Constraints on Two Active Galactic Nuclei in the Merger Remnant COSMOS J100043.15+020637.2

0.37 MB·

by  J. M. Wrobel

Accepted by ApJ on 2014January15
PreprinttypesetusingLATEXstyleemulateapjv.04/17/13
CONSTRAINTS ON TWO ACTIVE GALACTIC NUCLEI IN THE MERGER REMNANT
COSMOSJ100043.15+020637.2
J. M. Wrobel1,2, J. M. Comerford3, and E. Middelberg4
Accepted byApJ on 2014 January 15
ABSTRACT
COSMOSJ100043.15+020637.2is a merger remnant at z = 0.36 with two optical nuclei, NW and
4
SE, offset by 500 mas (2.5 kpc). Prior studies suggest two competing scenarios for these nuclei: (1)
1
SE is anactive galactic nucleus (AGN) lostfromNW due to a gravitational-waverecoil. (2) NW and
0
2 SE each contain an AGN, signaling a gravitational-slingshotrecoil or inspiralling AGNs. We present
new images from the Very Large Array (VLA) at a frequency ν = 9.0 GHz and a FWHM resolution
n
θ = 320 mas (1.6 kpc), and the Very Long Baseline Array (VLBA) at ν = 1.52 GHz and θ = 15
a
mas (75 pc). The VLA imaging is sensitive to emission drivenby AGNs and/orstar formation, while
J
the VLBA imaging is sensitive only to AGN-driven emission. No radio emission is detected at these
9
frequencies. Folding in prior results, we find: (a) The properties of SE and its adjacent X-ray feature
1
resemble those of the unobscured AGN in NGC4151,albeit with a much higher narrow emission-line

luminosity. (b)ThepropertiesofNWareconsistentwithithostingaCompton-thickAGNthatwarms
]
A ambientdust,photoionizesnarrowemission-linegasandisfree-freeabsorbedbythatgas. Finding(a)
isconsistentwithscenarios(1)and(2). Finding(b)weakensthe caseforscenario(1)andstrengthens
G
the case for scenario (2). Follow-up observations are suggested.
h. Subject headings: galaxies: active — galaxies: individual (COSMOSJ100043.15+020637.2)— galax-
p ies: interactions — galaxies: nuclei
-
o
r 1. MOTIVATION Comerford et al.2012). Moreover,some candidates may
t
s Simulations suggest that galaxy mergers can pro- turn out to be imposters because they harbor a recoil-
a ing AGN, a situation that both complicates and en-
duce remnants with two or more massive black
[
riches matters (e.g., Blecha et al. 2011; Guedes et al.
holes (Hoffman & Loeb2007; Amaro-Seoane et al.2010;
1 Kulkarni & Loeb 2012). When these black holes ac- 2011; Eracleous et al. 2012). This paper focuses on one
v crete,they canappearastwoormoreactivegalacticnu- such case, COSMOSJ100043.15+020637.2(J1000+0206
6 clei (AGNs) on kiloparsec scales (Van Wassenhove et al. hereafter). Early reports on J1000+0206 (Smolcic et al.
5 2012; Blecha et al. 2013b). Systematic surveys for 2008; Elvis 2009; Comerford et al. 2009b; Civano et al.
7 2010) were prompted by its unusual optical nature, in-
such multiple AGNs could provide observational con-
4 volving two apparent nuclei in a merger remnant with a
straints on AGN activation and tidally enhanced star
. tidal tail (Scoville et al. 2007).
1 formation (e.g., Liu et al. 2012; Koss et al. 2012), and
The integrated optical emission-line redshift of
0 on the black hole merger rate, with consequences
J1000+0206isz =0.36(Lilly et al.2007). Adoptingthe
4 for the signals expected for pulsar timing arrays and
1 gravity-wave detectors (Hobbs et al. 2010; Dotti et al. nomenclature and cosmology5 of Civano et al. (2012),
: 2012). Guided by early serendipitous discoveries the two nuclei, SE and NW, have a projected separation
v
of dual AGNs like 3C75 (Owen et al. 1985) and of 500 mas (2.5 kpc). Two scenarios have emerged for
i
X NGC6240 (Beswick et al. 2001; Komossa et al. 2003; these optical nuclei: either NW and SE each contain an
r Gallimore & Beswick 2004), recent systematic surveys AGN with its own narrow emission-line region, or SE is
a are now yielding spectroscopic samples of dual AGN anAGNrecoilingfromNWduetotheasymmetricemis-
candidates (e.g., Comerford et al. 2009a; Wang et al. sionofgravitationalwavesduringblack-holecoalescence.
