Discovery of a Small Central Disk of CO and HI in the Merger Remnant NGC 34

0.32 MB·

by  Ximena Fernández

Draftversion January10,2014
PreprinttypesetusingLATEXstyleemulateapjv.5/2/11
DISCOVERY OF A SMALL CENTRAL DISK OF CO AND H IIN THE MERGER REMNANT NGC 34
Ximena Ferna´ndez1, A.O. Petric2, Franc¸ois Schweizer3, J.H. van Gorkom1
Draft version January 10, 2014
ABSTRACT
We present CO(1-0) and H I(21-cm) observations of the central region of the wet merger remnant
NGC 34. The Combined Array for Research in Millimeter-wave Astronomy (CARMA) observations
4 detect a regularly rotating disk in CO with a diameter of 2.1 kpc and a total molecular hydrogen
1 mass of (2.1±0.2)×109 M⊙. The rotation curve of this gas disk rises steeply, reaching maximum
0 velocitiesat1′′ (410pc)fromthe center. Interestingly,HIobservationsdonewiththeKarlG.Jansky
2 VeryLargeArrayshowthatthe absorptionagainstthe centralcontinuumhastheexactsamevelocity
n range as the CO in emission. This strongly suggests that the absorbing H I also lies within 1′′ from
a the center, is mixed in and corotateswith the molecular gas. A comparisonof H Iabsorptionprofiles
J taken at different resolutions (5′′ − 45′′) shows that the spectra at lower resolutions are less deep
8 at the systemic velocity. This provides evidence for H I emission in the larger beams, covering the
regionfrom1 kpc to 9 kpc fromthe center. The centralrapidlyrotatingdisk waslikelyformedeither

] during the merger or from fall-back material. Lastly, the radio continuum flux of the central source
A at mm wavelengths (5.4±1.8 mJy) is significantly higher than expected from an extrapolation of
G the synchrotron spectrum, indicating the contribution of thermal free-free emission from the central
starburst.
.
h Subject headings: galaxies: evolution—galaxies: individual(NGC34,NGC17,Mrk938)—galaxies:
p interactions—galaxies: Seyfert—galaxies: starburst—radiocontinuum: galaxies
-
o
tr 1. INTRODUCTION Soifer et al.(1987)). Ingeneral,highIRluminosities are
s duetodustgrainsbeingheatedbyacombinationofstar-
Gas-rich mergers are key to understanding the hierar-
a
burst and AGN, with higher values indicating a greater
[ chical growth and evolution of galaxies (White & Rees
1978). Much progress has been made to understand AGN contribution (Sanders & Mirabel 1996). The IR
1 the overallrole ofmergersin the formationofearly-type luminosity of NGC 34 suggests that the starburst is the
v dominantphenomenonin the center, with a minor AGN
galaxies,anditisnowacceptedthatmanyareaproduct
1 component (Gon¸calves et al. 1999; Prouton et al. 2004).
ofgas-richmergers(e.g.,Schweizer1998). Therearestill
2 This has been confirmed by a detailed study of NGC
some uncertainties on how exactly the transformation
8 34’s IR spectrum done by Esquej et al. (2012), where
takes place, especially how a gas-rich system ultimately
1 the AGN bolometric contribution to the total infrared
rids itself of gas. These details are what possibly dic-
. luminosity is estimated to be 2+2%. In addition, the IR
1 tate whether the merger remnant becomes an elliptical −1
0 orwhetheranewdiskforms,asseeninsomeobservations images show that the starburst takes place in the inner
4 and simulations of gas rich mergers (Schiminovich et al. 0.5−2 kpc of the merger remnant. These authors find,
1 2013;Barnes2002). Thepossiblepresenceofastarburst however, that the presence of an AGN is needed to ex-
: plain the hard X-ray luminosity.
v or AGN further complicates the picture. While gas can
An optical study by SS07 showed that NGC 34 is the
i fuel both starbursts and AGNs, these phenomena may
X cause outflows and remove the gas from the inner parts. likely product of an unequal-mass merger, with a rich
system of young globular clusters. Deep optical images
r It is thus important to study the gas in the central re-
a gions of mergers in its molecular and atomic phases to show two long tidal tails and structures such as ripples,
jets,andshells,alltypicalofmergerremnants(Schweizer
help predict the ultimate fate of the system.
