Physical Properties of Molecular Clouds at 2 parsec Resolution in the Low-Metallicity Dwarf Galaxy NGC 6822 and the Milky Way

5.2 MB·

by  Andreas Schruba

Accepted for publication in The Astrophysical Journal
PreprinttypesetusingLATEXstyleemulateapjv.01/23/15
PHYSICAL PROPERTIES OF MOLECULAR CLOUDS AT 2 PARSEC RESOLUTION
IN THE LOW-METALLICITY DWARF GALAXY NGC 6822 AND THE MILKY WAY
Andreas Schruba1, Adam K. Leroy2, J.M. Diederik Kruijssen3,4, Frank Bigiel5, Alberto D. Bolatto6,
W.J.G. de Blok7,8,9, Linda Tacconi1, Ewine F. van Dishoeck1,10, Fabian Walter4
Accepted for publication in TheAstrophysicalJournal
ABSTRACT
We present the ALMA survey of CO(2-1) emission from the 1/5 solar metallicity, Local Group dwarf
7 galaxy NGC 6822. We achieve high (0.9(cid:48)(cid:48) ≈ 2pc) spatial resolution while covering large area: four
1 250pc × 250pc regions that encompass ∼2/3 of NGC 6822’s star formation. In these regions, we
0 resolve ∼150 compact CO clumps that have small radii (∼2−3 pc), narrow line width (∼1 km s−1),
2 and low filling factor across the galaxy. This is consistent with other recent studies of low metallicity
galaxies, but here shown with a 15× larger sample. At parsec scales, CO emission correlates with 8µm
n
emissionbetterthanwith24µmemissionandanti-correlateswithHα,sothatPAHemissionmaybean
a
J effective tracer of molecular gas at low metallicity. The properties of the CO clumps resemble those of
similar-size structures in Galactic clouds except of slightly lower surface brightness and CO-to-H ratio
0 2
∼1−2× the Galactic value. The clumps exist inside larger atomic-molecular complexes with masses
1
typical for giant molecular cloud. Using dust to trace H for the entire complex, we find CO-to-H to
2 2
] be ∼20−25× the Galactic value, but with strong dependence on spatial scale and variations between
A complexes that may track their evolutionary state. The H -to-Hi ratio is low globally and only mildly
2
above unity within the complexes. The SFR-to-H ratio is ∼3−5× higher in the complexes than
G 2
in massive disk galaxies, but after accounting for the bias from targeting star-forming regions, we
.
h conclude that the global molecular gas depletion time may be as long as in massive disk galaxies.
p Subject headings: galaxies: individual (NGC 6822) – ISM: clouds – Hii regions – radio lines: ISM
-
o
r
t 1. INTRODUCTION of log-normal shape (see review by Mac Low & Klessen
s 2004). In molecular clouds forming stars, observations
a Observationsshowthatstarsformincold,denseclouds
also find that this pdf exhibits a power-law tail at high
[ composed of molecular (H ) gas. However, our under-
2 columndensities. Thistailcorrespondstosmall,pc-sized,
standing of the physical processes of molecular cloud
1 high (column) density gas clumps likely to collapse under
and star formation is still limited (see reviews by McKee
v their self-gravity and form a new generation of stars (e.g.,
& Ostriker 2007; Kennicutt & Evans 2012; Tan et al.
8 Kainulainen et al. 2009; Rathborne et al. 2014; Abreu-
2014). In particular, our knowledge of how molecular
4 Vicente et al. 2015). We also observe that the structure
cloud structure relates to star formation is rapidly evolv-
7 of molecular clouds has an imprint on the output stellar
ing. Overthepastdecade,thislinkhasbeenexploredvia
2 population: the shape of the clump mass function and
detailed observations of clouds in our own Galaxy. These
0
the stellar initial mass function are similar (e.g., Alves
. show that the density structure inside molecular clouds
1 is governed by super-sonic turbulence, which creates a et al. 2007; Rathborne et al. 2009); there is an apparent
0 column density threshold for high mass star formation
(column) density probability distribution function (pdf)
7 (Kauffmann&Pillai2010); andthemaximumcloudmass
1 and maximum stellar cluster mass in galaxies correlate
Emailforcorrespondingauthor: [email protected]
v: 1Max-Planck-Institutfu¨rextraterrestrischePhysik,Giessen- and both increase with the gas pressure (Kruijssen 2014).
i bachstraße1,85748Garching,Germany The number of clouds and the diversity of physical en-
X 2DepartmentofAstronomy,TheOhioStateUniversity,140 vironmentsandevolutionarystatesthatcanbeprobedin
W18thSt,Columbus,OH43210,USA
the Milky Way remain limited. With the Atacama Large
r 3AstronomischesRechen-Institut,Zentrumfu¨rAstronomieder
a Millimeter/submillimeter Array (ALMA), we can now
Universita¨tHeidelberg,Mo¨nchhofstraße12-14,69120Heidelberg,
Germany resolve molecular cloud structure in the nearest galaxies
4Max-Planck-Institutfu¨rAstronomie,K¨onigstuhl17,69117 (e.g., Indebetouw et al. 2013). This allows the prospect
Heidelberg,Germany
to measure the link between galactic environment, cloud
5Institutfu¨rtheoretischeAstrophysik,Zentrumfu¨rAstronomie
derUniversit¨atHeidelberg,Albert-UeberleStr.2,69120Heidel- structure,andstarformationbeyondonlytheSolarNeigh-
berg,Germany borhood. Amainfirsttargetofsuchstudiesarelowmass,
6DepartmentofAstronomy,LaboratoryforMillimeter-Wave low metallicity galaxies. These “primitive” systems are
Astronomy, University of Maryland, College Park, MD 20742,
of interest because they appear so different from large
USA
7NetherlandsInstituteforRadioAstronomy(ASTRON),Post- spiral galaxies like the Milky Way. They have large reser-
bus2,7990AADwingeloo,TheNetherlands voirs of atomic gas in extended distributions, with long
8Astrophysics,CosmologyandGravityCentre,Univ.ofCape total gas consumption times and high gas mass fractions
Town,PrivateBagX3,Rondebosch7701,SouthAfrica
9Kapteyn Astronomical Institute, University of Groningen, compared to present-day large spiral galaxies. Their low
POBox800,9700AVGroningen,TheNetherlands abundanceofmetalsaffectstheirobservedpropertiesand
10LeidenObservatory,LeidenUniversity,P.O.Box9513,2300 may be expected to influence the abundance of cold gas,
RA,Leiden,TheNetherlands
2 Schruba et al.
thestructureofcoldclouds,andtheabilityofgastoform (see review by Bolatto et al. 2013).