2009;Liu et al.2010;Smith et al.2010;Koss et al.2012; Hydrodynamic simulations, with radiative transfer cal-
Barrows et al. 2013; Comerford et al. 2013). culations, of a gas-rich major merger show that each
Candidate dual AGNs can be difficult to confirm due scenario is consistent with available data (Blecha et al.
to obscuration and resolution issues (Comerford et al. 2013a).
2011; Fu et al. 2011a; Shen et al. 2011; Fu et al. 2012; Smolcic et al. (2008) assumed that the emission
from J1000+0206 at a frequency ν = 1.4 GHz
(Schinnerer et al.2007)isdrivenbyitsspectroscopically-
1National Radio Astronomy Observatory, P.O. Box O,
identified AGN. But this is far from certain. Figure 1
Socorro,NM87801;[email protected]
2The National Radio Astronomy Observatory (NRAO) is shows that the radio emission is localized to the inner
a facility of the National Science Foundation, operated under portions of the merger remnant (Schinnerer et al. 2007,
cooperative agreementbyAssociatedUniversities,Inc. 2010),wheresimulationsindicatethatstarformationcan
3Department of Astrophysical and Planetary Sci-
occur (Blecha et al. 2013a). Also, Civano et al. (2012)
ences, Universityof Colorado, Boulder, CO 80309;
[email protected] recently discovered an extended X-ray feature offset by
4Astronmisches Institut, Ruhr-Universitat Bochum,
Universitatsstr. 150, 44801, Bochum, Germany; middel-
[email protected] 5 H0 =70kms−1 Mpc−1,ΩM =0.3,ΩΛ =0.7
2 Wrobel, Comerford & Middelberg
02 06 39.0
02 06 39.0
38.5 38.5
0) 38.0 0) 38.0
0 0
0 0
2 2
ON (J 37.5 NW ON (J 37.5 NW
TI TI
NA 37.0 SE NA 37.0 SE
LI LI
C C
DE 36.5 DE 36.5
SW feature SW feature
36.0 36.0
35.5 35.5
10 00 43.25 43.20 43.15 43.10 43.05 10 00 43.25 43.20 43.15 43.10 43.05
RIGHT ASCENSION (J2000) RIGHT ASCENSION (J2000)
Fig. 1.— HST/ACS image of the F814W emission from Fig.2.—VLAimageofStokesI emissionfromJ1000+0206atν
J1000+0206spanning4.2′′. Theplussignsmarkthepositionsand =9.0GHzandspanning4.2′′. Thermsnoiseis0.006mJybeam−1
3σastrometricerrorsoftheSEandNWopticalnuclei(Scovilleet (1σ)andthebeamdimensionsatFWHMare400mas×260mas
al.2007),andtheapproximatepositionoftheX-rayfeatureoffset with an elongation PA = -32◦ (boxed hatched ellipse). Contours
byabout500mastotheSWofSE(Civanoetal.2012). Thecircle areat-6,-4,-2,2,4,and6times1σ. Negativecontoursaredashed
andellipsearederivedfromVLAimagesatν =1.4GHzwithθ= and positive ones are solid. No emission was detected above 3 σ.
