Misalignment between cold gas and stellar components in early-type galaxies

1.2 MB·English

by  O. Ivy Wong

Mon.Not.R.Astron.Soc.000,1–??(2011) Printed2February2015 (MNLaTEXstylefilev2.2)
Misalignment between cold gas and stellar components in
early-type galaxies
5
1 O. Ivy Wong,1⋆ K. Schawinski,2 G.I.G. J´ozsa,3,4,5 C.M. Urry,6,7 C.J. Lintott,8
0
2 B.D. Simmons,8 S. Kaviraj,8,9 and K.L. Masters10,11

n 1 International Centre forRadio Astronomy Research, The Universityof WesternAustralia M468, 35 Stirling Highway, Crawley,
a WA 6009, Australia
J 2 Institute for Astronomy, ETH Zrich, Wolfgang-Pauli-Strasse 27, 8093 Zu¨rich, Switzerland
0 3 SKA South Africa, Radio Astronomy Research Group, 3rd Floor, The Park, Park Road, Pinelands, 7405, South Africa
3 4 Rhodes University,Department of Physics and Electronics, Rhodes Centre for Radio Astronomy Techniques & Technologies,
P.O. Box 94, Grahamstown 6140, South Africa
] 5 Argelander Institut fu¨r Astronomie (AIfA), University of Bonn, Auf dem Hu¨gel 71, 53121 Bonn, Germany
A 6 Yale Centerfor Astronomy and Astrophysics and Department of Physics, Yale University,P.O. Box 208120, NewHaven,
G CT 06520-8120, USA
7 Department of Astronomy, Yale University,P.O. Box 208101, New Haven, CT 06520-8101, USA
h. 8 Oxford Astrophysics, DenysWilkinson Building, Keble Road, Oxford OX13RH, UK
p 9 Centre for Astrophysics Research, Universityof Hertfordshire, College Lane, Hatfield, Herts, AL10 9AB, UK
- 10 Institute of Cosmology & Gravitation, Universityof Portsmouth, DennisSciama Building, Portsmouth, PO1 3FX, UK
o 11 South East Physics Network (SEPNet), www.sepnet.ac.uk
r
t
s
a
Released2011XxxxxXX
[

1
v ABSTRACT
3 Recent work suggests blue ellipticals form in mergers and migrate quickly from the
5 blue cloud of star-forming galaxies to the red sequence of passively evolving galaxies,
6 perhapsasaresultofblackholefeedback.Suchrapidreddeningofstellarpopulations
7
implies that large gas reservoirsin the pre-mergerstar-formingpair must be depleted
0
onshorttimescales.Herewepresentpilotobservationsofatomichydrogengasinfour
.
1 blue early-type galaxies that reveal increasing spatial offsets between the gas reser-
0 voirs and the stellar components of the galaxies, with advancing post-starburst age.
5 Emission line spectra show associated nuclear activity in two of the merged galax-
1 ies, and in one case radio lobes aligned with the displaced gas reservoir. These early
:
v results suggest that a kinetic process (possibly feedback from black hole activity) is
i driving the quick truncation of star formation in these systems, rather than a simple
X
exhaustion of gas supply.
r
a Key words: galaxies: elliptical and lenticular, cD, galaxies: evolution, galaxies: for-
mation
1 INTRODUCTION slowlyexhaustitsgasreservoirbyformingstarsatadeclin-
ing rate. If instead the gas reservoir fuelling star formation
Star-forming galaxies show a correlation between stellar
isdestroyedeffectively instantaneously,thenstar formation
mass and star formation rate whose normalisation varies
willcease andthegalaxy will reddenrapidlywithin 1Gyr,
with redshift, but whose scatter remains tight out to high
as is observed in early-type galaxies (ETGs; Kaviraj et al.
redshift (Noeske et al. 2007; Peng et al. 2010; Elbaz et al.
2011; Wonget al. 2012; Schawinski et al. 2014). Interest-
2011). This correlation can be explained as a balance be-
ingly, this timescale of 1 Gyr is consistent with that pro-
tween gas inflows from cosmological scales and outflows
posed for the transition of a merger-driven Ultraluminous
drivenbysupernovae(Bouch´eet al.2010;Lilly et al.2013).
Infrared Galaxy (ULIRG)toan ETG(Emonts et al.2006).
Quenchingof star formation causes galaxies to depart from
Also, it has been known for a while that ULIRGs have a
thissteady-state along pathways that dependon theevolu-
high merger fraction (Sanderset al. 1988).
tionofthegassupplyandreservoir(Schawinskiet al.2014):
if thecosmological gas supply toagalaxy is shutoff, it will In addition, negative feedback from a galaxy’s central
active galactic nuclei has often been invoked in galaxy for-
mation models to slow down the star formation history
⋆ E-mail:[email protected] of simulated galaxies (e.g. Croton et al. 2006). However,
2 O. I. Wong et al.
[h]
Figure1.Toprow:opticalgriSDSS(Ahnetal.2014)compositeimagesofthefourblueearly-typegalaxiesinoursample.Thegalaxies
are arranged from left to right in terms of NUV −u colour from blue to red as a tracer of time. The white scale bars represent 10
arcseconds. Bottom row: zoomed-out optical images with the WSRT radio observations overlaid. The white contours (asinh-stretch)
represent the Hi and the green contours (logarithmic stretch) the radio continuum. The lowest radio and Hi contours begin at 3σ and
thehighestcontourismatchedtothepeakfluxdensityof0.87mJybeam−1,1.51mJybeam−1,0.86mJybeam−1,44.79mJybeam−1
forJ1237,J1117,J0900andJ0836;respectively.Itshouldbenotedthatthenuclear1.4GHzcontinuumpointsourcesarenotshownfor
J0900andJ1117toavoidconfusionwiththeHicontours.