1998). Inadditiontothis, the surface-brightnessprofiles
NGC 34 (= NGC 17 = Mrk 938) is an ideal candi-
in B, V, and I indicate the presence of a blue expo-
date to study these processes since it is gas-rich and
nential stellar disk with a scale length of a = 3.3 kpc,
hosts both a starburst and an AGN. Throughout the
possibly formed from gas falling back into the merger
paper, we adopt a distance of 85.2 Mpc, derived for
H = 70 km s−1 Mpc−1 by Schweizer & Seitzer (2007, remnant. The optical spectrum features a blueshifted
0 Na I D doublet in absorption, indicative of a mean gas
hereafter SS07). The infrared luminosity of this ob-
outflow velocity of −620±60 km s−1 and a maximum
ject is log(LIR/L⊙) = 11.61 (Chini et al. 1992), plac- velocity of −1050±30 km s−1.
ing it in mid-range of the classically defined Luminous
Infrared Galaxies (LIRGs; 11.0 < logLIR/L⊙ < 12.0; Earlier, we performed a first imaging study of the H I
andradiocontinuumtobetterunderstandthefateofcold
gas during the merger (Ferna´ndez et al. 2010, hereafter
1DepartmentofAstronomy,ColumbiaUniversity,NewYork, Radio I). We detected 7.2×109 M⊙ of H Igas, which is
NY2D10e0p2a7r,tmUeSnAt;oxfimAesntrao@naosmtryo,.cCoalulimfobrniaia.edInustitute of Technol- mostly distributed along the tails, and saw evidence for
gassettlingontotheblueexponentialdiskfoundbySS07.
ogy,Pasadena, CA91125, USA
3Carnegie Observatories, 813 Santa Barbara Street, Themostpuzzlingpartofourstudywasthedetectionof
Pasadena, CA91101, USA
2
astrongHIabsorptionfeatureof∼500kms−1 widthat
the systemic velocity,coveringboth blueshifted and red- Table 1
ObservationParameters
shifted velocities and seen against the strong unresolved
centralcontinuumsourceof62.4mJy. Wepresentedtwo CO(1–0)1 HI2
possiblescenariostoexplainthe puzzlingabsorption: ei-
ther we were probing an H I disk in absorption against FluxCalibrator Uranus 3C48
PhaseCalibrator 3C454.3 J2357-1125
an extended continuum source so that we saw both the
NumberofChannels 150 256
blueandredshiftedpartinabsorption,orwewereseeing Bandwidth(MHz) 1000 32
gas associated with the tidal tails in projection against Synthesized Beam 2.48′′×2.14′′ 8.31′′×4.46′′
the continuum. FieldofView 1−1.7′ 32′
VelocityCoverage 3000 7000
Inthisfollow-upradiostudy,wepresentnewCOimag-
(kms−1)
ing and H I absorption observations obtained at higher VelocityResolution 20 27
angular resolution and with a wider velocity coverage. (kms−1)
We seek to determine which of the above two scenarios Noise 8.4 0.21
explainsthepuzzlingabsorptionbetterbycomparingthe (mJybeam−1chan−1)
distributionandkinematicsoftheCOtothe HIabsorp- Colum(ncmde−n2sicthyasne−n1si)tivity3 1.1×1020(H2) 1.7×1020
tion data. In addition, the new continuum observations
at mm-wavelengths allow us to probe the spectral index 1 Observedin May 2011
2 Observedin April 2011
at high frequencies. Lastly, both sets of new observa-
3 In emission
tions have a much wider velocity coverage, enabling us
2.2. VLA Observations
to search for an atomic or molecular counterpart to the
outflow seen optically. NGC 34 was observed with the Karl G. Jansky Very
Thestructureofthepaperisasfollows. Wesummarize LargeArray(VLA)4 inApril2011inthe B arrayconfig-
theobservationsanddatareductioninSection2,present uration(spacingsof0.21−11.1km)duringtworunsfora
our results in Section 3, discuss what we have learned totalobservingtime of7 hours. The observationsuseda
from these observations in Section 4, and put forth our 32 MHz bandwidth centered at the heliocentric velocity
conclusions in Section 5. of5870kms−1,with atotalof256channelswithareso-
lution of 27 km s−1. This results in a velocity coverage
of 7000 km s−1, from 2426 km s−1 to 9421 km s−1.
