CO(1-0) detection of molecular gas in the massive Spiderweb Galaxy (z=2)

0.64 MB·English

by  B. H. C. Emonts

Mon.Not.R.Astron.Soc.000,1–??(2010) Printed28January2013 (MNLATEXstylefilev2.2)
CO(1-0) detection of molecular gas in the massive
z ⋆
Spiderweb Galaxy ( =2)
B. H. C. Emonts1†, I. Feain1, H. J. A. Ro¨ttgering2, G. Miley2, N. Seymour1, R. P.
Norris1, C. L. Carilli3, M. Villar-Mart´ın4, M. Y. Mao3, E. M. Sadler5, R.D. Ekers1,
3
1 G.A. van Moorsel3, R.J. Ivison6,7, L. Pentericci8, C.N. Tadhunter9, D.J. Saikia10,11
0
2 1CSIRO Astronomy and Space Science, Australia Telescope National Facility, PO Box 76, Epping NSW, 1710, Australia
2Leiden Observatory,University of Leiden, P.O. Box 9513, 2300 RA Leiden, Netherlands
n 3National Radio Astronomy Observatory, P.O. Box 0, Socorro, NM 87801-0387, USA
a 4Centro de Astrobiolog´ıa (INTA-CSIC), Ctra de Torrej´on a Ajalvir, km4, 28850 Torrej´on de Ardoz, Madrid Spain
J 5School of Physics, Universityof Sydney, NSW 2006, Australia

5 6UK Astronomy Technology Centre, Science and Technology FacilitiesCouncil, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ
2 7Institute for Astronomy, University of Edinburgh, Blackford Hill, Edinburgh EH9 3HJ
8INAF Osservatorio Astronomico di Roma, Via Frascati 33,00040 Monteporzio (RM), Italy

] 9Department of Physics and Astronomy, University of Sheffield, Sheffield S3 7RH, UK
O 10Cotton College State University,Panbazar, Guwahati 781 001, India
C 11National Centre for Radio Astrophysics, TIFR, Ganeshkhind, Pune 411 007, India
.
h
p
-
o
r ABSTRACT
t
s The high-redshift radio galaxy MRC1138-262 (‘Spiderweb Galaxy’; z = 2.16), is one
a
of the most massive systems in the early Universe and surrounded by a dense ‘web’
[
of proto-cluster galaxies. Using the Australia Telescope Compact Array, we detected

1 CO(1-0) emission from cold molecular gas – the raw ingredient for star formation –
v across the Spiderweb Galaxy. We infer a molecular gas mass of MH2 = 6×1010 M⊙
2 (for MH2/L’CO=0.8). While the bulk of the molecular gas coincides with the central
1 radio galaxy, there are indications that a substantial fraction of this gas is associated
0
with satellite galaxies or spread across the inter-galactic medium on scales of tens
6
of kpc. In addition, we tentatively detect CO(1-0) in the star-forming proto-cluster
.
1 galaxy HAE 229, 250 kpc to the west. Our observations are consistent with the fact
0 that the Spiderweb Galaxy is building up its stellar mass through a massive burst of
3 widespread star formation. At maximum star formation efficiency, the molecular gas
1 willbeabletosustainthecurrentstarformationrate(SFR≈1400M⊙yr−1,astraced
:
v by Seymour et al.) for about 40 Myr. This is similar to the estimated typical lifetime
i of a major starburst event in infra-red luminous merger systems.
X
r Key words: galaxies:active – galaxies:high-redshift – galaxies:clusters:individual:
a Spiderweb– galaxies:individual: MRC1138-262– galaxies:formation– galaxies:ISM
1 INTRODUCTION 1996). HzRGs are typically the massive central objects in
theseproto-clusters.
High-z Radio Galaxies (HzRGs) are signposts of large
over-densities in the early Universe, or proto-clusters, One of the most impressive HzRGs is MRC1138-
which are believed to be the ancestors of local rich clus- 262, also called the ‘Spiderweb Galaxy’ (z = 2.16;
ters (e.g. Miley & DeBreuck 2008; Venemanset al. 2007). Pentericci et al. 1997; Miley et al. 2006). It is one
Historically, HzRGs were often identified by the ultra- of the most massive galaxies in the early Universe
steep spectrum of their easily detectable radio contin- (M⋆∼2×1012M⊙; Seymouret al. 2007; DeBreuck et al.
uum, which served as a beacon for tracing the surrounding 2010). The Spiderweb Galaxy is a conglomerate of star
faint proto-cluster (Ro¨ttgering et al. 1994; Chambers et al. forming clumps (or ‘galaxies’, following Hatch et al. 2009)
that are embedded in a giant (>200kpc) Lyα halo, located
in the core of the Spiderweb proto-cluster (Pentericci et al.