stars. These targets are of particular interest because Despite these concerns, CO remains the second most
early galaxies share many of these properties. Although abundant molecule in metal-poor galaxies and CO emis-
present day dwarf galaxies are not perfect analogs to sion is an indispensable tool to detect cold, dense clouds
early-universe systems, the first galaxies were certainly and map their structure. Other indirect tracers of H ,
2
born with few metals and few stars, so that the physics including optical extinction, dust emission, ionized or
that we measure in these local galaxies would have also neutral atomic lines, and other molecular lines also suffer
been at play there. from metallicity effects. More practically, the resolution
Metals, both in the gas phase and in the form of inter- and sensitivity of ALMA still makes CO the fastest way
stellar dust, should affect the structure of star-forming to map molecular cloud structure at low metallicity.
clouds. Gas-phasemetalsactasimportantcoolants,while Observations do show CO emission to be faint in low
dustshieldscloudinteriorsfromexternalradiation,which metallicity galaxies. The ratio of CO emission to star
would heat the gas and dissociate molecules. Interstellar formation is a strong function of metallicity, with more
dust also facilitates molecule formation via reactions on metal-poor galaxies showing much less CO per unit star
grain surfaces. Because low temperatures make the gas formation than metal-rich galaxies (Schruba et al. 2012).
clouds susceptible to gravitational collapse, both cooling ObservationsofmolecularcloudsintheMagellanicClouds
and shielding are important to the ability of gas to form at ∼10 pc resolution show that their CO luminosities are
stars. If the formation of cold, dense gas depends on much lower than those of Galactic clouds of comparable
the abundance of metals, then pristine (low metallicity) size (e.g., Fukui et al. 2008; Hughes et al. 2010). On the
environments may be rendered inefficient or unable to other hand, dust emission indicates significant amounts
form stars. On the other hand, recent theoretical work of H gas (i.e., excess IR emission for their Hi mass) so
2
suggeststhatmoleculesmaynotbeessentialtoformcold, that the CO-to-H conversion factor can be orders of
2
dense gas because neutral atoms or ions (e.g., carbon magnitudehigherthaninourGalaxy(Bolattoetal.2011;
and oxygen) can act as effective coolants in pristine gas Leroy et al. 2011; Jameson et al. 2016; Shi et al. 2015).
(Glover & Clark 2012a). However, these regions may also Following the physical scenario above, one popular inter-
host H molecules because of the absence of dissociating pretation of these observations is that CO is selectively
2
radiation. In this case, the main effect of a lack of metals photodissociated compared to H over a large area in low
2
maybeobservational,renderingmoleculargashardtosee metallicity molecular clouds; in this case, CO molecules
or changing the balance of atomic and molecular gas in persist only in the most opaque, densest gas clumps (e.g.,
coldregionsbutnottheoverallabilityofgastoformstars. Pak et al. 1998; Bolatto et al. 2013).
An observational confirmation of this picture remains to The most direct test of this scenario is to resolve the
be made. structure of individual low metallicity molecular clouds.
We need observations to test how metals affect molec- For a long time, this was only possible in the Large and
ular cloud structure and star formation, but the lack Small Magellanic Clouds (LMC and SMC with ∼1/2
of metals complicates the process of observing molecu- and ∼1/5 solar abundance, respectively), and even then
lar clouds. In our Galaxy, molecular cloud structure is only at (cid:38)10 pc resolution (e.g., Mizuno et al. 2001; Bo-
mapped by observing dust extinction, dust emission, or latto et al. 2003; Wong et al. 2011). ALMA changes
molecular line emission. Unfortunately, H molecules are this, allowing ∼pc scale measurements of cloud structure
2
a very inefficient emitter in the cold interstellar medium in low metallicity galaxies throughout the Local Group.
andabsorptionmeasurementsrequireabrightbackground Indebetouw et al. (2013) demonstrated this capability,
source. Less common, but more visible molecules, most presenting a sub-pc-resolution view of a (small) part of
commonly carbon monoxide (CO), are used to trace H . the 30 Doradus region in the LMC. Recently, Rubio et al.
2
Of course, both dust and CO are made of metals, com- (2015) presented CO(1-0) observations from the Local
plicating their use to trace gas in metal-poor systems. Group galaxy WLM, which has only ∼1/8 solar abun-
For CO the problem is even more complex, because the dance. They found CO emission to originate from small
abundance of CO depends on shielding by dust or H ((cid:46)4 pc across) structures that fill only a tiny fraction of
2
from dissociating radiation and the conditions for CO to the molecular cloud area. Despite strong differences in
survive differ somewhat from those for H to survive. In CO morphology, Rubio et al. estimated that the physical
2
the Solar Neighborhood, this is only a modest concern properties (density, pressure, and self-gravity) of these
becausedustabsorbsenergeticphotonsoverabroadwave- CO-emitting structures are comparable to clumps of sim-
length range. As result, CO and H are well-mixed and ilar size in metal-rich clouds as observed in the Solar
2
COobservationsprovideanefficientandreliabletracerof Neighborhood. Thisresultarguesthatthestarformation
the molecular gas. In low metallicity environments this is process and the resulting stellar population (e.g., stellar
no longer the case. With decreasing metal and thus dust initial mass function and star cluster properties) may be
abundance, H self-shielding becomes the primary shield- only weakly affected by changing metallicity, with the
2
ing mechanism against dissociating radiation. Due to its main influence of metallicity to be changing the distribu-
low abundance CO cannot (effectively) self-shield and tion of the CO tracer molecules.
persists onlyinregions whereH hasabsorbedalldissoci- Rubio et al. (2015) found ten CO-emitting clumps in
2
atingradiationintheLyman-Wernerbands(Wolfireetal. two molecular clouds in one galaxy and Indebetouw et al.