2. OBSERVATIONSANDDATAREDUCTION
Each run was reduced with the Astronomical Imag-
2.1. CARMA Observations ingProcessingSystem(AIPS)usingstandardcalibration
NGC 34 was observed in the CO J =1−0 transition procedures. Thedatawerecombinedintheuv plane,and
at 115.2712GHz with the Combined Arrayfor Research then severaliterations of self-calibrationwereperformed
in Millimeter-wave Astronomy (CARMA) in May 2011. tocorrectforamplitude andphaseerrors. Wefirstmade
The observations were done in two configurations: in D a continuum image by averaging 80 line-free channels,
arrayfor4.1hoursonsourceandinCarrayfor5.8hours. using a robustness parameter of 1 and setting CLEAN
The spacings for these two arrays are 11 − 150m and boxes around the point source. This image served as
30−350m,respectively. Theprimaryfluxcalibratorwas the initial input model for the first run of phase self-
Uranus, and 3C454.3 was used as the passband calibra- calibration. After eachrun, we made a new input model
tor. Both sets of observations used a 1 GHz bandwidth by imaging the continuum applying the new solutions
centeredat113.11GHz,correspondingtothecentralfre- for subsequent self-calibration runs. After three runs of
quencyofthesingle-dishobservations. Thistranslatesto phase self-calibrationand one of amplitude, we achieved
a heliocentric velocity of 5731 km s−1, using the optical a dynamic range of about 760:1 and an rms sensitiv-
definition. This setup results in a velocity coverage of ity of 0.07 mJy beam−1. We then applied the solutions
2980kms−1,startingat4238kms−1 andendingat7218 to the whole data set, and subtracted the continuum in
km s−1, with 150 channels with a velocity resolution of the uv plane by making a linear fit through the line-free
20 km s−1. channels. The HIcube wasmadeusing arobustnesspa-
ThedatawerereducedwiththesoftwarepackageMul- rameter of 1, and cleaning the channels containing H I
tichannelImageReconstructionImageAnalysisandDis- absorption. The resulting images have an rms noise of
play (MIRIAD) using standard calibration procedures 0.2mJybeam−1 andasynthesizedbeamof8.31′′×4.46′′
foreachconfiguration(Sault et al.1995). Foreachband, (∼3.5×2 kpc).
we flagged the first and last 2 channels due to poor sen- Table 1 summarizes all parameters used for the
sitivity at the edge of the band. The phase stability CARMA and VLA observations.
wasinspectedandpoordatawereflaggedforeachtrack.
The CandDarraycalibratedvisibilities werethencom- 3. RESULTS
bined and the continuum was estimated from 696 (un- 3.1. A Central Disk of CO
smoothed) signal free channels. The deconvolution task
The CO observations reveal a disk of molecular gas in
uses a Steer CLEAN algorithm (Steer et al. 1984), and
thecentralregionsoftheremnant. Figure1iscomposed
we selected natural weighting (robustness parameter set
of various panels showing the gas distribution and kine-
to 2) to maximize the sensitivity to broad, faint struc-
matics. ThisCOdiskliesatthecenterofNGC34andis
tures. The final combined data cube has a synthesized
beam of 2.48′′×2.14′′ (∼ 1 kpc) at FWHM and a noise 4 The National Radio Astronomy Observatory is a facility of
of 8.4 mJy beam−1. theNationalScienceFoundationoperatedundercooperativeagree-
mentbyAssociatedUniversities,Inc.