observational evidence for such feedback is currently only consisted of an interacting pair of galaxies (Chang et al.
available forafewindividualgalaxies (Harrison et al. 2014; 2001). All subsequent campaigns to detect Hi in post-
Nyland et al. 2013; Alatalo et al. 2011; Hota et al. 2011; starbursts have been optimised towards detecting diffused
Croston et al. 2008; Kharb et al. 2006). With the advent low-surface brightness gas, while compromising on the an-
of very large surveys; much progress has been made to- gularresolutionrequiredtomapthelocationofthegas(e.g.
wards understanding of the co-evolution between galaxies Zwaan et al. 2013).
and their central AGN, in particular, the connection be-
tween classical bulges and central supermassive black holes
Todeterminethephysicalprocess responsible forshut-
(Heckman & Best 2014; Kormendy & Ho 2013).
tingdownstarformation,wepositthatpost-starburstgalax-
Blue ETGs do not fit the canonical bimodal scheme iesaretooevolvedandthatthesmokinggunforquenchingis
of red ellipticals versus blue spirals and are unlikely to be morelikelytobefoundinitspredecessorpopulation,namely,
thedescendantsofbluespirals(Tojeiro et al.2013).Rather, the blue early-type galaxies.
the blue ETGs appear to be transition-type galaxies and
are probable predecessors of local post-starburst galaxies
(Schawinski et al.2009;Wonget al.2012).PreviousCOob- Using the Westerbork Syntheses Radio Telescope, we
servations of blue ETGs show that themolecular gas reser- studytheHicontentofapilotsampleoffourblueearly-type
voirs are being rapidly destroyed duringthe composite star galaxies—theprogenitorstopost-starburstgalaxiesbecause
formation and AGN phase (Schawinskiet al. 2009). Local 1) these galaxies are still currently star-forming and there-
post-starburstgalaxiesaredefinedtobegalaxieswhichhave forethereshouldbegaswheretherearestarsbeingformed;
recentlyceasedformingstars.Thesegalaxiestypicallyshow and 2) the state of the gas morphology and dynamics will
little Hα or [OII] emission (indicative of current star for- shedlightonwhythesegalaxieswillsoonstopformingstars.
mation), but have strong absorption line signatures indica-
tive of a young stellar population. In addition, local post-
Section 2 describes theblueearly-typesample. The pi-
starbursts consists mostly of galaxies which have similar
lot observationsanddataprocessing methodsaredescribed
early-type structural properties to red sequence galaxies of in Section 3. We discuss the radio continuum and Hi imag-
comparable masses (Wonget al. 2012).
ing results in Section 4 and Section 5, respectively. Section
Earlier attempts to map the Hi content of post- 6 provides a summary of our results. The AB magnitude
starburstgalaxiesonlyfoundHiinoneoffivetargetswhich system is used throughout this work.
Misalignment between cold gas and stars in early-type galaxies 3
ur5
d) 33..00 colo
e u 4
ct 22..55 V-
dust corre 22..00 J1J2038736 J0900J1117 ation) NU3 J090J00 8- 3H6I -e HjeIc etejedcted
colour ( 11..55 ar form2 J1117 - HI displaced
u-r 11..00 nt st1 J1237 - HI rotating disk
e
00..55 ec
(r0
99..00 99..55 1100..00 1100..55 1111..00 1111..55 1.0 1.5 2.0 2.5 3.0
Stellar Mass log M (M ) (intermediate age stars) u-r colour
* O •
Figure2.Colour-stellarmassdiagramofETGsfoundfromSDSS Figure 3.TheUV/opticalcolour-colourdiagramsofblueearly-
and Galaxy Zoo. The contours represent the number density of type galaxies (green density contours) and the entire early-
ETGsoccupyingthecolour-stellarmassregionknownasthe‘red typegalaxypopulation(blackshadedcontours).Themodelstel-
sequence’. All four pilot sample galaxies are located in the op- lar population tracks with varying quenching timescales from
tical ‘green valley’ of the colour-mass diagram (as demarcated Schawinskietal.(2014)arealsooverlaid.Thelargeandsmallcir-
between the solidgreen lines). Allgalaxies show the presence of cles alongthe evolutionary tracks represent1Gyr and 100Myr,
intermediateagestellarpopulationsbytheir‘green’u−rcolour. respectively. Two out of four pilot sample galaxies show an ab-
senceofveryyoungstellarpopulationsindicatedbyredNUV−u
colours.ThosethreegalaxiesalsoshoweitheradisplacedHIreser-
2 BLUE EARLY-TYPE GALAXIES voir (J1117+51) or acompletely ejected HIreservoir (J0836+30
andJ0900+46).ThegalaxywhichisbluestinNUV −ufeatures
We obtain the photometric and spectroscopic data from
arotatingHIdisk.
the Sloan Digital Sky Survey (SDSS) DR7 (York et al.
2000; Abazajian et al. 2009) for all objects classified as
‘galaxy’(Strauss et al.2002).Thesampleof204low-redshift havequenchingtimescalesofseveralGyrs(Schawinski et al.
(0.02 < z < 0.05) blue ETGs were identified using 2014). Figure 3 illustrates the different evolutionary stages
the Sloan Digital Sky Survey (Adelman-McCarthy et al. representedbyeach ofthefour ETGs selected for thispilot
2008, SDSS; ) and the Galaxy Zoo project (Lintott et al. study on a UV/optical colour-colour diagram. The colour
2008; Schawinski et al. 2009; Lintott et al. 2011). The blue of model stellar populations along different star formation
ETGs account for 5.7% of all ETGs found within the quenching tracks are shown as solid lines where each large
same redshift range and are the most actively star- circle marks 1 Gyr and intervals of 100 Myr being shown
forming population of ETGs (Schawinskiet al. 2009). Fur- bysmallcirclesalongthestellarevolutiontracks.Theblack
ther description of the blue early-type galaxy selec- shaded contours represent the low-redshift early-type pop-
tion can be found in Schawinskiet al. (2009) and at ulation and the green contour represents the population
http://data.galaxyzoo.org. of blue early-type galaxies selected from Schawinski et al.