2010). Thus, CO emission traces only the densest, most (2013) studied a single region. With the goal of a statisti-
opaque parts of molecular clouds, while H remains to calmeasurementofthestructureofCOinlowmetallicity
2
fill most of the cloud volume. These physics are thought clouds over a wide area, we used ALMA to map CO(2-1)
to give a strong metallicity dependence for the CO-to-H emission across five star-forming complexes in the Local
2
conversion factor averaged over whole clouds or galaxies Group dwarf galaxy NGC 6822. These five regions con-
Physical Properties of Molecular Clouds 3
tain the bulk of the ongoing star formation activity in
TABLE 1
NGC6822. Byusingasetoflargemosaics(260pointings
Global Properties of NGC6822
in total) at λ=1.3 mm, we are able to cover the whole
area of each complex, from cloud core to outskirts, while
Property Value Reference
still achieving the highest spatial resolution (2 pc) yet
reached to study cloud structure in any galaxy beyond HubbleType IB(s)m(9.8) NED/LEDA
R.A.a 19h44m57.74s NED
the Magellanic Clouds. DEC.a −14d48m12.4s NED
In this paper, we present this new ALMA survey of Distance 474±13kpc Richetal.(2014)
NGC6822(section2)anduseittomeasurethestructure SystemicVel. −57±2kms−1 Koribalskietal.(2004)
ofthestar-formingISMatlowmetallicity(section3). Our Inclination 60±15deg Weldrakeetal.(2003)
PositionAngle 115±15deg Weldrakeetal.(2003)
results are presented in section 4. We estimate the global
E(B-V) 0.21mag Schlafly&Finkbeiner(2011)
CO luminosity of NGC 6822 (subsection 4.1). Then we foregrnd
E(B-V) 0.0−0.3mag Efremovaetal.(2011)
internal
measure the large-scale properties — size, mass, density, 12+logO/H 8.02±0.05dex Garc´ıa-Rojasetal.(2016)
and phase balance — of the atomic-molecular complexes R25 8.69arcmin LEDA
thathosttheCOemission(subsection4.2). Subsequently MV −15.2±0.2mag Daleetal.(2007)
wecharacterizethespatialandspectralintensitydistribu- Mstar 1.5×108 M(cid:12) Maddenetal.(2014)
Matom 1.3×108 M(cid:12) Weldrakeetal.(2003)
tion of CO emission from these complexes by comparing Mmol <1×107 M(cid:12) Gratieretal.(2010)
it to other PDR tracers (subsection 4.3) and to those Mdustb 2.9+−20..98×105 M(cid:12) R´emy-Ruyeretal.(2015)
in Galactic molecular clouds (subsection 4.4). Finally, GDRb 480+170 forabovevalues
−240
we derive the small-scale properties — size, line width, SFR(mix) 0.015M(cid:12) yr−1 Efremovaetal.(2011)
mass, and gravitational boundedness — of the CO-bright
Note. —Allmassesscaledaccordingtoouradopteddistance.
clumpsinourdataandcomparethemtocomparable-size
a Opticalcenter,theHidynamicalcenterisnearby.
structures in WLM and our own Galaxy (subsection 4.5).
b Dustmassderivedforanamorphouscarbonaceouscomponent.
We conclude by discussing these results (section 5) and
providing a brief summary (section 6).
M ≈ 3×105 M (R´emy-Ruyer et al. 2015). The
dust (cid:12)
1.1. The Low Metallicity Dwarf Galaxy NGC 6822
implied gas-to-dust ratio is GDR ≈ 480, which is ∼3
Table 1 summarizes the global properties of NGC 6822. times the Solar Neighborhood value of GDR = 162
(cid:12)
In many ways NGC 6822 resembles a two times less (Zubko et al. 2004; R´emy-Ruyer et al. 2014), but with a
massive version of the SMC (e.g., Jameson et al. 2016) factor of ∼2 uncertainty.
except that the SMC is currently undergoing a strong Figure 1 shows the morphology of NGC 6822 in atomic
interactionwiththeLMCandMilkyWay. Theproximity gas (gray scale) and recent star formation traced by Hα
(D =474±13 kpc; Rich et al. 2014) makes NGC 6822 an (orangecolor)withourALMAsurveyfieldsmarked(blue
ideal target to study cloud structure and star formation boxes). Starformationisconcentratedintheinnerpartof
at high resolution; at this distance, 1(cid:48)(cid:48) ≈ 2.3pc so that the Hi distribution, coincident with the main stellar disk.
ALMA easily resolves cloud sub-structure. Like other OurfourinnerALMAfieldstargetprominentstar-forming
comparatively isolated dwarf irregular galaxies, NGC (Hii) regions in this active area. Together, these harbor
6822 is rich in gas with an atomic gas mass11 of M ≈ 63% of the global Hα flux and 65% of the global Spitzer
atom
1.3×108 M (Weldrake et al. 2003; de Blok & Walter 24µmflux(atracerofembeddedstarformation), sothat
(cid:12)
2006a). This is comparable to the galaxy’s stellar mass, with our ALMA survey we probe the cloud complexes
M ≈1.5×108 M (Madden et al. 2014), so that the responsiblefor∼2/3ofthecurrentstarformationactivity
star (cid:12)
gas mass fraction is ∼50%. NGC 6822 is actively forming inNGC6822. Ourfifthfieldtargetsaregioninthenorth-
stars, the star formation rate (SFR) derived from various west part of the Hi disk, selected to search for cold gas
tracers is SFR ≈ 0.015 M yr−1 (Efremova et al. 2011, associatedwithlow-levelstarformationactivityevidenced
(cid:12)
and references therein), giving it a specific star formation inoptical,ultraviolet,andHαimaging(deBlok&Walter
rate,sSFR ≈10−10 yr−1,typicalofastar-forminggalaxy. 2003, 2006b).
