarapproach andthesummationisdonefortheCOtransitionsfromJ =1to
CO,norm u
was recently adopted in Rosenbergetal. (2015), where the pa- 13(J =2notincluded)withS/N >5.Westartwithfittingthe
u s
rameter α = [L (J = 11−10)+ L (J = 12−11)+ L (J = COlineswithasingletemperaturecomponent,whichissimplis-
CO CO CO
13−12)]/[L (J = 5−4) + L (J = 6−5) + L (J = 7−6)] ticconsideringthatmultipleISMphasesarelikelymixedatour
CO CO CO
was used to categorize the CO SLEDs of 29 (ultra)luminous spatialresolutionof∼10pc.Nevertheless,usingonecomponent
infrared galaxies ((U)LIRGs). Almost an order of magnitude wouldstillprovideaveragephysicalconditionsofthephasesin
variation in the α parameter was found, confirming the diverse thebeam.
A85,page9of25
A&A596,A85(2016)
Fig.10. Histogramsof“good”parametersforthecentralpixelofourFTSmaps([0,0]inFig.8;[RA,Dec]=[05h39m37s,−69◦46(cid:48)13(cid:48)(cid:48)]).To
derive these histograms, the threshold of χ2 ≤ minimum χ2 + 4.7 is applied to the calculated χ2 distribution. Along with the log value of the
best-fitparameter,thestandarddeviationoftheχ2 distributionmeasuredinlogspaceareshowninblueinthetoprightcornerofeachplot.The
correspondingbluelineofeachhistogramrepresentsthebest-fitparameter.Finally,thenumberinblackintheT plotistheratioofthenumber
k
of“good”modelstothetotalnumberofRADEXmodelsinourstudy.SeeSect.4.2fordetails.
Wedeterminethebest-fitRADEXmodelsforindividualpix- region.Forexample,Ωvariesfrom0.02to0.5,butagainmostly
els and show them with the observed CO SLEDs in Fig. 8. To around the median of 0.1 (Fig. 11). The estimated beam filling
illustratehowtoestimatetheuncertaintiesinthebest-fitparam- factorof∼0.1suggeststhattheCOclumpsinN159Wareonav-
eters,wethenshowthehistogramsof“good”parametersforthe erage∼9(cid:48)(cid:48)insize(∼2pcattheLMCdistance).Thisisconsistent
central pixel of our FTS maps ([0,0] in Fig. 8) in Fig. 10. To with what high-resolution ALMA observations of 13CO(1–0),
derivethesehistograms,weapplyathresholdofχ2 ≤minimum 13CO(2–1),and12CO(2–1)inN159Wfound(Fukuietal.2015):
χ2 +4.7 to the calculated χ2 distribution: ∆χ2 = 4.7 is chosen thespatiallyresolvedCOstructuresareroughly∼10(cid:48)(cid:48)insize.
for the 1σ confidence interval with four parameters (Tk, n(H2), Aninterestingtrendisthatthepixelswithhighkinetictem-
N(CO),andΩ;Pressetal.1992).Inadditiontothehistograms peratures(∼500–754K)andbeamfillingfactors(∼0.1−0.2)co-
of the four primary parameters, those of secondary parame- incidewiththedatapointswiththeflattestCOSLEDs(Fig.9),
ters, thermal pressure P = nTk (cid:39) n(H2)Tk and beam-averaged essentiallytracingthemassivestar-formingregions.Ontheother
CO column density (cid:104)N(CO)(cid:105) = ΩN(CO), are presented. We hand,theCOcolumndensitydistributionshowstheopposite:the
calculate the upper and lower error bounds by measuring the peak(∼(0.6–1)×1018cm−2)insteadoccursatthesoutheastand
standarddeviationofeach“good”parameterdistributioninlog northwest edges of our FTS coverage. We note, however, that
space, while noting that the distributions of “good” parameters theprimaryparametersaredegenerateinourRADEXmodeling
are not always symmetric around the best-fit values. Since the andhaverelativelylarge1σuncertainties(e.g.,Table2),limiting
primary parameters are degenerate in our modeling (Tk with ourabilitytoexaminethespatialdistributionsoftheparameters
n(H2)and N(CO)withΩ)5,theirproducts, Pand(cid:104)N(CO)(cid:105),are moreaccurately.
betterconstrainedingeneral.Theimagesofthebest-fitparame- Astheprimaryparametersarenotindependentofeachother,
tersarepresentedinFig.11andtheirrangesand1σuncertainties itisthereforetheirproducts(secondaryparameters)thatcanbe
aresummarizedinTable2. better constrained and interpreted more straightforwardly. We
findthatboththethermalpressureandbeam-averagedCOcol-
umn density distributions are quite uniform with only a factor
4.3. Spatialdistributionsofthebest-fitparameters
offourvariationsoverourwholecoverage.Theirmedianvalues
Figure11showshowthepropertiesoftheCO-emittinggasvary are9.5×105Kcm−3and3.0×1016cm−2respectively.
acrossN159W.Wesummarizeourfindingsasfollows. In summary, the excellent agreement between the RADEX
Asfortheprimaryparameters,T andn(H )showquiteuni- models and our CO SLEDs (Fig. 8) suggests that the CO lines
k 2
form distributions (only a factor of two variations around the upto J = 12–11observedon∼10pcscalesareonaveragepro-
medianvaluesof429Kand2.2×103 cm−3),whileN(CO)and ducedbyasingletemperaturecomponent.Ourmodelingshows
Ω change by a factor of ∼30 across the entire ∼40 pc × 40 pc that this component can be characterized by high thermal pres-
suresof∼9.5×105Kcm−3andmoderatebeam-averagedCOcol-
5 Highertemperaturesandlowerdensitiescanproducethesameinten- umndensitiesof∼3.0×1016 cm−2 whosedistributionsareuni-
sityaslowertemperaturesandhigherdensities.Thesamethingapplies formacrossthe∼40pc×40pcregion.Consideringthegoodfit
toCOcolumndensitiesandbeamfillingfactors. withthesingletemperaturecomponent,wedonotattempttoadd
A85,page10of25