2.5′′ and 1.5′′, respectively (Schinnerer et al. 2010). A point-like Thecircle,ellipse,plussignsandscalearethesameasforFig.1.
source was detected in the former image, and the circle shows its
localizationasθ=2.5′′. Aresolvedsourcewasdetectedinthelat- ion 6. After further minor edits, the CASA task imagr
ter image, and the ellipse shows its deconvolved Gaussian extent was used to form and deconvolve a naturally-weighted
=at9F1W◦.HSMcaloefi2s.015′′′′=×51.0.139′k′pwc.ithanelongationpositionanglePA image of the Stokes I emission from J1000+0206. This
imaged spans 2561 × 60 mas in each coordinate, has an
rms noise level σ = 0.006 mJy beam−1 and a geometric
about500mas(2.5kpc)tothesouthwest(SW)oftheSE resolutionatFWHMθ =320mas. Itsinnerportionsare
optical nucleus, further complicating the observational showninFigure2. Withadequatesignal-to-noise,struc-
picture. turesas largeasabout10 θ ∼ 3.2′′ couldbe represented
If the radio emission from J1000+0206 arises from inFigure2. No emissionwasdetectedaboveathreshold
the AGNs and star formation, Figure 1 makes it clear of 0.018 mJy beam−1 (3 σ).
that disentangling these contributions requires imaging
atsubarcsecondresolution. §2presentsnewimagesfrom 2.2. VLBA
theKarlG.JanskyVeryLargeArray(VLA;Perley et al.
J1000+0206was observedwith the VLBA during four
2011) at ν = 9.0 GHz and a FWHM resolution θ = 320
6-hour sessions spanning 2012 June 24 to 2013 January
mas (1.6 kpc), and from the Very Long Baseline Array
11 under proposal code BM360. A coordinate equinox
(VLBA; Napier et al. 1994) at ν = 1.5 GHz and θ =
of 2000 was used. J1011+0106, with a one-dimensional
15 mas (75 pc). § 3 explores the implications of the new
position error of 2 mas (1 σ), was used as a phase cali-
imaging,whileasummaryandconclusionsappearin§4.
brator. The switching time between it and J1000+0206
2. IMAGING was 360 s, and involved a switching angle of 2.7◦. Ev-
ery4.1s the correlator(Deller et al.2011) produced128
2.1. VLA
contiguous 2-MHz channels, yielding a total bandwidth
J1000+0206was observed with the VLA in its A con- of 0.256 GHz per circular polarization centered at ν =
figuration on 2012 October 12 UT under proposal code 1.52 GHz. VLBA system temperatures and gains were
12B-072. Acoordinateequinoxof2000wasused. J1024- used to set the amplitude scale to an accuracy of about
0052, with a one-dimensional position error of 10 mas 5%. Atotalofabout450baselinehourswereaccruedon
(1 σ), was employed as a phase calibrator. The switch- J1000+0206.
ing time between it and J1000+0206 was 480 s, and in- For each 6-hour session, NRAO’s Astronomical Image
volved a switching angle of 6.6◦. The a priori point- ProcessingSystem(AIPS;Greisen2003)wasusedtocal-
ing position for J1000+0206 was shifted 0.6′′ south of ibrateandeditthedatafollowingtheapproachdescribed
the SDSS position(Comerford et al.2009b)to avoidthe by Middelberg et al. (2013). The AIPS task imagr was
risk of phase-center artifacts. Every 1 s the correlator used to form and deconvolve a naturally-weighted im-
produced 1024 contiguous 2-MHz channels, yielding a age of the Stokes I emission from J1000+0206. This
total bandwidth of 2.048 GHz per circular polarization image spans 2048 × 1 mas in each coordinate, has a
centered at ν = 9.0 GHz. Observations of 3C138 were geometric resolution at FWHM θ = 15 mas, and was
used to set the amplitude scale to an accuracy of about correctedfor primary beam attenuation. Combining the
1% (Perley & Butler 2013). The net exposure time on u-v data from the four sessions resulted in an rms noise
J1000+0206was about 2150 s. level σ = 0.013 mJy beam−1. With adequate signal-to-
Release4.1.0oftheCommonAstronomySoftwareAp- noise,structures as large as about 10 θ ∼ 150 mas could
plications (CASA) package (McMullin et al. 2007) was
usedtocalibrateandeditthedatainanautomatedfash- 6 science.nrao.edu/facilities/vla/data-processing/pipeline
Two AGNs in Merger Remnant COSMOSJ100043.15+020637.2 3
be imaged. No emission was detected above a threshold linecomponents,withlinefluxratiosofbothcomponents
of 0.078 mJy beam−1 (6 σ). falling within the AGN region of the Baldwin-Phillips-