Table2summarizesthepropertiesofourtargetregions.
Despite abundant atomic gas and signatures of high
All four inner regions have been studied extensively and
mass star formation, NGC 6822 has a modest reservoir of
moleculargas. ThemoleculargasmassisM (cid:46)1.0×107 twoofthem,HubbleV&X(ourALMAFields2&1),are
mol classic targets for extragalactic studies of young stellar
M (based on IRAM 30-m observations by Gratier et al.
(cid:12) clusters. As a result, they have been studied via ground-
2010, but also see subsection 4.1 & 5.5 below). Like
based optical broadband (Bianchi et al. 2001; de Blok
other low mass galaxies, NGC 6822 is poor in metals,
with metallicity ∼1/5 the Solar value12 (12+logO/H= & Walter 2000) and narrowband (Hα) imaging (Hodge
et al. 1989; de Blok & Walter 2006b), as well as space-
8.02±0.05,Garc´ıa-Rojasetal.2016;Herna´ndez-Mart´ınez
et al. 2009). It is also poor in dust, with dust mass13 basedbroadband(Bianchi&Efremova2006;Bianchietal.
2012; Efremova et al. 2011) and narrowband (various
11 Throughoutthepaper,allgasmassesincludeafactorof1.36 nebular lines) imaging (O’Dell et al. 1999). These studies
toaccountforheavyelementsandliteraturevaluesarere-scaledto show that all four inner regions are actively forming
ouradopteddistancewherenecessary. massive stars and have likely done so for the past ∼10
12 Throughoutthepaper,weassumeasolaroxygenabundance
Myr. They contain up to ∼100 OB-type stars each and
12+log(O/H) =8.69 and a total solar mass fraction of metals
(cid:12) have Hα luminosities of a few times 1038 erg s−1, which
Z(cid:12)13=W0e.01a4do(pAtspthluenirddeutsatl.m2a0s0s9e).stimate derived from fitting the would rank them among the brightest and most massive
Gallianoetal.(2011)semi-empiricaldustmodelandassumingan star-forming regions in our Galaxy. There is no clear
amorphouscarboncomposition. evolutionary sequence established in the literature, but
4 Schruba et al.
Fig. 1.—OurfiveALMAsurveyfields(bluerectangles,each250pc×250pcinsize)overlaidonanHiimage(grayscale)withcontoursat
columndensitiesofNH=3, 10, 30×1020 cm−2 andanHαimage(orangecolor)highlightingthelocationofprominentHiiregions. The
ALMAsurveycovers∼2/3ofNGC6822’sglobalHαand24µmflux,implyingthatwemapthemolecularISMhosting∼2/3ofthecurrent
starformationactivity. Zoom-insforeachfieldshowingtheALMAdataalongwithancillarydataarepresentedinFigure2,3,4.
Hubble IV & V (our ALMA Fields 4 & 2) show more observing session. Each session contains observations of
compact CO and SFR morphologies than Hubble I&III & a bandpass calibrator. This was the quasar J1924-2914
X(ourALMAFields3&1)andhigherratiosofembedded (flux density ∼ 3.2 Jy at the time of observations) for
toexposedSFRtracerluminosities(8µmor24µmversus Fields 1, 2, 3, 5 and J1733-1304 (flux density ∼1.4 Jy at
Hα; Table 2). Based on this, we argue below that these the time of observations) for Field 4. For all sessions, the
two regions (Fields 4 & 2) may be currently more active phase calibrator was J1939-1525 (inferred flux density of
(i.e., younger) than the others. 0.23 Jy at the time of observation). Titan was observed
during three observing sessions (Fields 2, 3, 4) to set
2. ALMA SURVEY
the absolute flux scale with an estimated uncertainty of
We observed five fields in NGC 6822 with ALMA in 5% using the Butler-JPL Horizons-2012 model (ALMA
Cycle 1 using the 1.3-mm Band 6 receivers (project code: Memo #594). The same flux scale was imposed on the
2013.1.00351.S; PI. A. Schruba). Each field consists of othertwodatasetsthatlackobservationsofTitan(Fields
52 pointings distributed in a Nyquist-spaced hexagonal 1, 5) by requiring the two quasars to have the same flux
grid and covers a 110(cid:48)(cid:48)×110(cid:48)(cid:48) ≈250 pc×250 pc area at densities for all observations.
D =474 kpc. The fields are centered on prominent Hii The data were processed in the Common Astronomy
regions. TheyareshowninFigure1withtheirproperties Software Applications package (CASA, version 4.2.2;
listed in Table 2. Our spectral setup includes one “line” Petry & CASA Development Team 2012) using the
spectral window targeting CO(2-1). This window has “analyst-calibrated” data sets created with help of the
bandwidth0.938GHzwithchannelwidth244kHz(≈0.32 QA2 script-generator tool. The calibrated visibilities
km s−1) and is centered at 230.612 GHz. This “line” were imaged and deconvolved with the clean task using
spectral window covers the CO(2-1) line (rest frequency standard parameters. We chose an angular pixel scale of
230.538 GHz) over a velocity range of −660 to +553 0.15(cid:48)(cid:48) and a channel width of 0.635 km s−1 and imaged
km s−1 (LSRK), easily enough to capture all emission the data using natural weighting. We restored the decon-
from NGC 6822 (systemic velocity −48 km s−1 LSRK). volved image using a single fixed elliptical Gaussian for
each mosaic, so that each image has a single beam.