(2009). The observed stellar populations of J0836+30 and
J0900+46 appear to be relatively evolved along the fastest
2.1 The pilot sample
quenchingevolutionarypathwaythatoccursontimescalesof
To investigate the fate of the gas reservoir when star for- several hundred Myr to 1 Gyr (Schawinski et al. 2014). On
mation is quenched, we select four ETGs with compa- the other hand, the stellar population colours of J1117+51
rable stellar masses and UV/optical colours indicative of and J1237+39 are less evolved and consistent with earlier
rapid quenching: all four lie in the optical u−r green val- stages ofquenching.Wenotethat J1117+51 andJ1237+39
ley, indicating they have significant intermediate-age stel- appear to favour slower quenching pathways in Figure 3.
lar populations, and their NUV − u UV/optical colours However, this could be dueto differences between the start
range from very blue, consistent with ongoing star forma- timesforquenchinginthesegalaxies relativetothefiducial
tion (J1237+39), tothereddercoloursof passive,quenched starting point for quenching in the models. The Hi proper-
galaxies (J0836+30)—see Figure 2. Due to the diffused na- tiesofeachsampledgalaxyappearstobeconsistentwithits
ture of Hi gas, we have also selected the nearest galaxies respectiveevolutionarystage.SeeSection5formoredetails
which can be observed by the Westerbork Synthesis Radio of theobserved Hi properties of thissample.
Telescope(WSRT).Theoptical/stellarmorphologiesofthis Inaddition,allfourgalaxiesshowopticalemissionlines
sample are presented in the top row of Figure 1. dominatedbyactivityotherthanstarformation:J0836and
Stellar population modelling find two distinct star for- J0900 show Seyfert-like lines while J1117 and J1237 show
mationtruncationtimescalesforearly-andlate-typegalax- low-ionization lines which could be due to black hole ac-
ies whereby the early-type galaxies appear to have fast cretion,shocksorevolvedstars (Sarzi et al. 2010).Figure4
quenchingtimescales (61Gyr),whilethelate-typegalaxies presentstheBaldwin,Phillips &Televich (BPT) diagnostic
4 O. I. Wong et al.
1.5
Kewley et al. (2001) Seyfert Seyfert Seyfert
Kauffman et al. (2003)
1.0 Schawinski et al. (2007) J0900
J0836
b) 0.5 LINER
H
OIII]/ J12J317117
g ([ 0.0 LINER
o LINER
l
-0.5
-1.0
Star-forming Star-forming Star-forming
-1.0 -0.5 0.0 -1.0 -0.5 0.0 -1.5 -1.0 -0.5 0.0
log ([NII]/Ha) log ([SII]/Ha) log ([OI]/Ha)
Figure 4. Optical nebular emission line diagrams. The nebular emission line ratios of the entire Schawinski et al. (2009) sample of
blue early-type galaxies areplotted as black points. The bluesolidpoints represent the objects selected inthis pilot sample. The solid
and dashed line are two lines used to separate the nebular emission originating from star formation rather than that of Seyfert or
LINER activity (Kewleyetal. 2001; Kauffmannetal. 2003). The dotted-dashed line differentiates between ionisation levels obtained
fromLINERversusSeyfertcores(Schawinski etal.2007).
diagram(Baldwin et al.1981)whichshowstheemissionline
ratiosobservedforblueearly-typegalaxies.Table1liststhe
optical properties of this pilot sample. Stellar masses were
obtained from spectral fitting(Schawinski et al. 2009).
3 PILOT SURVEY
3.1 Radio observations
Weperformed imagingofthe1.4GHzradiocontinuumand
theatomicHydrogen(Hi)emissionofourpilotsampleusing
theWSRTintheNetherlandsbetweenJuneandNovember
2012. As the WSRT is an east-west interferometer, our ob-
servations were divided into several epochs to optimise the
uv-coverage. The single 20 MHz band is sampled over 1024
channelsinthemaxi-shortbaselineconfigurationinorderto
obtainaspectralresolution of4kms−1 spanningtherange
of 4000 km s−1.
Usingthissetup,a24-houron-sourceintegrationshould
resultinanRMSnoiselevelof0.54mJybeam−1 atfullan-
gularresolution overaFWHMof8.25 kms−1 in velocity—
correspondingtoacolumn densitysensitivityof1.46×1019 Figure 5.MulticolourcompositeoftheJ0836+30field(centred
atoms cm−2 per resolution assuming a uniform weighting1. onJ0836+30)wherethebackgroundcolourimageisamulticolour
composite of the five SDSS optical ugriz bands. The 1.4 GHz
Due to timetabling constraints, we were not able to obtain
radio continuum and Hi emission from these pilot observations
24-hour on-source integration for each of the four galaxies.
areshowninmagenta andcyan, respectively. Theyellow ringin
Totalintegrationtimesforeachgalaxy,theresultantsynthe-
the bottom-left of the figure represents the beam of the radio
sisedbeam propertiesandsensitivitiesarelisted inTable2.
observations.
3.2 Data processing
1 Noise estimations are obtained from the
WSRT exposure time calculator found at The observations are calibrated and processed in the
http://www.astron.nl/ oosterlo/expCalc.html standard manner using the Miriad Reduction software
Misalignment between cold gas and stars in early-type galaxies 5
Table 1.Opticalpropertiesofourlow-redshiftblueearly-typegalaxysample.