3
(b) (b)
(c)
(a)
m
a
e
B
y/
J
Velocity (km/s)
(d) (e)
Figure 1. MapsofthedistributionandkinematicsoftheCOdiskfoundinNGC34. (a)COemissionascomparedtotheopticalandHI
emission;thecyancontours showtheCOemission,andtheHIcontours(inblue;fromRadioI)aredrawninlevelsof(8,28,48,68,108)
×1019 cm−2 overlaid on an optical image from SS07. (b) CO distribution map (moment 0) contours drawn starting at 5% of the peak,
inintervalsof10%. (c)Thespectrum oftheCOemissiondonebysettingaboxaroundtheemissionshowinghowmuchCOthereisata
givenvelocitybin. (d)Velocitymap(moment 1)overlaidwithisovelocitycontours drawninintervalsof50kms−1. (e) Position–velocity
diagram along the major axis of the optical disk(PA −9◦; SS07). Note that the fullrange of velocities arenot shown inthese maps, we
limitthem toshowabettercontrastortozoom-ininaregionofinterest.
muchsmallerthanthe opticaldisk (Figure1a). Itshows spacings could have detected this more extended gas.
a regular rotation pattern as seen in Figures 1d and 1e. Asseeninthe spectrumofFigure1c,theCO emission
There are hints of extended emission towards the north- spanscloseto500kms−1. Theintensity-weightedveloc-
west, northeast, and south (Figure 1b). We estimate a ity field (Figure 1d) and the position-velocity (PV) dia-
diameterof2.1kpc bymeasuringthe extentofthe emis- gram(Figure1e)alongthemajoraxisoftheopticaldisk
sionenclosedbythesecondlowestdensitycontour,which (PA −9◦; SS07) show that the kinematics are consistent
corresponds to 4.31 Jy km s−1 beam−1. with rotation, with the north side of the disk receding,
We calculate the CO luminosity with the following and the south side approaching. The extended emission
equation (Solomon et al. 1997): features mentionedabove havevelocities consistentwith
the gas in the main body of the disk, suggesting they
LCO =3.25×107 Sdv νo−b2sDL2(1+z)−3, (1) are real and possibly part of a fainter component not
detected in our observations. The PV diagram demon-
where Sdv is the integrated CO flux in units of Jy km stratesthattherotationcurverisesverysteeply,reaching
s−1, νobs is the observed frequency in GHz, and DL is peak velocities within 1′′ fromthe center, andshowinga
theluminositydistanceinMpc. WecomputetheCOflux slight decrease towards the edges.
usingchannelmapsandgetavalueof151.9±11.3Jykm
s−1,whichresultsinaCOluminosityof(2.6±0.2)×109 3.2. High Resolution H I Absorption
K km s−1 pc−2. We multiply this value by αCO to get Our new VLA B-array observations probe the previ-
a molecular hydrogen mass in solar masses (M⊙). Here ously known H I absorption feature (Radio I) with a
we use α = 0.8, which is the standard conversion factor wider velocity coverage and at higher resolution. Figure
forstarburstingsystems(Downes & Solomon1998),and 2consistsofthree panelsshowingtheseobservations,in-
get an H2 mass of 2.1±0.2×109 M⊙. cludingacomparisontotheabsorptionprofilespresented
Single-dish CO(1–0) data (Kandalyan 2003; in Radio I, and to the CO observations in emission pre-
Chini et al. 1992; Kruegel et al. 1990) yield a flux sented in Section 3.1. The top panel represents the H I
of 170 ± 12 Jy km s−1, which is 12% higher than absorption feature seen against the peak of the contin-
what we detect. This is a marginal difference, but—if uum emission and showing the full velocity coverage of
significant—it might be indicating that our observations 2426–9421kms−1,withnohintofabsorptionseenatthe
miss an extended component of the CO disk. This outflow velocity (∼5000 km s−1) detected by SS07 (see
component is hinted at by the weak extensions seen next section).
in Figure 1b, suggesting that observing with shorter The second panel compares these higher resolution (2
4
0.0010 sourceplusemissionfurtheroutwithinthecentralbeam
in C and D array. We can estimate how much emission
0.0005
there is on different scales by looking at the difference
0.0000 in the absorption profiles. We do this by summing the
differencebetweenthetwoprofilesatdifferentresolutions
−0.0005
for each velocity bin to get a flux. As shown in Radio I,
Jy/beam−−00..00001105 t6hearnedis178.3k×pc10o8fMth⊙eorfeHmnIainnte.miNssoiwonwbeetcwaenendtehteerimnnineer
how much H I in emission there is between the inner 2
−0.0020 and 6 kpc by comparing the CnB and the B array data.