Table 3 reports dates and weather conditions for each
Physical Properties of Molecular Clouds 5
TABLE 2
Properties of Target Regions
Property Unit Field1 Field2 Field3 Field4 Field5
HiiRegionName ··· HubbleX HubbleV HubbleI&III HubbleIV ···
ClusterName ··· OB13 OB8 OB1&3 OB5 ···
ClusterMass M(cid:12) 7×103 4×103 ··· ··· ···
No.O-typestars ··· 35O−B2V >40 ··· ··· ···
No.OB-typestars ··· 70O−B5V 80O−B5V ··· ··· ···
HαLuminosity 1038 ergs−1 3.362(18%) 4.100(22%) 3.409(18%) 0.941(5.0%) 0.010(0.1%)
8µmFluxDensity microJy 0.087(2.3%) 0.211(5.5%) 0.055(1.5%) 0.150(3.9%) ···
24µmFluxDensity microJy 0.241(9.4%) 0.931(36%) 0.155(6.0%) 0.332(13%) ···
ALMACycle1CO(2-1)Data(NaturalWeighting)
BeamMajorAxis arcsec 0.97 1.03 1.04 1.55 1.16
BeamMinorAxis arcsec 0.71 0.69 0.69 0.68 0.74
BeamSize arcsec 0.83 0.85 0.85 1.03 0.92
··· parsec 1.90 1.94 1.96 2.36 2.13
RmsNoise millyJy 18.7 14.1 14.9 12.4 13.1
··· K 0.63 0.45 0.47 0.27 0.35
Sensitivity Kkms−1 1.1 0.8 0.8 0.5 0.6
··· M(cid:12) pc−2 4.9 3.5 3.7 2.1 2.7
··· M(cid:12) 17.5 13.3 14.0 11.7 12.3
Flux 103 Kkms−1 15.1 117.6 24.3 123.5 ···
Luminosity 103 Kkms−1 pc2 1.8 14.0 2.9 14.7 ···
Mass(forαCO,MW) 103 M(cid:12) 7.8 60.8 12.6 63.8 ···
ALMACycle11.3mmContinuumData(NaturalWeighting)
BeamSize arcsec 0.87 0.89 0.90 1.08 0.96
RmsNoise millyJy 0.19 0.16 0.17 0.14 0.13
Atomic-MolecularComplexMassesasderivedfromHerschelDustModelling
Himass 105 M(cid:12) 3.97±0.04 3.50±0.04 0.76±0.04 6.34±0.04 ···
Dustmass 103 M(cid:12) 3.1±2.2 5.5±3.8 1.7±1.1 4.0±4.0 ···
GDR ··· 420±345 275±127 410±215 420±293 ···
InferredHi+H2 mass 105 M(cid:12) 13.1±1.9 15.0±4.2 6.8±1.8 16.9±1.9 ···
InferredH2 mass 105 M(cid:12) 9.1±1.3 11.5±4.2 6.0±1.8 10.5±2.8 ···
InferredαCO M(cid:12)pc−2(Kkms−1)−1 572±93 90±33 235±72 83±22 ···
Note. —AdopteddistanceD=474kpc; CObrightnesstemperatureratioR21 =1.0; andCO-to-H2 conversionfactor
αCO=4.35M(cid:12) pc−2 (Kkms−1)−1. Sensitivity(1σ)determinedover5.0kms−1. Percentages(giveninparantheses)for
Hα,8µm,24µmfluxesstatefractionsofNGC6822’sglobalfluxes.
TABLE 3
ALMA Observations
TargetField ExecutionBlock StartTime No.ofAntennas AverageElevation PrecipitableWaterVapor
(UTC) (degrees) (mm)
Field1 uid://A002/X7d44e7/X1d11 2014-03-2310:06:34 32(2flagged) 63 2.6
Field2 uid://A002/X7d727d/X63 2014-03-2409:44:57 33(1flagged) 60 1.4
Field3 uid://A002/X7d727d/X1d6 2014-03-2410:29:44 33(2flagged) 70 1.3
Field4 uid://A002/X7d76cc/X19aa 2014-03-2508:51:45 32(0flagged) 49 ···
Field5 uid://A002/X7d76cc/X1e03 2014-03-2511:22:38 31(0flagged) 80 0.6
The properties of the final data cubes are listed in the amount of missing flux from existing IRAM 30-m
Table 2. The average synthesized beam size has FWHM CO(2-1)mappingat15(cid:48)(cid:48) ≈36pcresolution(Gratieretal.
0.9(cid:48)(cid:48) ≈2.0 pc and the achieved rms brightness sensitivity 2010), which covers the ALMA Fields 1, 2, 3. However,
is ∼0.5 K. This translates to a 1σ surface brightness only Field 2 has been robustly detected in the IRAM
sensitivity of ∼0.9 K km s−1 over 5 km s−1 (about 2× 30-m data at a noise level of ∼50 mK over 0.4 km s−1.
the FWHM for a typical CO structure; see below). For For this field, the ALMA observations recover (73±20)%
an appropriate choice of conversion factor (see below), of the single-dish flux. The large uncertainty reflects the
the implied 1σ sensitivity in molecular gas mass surface difficulty to determine the total flux in the IRAM 30-m
densityis∼8M pc−2 andthe1σpointsourcesensitivity cubeduetobaselineinstabilities. Thespatialdistribution
(cid:12)
is ∼30 M . of CO in our other fields is comparable to that in Field 2
(cid:12)
TheALMAobservationsincludebaselinesof15−438m or more compact, so we expect a similar level of flux
length. Thus, emission extending over scales larger than recovery throughout the survey.