SDSSID Galaxy RA(J2000) Declination(J2000) Redshift Distance r u−r log(M⋆)
(1) (2) (3) (4) (5) (6) (7) (8) (9)
587735044686348347 J0836+30 08:36:01.5 +30:15:59.1 0.02561 105 -21.1 2.27 10.7
587731887888990283 J0900+46 09:00:36.1 +46:41:11.4 0.02748 113 -21.3 2.20 10.5
587732135382089788 J1117+51 11:17:33.3 +51:16:17.7 0.02767 115 -21.4 2.33 10.6
587738946685304841 J1237+39 12:37:15.7 +39:28:59.3 0.02035 84 -20.9 2.16 10.3
Col.(1):SDSSobjectidentification. Col.(2): Galaxyidentification usedinthispaper.Col.(3): Galaxycenter’srightascension.Col.
(4):Galaxycenter’sdeclination.Col.(5):Redshift.Col.(6):DistanceinMpc.Col.(7):SDSSr-bandmagnitude.Col.(8):Opticalu−r
colour.Col.(9):log(stellarmass)insolarmassasestimatedbyspectralfitting.
Table 2.WSRT observation summary.WSRT observationintegrationtimesandresultingbeam propertiesaredetailedforeachofour
pilotbluegalaxies.
Galaxy Observationsdates Totalintegrationtime Beamsize BeamPositionangle RMSHi RMS1.4
(1) (2) (3) (4) (5) (6) (7)
J0836+30 Nov15,25 20.3 29.5,13.9 -2.0 4.1 0.083
J0900+46 Oct23,Nov11 10.3 24.0,12.1 -1.7 11.0 0.062
J1117+51 June30,Nov18 24.0 19.8,14.5 9.6 3.7 0.082
J1237+39 Aug24,Oct25,Nov21 24.2 27.6,17.5 4.7 32.6 0.038
Col.(1):Galaxyidentification. Col.(2):Observationdates.Col.(3):Totalon-sourceintegrationtimeinhours.Col.(4): Synthesised
beamdiameterinarcseconds.Col.(5):Synthesisedbeam positionangleindegrees.Col.(6):RMSnoiselevelfortheHiemissionin
mJybeam−1.Col.(7): RMSnoiselevelforthe1.4GHzradiocontinuum inmJybeam−1.
(Sault et al.1995).Aftertheinitialbandpasscalibrationus- sible mechanism for the displacement of the Hi reservoir in
ingaWSRTstandardcalibrator,continuumimagesarepro- J0836+30 can befound in Section 5.1.
duced and improved by iteratively applying a continuum We argue that the two radio lobes flanking both sides
self-calibration. The resulting calibration tables are then of J0836+30 are likely to be faded radio lobes belonging to
used for theline imaging as well. Afterasecond continuum J0836+30 and not likely to be due to a chance alignment
subtraction, the resulting Hi image cubes are produced by with background sources. Although there is some overlap
usingarobustweightingof0.4inordertoproducetheopti- between theSouth-Eastern radio lobe with a WISE3.6 mi-
mal balance between angular resolution and surface bright- cronsource,wedonotfindanoptical/IRcounterpartforthe
ness sensitivity. We average across every second channel to North-Western radio lobe. In addition, the projected sepa-
improvethesignal-to-noisesensitivitywhichresultsinave- rationbetweenbothlobesandJ0836+30aresimilarenough
locityresolutionof 8.5kms−1(afterHanningsmoothing).It for the differences in angular separation (as well as bright-
should be noted that these final reduction parameters were ness) tobe accounted for by projection effects.
determinedafterwehadalreadytestedvariouscombinations Thenuclearradioluminositiesrangefrom 2.0×1035 to
ofweightingschemes(includingnaturalanduniformweight- 7.4×1035 ergss−1 (or1.0×1020 to3.7×1021 WHz−1)and
ings)withdifferentangularandfrequencytaperingforeach
areconsistentwiththosefoundmostlyinotherradiostudies
setofobservations.Table2liststheRMSlevelsobtainedfor
oflow-redshiftstar-forminggalaxies(Mauch & Sadler2007;
each galaxy in oursample.
Jarvis et al. 2010). For J0836+30, the luminosities of the
radio lobes in the north-west and south-east directions are
9.26×1035and1.22×1037ergss−1,respectively.Table3lists
themeasured1.4GHzradiocontinuumproperties.Figure1
shows the radio continuum emission as green contours and
4 RADIO CONTINUUM PROPERTIES OF theHiemissionaswhitecontoursfortheentirepilotsample.
BLUE EARLY-TYPE GALAXIES
Assuming that the nuclear radio continuum emission
Weobserve1.4GHzradiocontinuumemissioninthenuclear is due to star formation, we can estimate the star forma-
regions of J0900+46, J1117+51 and J1237+39. No nuclear tion rate upper limits in our pilot sample. Using the star
radio emission was found for J0836+30, but we do observe formation rate calibrations from Bell (2003);Hopkins et al.
tworadiolobesextending88.4kpcnorth-westand102.5kpc (2003), we find that our galaxies have star formation rates
south-east of J0836+30 along the same direction as the ex- thatarefewerthan2solarmassperyear.J1117+51hasthe
tragalacticHicloud.Figure5showsamulticolourcomposite highest estimated star formation rate with 1.7 solar mass
imageoftheJ0836+30 fieldwheretheradiocontinuumand peryear.
the integrated Hi emission is shown in magenta and cyan, Archival IRAS 60 µm observations (Moshir & et al.
while the background three-colour image is produced from 1990) were found for J1117+51 and J1237+39. Using the
the five SDSS ugriz bands. Further discussion of the pos- far-infrared to radio correlation (FRC) of low-redshift star-
6 O. I. Wong et al.
forminggalaxiesYunet al.(2001),wefindthattheexpected is approximately 373 Myr. Similarly, the Hi emission de-
far-infrared 60 µm luminosity that correlates to our mea- tectednearJ0900+46 spansbetween14kpcto86kpcfrom
sured radio luminosity if the radio emission is due solely theopticalcenterofJ0900+46andappearstoberedshifted
to star formation is log L60 of 9.6 and 9.0 for J1117+51 by approximately 70 kms−1 (Figure 8b). We estimate the
and J1237+39, respectively.In comparison to theIRASlog displacement timescale to range between 0.2 to1.2 Gyr.