−0.0025 We calculateanHImassof2.5×108 M⊙ inthe velocity
interval5700−6000kms−1. Sincethetwodatasetshave
−0.0030 different velocity resolutions, we created velocity bins of
−0.0035 50 km s−1 and then averaged the values that were in a
3000 4000 5000 6000 7000 8000 9000
Velocity (km/s) given bin.
0.0010 Thebottompanelcomparestherangeofvelocitiesseen
in the H I absorption and the CO emission. The plot
0.0005
shows that the velocity ranges for the emission and the
0.0000 absorption are almost identical. The range covered for
the H Iabsorptionprofile is 466±27 km s−1 and for the
−0.0005
COprofileis460±20kms−1,asmeasuredatthe profile
m
a−0.0010 base. Thisremarkableagreementinvelocityrangemakes
e
Jy/b−0.0015 istampleaucseibntleratlhagtasthdeisHk,IiamnadgCedOienmCisOsioantsa2r.4is8e′′fr×om2.1th4′e′
−0.0020 resolution (Figs 1b and 1d). In addition, it implies that
−0.0025 the H Iis intermixed with the molecular gas and that it
B also reaches maximum velocities within ∼ 1′′ (0.4 kpc)
−0.0030 C from the center.
D
−0.0035
5200 5400 5600 5800 6000 6200 6400
Velocity (km/s) 3.3. Molecular and Atomic Gas Outflow
We can use the current observations to search for a
0.10 0.003 counterpart in H I and/or CO(1–0) to the outflow de-
tected by SS07 in the NaD I doublet. We do not detect
0.002
theoutflowineitherline,butcanplaceupperlimits. We
0.05
calculate the noise near the central regions of NGC 34
0.001 andaround∼5000kms−1,consistentwiththe velocities
m
a of the optical outflow. We use the following equation to
be 0.00 0.000
y/ getamassoutflowrateupperlimit(Heckman et al.2000;
J
−0.001 Rupke et al.2002)forbothHIabsorptionandemission:
−0.05
−0.002 M˙ (H)=21 Ω C r∗ N(H)
CO (cid:18)4π(cid:19) f(cid:18)1 kpc(cid:19)(cid:18)1021 cm−2(cid:19)
−0.10 HI −0.003
∆v
5200 5400 5600Veloc5it8y0 (0km/s)6000 6200 6400 ×(cid:18)200 km s−1(cid:19)M⊙ yr−1, (2)
Figure 2. HIabsorptionathigherresolutionandwidervelocity
where Ω is the solid angle subtended by the outflow, C
coverage. Top: B-arrayobservationsshowingtheHIabsorptionat f
bothblueshiftedandredshiftedvelocities. Middle: Comparisonof is the line-of-sight covering fraction, r∗ is the size of the
theabsorptionprofileobservedatthreedifferentresolutions(DnC: region from where the outflow is produced, N(H) is the
18kpc,CnB:6kpc,andB:2kpc). Bottom: ComparisonoftheHI columndensity,and∆visthevelocityoftheoutflow. For
absorptionprofilewiththeCOemissionspectrum. They-axesare
the first three quantities, we assume Ω = 4π, C = 0.5,
different for the two, but are shown in the same plot to compare f
thevelocities. andr∗=1kpc,whicharevaluesconsistentwithoutflows
seen in LIRGs (Rupke et al. 2002). For the absorption,
we calculate the H I column density assuming 5σ and
kpc) data to the previous H I observations in the DnC get 4.2×1020 cm−2 (assuming Ts = 100 K). Lastly, we
(18 kpc) and CnB (6 kpc) configurations. Note that the take∆v =620kms−1,whichisthemeanvelocityofthe
depthofabsorptionis similarinthe three configurations outflow detected by SS07. The resulting upper limit for
at the extreme velocities around 5600 km s−1 and 6000 themassoutflowrateis14M⊙yr−1inabsorption. Using
km s−1, but is very different near the systemic velocity thesamevalues,butnowcalculatingN(HI)inemission,
of5870kms−1. Adeepfeatureis seeninBarrayatthis theupperlimitisabout130M⊙yr−1. Thislargenumber
velocity, which is shallower in C and nearly absent at isaconsequenceofthepoorsurfacebrightnesssensitivity
some velocities in D. We can understand this as follows: in B array.