the maximum recoverable scale of about 0.6λ/L ≈ We identify genuine emission in the cubes by searching
min
11(cid:48)(cid:48) ≈25 pc is missing in our data sets. We can estimate for emission peaks above 5σ in two adjacent channels
6 Schruba et al.
which we then grow to include all neighboring pixels
TABLE 4
above 1.5σ. This method is somewhat conservative and
Properties of Milky Way Clouds
holds the potential to miss real but low signal-to-noise
emission. However, the comparison to the IRAM 30-m
Property Orion W3/W4 Carina
data suggests that the amount of signal missed has to be
small. Distance(kpc) 0.45 2.0 2.3
No.O-typestars 3 10 70
In addition to CO(2-1), we also observed three “con-
No.OB-typestars 43 105 135(200)
tinuum” spectral windows, each with 2 GHz bandwidth CloudMass(M(cid:12)) 2×105 4×105 6×105
and 15.625 MHz channel width, centered at 229.196 GHz,
References. — Orion: Muench et al. (2008),
215.063 GHz, and 213.188 GHz. We imaged these using
Wilsonetal.(2005);W3/W4: Kiminkietal.(2015),
thecleantaskwiththe“multi-frequencysynthesis”(mfs) Polychroni et al. (2012); Carina: Smith & Brooks
mode. The synthesized beam sizes (0.9(cid:48)(cid:48) ≈ 2.0pc) and (2008),Roccatagliataetal.(2013)
achieved rms sensitivity (∼0.17 mJy) are also listed in
Table 2. Despite a few bright pixels, widespread contin-
uum emission is not detected and we defer discussion of
been kindly provided by M. Galametz and S. Madden
this part of the data to future work.
(private communication). The agreement between the
3. OTHER DATA & METHODOLOGY three dustmaps isgenerally poor. The Galactic cirrus to-
wardNGC6822hassimilarbrightnessasNGC6822itself
Weanalyzetheresultsofoursurveyinfourways. First,
which severely complicates the usability of the infrared
we extrapolate the results of our survey to make some ob-
data. We have tested several methods to remove the
servationsaboutNGC6822asawhole. Thenweconsider
Galactic cirrus but failed to converge on robust results.
thelarge-scalestructureofstar-formingatomic-molecular
In addition, we suspect problems in the Herschel PACS
complexes, comparing the distributions of atomic gas,
and SPIRE data products themselves (e.g., potentially
moleculargas,anddustacrosseachfield. Becauseatomic
relatedtothedynamicrangeorrecoveryofextendedemis-
gas and dust are observed at much coarser resolution
sion; see also Abreu-Vicente et al. 2016) as individual
thanourCOsurvey, thiscomparisonisrestrictedtolarge
bands show inconsistencies in their intensities (by a fac-
spatial scales, reducing each of our fields to a ∼ 5×5
tor of a few) with any reasonable infrared SEDs in either
element grid. After this, we examine the distribution of
faintorbrightregions. Forthatreason,werevertedusing
CO at high resolution, including cross-comparison with
SPIRE’s350µmbandasdustproxybutnotethatresults
tracers of hot dust and recent star formation. Finally, we
derived from more complex dust modeling (i.e., modified
report the detailed properties of the ∼150 compact CO
blackbody, Galliano et al., or Draine & Li model) agree
clumps in our maps.
within a factor of a few. This uncertainty dominates the
3.1. Tracers of Recent Star Formation error budget in our analysis of dust-inferred gas masses.
The Hi and dust maps are evaluated at a common
We compare the CO emission to Hα and dust emission.
resolution of 25(cid:48)(cid:48) ≈57pc, set by the diffraction limit of
At low resolution, the Hα map (de Blok & Walter 2006b)
SPIRE’s 350 µm band. At this resolution, each of our
traces the distribution of recent star formation. At the
fields has 5×5 almost independent 50×50 pc elements.
high, 0.9(cid:48)(cid:48) ≈ 2 pc, resolution of our ALMA data, Hα
The outer 50 pc wide ring lacks signatures of high mass
traces the structure of Hii regions. We compare CO to
star formation and molecular gas; we use it to measure
dust emission at 8 µm and 24 µm, both from the Spitzer
the local gas-to-dust ratio and to measure the amount of
Infrared Nearby Galaxy Survey (SINGS, Kennicutt et al.
foreground and background emission from dust and Hi.
2003) and first presented in Cannon et al. (2006). The
8 µm map is dominated by emission from polycyclic aro-
3.3. Matched Resolution Comparison Data from
matic hydrocarbons (PAH) and traces photon dominated
the Milky Way and WLM
regions (PDR); the 24 µm map measures hot dust and
traces embedded star formation. In order to interpret our results, we use matched reso-
lution CO data from the Milky Way and WLM. Matched
3.2. Atomic Gas and Dust
spatial resolution measurements of CO emission from
We use Hi and dust surface density maps to measure Milky Way clouds offer a view on cloud structure at high
the large scale structure of the star-forming complexes. (solar)metallicity,whileWLMistheonlyotherlowmetal-
The Hi data are from the Very Large Array (VLA) taken licity galaxy observed so far with data quality matching
as part of the LITTLE THINGS survey (Hunter et al. our own. The contrast of these clouds with our results in
2012). Due to calibration complications, these were not NGC6822illuminateshowconditionsinourtargetgalaxy
released with the rest of the survey, however, will be affectcloudstructureandthedegreetowhichconclusions
presented in I. Bagetakos et al. (in preparation) and have about the impact of metallicity may be general (if they
been kindly provided for use in this paper. apply to both WLM and NGC 6822).