⋆ FromobservationswiththeAustraliaTelescopeCompactArray 1997;Carilli et al.1998,2002).Thecentralgalaxyhoststhe
† E-mail:[email protected] ultra-steep spectrum radio source MRC1138-262 (−1.2 6
(cid:13)c 2010RAS
2 B. H. C. Emonts et al.
α48..15GGHHzz 6 −2.5 for the various continuum components; 2012 in the compact hybrid H75 and H168 array configu-
Pentericci et al. 1997; Carilli et al. 1997). This source ex- rations. The total on-source integration time was 22h (af-
erts dramatic feedback onto the Lyα gas (Nesvadbaet al. ter discarding data taken in poor weather, i.e. atmospheric
2006). path length rms fluctuations >400µm; Middelberg et al.
Emission-line surveys identified tens of galaxies to 2006). Both 2GHz ATCA bands were centred close to
be associated with the Spiderweb Galaxy and its sur- ν =36.5GHz(T ∼70−100K),correspondingtothered-
obs sys
rounding proto-cluster (Pentericci et al. 2000; Kurk et al. shifted CO(1-0) line. The phases and bandpass were cal-
2004; Croft et al. 2005; Doherty et al. 2010; Kuiper et al. ibrated every 5-15 min with a 2 min scan on the nearby
2011). These galaxies harbour a significant fraction of the bright calibrator PKS1124-186. Fluxes were calibrated us-
system’s unobscured star formation (Hatch et al. 2009). ing Mars.
Over-densities of red sequence galaxies were found by For the data reduction we followed Emonts et al.
Kodama et al.(2007)andZirm et al.(2008).AtMpcscales, (2011b).Therelativefluxcalibrationaccuracybetweenruns
an overdensity of X-ray AGN trace a filamentary structure was.5%,whiletheuncertaintyinabsolutelyfluxaccuracy
roughly aligned with the radio axis (Pentericci et al. 2002). was up to 20% based on the flux-model for Mars (version
Kuiperet al. (2011) showed that the proto-cluster is dy- March2012).Thebroad2GHzband(∆v≈16,000kms−1)
namically evolvedandapossible mergeroftwosubclusters. allowedustoseparatethecontinuumfromthelineemission
Hatch et al.(2009)predictedthatmostofthecentralproto- in the uv-domain by fitting a straight line to the line-free
cluster galaxies will merge with MRC1138-262 and double channels. We Fourier transformed the line data1 to obtain
its stellar mass by z = 0, but that gas depletion will have a cube with robust weighting +1 (Briggs 1995), beam-size
exhaustedstarformationlongbefore,sothattheSpiderweb 9.54”×5.31”(PA63.3◦)andchannelwidth8.6kms−1.The
Galaxy will evolve into a cD galaxy found in the centres of linedatawerebinnedby15channelsandsubsequentlyHan-
present-dayclusters. ningsmoothedtoavelocityresolutionof259kms−1,result-
Sub-millimeterobservationsshowedthattheSpiderweb ing in a noise level of 0.085 mJybm−1 per channel.
has massive star formation extended on scales of >200kpc The spectra presented in this paper were extracted
(Stevenset al. 2003), in agreement with PAH (Polycylic against the central pixel in the regions descibed in the text
Aromatic Hydrocarbon) emission from MRC 1138-262 (pixelsize2.3′′×2.3′′),unless otherwise indicated. Total in-
and two Hα-emitting companions (Ogle et al. 2012). De- tensity images of the CO(1-0) emission were made by sum-
spitesignificantjet-inducedfeedback(Nesvadbaet al.2006; ming the channels across which CO(1-0) was detected. All
Ogle et al. 2012), Seymouret al. (2012) derived a high star estimatesofL′ inthispaperhavebeenderivedfromthese
CO
formation rateof SFR≈1400 M⊙yr−1 andan AGNaccre- totalintensity images. Thedata werecorrected forprimary
tion rate of 20% of the Eddington limit for MRC1138-262, beam attenuation (FWHM = 77arcsec) and are
PrimBeam
indicatingthatitisinaphaseofrapidgrowthofbothblack presented in optical barycentric velocity with respect to
hole and host galaxy. z=2.161.
Howlongthisphasewilllastandhowmuchstellarmass
will be added depends on the available fuel for the ongoing
star formation. The raw ingredient for star formation (and
potentialAGNfuel)ismolecularhydrogen(H ).Extremely 3 RESULTS
2
luminous mid-IR line emission from warm (T > 300K),
Wedetect CO(1-0) emission in theSpiderweb Galaxy (Fig.
shocked H gas has been detected in MRC1138-262 with
2 1). The CO(1-0) profile appears double-peaked, with a
Spitzer(Ogle et al.2012).Thisindicatesthattheradiojets
firm 5σ ‘red’ peak and tentative 3σ ‘blue’ peak, sepa-
may heat large amounts of molecular hydrogen, possibly
rated by ∼1000 km s−1(with σ derived from the integrated
quenchingstarformationinthenucleus(seeOgle et al.2012
line profile). Figure 1 (left) shows a total intensity map
for a discussion). However, in order to fuel the observed
of the red and blue component. The total (‘red+blue’)
large star formation rate, an additional extensive reservoir
CO(1-0) emission-line luminosity that we derive is L′ =
ofcoldmoleculargasmustbepresent.Anexcellenttracerof CO
7.2 ± 0.6 × 1010 K km s−1 pc2 (following equation 3 in
the cold component of H is carbon-monoxide, CO(J,J-1).