L60 values of 9.4 and 9.3, we find that the predicted val-
ues from the FRC are consistent with the idea that the 1.4
5.1 Mechanisms for the removal of Hi
GHzcontinuumemission originates from star formation for
both J1117+51 and J1237+39. It should be noted that the Support for the hypothesis that the observed Hi clouds in
FRChasascatter ofapproximately 0.25 dexandtheIRAS thevicinity of J0836+30 and J0900+46 havebeen removed
measurement uncertainties are at the20% level. from their parent galaxy come from two main arguments.
Although the radio continuum emission observed from Firstly, the proximity of the gas clouds are very similar to
J1237+39appearstobeextended,itisinfacttheresultofa the distances between the Milky Way and its high velocity
chance alignment between J1237+39 and a projected back- clouds (HVCs). Secondly, neither J0836+30 nor J0900+46
ground galaxy, SDSS J123716.92+392921.9, which is 8.5 resideindenseenvironmentswithcloseneighbouringgalax-
times more distant (at redshift z = 0.176) than J1237+39. ies which could be responsible for stripping the gas reser-
Figure6showsthattheobservedradioemissioncanbesim- voirsfromeachofthesegalaxies.Infact,Verley et al.(2007)
ply modelled by two point sources. The lack of any regions has that found both J0836+30 and J0900+46 to be iso-
which display a flux excess or a flux deficit in our resid- lated galaxies using different galaxy isolation metrics. Sev-
ual image (panel c of Figure 6) is consistent with the idea eral other previous studies havealso classified J0836+30 to
that the radio emission from J1237+39 does not have an beaprototypicalisolated galaxy (Verdes-Montenegro et al.
extendeddiffuse component. 2005; Stockeet al. 2004; Karachentseva 1973).
In dense galaxy environments, such as clusters or
compact galaxy groups, intergalactic gas clouds are not
5 Hi CONTENT OF BLUE EARLY-TYPE uncommon (Kilborn et al. 2000; Oosterloo & van Gorkom
2005; Borthakur et al. 2010) due to various gravitational
GALAXIES
and hydrodynamic interaction processes such as galaxy-
As the four blue ETGs in our sample represent four sepa- galaxy interactions (Mihos 2004),harassment (Moore et al.
ratestagesofevolution(astracedbytheUV/opticalcolours 1996, 1998),ram pressure and turbulent viscous stripping
and the [OIII]/[Hβ] ratios), we observe a correlation with (Vollmer et al. 2001).
theirrespectiveHigas-to-stellarmassfractionswherebythe Ram pressure stripping is unable to shift such large
gas fractions decrease as a function of time with increasing amounts of gas using simple Toomre & Toomre pressure
UV/optical colours and increasing [OIII]/[Hβ] ratios. Ta- arguments (e.g. Chunget al. 2007). Also, the galaxies in
ble 3 lists the measured Hi properties. The galaxy at the thispilot sampledonot havenearbygalaxy neighboursnor
earlieststageofquenchinginoursample(i.e.thegalaxywith dotheyresidein densegalaxy environmentsuchasclusters
thebluest NUV/optical colour and theweakest [OIII]/[Hβ] or compact groups.
ratio), J1237+39, features a central, undisturbed, rotating Nearbyexampleswherethemajorityofthegalaxy’sgas
HI disk. Both spatial and kinematic asymmetries are not is displaced via tidal stripping also results in strong mor-
observed in this galaxy (see Figure1 and Figure 7). phological distortions (e.g. NGC 4438 in the Virgo Cluster
In the second bluest galaxy, J1117+51, the Hi gas ap- (Hotaet al. 2007)). Hotaet al. (2012) has also identified a
pearstobemostly within J1117+51. Howeverthegasmor- post-merger system, NGC 3801, where the extraplanar Hi
phology is fairly asymmetric and appears offset from the resulting from the merger has yet to be encountered by a
optical centre of the galaxy. Assuming the optical velocity newly-triggered radio jet. As we do not observe any strong
tobethesystemicvelocityofeachgalaxy,wefindtheHigas opticalsignaturesfromourgalaxiessuchasstellartidaltails,
in J1117+51 to be largely blue-shifted by up to 170 kms−1 we can only infer that (1)the tidal features resulting from
from the galaxy’s systemic velocity (see Figure 7). thestronginteractionthatremovedthelargegasmasshave
In the redder galaxies and the galaxies with stronger faded; or (2)another physical mechanism is responsible for
[OIII]/[Hβ] ratios—indicative of ionisation levels consistent thegas removal.
with Seyfert activity, the Hi reservoirs of J0836+30 and Inthelattercase,wepositthatanactivecentralengine
J0900+46 are completely displaced from their respective has the required energy to expel a galaxy’s gas. In thecase
galaxy centres by 14–86 kiloparsecs. In addition, we find of J0836+30, we observe two radio lobes in the same direc-
asignificantamountofred-shiftedHi(by≈70−80kms−1) tion as the ejected Hi cloud. This is suggestive that at ear-
withrespecttothesystemicvelocityoftheindividualgalaxy lierepochs, J0836+30 hosted aradio AGNwhich may have
(see Figure 1). blownoutthegascloudthatweobserve.Thishypothesisis
In J0836+30, the entire gas reservoir is offset spatially consistent with recent studies which havefound nuclearac-
by1arcminute(projecteddistanceof30.5kpc).Toapprox- tivityto bemore prevalent in star-forming galaxies relative
imate the timescale for the displacement of the gas reser- to weakly star-forming or quenched galaxies (Rosario et al.