Since the continuum source is unresolved in B array, we ¿From the CO emission, we can calculate an upper
see a combination of absorption against the continuum limitforthe outflowratebydividing themolecularmass
5
4. DISCUSSION
4.1. The Gaseous Disk in the Inner Regions of the
Remnant
OurCO observationsconfirmthe presenceof a disk in
the inner regions of the remnant, which was one of the
twoscenariossuggestedinRadioI to explainthe HIab-
y) sorption observations. The match in velocities for both
mJ
sets of observations (Fig. 2) strongly suggests that the
Flux (101 central H I absorption is due to the presence of a rotat-
ing disk. As shown in the PV diagram of Figure 1e, the
maximumvelocitiesarereachedwithinonly1′′ (0.4kpc)
fromthecenter. Theclosematchinvelocitybetweenthe
moleculargasandHIabsorptionarguesthattheabsorb-
ing gas also reaches maximum velocities within 0.4 kpc
fromthecenter. Thisisconsistentwithhigher-resolution
continuum images by Condon et al. (1991) that resolve
100 101 102 the central source and show it to be a dominant central
Frequency (GHz)
component with an extension to the south, with a total
Figure 3. Flux measurements at different frequencies, with the size of 2.5′′ (1 kpc). This extended continuum makes it
firstfourmeasurements taken fromClemensetal.(2008)and the
measurementat113GHzshowingthenewobservationspresented possible to trace out the H Iabsorptionto its maximum
here. Thelinemarkstheleast-squaresfittothefirstfourmeasure- velocity at 1′′ from the center.
ments (slope of −0.8), but is extended to demonstrate that the The question now is whether any of the continuum
spectralindexchanges athigherfrequencies. emission could be due to the AGN. Several recent stud-
ies have calculated the AGN contribution to be small
by the dynamical timescale (e.g., Alatalo et al. 2011). (1−10%)inNGC34(e.g.,Vega et al.2008;Esquej et al.
We compute an upper limit for the molecular-gas mass 2012; Murphy 2013). In addition to this, VLBI observa-
using5σandthevelocityoftheopticaloutflow,resulting tions report a single-baseline detection (Lonsdale et al.
in 6.2×107 M⊙. We can estimate the dynamical time 1993),whichsuggeststhatthismergerremnantdoesnot
by assuming a size of 1 kpc for the wind emitting region host a strong point source due to the AGN, and its con-
and the optical velocity, which results in a timescale of tinuum is likely dominated by the starburst.
only 1.6 Myr. These two values yield an upper limit for
the CO mass outflow rate of about 40 M⊙ yr−1. 4.2. The Fate of Gas in the Merger
The optical study by SS07 and the radio observations
3.4. Central continuum
presentedhere and in RadioI haveshowndifferent com-
This study includes new continuum images ofthe cen- ponents of the new disk formed after the merger. SS07
tral components of the merger remnant. The VLA B- find a blue stellar disk, with spiral structure extending
array observations show a slightly resolved point source out to a radius of ∼3.3 kpc, that seems to have formed
with a flux of 60.6±2.0 mJy, consistent with the mea- ∼400Myragoandisembeddedinaredspheroid. InRa-
surement presented in Radio I. dio I, we found that the H Ifrom the northern tidal tail
The CARMA observations reveal a faint central con- isfallingbackandcontinuingtofeedtheouterregionsof
tinuum source at millimeter wavelengths with a flux of this optical disk, favoring the idea of inside-out growth.