2 Solomon & Vanden Bout 2005).2 Table 1 summarises the
Particularly efficient is the study of the ground-transition
CO(1-0) emission line properties.
CO(1-0), which is the most robust tracer of the overall
Figure1(left)showsthatthebulkoftheCO(1-0)coin-
H gas, including the widespread, low-density and sub-
2 cideswith theradio galaxy (region ‘B’). However,thereare
thermally excited component (Papadopoulos et al. 2000,
strong indications from both the gas kinematics and distri-
2001; Papadopoulos & Ivison 2002; Dannerbaueret al.
butionthatasignificant fraction of theCO(1-0) emission is
2009;Carilli et al. 2010; Ivison et al. 2011).
spread across tensof kpc.
In this paper,we present the detection of CO(1-0) in
the Spiderweb Galaxy. We assume H = 71km s−1Mpc−1,
0
Ω =0.27 and Ω =0.73 (i.e. angular distance scale of 8.4
M Λ
kpcarcsec−1 and luminosity distance D =17309 Mpc). 1 Similarly,wemadeacontinuummap(10.39”×6.55”,PA75.5◦).
L
Wedetect the36.6GHzradiocontinuum withanintegrated flux
ofS36.6GHz =10.7mJy(P36.6GHz=3.6×1026 WHz−1)across
three beam-sizes, following the morphology of high resolution
2 OBSERVATIONS 4.7/8.2GHz data of Carillietal. (1997). A detailed discussion
onthe36.6GHzradiocontinuum isdeferredtoafuturepaper.
CO(1-0) observations were performed with the Australia 2 ThemeasurementerrorinL′ doesnotincludea20%uncer-
CO
Telescope Compact Array (ATCA) during Aug 2011 - Mar taintyinthemodelofourusedfluxcalibratorMars(Sect. 2).
(cid:13)c 2010RAS,MNRAS000,1–??
CO(1-0) in the Spiderweb Galaxy 3
Figure 1.Left:ContoursoftheCO(1-0)emissionoverlaidontoanHST/ACS(g475+I814)imageoftheSpiderwebGalaxy(Mileyetal.
2006). The plot roughly equals the size of the giant Lyα halo (see Fig.3), hence the individual galaxies shown here are part of the
Spiderweb Galaxy. Contour levels of the CO(1-0) emission are: 2.8, 3.5, 4.2, 5.0σ (no negative contours are present at the same -2.8σ
level,butseeFig.3foralargerfield-of-view).Theredandbluecontours representtheredandbluepartofthedouble-peakedCO(1-0)
profile, as indicated by the horizontal bar in the right-middle plot (velocity ranges: red: −453 < v < +453km s−1 with σ = 0.046
Jybm−1×km s−1; blue: −1228 < v < −453km s−1 with σ = 0.043 Jybm−1×km s−1). Note that the blue signal is too weak to be
detectedacrosstheFWHMofthesythesizedbeam,hencetheemissionanditsexactlocationneedtobeverifiedwithfutureobservations.
Theblackcontoursrepresentthe8.2GHzradiocontinuumfromCarillietal.(1997).Thedashedellipseinthebottom-leftcornershows
the ATCA beam-size. The red and blue numbers indicate the velocities of individual galaxies as derived from optical emission lines
(adjustedforz=2.161;Kuiperetal.2011).Right:CO(1-0)emission-lineprofilesatfourdifferentregionswithintheSpiderwebGalaxy.
TheaperturesofregionsA,BandCaredescribedinSect2.ThespectrumofthetentativeNEtailwastakenbyaveraging4pixelsalong
thedashedlineintheleftplot.Althoughtheprofilesarenotmutuallyindependent, theysuggestthatthereisachangeintheCO(1-0)
kinematicsacrosstheregionsA,BandC.Thevelocityiscentredonz=2.161(Sect.3).Thedottedhistograminthebottomplotshows
the distributionof optical and UV rest-framelineemitters inthe Spiderweb Galaxy and surrounding proto-cluster (from Kuiperetal.
2011).[Seetheelectronicversionofthepaperforcolorimages.]