voirs from J0836+30 and J0900+36, we simply divided the 2013).
projected distances of the gas clouds by the observed Hi As the displacement timescale for this cloud is esti-
velocity offsets from the systemic velocities taken from the matedtobe≈373Myr,itisconceivablethatJ0836+30has
optical observations. For J0836+30, the timescale for the since evolved to a more Seyfert state (as observed via the
gas reservoir to have reached its current projected distance optical nebular emission). A test of this hypothesis would
Misalignment between cold gas and stars in early-type galaxies 7
Figure 6. The 1.4 GHz radio continuum emission from J1237+39. Centred on J1237+39, we model the emission as two overlapping
pointsources.Thesolidbluelineinallthreepanelsmarkthe3σfluxleveloftheobserved1.4GHzemission(seePanela).Panelsband
cshowthemodeloftwopointsourcesandtheresidualmaps,respectively.
Figure 7.Velocity maps and spectra of J1117+51 (panel a) and J1237+39 (panel b). The 3-color velocity maps correspond to the Hi
emissioninthe velocity rangehighlighted inthe same color onthe corresponding velocity spectrum. It shouldbe noted that the green
representstheopticalvelocitywhichweassumetobethesystemicvelocityofeachoftheblueearly-typessampledinthispaper.
8 O. I. Wong et al.
Table 3.Radio(1.4GHz)andHipropertiesofblueearly-typegalaxies.
Galaxy S1.4 L1.4 L1.4 SFR1.4 SHi vHi wHi MHI
(1) (2) (3) (4) (5) (6) (7) (8) (9)
J0836+30 <2.5×10−4 <3.3×1020 ∗∗ <6.6×1034 <0.4 0.078(0.025) 7705 140 2.0(0.7)
J0900+46 9.7(2.4)×10−4 1.5×1021 3.0×1035 1.1 0.410(0.182) 8277 62 12.3(5.5)
J1117+51 23.4(3.8)×10−4 3.7×1021 7.4×1035 1.7 0.100(0.020) 8185 140 3.1(0.6)
J1237+39 12.0(1.6)×10−4 1.0×1021 2.0×1035 0.7 2.448(0.403) 6096 245 40.8(6.7)
∗∗Notethattheluminositiesoftheradiolobesinthenorth-westandsouth-eastdirectionsare4.63(±0.44) ×1021 WHz−1 and
6.12(±0.03) ×1022 WHz−1,respectively.Col.(1): Galaxyidentification. Col.(2):Total 1.4GHzradiocontinuum emissioninJansky.
Col.(3):Radiocontinuum luminosity(1.4GHz)inWHz−1.Col.(4): Radiocontinuum luminosity(1.4GHz)inergss−1.Col.(4):
EstimatedstarformationrateusingthecalibrationbyBell(2003);Hopkinsetal.(2003)inM⊙ year−1.Col.(5):Integrated Hi
emissioninJansky.Col.(6): Hiradialvelocityasmeasuredfromthemid-pointatfull-widthhalf-maximuminkms−1.Col.(7):The
50%Hivelocitywidthinkms−1.Col.(8):Total Himassinunitsof×108M⊙.Valuesinbrackets givetheestimatederrors.
Figure 8.Velocity maps and spectra of J0836+30 (panel a) and J0900+46 (panel b). The 3-color velocity maps correspond to the Hi
emission in the velocity range highlighted in the same color on the corresponding velocity spectrum. The white solid circle represents
thepositionoftheblueearly-typegalaxyrelativetotheobservedHiemission.Itshouldbenotedthat thegreenrepresents theoptical
velocitywhichweassumetobethesystemicvelocityofeachoftheblueearly-typessampledinthispaper.
be the presence of ionized gas adjacent to and between the so far been from studies where the outflows are determined
observed Hi cloud and the galaxy. Therefore, in the case of from the kinematic structures of the emitting or absorb-
J0836+30, we think that the alignment between the radio inggas(Gopal-Krishna & Irwin2000;Nesvadbaet al.2009;
lobes and the extragalactic Hi cloud is more suggestive of Mahony et al. 2013; Morganti et al. 2013). However, recent
gas ejection via radio jets than a previous tidal encounter observationsof12CO(1-0)emission from high-redshiftradio
whosetidalsignatureshavesincefadedinthepast373Myr. galaxies have found spatial offsets between the molecular
PreviousobservationsofAGN-drivengasoutflowshave gas reservoirand theirhost galaxies on thescales ofseveral
Misalignment between cold gas and stars in early-type galaxies 9
tens of kiloparsecs and aligned in the same direction to the NUV−rcolors toE+Agalaxies andstraddletheregion be-
radiohotspots/lobes(Emonts et al.2014).Weproposethat tween star-forming galaxies with low gas fractions and that
wearewitnessingasimilarejectionofcoldgasinJ0836+30. of gas-rich early-typegalaxies.
InthecaseofJ0900+46,weneitherobservestrongradio Figure 9 compares the Hi-stellar mass ratio and the
continuumemissionnornearbygalaxieswhichcouldremove integrated NUV−r colour of J1117+51 and J1237+39
thegas from this galaxy. (marked as black stars) to those of E+A galaxies (rep-
We posit that the observed extragalactic gas clouds in resented by black crosses) where Hi has been previously
J0836+30 and J0900+46 are outflows and not inflows for detected (Chang et al. 2001; Buyleet al. 2006; Helmboldt
two reasons. Firstly, if the observed cold clouds were be- 2007; Zwaan et al. 2013).galaxies from ALFALFA are rep-
ing accreted, the source of this gas is either a neighbour- resentedbybluesolidpointsandthesolidlinemarkstheav-
ing galaxy or primodial gas leftover from galaxy formation. erage colours and average gas fractions found from GASS.