Terlevichdiagram,withinthe1σ errorsonthefluxmea-
3. IMPLICATIONS surements.
3.1. Overview Projected spatial separations between the double
emission-line components of 2.1±0.7 kpc, 1.4±0.6 kpc,
ThenewVLBAimageatν =1.52GHzandθ=15mas
and 2.3±0.6 kpc were measured for [OIII] λ5007, Hα,
(75 pc) filters for emission with a rest-frame brightness
and [NII] λ6584, respectively. The mean of the three
temperature Tb >3.5×105 K, not achieved by even the spatial separations measurements, weighted by their in-
mostcompactstarbursts(Condon 1992). Moreover,the
versevariances,is1.9±0.4kpc. However,theHαspatial
VLBAimageistooshallowtodetecteventhemostlumi-
separationmeasurement is not consistent with the sepa-
nousradiosupernovabeyondz ∼0.1(e.g.,Garrett et al. rationsmeasuredin[OIII] λ5007and[NII] λ6584andis
2005; Middelberg et al. 2013). Thus the VLBA image
skewed low because of partial obscuration by an imper-
is insensitive to emission driven by star formation. In
fectly subtracted night-sky line. Using only the spatial
contrast, the new VLA image (Figure 2) at ν = 9.0 separations measured in [OIII] λ5007 and [NII] λ6584
GHz and θ = 320 mas (1.6 kpc) is sensitive to emission
(which are consistent with one another, to within the
driven by star formation, and to AGN-driven emission
1 σ errors), the inverse-variance-weighted mean spatial
in the SE or NW optical nuclei or associated with the
separation is 2.2±0.5 kpc.
X-ray feature offset to the SW of the SE optical nucleus
In projection, the spatial separation between the dou-
(Comerford et al. 2009b; Civano et al. 2010, 2012).
bleemission-linecomponents(2.2±0.5kpc)isconsistent
Applying a K-correctionfor a spectralindex α=−0.7
with the spatial separation between the double nuclei
(S ∝ να), the VLA nondetections in Figure 2 imply
in the HST image (2.50±0.05 kpc), to within their er-
that on scales of 1 θ = 0.32′′ (1.6 kpc) to 10 θ = 3.2′′ rors. There is a 3.7◦ difference between the PA of the
(16 kpc) the merger remnant has a spectral luminosity
DEIMOS slit and the PA of the two nuclei in the HST
L9.0 GHz <7.3×1021 W Hz−1. Given these sub-galactic image, but this results in a negligible (0.2%) difference
scales, we adopt local indicators of star formation rates between the spectrum-based and image-based measure-
(SFRs) (e.g., Calzetti 2012). We apply equation (15) ments of spatial separations. These similar spatial sepa-
of Murphy et al. (2011) at a rest-frame frequency ν = rations,combinedwiththeexcitationstateofthenarrow
9.0(1+z) GHz = 12.2 GHz, include a thermal contribu- emission-line components,are takenas strong indicators
tion for an electron temperature Te =104 K and a non- for the presence of two AGNs, likely associated with SE
thermalcontributionwithaspectralindexα=−0.8,and and NW.
findaSFR<17M⊙ yr−1foraKroupainitialmassfunc- The exposure time of the DEIMOS spectrum was too
tion(IMF).Thisisaninterestingregimetobeexploring, shorttoallowustolocalize,withconfidence,eachnarrow
giventhatthehostgalaxiesofoptically-selectedAGNsat emission-line component with each optical nucleus. For
z <0.3couldhaveSFR∼1-20M⊙yr−1,spanningvalues nowwemaketheplausibleassumptionthat[OIII] λ5007
for normal spiral galaxies to those for moderately lumi- luminosities L([OIII])=4.8−9.9×1042 erg s−1 and ex-
nous starbursts (Kim et al. 2006; Condon et al. 2013). citation ratios [OIII]λ5007/Hβ = 3.4−4.4 for the two
Prior VLA imaging at ν = 1.4 GHz (Schinnerer et al. AGNs (Comerford et al. 2009b). In the near future we
2010) localized the emission to the circular and ellipti- will obtainKeck/OSIRISdata to test this important as-
cal regions, outlined in Figures 1 and 2, that encom- sumption (F. Muller-Sanchez, priv. comm.).