First, despite the limited spatial resolution of our ob- In addition, thedouble-peakedCO(1-0) profile spreads
servations,thereappearstobeavelocitygradientinthegas over 1700 km s−1 (FWZI). This is extreme compared
kinematics across the inner 30-40 kpc of the Spiderweb. As to what is found for quasars and submm-galaxies (see
shown in Fig.1, the redshift of the CO(1-0) peak emission Coppin et al. 2008; Wang et al. 2010; Ivison et al. 2011;
decreaseswhengoingfromregion‘A’toregion‘B’toregion Riecherset al.2011;Bothwell et al.2012;Krips et al.2012,
‘C’. Most prominent is the apparent spatial separation be- andreferencestherein).Afewnotableexceptionsarehigh-z
tweenthepeakofthebluecomponentofthedouble-peaked systems in which the broad CO profiles arise from merging
CO(1-0) profilein region Candthepeakof theredcompo- galaxies(Salom´e et al.2012,andreferencestherein).Ascan
nentinregionB.Fig.2visualisesthatalsobetweenregionsB beseeninFig.1(bottomright),thedouble-peakedCOpro-
andAthereisaclearvelocitygradientintheCO(1-0)emis- file resembles the velocity distribution of optical line emit-
sion.ThedecreaseinthevelocityoftheCO(1-0)peakemis- ters detected in the Spiderweb proto-cluster (Kuiper et al.
sion from region A → B → C is consistent with a decrease 2011), be it with a lower velocity dispersion of the CO gas,
in redshift of optical line emitting galaxies found in these in particular on the blue-shifted side.
regions (see Fig.1, though note that the large ATCA beam
Theseresultsthusindicatethatasignificantfractionof
prevents us from spatially resolving the individual galax-
the CO(1-0) detected in the Spiderweb Galaxy likely origi-
ies in our CO data). Fig.1 also shows indications for an
nates from (merging) satellites of the central radio galaxy,
extension (‘tail’) of the CO(1-0) emission beyond region A
or theinter-galactic medium (IGM) between them.
(stretchingupto100kpcNEoftheradiocore),butthistail
is detected only at a 3σ level and thusneeds to be verified. Ourresultsalsosuggeststhattheredshiftofthecentral
Wehavestartedanobservationalprogramtoverifythedis- radio galaxy is associated with the red peak of the CO(1-
tribution and kinematics of theCO(1-0) emission at higher 0) profile, giving z = 2.161 ± 0.001. Kuiper et al.
CO(1−0)
sensitivityandspatialresolution(resultswillbereportedin (2011) discuss that determining the redshift from optical
a futurepaper). andUVrest-frameemission linesisboundtoamuchlarger
(cid:13)c 2010RAS,MNRAS000,1–??
4 B. H. C. Emonts et al.
v = -130 km/s v = 0 km/s v = 130 km/s v = 259 km/s
16 kpc
Figure 2.Channelsmaps oftheCO(1-0)emissioninregionsAandB,highlightingagradientinthegas kinematics.Contour levelsof
theCO(1-0)emission(red)are:3.0,3.8,4.5σ (withσ=0.085mJybm−1;Sect.2).Nonegativefeaturesarepresentinthesechannels at
thesamelevel(-3σ). Theradiocontinuum (blackcontours),beam-sizeandredshiftarethesameasinFig.1.
Table 1. CO(1-0) properties in the various regions of the Spiderweb Galaxy and HAE229: zCO(1−0)=redshift, vCO(1−0)=velocity of
thepeakemissionw.r.tzCO(1−0)=2.161;FWHM=CO(1-0)linewidthathalfthemaximumintensity;FWZI=fullwidthoftheCO(1-0)
profile at zero intensity; Sν(peak)=CO peak flux; ICO(1−0)=CO integrated flux; L′CO(1−0)=CO total intensity: MH2=H2 mass (Sect.
4.1).z,v,FWHMandSν havebeenderivedfromGaussianfitstotheCOprofilesinFigs.1and3.ICO(1−0),L′CO(1−0) andMH2 have
beenderivedfromthetotalintensityimagesinFigs.1and3.
Spiderweb Galaxy (SG) HAE229
SGregionA SGregionB SGregionC
zCO(1−0) 2.163±0.001 2.161±0.001 2.150±0.001 2.147±0.001
vCO(1−0) (kms−1) 175±75 0±30 -1060+−14805 -1355±65
FWHM (kms−1) 550+165† 540±65 550+150† 395±75
−210 −300
FWZI (kms−1) 905±130(A+B)‡ 775±130 520±130
Sν(peak) (mJybeam) 0.44±0.06 0.44±0.06 0.30±0.09 0.32±0.05
ICO(1−0) (Jykms−1) 0.28±0.03(A+B)‡ 0.03±0.01 0.14±0.01
L′CO(1−0) (×1010 Kkms−1pc2) 6.5±0.6(A+B)‡ 0.7±0.2 3.3±0.2
MH2 (×1010 M⊙) 5±1(A+B)‡ 0.6±0.2 3±1
† ValuesarequotedforasingleGaussianprofilefit,witherrorsreflectinguncertainties duetoassymmetryofthecorrespondingprofile
component.
‡ RegionsAandBarespatiallyunresolvedandonlymarginallyresolvedkinematically,henceasinglevalueisderivedfromtheentire
‘red’partofthetotalintensityimageofFig.1.
uncertainty, but they derive 2.158 <z <2.170, which is in kms−1 pc2)−1](whereM includesa helium fraction; e.g.