However, neither galaxies have neighbouring galaxies and Near-ultraviolet (NUV) measurements of our sample were
the accretion of primodial gas at the current epoch is in a obtained from the Galex satellite telescope via the NASA
warm-hot phase rather than that of the observed cold Hi. Extragalactic Database2.
Secondly, Bouch´eet al. (2012) found that gas outflows are We also compare our pilot sample of blue ETGs to re-
usually found at angles nearly orthogonal (> 60 degrees) centHiobservationsofnearbyETGsfromtheAtlas3Dsur-
to the plane of the galaxy, whereas inflows have small az- vey (Serra et al. 2012). The Atlas3D ETGs with regularly-
imuthalangles(<30degrees)thegalacticplane.Therefore, rotating undisturbed Hi morphologies are represented by
wethinkthatit’smorelikelythatourobservedextragalactic open circles, while, Atlas3D galaxies with very disturbed
Hi clouds originated from our target galaxies. and unsettled Hi morphologies and kinematics are repre-
The mass of our gas clumps are approximately the to- sented by solid diamonds in Figure 9. It should be noted
talmassof HVCsaroundtheMilky Way,theLMC andthe that all theAtlas3D ETGs with disturbed Himorphologies
SMC.Assumingthatourobservedgasclumpsareanalogous reside in group or Virgo Cluster environments. Also, for a
totheseHVCs,wecanexpectasimilargasdissipationtime given NUV−r colour, these Atlas3D galaxies have higher
on the order of several hundred million years if there isn’t gas fractions than that of the E+A or blue ETG samples.
somesortofsupportmechanismthatpreventsorslowdown Thissuggeststhateither(1)theAtlas3DETGswhereHiis
against cloud dissipation (Putman et al. 2012). The cloud observed are at earlier stages of evolution than the sample
survival time is linked closely to its total mass, cloud den- of E+A galaxies or blue ETGs; or that (2) the merger sce-
sity,relativehalodensityandvelocity(Putman et al.2012). narioisaslikelytoresultinanincreaseingasfraction asit
Previousstudieshavefoundthatwhiletheleastboundma- istoareduction.Upperlimitsareshown forJ0836+30 and
terialislikelytoexpandoutintotheIGM,boundstructures J0900+46 dueto thelack of Hi within the host galaxies.
are likely to fall back onto the galaxies in less than 1 Gyr We find the Hi gas fraction and NUV−r colour of
(Hibbard & Mihos 1995; Hibbard & vanGorkom 1996). J1237+39 (thegalaxy at theearliest stage of quenching)to
be very comparable to those of star-forming galaxies from
boththeALFALFAandGASSsurveys.Ontheotherhand,
5.2 Hi gas fractions
the gas fractions for the three more evolved blue ETGs
Using thescaling relations from the α40 catalog of theAL- in our sample are significantly lower than the average gas
FALFA survey (Huanget al. 2012), star-forming galaxies fractions expected from star-forming galaxies, transitioning
with the stellar masses of J0836+30 and J0900+46 would galaxies or even gas-rich early-type galaxies with similar
be expected to have average Hi masses of approximately NUV−r colours. Similar to our results, the gas fractions of
1×1010 M⊙ and8.7×109 M⊙,yetweobserveonly 2×108 post-starburstE+Agalaxies arelowerthantheaveragegas
M⊙ and1.2×109 M⊙, respectively.This implies that some fractionsforanygivenNUV−rcolours.Thecombinationof
fraction of the gas may have been ionized and heated and the Hi mapping and the radio continuum observations of
we are currently only observing the remaining fraction of J0836+30 and J0900+46 suggest that thestar formation in
Hi. The external Hi clouds we observe represent the rem- thesetwogalaxieswillbetruncatedsoon.Thisresultiscon-
nants of the original gas reservoirs swept out of the host sistent with those of Schawinski et al. (2009) who were un-
galaxy,andfromthepresentdataitisnotclearwhetherthe abletodetectanysignificantmoleculargasreservoirswithin
missingHiwasalsoexpelledorheatedorsimplydissipated. Green Valley galaxies with Seyfert ionisation properties.
Regardless, the displaced gas reservoir is required to com- As previously seen in Figure 3, these pilot Hi observa-
plete the quenching of these galaxies and prevent further tionsshow thatthemain mechanism for thefast quenching
star formation. of star formation in blue early-types is due to the physical
Extrapolating from theKennicutt-Schmidtstar forma- displacement ofthemaingas reservoirfrom whichstars are
tionlaw,thecorrelationbetweentheHigasfractionandthe formed. Hence, it is likely that the depressed gas fractions
starformationhistorytracedbytheNUV−rcolorprovidesa from our sample (see Figure 9) are due to a fast quenching
roughindicatorforthecoldgassurfacedensity(Huanget al. process (relative to other evolving galaxies) which removes
2012; Zwaan et al. 2013). Relative to gas-rich star-forming a significant fraction, if not the entire gas reservoir, from
galaxies (mostly spirals) (ALFALFA; Huanget al. 2012), which stars are formed.
low-redshift transition galaxies with stellar masses greater Should the gas fall back, it may restart minor star
than 1010 M⊙ (GASS; Catinella et al. 2012) and nearby formation which may be visible as the frosting ≈ 1%
early-type galaxies from the Atlas3D survey (Serra et al.