passtheinnermergerremnant,theopticalnucleiSEand
NW, and the X-ray feature SW of the SE nucleus. The 3.2. The SE Optical Nucleus and its SW X-ray Feature
circular region marks a point-like source with an inte-
grated flux density S1.4 GHz = 0.113± 0.010 mJy and Civano et al.(2010,2012)providestrongevidencethat
a K-corrected spectral luminosity L1.4 GHz ∼ 4.6×1022 the SE optical nucleus is an unobscured AGN of Type
W Hz−1. If driven by star formation only, the corre- 1, thus featuring a broad emission-line region. SE has
sponding SFR = 34 M⊙ yr−1 (Murphy et al. 2011, eqn. anX-raycounterpartwithaluminosity L(2−10keV)=
15). This is at least a factor of two higher than the new 1.14×1043ergs−1(Civano et al.2010,2012). ItsEdding-
VLA limit derivedabove,meaning thatup to halfof the tonratiois 0.04(Civano et al. 2010;Trump et al.2011),
integrated flux density could arise from star formation. typical for a Type 1 AGN in COSMOS (Trump et al.
Conversely, any AGN-driven emission could have an in- 2009). The new VLA and VLBA nondetections of SE
tluemgriantoesditSy1.ν4LGνH(z1.∼4 G0.H05z6)−=0.(131−3m6)Jy×a1n0d38aeKrg-cos−rr1e,ctthede immipnloysitthieastνaLnyν(A9.G0NG-dHrziv)e<ne6m.6is×sio1n03h8aesrKg-sc−o1rr(e3ctσed)loun-
regime of low-luminosity AGN (LLAGN) at z = 0 (Ho scales below 1.6 kpc and νLν(1.52 GHz) < 4.7×1038
2008). ergs s−1 (6 σ) on scales below 75 pc. From § 3.1, the
Analyses of the HST F814W ACS image and the luminosity of any AGN-driven emission from SE cannot
Keck/DEIMOSspectrum of the galaxy were reported in exceed νLν(1.4 GHz)=6×1038 erg s−1.
Comerford et al.(2009b),andcanbesummarizedasfol- Following Terashima & Wilson (2003), the new VLA
lows. UsingSourceEXtractorontheHSTimage,theSE photometry for SE yields log νLν(9.0 GHz)/L(2 −
and NW optical nuclei were found to be aligned along a 10 keV) < −4.2. This value is consistent with SE’s
PA=139.6◦,with aprojectedseparationof500±9mas modest Eddington ratio (Ho 2008). From § 3.1, the
(2.50±0.05kpc). TheDEIMOSspectrumwastakenwith [OIII] λ5007 luminosity of SE is L([OIII]) = 4.8 −
a slit PA = 143.3◦ and spanning 4730–9840˚A in wave- 9.9×1042 erg s−1 (Comerford et al. 2009b). The VLBA
length. This spectrum revealeddouble narrowemission- luminosity, free from contamination from star forma-
4 Wrobel, Comerford & Middelberg
tion, thus implies that SE is radio quiet as defined by The NW optical nucleus could harbor an intrinsically
Zakamska et al. (2004) or Trump et al. (2011, 2013), a faint AGN, with a narrow emission-line region only.
trait that is also consistent with its modest Eddington However, evidence for a very obscured Type 2 AGN
ratio. in NW is beginning to emerge. Recent prescriptions
For context, several key properties of SE resemble or for identifying obscured AGNs (Juneau et al. 2011) can
are consistent with those of the prototypical Type 1 be applied. From § 3.1, the excitation ratio of NW is
AGN NGC4151. Adopting a distance of 13.3 Mpc for [OIII]λ5007/Hβ = 3.4 − 4.4 (Comerford et al. 2009b).