H2
agreement with our estimated z . Solomon & Vanden Bout 2005). This is consistent with α
CO(1−0) x
found in ultra-luminous infra-red galaxies (LIR > 1012L⊙;
Downes & Solomon1998),butwestressthattheconversion
3.1 HAE 229 fromL′ intoM isnotyetwellunderstood(Tacconi et al.
CO H2
2008; Ivison et al. 2011) and that α crucially depends on
Fig.3 shows that also the dusty star-forming galaxy HAE x
the properties of the gas, such as metallicity and radiation
2202190)(Mis⋆d∼ete5c×te1d0i1n1MCO⊙(;1-K0)urakt3et.7aσl.sig2n0i0fi4c;anDceo.hWerteydeetriavle. field (Glover & Mac Low 2011). Adopting αx =0.8 M⊙ (K
kms−1 pc2)−1resultsinanestimatedmoleculargasmassin
L′ =3.3±0.2×1010 K km s−1 pc2 for HAE229. Table1
suCmOmarisestheCO(1-0)properties.HAE229islocated250 theSpiderwebGalaxyofMH2 ∼6×1010 M⊙.Wearguethat
this is likely a conservative estimate, based on the adopted
kpc(30”) west of MRC1138-262, i.e. outsidethe gaint Lyα
conversionfactorandthefactthatalargeamountofshock-
halo.TheCO(1-0)signalpeaksatv=−1354kms−1,which
heatedmoleculargasresidesinthewarm(T>100K)phase
agrees with the Hα redshift from Kurk et al. (2004). None
(see Ogle et al. 2012). The putativeH mass of HAE229 is
oftheotherline-emittinggalaxiesoutsidetheLyαhalo,but 2
within the field-of-view of our observations, is reliably de- MH2∼3×1010M⊙.
tected in CO(1-0).
4.1.1 Nature of the molecular gas
In Sect. 3 we saw that, while the CO(1-0) distribution is
4 DISCUSSION concentrated on the central radio galaxy, the CO emission
spreads across the inner 30-40 kpc (a region which is rich
4.1 Molecular gas in the Spiderweb
insatellitegalaxies).Basedonitsdistributionandkinemat-
We can estimate the mass of molecular gas by adopting a ics, we argued that part of the molecular gas is thus most
standard conversion factor αx = MH2/L′CO = 0.8 [M⊙ (K likely associated with these satellite galaxies or the IGM
(cid:13)c 2010RAS,MNRAS000,1–??
CO(1-0) in the Spiderweb Galaxy 5
between them. The extreme FWZI of the CO(1-0) emis- morerapidlydepletedbyfeedbackprocesses,suchasshock-
sion (Sect. 3) is another indication that the double-peaked heating(Ogle et al. 2012) or jet-inducedoutflows (found to
profile is not likely caused by a nuclear disc in the central occur at rates of ∼400M⊙yr−1 in the optical emission line
radio galaxy. This is consistent with earlier speculation by gasbyNesvadbaet al.2006).Nevertheless,boththecurrent
Ogle et al. (2012) that the high star formation rates could largestarformationrateandcoldmoleculargascontentim-
be the result of the accretion of gas or gas-rich satellites ply that we are witnessing a phase of rapid galaxy growth
(which were found by Hatch et al. 2009 to contain most of though massive star formation, coinciding with the AGN
the dust-uncorrected, instantaneous star formation), while activity.
nuclearstarformationmaybequenchedbyjet-inducedheat- TheH massismuchlargerthantheestimatedmassof
2
ing of the molecular gas. theemission-linegasintheLyαhalo(Memis =2.5×108M⊙;
The tentative NE tail (see Fig.1) spreads further out, Pentericci et al. 1997). However, Carilli et al. (2002) show
beyond region A (which is rich in satellite galaxies) into a thattheradiosourceisenvelopedbyaregionofhot,shocked
regionwithnoknowncompaniongalaxiesordetectableLyα X-raygasofpotentiallyMhot =2.5×1012M⊙.Thissuggests
emission (Fig. 3). However, a Mpc-scale filamentary struc- that,evenwhenthecurrentreservoirofcoldmoleculargasis
tureexistsineast-westdirection(Pentericci et al.2002).We consumed,thereisapotentialgasreservoiravailableforfu-
speculate that – if confirmed – this tentative tail might in- tureepisodesofstarburst(andAGN)activity,providedthat
dicate that cold gas is found, or being accreted along, this thegascancooldowntoformmolecularclouds(e.g.Fabian
filament. 1994)andthisprocessisnotentirelycounter-actedbyongo-
Two alternative scenarios that should be considered ingAGNfeedback(Carilli et al.2002;Nesvadbaet al.2006;
to explain the CO characteristics are AGN driven out- Ogle et al.2012).Themerger ofproto-clustergalaxies with
flows and cooling flows. Cooling flows have been detected the Spiderweb Galaxy may also trigger a new burst of star
in CO in giant central cluster (cD) galaxies at z<0.4, formation, dependingontheavailablegas reservoirinthese
some of which contain H masses similar to that of the systems.