5.4±1.8 mJy. This measurement allows us to probe the Onsmallerscales,wehaveshowninthisstudythatthere
spectral index at high frequencies. At low frequencies is a central disk of CO and H I of about 2.1 kpc in di-
(1−10 GHz), the spectrum for star-forming galaxies is ameter. In addition, we find evidence for more extended
usuallydominatedbysynchrotronradiation,withtypical H Iemission further out as shown by the comparison of
valuesofα=−0.8. Athigherfrequencies(10−200GHz), the absorption profiles at different resolutions.
free-freeemissionstartsto dominateandthe spectralin- Observations and simulations of mergers suggest that
dexis α=−0.1. Dustemissiontakesoveratfrequencies about 50% of the gas gets quickly moved to the cen-
above 200 MHz with α=1.5 (Condon 1992). tral regions, losing its angular momentum due to cloud-
Figure3showstheradiospectrumforNGC34between cloud collisions, while the other 50% gets moved to
1.4 GHz and 113.11 GHz. The first four values in the large distances from the center, but remains bound
rangeof 1.4−22.5 GHz were reportedby Clemens et al. and will eventually fall back (Hibbard & Mihos 1995;
(2008), and the one at 113.11 GHz is the measurement Hibbard & van Gorkom1996; Barnes2002). In NGC 34
presented here. We fit a line through the first four mea- we see evidence for both processes, and at the present
surements and extend it to 113.11 GHz to better show time there are comparable amounts of cold gas in the
that the spectrum flattens at higher frequencies. The fit tidal tails and the center. We know the blue stellar disk
ofthelineyieldsaslopeof−0.8,indicatingthattheemis- formed∼400Myragofromgassettlingtowardthecenter
sionatthese frequenciesis dueto synchrotronradiation. during the merger. The central gaseous disk presented
The spectral index between 22.5 GHz and 113.11 GHz here could have formed during the merger or more re-
is α = −0.20±0.07. The flattening of the spectrum at cently from fall-back material. The major axes of the
higherfrequenciesshowsthatsynchrotronemissionisno gaseous disk and the blue stellar disk have almost iden-
longerdominant, andfree-free emissionfromHII regions ticalpositionangles,indicating thatthe twodisks either
takes over. have the same formation mechanism or are dynamically
6
linked. ThecurrentSFRintheinner2kpccalculatedby states of California, Illinois, and Maryland, and the Na-
Esquej et al. (2012) from the 24µm luminosity is 42 M⊙ tional Science Foundation. Ongoing CARMA develop-
yr−1, which implies a consumption timescale of ∼ 50 ment and operations are supported by the National Sci-
Myr for the molecular hydrogen presented here. This ence Foundation under a cooperative agreement, and by
indicatesthatthe gasinthe centraldiskmustbe contin- the CARMA partner universities. This work was par-
uously replenished to sustain the current star formation tiallysupportedbyNSFgrantNo. 1009476toColumbia
rate. University.
Lastly, both sets of observations have enough velocity
coveragetopermitsearchingforapossibleradiocounter-
part to the outflow seen optically in Na D I absorption REFERENCES
by SS07. However, we do not find any evidence for such
acounterpartintheatomic-ormolecular-gasphase. We
are able to place approximate upper limits on the HI
mass outflow rate in both emission and absorption, and Alatalo,K.,etal.2011,ApJ,735,88
for the CO in emission. Barnes,J.E.2002, MNRAS,333,481
Chini,R.,Kruegel,E.,&Steppe, H.1992,A&A,255,87
5. SUMMARYANDCONCLUSIONS Clemens,M.S.,Vega,O.,Bressan,A.,Granato,G.L.,Silva,L.,
&Panuzzo,P.2008,A&A,477,95
Wehavepresentednewhigh-resolutionHIandCOob- Condon,J.J.1992,ARA&A,30,575
servationstostudythecentralregionsofthemergerrem- Condon,J.J.,Huang,Z.-P.,Yin,Q.F.,&Thuan,T.X.1991,
nant NGC 34. Our main findings are as follows: ApJ,378,65
Downes,D.,&Solomon,P.M.1998,ApJ,507,615
• WedetectarotatingCOdiskof2.1kpcindiameter Esquej,P.,etal.2012,MNRAS,423,185
with a molecular-hydrogen mass of (2.1 ± 0.2)× Fern´andez,X.,vanGorkom,J.H.,Schweizer,F.,&Barnes,J.E.