NGC4151, it has (a) L(2− 10 keV) = 5.0 × 1042 erg Sucha ratio,whencombinedwith estimates for the host
s−1 (Ho 2009); (b) when integrated within a radius galaxy’s stellar mass M⋆ ∼ 1010−11M⊙ (Civano et al.
of 2′′ (130 pc), νLν(1.425 GHz) = 1.0×1038 erg s−1, 2010; Kartaltepe et al. 2010), occupy the AGN locus of
νLν(8.465 GHz) = 1.4×1038 erg s−1 and α = −0.75 themass-excitationdiagram(Fig.4;Juneau et al.2011).
Moreover, as Civano et al. (2012) suggest, NW could
(Kukula et al. 1995; Ho & Ulvestad 2001); and (c) log
νLν(8.465 GHz)/L(2−10 keV) = −4.6. Mundell et al. lack an X-ray counterpart due to atomic absorption.
Comparing NW’s Compton-thickness parameter T =
(2003) and Wang et al. (2011a) reportevidence that the
L(2−10 keV)/L([OIII]) to its [OIII] λ5007 luminosity
radio source dynamically interacts with the interstellar
(Comerford et al. 2009b; Civano et al. 2012) identifies it
medium (ISM) at these radii in NGC4151.
as a Compton-thick candidate (Fig. 10; Juneau et al.
Civano et al. (2012) noted that the X-ray feature
2011).
offset by about 500 mas (2.5 kpc) to the SW of the
NW could also lack a counterpartat ν = 9.0 GHz due
SE optical nucleus could be driven by star formation.
to free-free absorption by its narrow emission-line gas,
Indeed, de La Fuente Marcos & de La Fuente Marcos
a concept applied to account for the flat radio spectra
(2008) suggest that the impulse from the passage of a
of some Type 2 AGN with L([OIII]) = 1042 erg s−1
recoiling black hole could could trigger star formation
(Lal & Ho 2010). For a characteristic electron temper-
in its wake. In the scenario where SE has recoiled from
NW, the SW X-ray feature might be its star formation ature Te = 104 K and density ne = 103 cm−3, a path
wake. That feature is about 18% as luminous as SE in length of about 380 pc gives a free-free optical depth
the passband of the Chandra High Resolution Camera of unity at ν = 10 GHz (Ulvestad & Ho 2001). Intrigu-
(HRC),indicatingthatithasL(2−10keV)=2.1×1042 ingly,thisfree-freepathlengthresemblesthescalelength
erg s−1 (Civano et al. 2010, 2012). of about 100 mas (500 pc) reported for NW’s starlight
From§3.1,thenewVLAupperlimittotheSFRatthe (Civano et al. 2010). This path length can be cross-
location of the SW X-ray feature is SFR < 17 M⊙ yr−1 checked in several ways. First, the Hβ luminosity for
for a Kroupa IMF. This correspondsto L(2−10 keV)< either the blueshifted or redshifted narrow emission-line
8.5 × 1040 erg s−1 according to Ranalli et al. (2003), gas(Comerford et al.2009b)shouldn’tbe exceeded;this
can be achieved by invoking a plausible volume-filling
who use a Salpeter IMF. Even with this caveat of dif-
factor f ∼1−3×10−3. Second, if the narrowemission-
ferent IMFs, the X-ray emission from the SW feature
line gashas a sphericalgeometrywith diameter 150mas
is much too luminous to be driven by star formation.
(2 × 380 pc), the source at ν = 9.0 GHz should be
This points to the SW X-ray feature being AGN-driven,
much smaller; this can be checked with sensitive VLBA
perhaps, as Civano et al. (2012) suggest, because SE
imaging. Third, at VLBA resolutions the spectrum of
photoionizes the ambient ISM on kiloparsec scales and
the source at frequencies ν < 9.0 GHz should be expo-
produces optical and X-ray emission lines analogous to
the situation at radii beyond 2′′ (130 pc) in NGC4151 nentially suppressed; this is also testable with sensitive
VLBA imaging.