2
Spiderweb Galaxy. However, compared to the Spiderweb For the dusty star-forming galaxy HAE229, L′
CO(1−0)
Galaxy, these cooling flow galaxies show much narrower is comparable to that of IR-selected massive star-forming
typical line widths (FWHM < 500km s−1, taking into galaxies at z=1.5 (Aravenaet al. 2010) and some high-z
CO
account an uncertain α and the use of narrow-band re- submm galaxies (e.g. Ivison et al. 2011). Our CO results
x
ceivers; Edge 2001; Salom´e & Combes 2003; Salom´e et al. are consistent with observations by Ogle et al. (2012) that
2006).Radio-jet drivenoutflowsofoptical emission-linegas HAE229isgoingthroughamajorandheavilyobscuredstar-
werefoundonscalesoftensofkpcintheSpiderwebGalaxy burstepisode.From theircalculated SFR∼880M⊙yr−1,we
by Nesvadbaet al. (2006). Similar to the CO distribution, derive t ≈ 30 Myr, i.e. similar to that of the Spiderweb
depl
theseopticalemission linesaresignificantlymoreredshifted Galaxy.
NW compared to SE of the radio core, though the ambi-
guity in optical redshifts makes a direct comparison diffi-
4.3 CO(1-0) in HzRGs
cult.Still,thereisaninterestingalignmentbetweenthered-
shiftedCO(1-0)emissioninregionAandtheregioninwhich MRC1138-262 is part of an ATCA survey for CO(1-0) in a
Nesvadbaet al. (2006) detect the fastest redshifted outflow southern sample of HzRGs (1.4 < z < 3; Emonts et al in
velocities in the optical emission-line gas (their ‘zone 1’, prep,seealsoEmonts et al.2011a,b).Sofar,itisoneofonly
whichstretchesinbetweenregionAandB).TheFWHMof veryfewsecureCO(1-0)detectionsamongHzRGs;twoother
theopticalemission-linegasis,however,significantlylarger examples being MRC0152-209 (z = 1.92; Emonts et al.
than that of the CO(1-0) emission, indicating that it has a 2011a) and 6C1909+72 (z=3.53; Ivison et al. 2012).
much larger velocity dispersion. Both the cooling flow and CO detections in HzRGs made with narrow-band
theradio-jetfeedbackscenariodeservefurtherinvestigation, receivers and/or higher transitions are also still lim-
once CO observations with higher resolution and sensitvity ited in number (Scoville et al. 1997; Alloin et al. 2000;
haveconfirmed the extentof theCO(1-0) emission. Papadopoulos et al. 2000, 2001; DeBreuck et al. 2003a,b,
2005;Greve et al.2004;Klamer et al.2005;Nesvadbaet al.
2009; Emonts et al. 2011b; Ivison et al. 2008, 2012, also re-
4.2 Evolutionary stage
view by Miley & De Breuck 2008). However, in some cases
Fromfittingthemid-tofar-IRspectralenergydistribution, CO is resolved on tens of kpc scales (Ivison et al. 2012),
Seymouret al. (2012) derive a starburst IR-luminosity of associated with various components (e.g. merging gas-rich
LIR = 8×1012L⊙ and star formation rate of SFR=1390 galaxies; De Breuck et al. 2005), or found in giant Lyα ha-
M⊙yr−1 forMRC1138-262. LIR/L’CO(1−0) agreeswellwith los that surround the host galaxy (Nesvadbaet al. 2009).
correlationsfoundinvarioustypesoflow-andhigh-zobjects This shows that detectable amounts of cold molecular gas
(e.g. Ivison et al. 2011). Assuming that all the H is avail- inHzRGsarenotrestrictedtothecentralregionoftheradio
2
able to sustain the high SFR, we derive a minimum mass galaxy.Ivison et al.(2012)discussthatCOdetectedHzRGs
depletion time-scale of t = MH2 ≈40 Myr. This is com- (often pre-selected on bright far-IR emission from a star-
depl SRF
parabletotheestimatedtypicallifetimeofamajorstarburst burst)aregenerally associated with mergeractivity.Thisis
episodeinIR-luminousmergersystems(Mihos & Hernquist also the case for the Spiderweb Galaxy, which is located in
1994; Swinbank et al. 2006), though the bulk of the in- an extreme merger environment.
tense star formation in the Spiderweb Galaxy may occur Using the JVLA, Carilli et al. (2011) mapped CO(2-1)
on scales of tens of kpc (Sect. 4.1.1). The mass deple- emission throughout a z=4 proto-cluster associated with
tion time-scale may be shorter if the cold molecular gas is the sub-millimeter galaxy GN20 (tracing a combined mass
(cid:13)c 2010RAS,MNRAS000,1–??