Thedistributionofdustmasssurfacedensityisinferred In the Milky Way, we use CO maps of Orion, Carina,
from infrared data from the Herschel Dwarf Galaxy Sur- andW3/W4. AsTable4shows,thesecloudshavemasses
vey(Maddenetal.2013). Wemodeltheinfraredspectral only a bit lower than the atomic-molecular complexes
energy distribution (SED) between 70−500 µm using targetedinNGC6822. Theyspanarangeofmassivestar
a modified blackbody, the Draine & Li (2007) model, formation activity, with Orion showing modest high mass
and the Galliano et al. (2011) model with an amorphous starformationandwithCarinaandW3/W4beingtwoof
carbonaceouscomponent(seeGalametzetal.2010;R´emy- the most active star-forming regions in the Galaxy. The
Ruyer et al. 2015, for details). The latter data set has CO(1-0) data for Orion and Carina are part of the CfA
Physical Properties of Molecular Clouds 7
Fig. 2.—ALMACO(2-1)peakbrightnessmapsforFields1−4;Field5showsnogenuinesignalandisomittedhere. Thefieldofviewof
eachmosaicis110(cid:48)(cid:48)×110(cid:48)(cid:48)≈250pc×250pcandtheresolutionis0.9(cid:48)(cid:48)≈2.0pc. Thepeakbrightnesslevelinsignal-freeregionscorresponds
to2.5timesthermsnoiselevelwhichvariesbetween0.3−0.6Kamongthefields(Table2). IntegratedintensitymapsareshowninFigure3.
1.2-m Galactic Plane Survey14 and have been presented NGC 6822 to those measured from the FCRAO Outer
in Wilson et al. (2005) and Grabelsky et al. (1987). The Galaxy Survey (Heyer et al. 2001). These data, as re-
CO(2-1) data for the molecular cloud complex W3/W4 processed by Brunt et al. (2003), have a resolution of
have been obtained with the 10-m Heinrich Hertz Sub- 100(cid:48)(cid:48)×0.98 km s−1. At the distance of the Perseus arm
millimeter Telescope (HHT) by Bieging & Peters (2011). (D ≈2 kpc), this corresponds to ∼1 pc. They are thus
For a rigorous comparison, we convolve the Orion and closely matched to the resolution of our ALMA data and
W3/W4 data to the same spatial (2 pc) and spectral provide an ideal Galactic point of reference.
(0.635 km s−1) resolution as our NGC 6822 data. We do We also compare to the ALMA observations of WLM
not match the sensitivities, which typically are a factor by Rubio et al. (2015). WLM is a Local Group dwarf
of a few better for the Galactic data. The Carina data galaxy with ∼1/8 solar metallicity. Its stellar mass and
from the CfA 1.2-m telescope have a native resolution current star formation activity are both ∼10 times lower
of 5.6pc×1.3km s−1; we compare these to our data at than NGC 6822. Rubio et al. have observed CO(1-0) at
their native resolution. 6.2pc×4.3pc and 0.5 km s−1 resolution in two atomic-
We compare the clump properties that we measure for molecular complexes and report the detection of ten CO-
emitting structures. In our analysis of the macroscopic
14 https://www.cfa.harvard.edu/rtdc/CO/
8 Schruba et al.
Fig. 3.—ALMACO(2-1)integratedintensity(moment0)mapsforFields1−4;Field5showsnogenuinesignalandisthusomittedhere.
Thefieldofviewofeachmosaicis110(cid:48)(cid:48)×110(cid:48)(cid:48)≈250pc×250pcandtheresolutionis0.9(cid:48)(cid:48)≈2.0pc. Themomentmapsaresignal-masked
(seetext)andemission-freeregionsareshowningray. Contourmapsofthemoment0mapswithphysicalunitsarepresentedinFigure4.
properties of the CO-emitting structures in NGC 6822 a signal-to-noise cut across several channels. Within this
(subsection 4.5) we include their measurements for WLM mask,wefindallsignificantlocalmaxima. Foreachmaxi-
as listed in their Table 1. mum,weidentifythenearbypixelsthatcanbeassociated
with only that peak (and no others) in an iso-intensity
3.4. Cloud Property Measurements contour. For each region of emission, we measure its size,
We identify discrete objects in our data set, measure line width, and luminosity. We use several methods to do
their size, line width, and luminosity, and compare them this: spatial and spectral moments, area measurements
to the properties of similarly sized objects in our com- at half-peak value or some threshold, and ellipse fitting.
parison data sets. To do this, we use an updated version Eachmeasurementiscorrectedforthefactthatitismade
of the CPROPS algorithm15 (Rosolowsky & Leroy 2006, at finite sensitivity and resolution. The sensitivity calcu-
and A.K.Leroy & E.Rosolowsky, in preparation). For lations assume either a Gaussian profile or use a curve
detailsonCPROPS,werefertoRosolowsky&Leroy(2006), of growth method to account for the finite sensitivity of
Leroyetal.(2015),andA.Schrubaetal.(inpreparation). the data. The resolution corrections are made after the
Briefly,weconsideronlysignificantemission,identifiedby correction for sensitivity and use quadratic subtraction of
thetwo-dimensionalbeamandchannelwidth. Thevalues
15 https://github.com/akleroy/cpropstoo/ reported here are the mean across all of these character-
Physical Properties of Molecular Clouds 9
Fig. 4.—ALMACO(2-1)integratedintensitymapsforFields1−4(fromtoptobottom)shownascontoursovergrayscalemapsofHα
andSpitzer 8µmand24µm(fromlefttoright);thecontourlevelsareatCOintegratedintensitiesofICO=2, 10, 20Kkms−1.
10 Schruba et al.
ization methods. We adopt the scatter in results from andH canbemorecomplex,especiallyatlowmetallicity
2
different measurement approaches as our best estimate where a large envelope of CO-poor H may exist. We
2
of the uncertainty of the size, line width, and luminosity, will consider three scales for α : the scale of CO-bright
CO
because this tends to be as large as any statistical or clumps, the scale of whole individual complexes, and the
calibration uncertainty. whole galaxy. At the small scales of clumps we disregard
the H -rich but CO-poor envelopes of molecular clouds.