2012; Cappellari et al. 2013; Younget al. 2013), we find
that J1117-51 and J1237+36 havesimilar gas fractions and 2 http://ned.ipac.caltech.edu
10 O. I. Wong et al.
thatthemajorityofgascloudsidentifiedwithnoknownop-
2 ticalcounterpartcanbeeasilyreproducedbygalaxy–galaxy
interactions (Bekki et al. 2005; Duc& Bournaud 2008).
1
6 SUMMARY
0
n) J1237+39 Wehaveperformed deep imaging of theHicontent and 1.4
o
cti GHzradiocontinuumemissionoffourblueearly-typegalax-
a -1
as fr iteiosnthuastinagrethaetWfouSrRdTiff.TerheentHsitamgoerspohfosltoagriefso,rkminateimonattircusnacnad-
g
g ( -2 J1117+51 relative Hi-gas fractions are excellent probes and measures
o
l J0900+46 of thequenchingevolutionary stages for each of oursample
galaxy. A summary of our results are as follows:
-3 Blue ETGs (this work) J0836+30
(i) Weobservenuclear1.4GHzradiocontinuumemission
ALFALFA gas-rich galaxies (Huang et al 2012)
E+A galaxies (Zwaan et al. 2013, Buyle et al. 2006, Chang et al 2001) that are consistent with emission from star formation from
-4 Atlas3D ETGs (Serra et al 2012, Capellari et al 2013, Young et al 2013) three galaxies at the three earliest stages of quenching evo-
Atlas3D ETGs with unsettled HI reside in pairs/groups/Virgo Cluster lution, namely, J1237+39, J1117+51 and J0900+46 (in the
0 1 2 3 4 5 6 7 order of earliest to later stages of evolution).
NUV-r
(ii) The galaxy at the earliest stage of quenching,
J1237+39, also has the bluest NUV−r colour, weakest
[OIII/Hβ]ratio,thehighestHigasfractionandasymmetric
Figure 9. Hi-to-stellarmass ratioas afunction of galaxy color.
rotating Hi disk.
Thesolidlineshows the average gas fractionsfoundfromGASS
(iii) Atmore advancedstages of quenchingevolution, we
surveyofmassivetransition-typegalaxies(Catinellaetal.2012).
observe increasingly asymmetric and increased spatial off-
sets between the Hi gas and the stellar component of the
mass fraction of young stars often observed in quenched galaxy. Non-rotating Hi gas kinematics are also observed.
early-type galaxies (Yi et al. 2005; Schawinski et al. 2006; Inthe case of thetwo galaxies at themost advanced stages
Kaviraj et al. 2007). While it may cause frosting, this re- of evolution where Seyfert ionisation signatures have been
turning gas however will not return thehost galaxy to self- observed (J0900+46 and J0836+30), the Hi gas reservoirs
regulated star-formation on the main sequence. An alter- have been entirely expelled by approximately 14–86 kpc
native fate for the ejected gas is that it persists at large fromtheirrespectivehostgalaxies.Duetothelackofneigh-
distances from the quenched galaxy, as is observed in the bouring galaxies, it is difficult to attribute thestripped gas
Milky Way (in the form of the starless Magellanic stream; to tidal or ram pressure interactions.
For et al. 2014); as well as in many quenched early-type (iv) Inthegalaxyatthemostadvancedquenchingstage,
galaxies (e.g. Serra et al. 2012). J0836+30, the expelled gas reservoir is observed to be in
alignment between the host galaxy and two radio lobes—
suggesting the gas reservoir may have been swept out of
5.3 Argument against coincident dark galaxies
thegalaxy by powerful outflows from thecentral AGN in a
Dark galaxies can be typically defined as the extreme end previousactivephase.Thisscenarioisconsistentwithrecent
oflowsurfacebrightnessgalaxieswherefewstarsarefound, observations of AGN-driven gas outflows both locally and
consistingmainlyofgas.Withrespecttoourobservedextra- at higher redshifts (Sturm et al. 2011; Mahony et al. 2013;
planar gas clumps, it is very unlikelythat thesegas clumps Morganti et al. 2013;Emonts et al. 2014).
are neighbouring dark galaxies. Current re-ionization mod- (v) We conclude that the rapid quenchingof star forma-
els of the Universe predict the latter as they find that the tioninlow-redshiftearly-typegalaxiesisduetothephysical
gasfrom 95percentofthelow-mass systems(Mvirial 6108 expulsionoftheentiregasreservoir,ratherthantheexhaus-
M⊙ or vcirc 620 km s−1) appears to have been photoevap- tionofgasviastarformation.AnAGNoutflowhasthenec-
orated duringthe epoch of re-ionization (Susa& Umemura essaryenergytoexpelthereservoirofcold gas.Inaddition,
2004). wedonotthinkthatAGNheatingofthegasisadominant
Apart from a few reported cases of extragalactic mechanism because the Hi gas would not remain visible as
gas clouds in group/cluster environments (Minchin et al. a coherent structure if this was the case.
2005; Davies et al. 2006)) with no known optical counter-
parts, current and previous Hi all-sky surveys such as AL-
FALFA(Hayneset al.2011)andHIPASS(Wonget al.2009;
Doyleet al. 2005) have not found any isolated extragalac- Acknowledgments. We thank the anonymous ref-
tic HI clouds devoid of stars. This is consistent with simple eree for their support of this project and for improving
gasequilibriummodels(Taylor & Webster2005)whichcon- the manuscript. OIW acknowledges a Super Science Fel-
cludedthatintheabsenceofaninternalradiationfield,dark lowship from the Australian Research Council and the He-
galaxies/gas clouds with masses greater than 109 M⊙ will lena Kluyver visitor programme at ASTRON/JIVE. KS is
become Toomre unstable against star formation and start supported by SNF Grant PP00P2 138979/1. The Wester-
forming stars. Galaxy evolution simulations demonstrated borkSynthesisRadioTelescopeisoperatedbytheASTRON