6 B. H. C. Emonts et al.
50 kpc FWHMPB CO(1−0)
Spiderweb Galaxy
HAE 229
v
opt
Figure 3. Left:TheSpiderwebGalaxyandHAE229.Shownisazoom-outofFig.1(left)includingtheCO(1-0)contours ofHAE229
(purple)250kpctothewest.Thelight-bluecontoursindicatetheextentoftheLyαhalothatencompassestheSpiderwebGalaxy(from
Mileyetal. 2006, see also Pentericci etal. 1997). The red and blue contours are the same as in Fig.1 (2.8, 3.5, 4.2, 5.0σ), with σ the
noiselevelatthephasecenter(centredontheSpiderwebGalaxy).Thedark-redanddark-bluecontoursarethecorrespondingnegative
signal (present only at -2.8σ at significant distance from the phase center, where the noise increases due to correction of the signal for
primary beam attennuation). Contour levels of the CO(1-0) in HAE229 are: -2.8 (grey), 2.8, 3.5σ (purple), derived across te velocity
range indicated by the purple bar in the spectrum on the right (−1618 < v < −1100km s−1; σ = 0.048 Jybm−1×km s−1). For
HAE229, σ is the local rms noise (i.e. 30 arcsec from the phase center) in the primary-beam corrected total intensity CO image. The
dashedpartialcircleontherightshowstheFWHMoftheprimarybeam (i.e.theeffective field-of-view)ofourobservations. Thesmall
arrowsindicatethelocationofapeakin24µmemissionthatcoincideswithHAE229(datafromDeBreucketal.2010).Right:CO(1-0)
spectrum ofHAE 229. The systemicvelocity, as derived fromoptical emissionlines,isalsoindicated (Kurketal.2004). Note that the
highernoiseinthisspectrum comparedtothespectraofFig.1isaresultoftheprimarybeamcorrection.
ofMH2 ∼2×1011M⊙).OurresultsontheSpiderwebGalaxy wealthofAustraliaforoperationasaNationalFacilityman-
andHAE229areanotherexampleofthepotentialforstudy- aged by CSIRO.
ingthelowest CO transitionsinproto-clusterenvironments
with theATCA and JVLA.
REFERENCES
AlloinD.,BarvainisR,GuilloteauS.,2000,ApJL,528,L81
5 CONCLUSIONS
AravenaM., et al., 2010, ApJ, 718, 177
WedetectCO(1-0) emission from cold molecular gas across Bothwell M. S., et al., 2012, ArXiv e-prints:1205.1511
themassiveSpiderwebGalaxy,aconglomerateofstarform- Briggs D.S., 1995, PhD thesis, New Mexico Tech
inggalaxies at z=2.16. Whilethebulkof theCO(1-0) coin- CarilliC.L.,RoettgeringH.J.A.,vanOjikR.,MileyG.K.,
cideswiththecentralradiogalaxy,partofthemoleculargas vanBreugel W. J. M., 1997, ApJS,109, 1
isspreadacrosstensofkpc.Weexplainthatthisgasismost Carilli C. L., Harris D. E., Pentericci L., Rottergering
likelyassociatedwithsatellitesofthecentralradiogalaxy,or H.J.A.,MileyG.K.,BremerM.N.,1998,ApJ,494,L143
the IGM between them (though other scenarios are briefly Carilli C. L., et al., 2010, ApJ,714, 1407
discussed). The extensive reservoir of cold molecular gas CarilliC.L.,HarrisD.,PentericciL.,Ro¨ttgeringH.,Miley
likely provides the fuel for the widespread star formation G., Kurk J., van Breugel W., 2002, ApJ,567, 781
that has been observed across the Spiderweb Galaxy. Con- Carilli C. L., Hodge J., Walter F., Riechers D., Daddi E.,
tinuous galaxy-merger and gas-accretion processes are the DannerbauerH.,Morrison G. E., 2011, ApJL,739, L33
likely triggers for the observed high star formation rates. ChambersK.C.,MileyG.K.,vanBreugelW.J.M.,Huang
The total mass of cold gas (MH2 = 6×1010[αx=0.8] M⊙) J.-S., 1996, ApJS, 106, 215
is enough to sustain the current high star formation rate Coppin K. E. K., et al., 2008, MNRAS,389, 45
in the Spiderweb Galaxy for ∼40 Myr, which is similar to CroftS.,KurkJ.,vanBreugelW.,StanfordS.A.,deVries
the typical lifetime of major starburst events seen in IR- W., Pentericci L., Ro¨ttgering H.,2005, AJ, 130, 867
luminousmergersystems.OurCOresultsontheSpiderweb DannerbauerH., et al., 2009, ApJL, 698, L178
Galaxyshowthepotentialforstudyingthecoldgasthrough- De Breuck C., Downes D., Neri R., van Breugel W., Reu-
outhigh-zproto-clusterswiththeATCA,JVLAandALMA. land M., Omont A., Ivison R., 2005, A&A,430, L1
De Breuck C., et al., 2003a, A&A,401, 911
DeBreuckC.,NeriR.,OmontA.,2003b,NewAR,47,285
De Breuck C., et al., 2010, ApJ,725, 36
ACKNOWLEDGMENTS
Doherty M., et al., 2010, A&A,509, A83
We thank Ernst Kuiper for sharing his HST imaging and Downes D., Solomon P.M., 1998, ApJ,507, 615
the anonymous referee for useful feedback that improved Edge A. C., 2001, MNRAS,328, 762
thispaper.WethanktheNarrabriobservatorystafffortheir Emonts B. H. C., et al., 2011a, ApJL, 734, L25
help. BE acknowledges the Centro de Astrobiolog´ıa/INTA Emonts B. H. C., et al., 2011b, MNRAS,415, 655
for their hospitality. NS is recipient of an ARC Future Fel- Fabian A.C., 1994, ARA&A,32, 277
lowship.TheAustraliaTelescopeisfundedbytheCommon- Glover S. C. O., Mac Low M., 2011, MNRAS,412, 337
(cid:13)c 2010RAS,MNRAS000,1–??