(Wang et al.2011b). Notably,SE’s[OIII]λ5007luminos-
Civano et al.(2012)assembledthespectralenergydis-
ity is about 37-76 times that of NGC4151 on kpc scales
tribution at far-infrared and shorter wavelengths, cor-
(Ho & Ulvestad 2001), perhaps because of SE’s location
recteditfor the contributionsofthe Type1 AGNinSE,
in a merger remnant rather than in a typical LLAGN
andinferredaninfrared(IR)luminosityL(IR)∼6×1044
host.
ergs−1. ThatluminositycorrespondstoaKroupa-based
3.3. The NW Optical Nucleus SFR ∼ 23 M⊙ yr−1 (Murphy et al. 2011, eqn. 4), al-
thoughthisvaluewouldbeoverestimatedifdustwarmed
The new VLA and VLBA nondetections of NW im-
by an obscured Type 2 AGN in NW contributed to the
ply that any AGN-driven emission has K-corrected lu-
minosities νLν(9.0 GHz) < 6.6×1038 erg s−1 (3 σ) on I2R013lu).minForosimty§(e3.g.1.,, dtheeGnriejwp eVtLaAl. 1u9p9p2e;rJluimneitautoettahle.
scales below 1.6 kpc and νLν(1.52 GHz) < 4.7×1038 SFRatthe locationofthe NW opticalnucleus is SFR<
ergs s−1 (6 σ) on scales below 75 pc. From § 3.1, the 17 M⊙ yr−1. This lower radio-based SFR suggests that
luminosity of any AGN-driven emission from NW can-
theIRphotometrycouldsuffersomecontaminationfrom
not exceed νLν(1.4 GHz) = 6 × 1038 erg s−1. From dust warmed by an obscured AGN.
§ 3.1, the [OIII] λ5007 luminosity of NW is L([OIII])=
4.8−9.9×1042 erg s−1 (Comerford et al. 2009b). The 4. SUMMARYANDCONCLUSIONS
VLBAluminosity,freefromcontaminationfromstarfor- We presented new radio imaging of J1000+0206, a
mation,thusimpliesthatNWisradioquietfollowingthe merger remnant at z = 0.36 with two optical nuclei,
Zakamska et al. (2004) definition. The X-ray emission NW and SE, offset by 500 mas (2.5 kpc). The new VLA
from the NW optical nucleus is less than 4.5% as lumi- imagingatν =9.0GHzissensitivetoemissiondrivenby
nous as SE in the HRC passband, suggesting that NW AGNs and/or star formation, and the new VLBA imag-
has L(2−10 keV) < 5.1×1041 erg s−1 (Civano et al. ing at ν = 1.52 GHz is sensitive only to emission driven
2010, 2012). by AGNs. No radio emission was detected at these fre-
Two AGNs in Merger Remnant COSMOSJ100043.15+020637.2 5
quencies. Combiningnewandpriorresults,thefollowing In the scenario where SE is an AGN that has recoiled
self-consistent picture emerged: from NW due to the asymmetric emission of gravita-
tional waves during black-hole coalescence, NW cannot
• The new VLA photometry at ν =9.0 GHz implies hostanobscuredAGN.Theemergingevidenceforavery
aSFR<17M⊙ yr−1fortheinnermergerremnant, obscuredType2AGNinNWthusbothweakensthecase
the optical nuclei SE and NW, and the X-ray fea- for SE being a gravitational-waverecoil and strengthens
ture adjacent to the SE nucleus. After correcting the case that each optical nucleus contains it own AGN.
the priorVLA photometry at ν =1.4GHz for this Two interpretations are consistent with the latter sce-
SFR, the AGN-driven emission has an integrated nario: (i) SE and NW mark inspiralling dual AGNs like
luminosity νLν(1.4 GHz)=(3−6)×1038 erg s−1, thosesoughtinsystematicspectroscopicsurveysfordual
in the realm of LLAGN in the local Universe. AGN candidates (Comerford et al. 2009a; Wang et al.