2010,AJ,140,1965
109 M⊙. Gon¸calves,A.C.,V´eron-Cetty,M.-P.,&V´eron,P.1999,A&AS,
135,437
• The velocity width of the CO(1–0) emitting gas Heckman,T.M.,Lehnert, M.D.,Strickland,D.K.,&Armus,L.
matches the H I absorption width, indicating that 2000,ApJS,129,493
Hibbard,J.E.,&Mihos,J.C.1995,AJ,110,140
the broad H I absorption is due to the central gas
Hibbard,J.E.,&vanGorkom,J.H.1996, AJ,111,655
disk and is co-spatial with the CO.
Kandalyan,R.A.2003,A&A,398,493
Kruegel,E.,Chini,R.,&Steppe, H.1990,A&A,229,17
• ThenewB-arrayobservationsoftheHIabsorption Lonsdale,C.J.,Smith,H.J.,&Lonsdale,C.J.1993,ApJ,405,
feature allow us to calculate that there is 2.5 × L9
108 M⊙ of H I in the annular region of 1−3 kpc Murphy,E.J.2013,ArXive-prints
Prouton,O.R.,Bressan,A.,Clemens,M.,Franceschini,A.,
from the center.
Granato, G.L.,&Silva,L.2004,A&A,421,115
• We do not detect a molecular- or atomic-gas out- Rupke,D.S.,Veilleux,S.,&Sanders,D.B.2002,ApJ,570,588
Sanders,D.B.,&Mirabel,I.F.1996, ARA&A,34,749
flow, but place upper limits on the mass outflow Sault,R.J.,Teuben, P.J.,&Wright,M.C.H.1995,in
rates for H I in emission and absorption, and for AstronomicalSocietyofthePacificConferenceSeries,Vol.77,
CO in emission. AstronomicalDataAnalysisSoftwareandSystemsIV,ed.
R.A.Shaw,H.E.Payne,&J.J.E.Hayes,433
• We have obtained new continuum images of the Schiminovich,D.,vanGorkom,J.H.,&vanderHulst,J.M.
2013,AJ,145,34
central source at 1.4 GHz and 113 GHz. We find
Schweizer,F.1998,inGalaxies:Interactions andInducedStar
that the spectrum flattens at the highest frequen-
Formation,ed.D.Friedli,L.Martinet,&D.Pfenniger(Berlin:
cies to a spectral index of α = −0.20±0.07. This Springer),105
indicates that the radio continuum at those fre- Schweizer,F.,&Seitzer,P.2007,AJ,133,2132
quencies is due to free–free emission from the cen- Soifer,B.T.,Sanders,D.B.,Madore,B.F.,Neugebauer,G.,
Danielson,G.E.,Elias,J.H.,Lonsdale,C.J.,&Rice,W.L.
tral starburst.
1987,ApJ,320,238
Solomon,P.M.,Downes,D.,Radford,S.J.E.,&Barrett,J.W.
1997,ApJ,478,144
Steer,D.G.,Dewdney,P.E.,&Ito, M.R.1984, A&A,137,159
Support for CARMA construction was derived from
Vega,O.,Clemens,M.S.,Bressan,A.,Granato,G.L.,Silva,L.,
the Gordon and Betty Moore Foundation, the Kenneth &Panuzzo,P.2008,A&A,484,631
T. and Eileen L. Norris Foundation, the James S. Mc- White,S.D.M.,&Rees,M.J.1978,MNRAS,183,341
Donnell Foundation, the Associates of the California In-
stitute of Technology, the University of Chicago, the