CO(1-0) in the Spiderweb Galaxy 7
GreveT.,Ivison R.,Papadopoulos P.,2004, A&A,419, 99
Hatch N. A., Overzier R. A., Kurk J. D., Miley G. K.,
Ro¨ttgeringH.J.A.,ZirmA.W.,2009,MNRAS,395,114
Ivison R.J., et al., 2008, MNRAS,390, 1117
IvisonR.J.,PapadopoulosP.,SmailI.,GreveT.,Thomson
A., Xilouris E., Chapman S.,2011, MNRAS,412, 1913
Ivison R.J., et al., 2012, MNRAS,425, 1320
KlamerI.J.,EkersR.D.,SadlerE.M.,WeissA.,Hunstead
R.W., De Breuck C., 2005, ApJL, 621, L1
Kodama T., et al., 2007, MNRAS,377, 1717
Krips M., NeriR., Cox P., 2012, ApJ,753, 135
Kuiper E., et al., 2011, MNRAS,415, 2245
Kurk J. D., Pentericci L., Overzier R. A., Ro¨ttgering
H.J. A.,Miley G. K., 2004, A&A,428, 817
MiddelbergE.,SaultR.,KestevenM.,2006,PASA,23,147
Mihos J. C., Hernquist L., 1994, ApJL,431, L9
Miley G., De Breuck C., 2008, A&ARv,15, 67
Miley G. K., et al., 2006, ApJL, 650, L29
NesvadbaN. P.H., Lehnert M. D., Eisenhauer F., Gilbert
A., Tecza M., AbuterR., 2006, ApJ,650, 693
NesvadbaN. P. H.,et al., 2009, MNRAS,395, L16
OgleP.,DaviesJ.E.,AppletonP.N.,BertincourtB.,Sey-
mourN., Helou G., 2012, ApJ, 751, 13
PapadopoulosP.,IvisonR.,CarilliC.,LewisG.,2001,Na-
ture,409, 58
Papadopoulos P. P., Ivison R.J., 2002, ApJL,564, L9
PapadopoulosP.,Ro¨ttgeringH,vanderWerfP,Guilloteau,
OmontA,vanBreugelW,TilanusR,2000,ApJ,528,626
PentericciL.,KurkJ.D.,CarilliC.L.,HarrisD.E.,Miley
G. K., Ro¨ttgering H. J. A., 2002, A&A,396, 109
Pentericci L., et al., 2000, A&A,361, L25
Pentericci L., Roettgering H. J. A., Miley G. K., Carilli
C. L., McCarthy P., 1997, A&A,326, 580
Riechers D. A., et al., 2011, ApJL,739, L32
Roettgering H. J. A., Lacy M., Miley G. K., Chambers
K.C., SaundersR., 1994, A&AS,108, 79
Salom´e P., Combes F., 2003, A&A,412, 657
Salom´e P., et al., 2006, A&A,454, 437
Salom´e P., Gu´elin M., Downes D., Cox P., Guilloteau S.,
Omont A., Gavazzi R., Neri R.,2012, A&A,545, A57
Scoville N. Z., Yun M. S., Windhorst R. A., Keel W. C.,
ArmusL., 1997, ApJL,485, L21
SeymourN., et al., 2007, 171, 353
SeymourN., et al., 2012, ApJ, 755, 146
SolomonP.M.,VandenBoutP.A.,2005,ARA&A,43,677
StevensJ. A.,et al., 2003, Nature, 425, 264
Swinbank A. M., et al., 2006, MNRAS,371, 465
Tacconi L. J., et al., 2008, ApJ,680, 246
VenemansB. P., et al., 2007, A&A,461, 823
Wang R., et al., 2010, ApJ,714, 699
Zirm A. W., et al., 2008, ApJ, 680, 224
(cid:13)c 2010RAS,MNRAS000,1–??