2
3.5. CO Excitation and CO-to-H At the scale of individual atomic-molecular complexes we
2
account for all gas (including CO-poor H ) but results
We observe CO(2-1) at high (0.9(cid:48)(cid:48) ≈2pc) spatial reso- 2
may reflect the local environment or evolutionary state
lution. Similar to CO(1-0), we find CO(2-1) emission to
of an individual region. At the scale of the whole galaxy
emergefromcolddensegas,especiallygaswithsignificant
we somewhat marginalize over these conditions.
optical depth, though it can also be emitted under other
conditions. However, the ratio of the brightness tem-
3.6. HI Opacity Correction
peratures of the two lines, R , can vary. If CO(2-1) is
sub-thermallyexcitedorwhen2t1heRayleigh-Jeansapprox- Throughout the paper we work with the Hi emission
imation breaks down at low temperatures, then R ≤1. withoutopacitycorrection. Galaxy-widestudiesconclude
21
On the other hand, gas of low opacity has R ≥1. Ob- that local opacity corrections to the column density can
21
servations in our Galaxy find R =0.65±0.1 at a few exceed an order of magnitude and add globally 20−30%
21
10’s of parsec spatial scale; this is an average of diffuse to the atomic gas mass (see Braun et al. 2009; Kalberla
emission from low density gas and opaque emission from & Kerp 2009; Bolatto et al. 2013, and references therein),
densegas(Yodaetal.2010,J.Mottrametal.,inprepara- but without providing clear quantitative prescriptions
tion). Observationsofnearbydiskgalaxiessuggestavery how to correct observed 21-cm Hi data sets for optical
similar line ratio of R =0.7 measured on ∼kpc spatial depth effects. Small-scale or pencil-beam studies within
21
scale (e.g., Leroy et al. 2009, 2013). On the other hand, the Milky Way suggest the cold neutral medium (that
values of R =1.0±0.3 are found in molecular clouds at causes the absorption) to be in compact clouds of parsec-
21
20 pc resolution in the LMC and SMC (Israel et al. 2003; size or narrow filaments and sheets with up to 10−100 pc
Bolatto et al. 2003) or on larger (∼100 pc) spatial scales length (Heiles & Troland 2003; Kalberla & Kerp 2009),
in IC 10 (L. Bittle et al., in preparation). More detailed but here it remains unknown how these findings extend
multi-transition studies of the CO emission from molec- to galactic scales. Recently, Bihr et al. (2015) presented
ular clouds in the SMC indicate that the CO emission work on the massive cloud complex W43 and advocated
originates from two gas components: a more tenuous and for opacity corrections as high as ∼2.4 over ∼100 pc
not very dense (n = 102 −103 cm−3) component of scales but the mass and surface density of W43 is a
high temperature (HT2 =100−300 K) and a population factor (cid:38) 5 higher than the cloud complexes studied in
kin
of much denser clumps (n = 104−105 cm−3) of low NGC 6822. Overall, these results highlight that optical
H2 depth effects are present in Hi observations but also
temperature (T = 10−60 K) (Israel et al. 2003; Bo-
kin
show that we lack a conclusive understanding how to
latto et al. 2005). As we will see in subsection 4.5, we do
correct 21-cm Hi observations. Therefore, we adopt the
not probe the very dense clumps with our ALMA data,
standard assumption of optically thin Hi emission and
and most of the admittedly scarce observations of low
work without opacity correction.
metallicity star-forming galaxies seem to favor R ≈1;
21
therefore, we adopt R = 1.0 throughout the paper.
21 4. RESULTS
This mainly affects comparisons to Galactic data and we
discuss possible variations when they become relevant. We consider CO emission and molecular gas in NGC
We measure CO emission but are often interested in 6822 moving from large to small scales. First we derive
the distribution of H . The metallicity of NGC 6822 an estimate of the galaxy-wide CO flux for NGC 6822.
2
and the high spatial resolution of our data both compli- Then we consider the structure of the atomic-molecular
cate the translation of CO to H . The CO abundance star-forming complexes that fill our survey fields. We
2
strongly depends on shielding of the dissociating radi- then analyze the local correspondence of CO emission to
ation field and thus is a strong function of metallicity tracers of the ISM and recent star formation and study
(Wolfire et al. 2010). So far, the exact metallicity de- the distribution of CO intensities in our survey fields.
pendence of α remains poorly known, as does any Finally, we characterize the compact structures seen in
CO
secondary dependence on radiation field, cloud structure, ourmaps,comparingthemtosimilarstructuresmeasured
and other quantities. We will derive our own estimates in our Galaxy and WLM.
for α for NGC 6822 using alternative ISM tracers
CO(2−1)
(dust) and dynamical methods. Doing so, we reference to 4.1. Total CO Luminosity of NGC 6822
thecommonlyadoptedMilkyWayvalue,αCO(1−0) =4.35 The whole area of NGC 6822 has not yet been mapped
M pc−2 (K km s−1)−1, which includes a factor of 1.36 in CO, but the galaxy-integrated CO luminosity is im-
(cid:12)
to account for heavy elements (Bolatto et al. 2013). portant to compare the galaxy to other systems. We
Because of our high resolution, the scale-dependence estimate this quantity via an “aperture correction” from
of the CO-to-H conversion factor will also be relevant. ourobservedfieldstothewholegalaxy. Todothis,wecon-
2
The conventional extragalactic definition of α is the sider several tracers of recent star formation, Hα, 24 µm,
CO
mass-to-lightratioofH masstoCOemissionoveralarge 70µm. Thesetracersshouldscalelinearlywithmolecular
2
partofagalaxy. Inthisdefinitioncloudsubstructureand gas, and so CO emission, over large, ∼kpc, spatial scales
even,tosomedegree,cloudpopulationsareaveragedover. (e.g., Schruba et al. 2011; Leroy et al. 2013). We mea-
Within an individual cloud, the relationship between CO sure CO luminosity in our fields, and then also measure