2009;Liu et al.2010;Smith et al.2010;Koss et al.2012;
• ThepropertiesofSEanditsadjacentX-rayfeature
Barrows et al. 2013; Comerford et al. 2013), or (ii) SE
match those of the unobscured prototype AGN in
is a gravitational-slingshotrecoilmovingawayfromNW
NGC4151. However, the luminosity of the narrow
andtheType2AGNthatitharbors(Civano et al.2010,
emission-line gas associatedwith SE is about forty
2012; Blecha et al. 2013a).
to eighty times higher than that associated with
It must be noted that SE’s broad Hβ emission line is
the AGN in NGC4151. Deeper radio, optical and redshiftedbymorethan1000kms−1 relativetothenar-
X-ray studies of SE will test its resemblance to,
row Hβ emission line (Civano et al. 2010). Such large
and difference from, the Type 1 AGN prototype
velocity offsets occur in less than 1 % of normal Type 1
in NGC4151. For example, the difference in nar-
AGN (Bonning et al. 2007; Eracleous et al. 2012), mak-
row emission-line luminosities could reflect atypi-
ing it very unlikely that they will occur by chance in
cal, merger-induced conditions near SE.
an inspiralling Type 1 AGN. In contrast, such large ve-
locity offsets are predicted for a gravitational-slingshot
• The properties of NW are consistent with it host-
recoil (Hoffman & Loeb 2007), a point in favor of that
ing a Compton-thick AGN that warms the ambi-
interpretation for J1000+0206.
entdustandphotoionizesthe narrowemission-line
We thank the referee for a timely and helpful report.
gas, and is free-free absorbed by that gas at ν =
WealsoacknowledgeusingNedWright’sCosmologyCal-
9.0 GHz. These findings enhance the prospects
culator (Wright 2006) and benefiting from helpful dis-
that AGN-driven emission could emerge from ra-
cussionswithMichaelCooper,FranciscoMuller-Sanchez,
dio and X-ray studies of NW at deeper levels and
Jonathan Trump and Craig Walker. EM acknowledges
X-ray studies of NW at harder wavelengths. For
financial support from the Deutsche Forschungsgemein-
example, deeper VLBA imaging could reveal a ra-
schaft through project FOR1254.
dio spectrum that is exponentially suppressed by
Facilities: VLA, VLBA.
free-freeabsorption. Thesefindingsalsounderscore
the urgency of localizing narrow emission-line gas,
whether blueshifted or redshifted, to NW.
REFERENCES
Amaro-Seoane,P.,Sesana,A.,Hoffman,L.,etal.2010,MNRAS, deLaFuenteMarcos,R.&deLaFuente Marcos,C.2008,ApJ,
402,2308 677,L47
Barrows,R.S.,SandbergLacy,C.H.,KennefickJ.,etal.2013, Deller,A.T.,Brisken,W.F.,Phillips,C.J.,etal.2011,PASP,123,
ApJ,769,95 275
Blecha,L.,Cox,T.J.,Loeb,A.,&Hernquist,L.2011, MNRAS, Dotti,M.,Sesana,A.,&Decarli,R.2012,Advances in
412,2154 Astronomy,940568
Blecha,L.,Civano,F.,Elvis,M.,&Loeb,A.2013a, MNRAS, Elvis,M.,BAAS,41,708
428,1341 Eracleous,M.,Boroson,T.A.,Halpern,J.P.,&Liu,J.2012,
Blecha,L.,Loeb,A.,&Narayan,R.2013b, MNRAS,429,2594 ApJS,201,23
Beswick,R.J.,Pedlar,A.,Mundell,C.G.,&Gallimore,J.F.2001, Fu,H.,Myers,A.D.,Djorgovski,S.G.,&Lin,Y.2011a,ApJ,733,
MNRAS,325,151 103
Bonning,E.W.,Shields,G.A.,&Salviander,S.2007,ApJ,666, Fu,Z.Zhang, Assef,R.J.,etal.2011b,ApJ,740,L44
L13 Fu,H.,Lin,Y.,Myers,A.D.,etal.2012, ApJ,745,67
Calzetti,D.2012,arXiv:1208.2991v1 Gallimore,J.F.,&Beswick,R.2004,AJ,127,239
Civano,F.,Elvis,M.,Lanzuisi,G.,etal.2010,ApJ,717,209 Garrett,M.A.,Wrobel,J.M.,&Morganti,R.2005,ApJ,619,105
Civano,F.,Elvis,M.,Lanzuisi,G.,etal.2012,ApJ,752,49 Greisen,E.W.2003,inInformationHandlinginAstronomy,ed.
Comerford,J.M.,Gerke,B.F.,Newman,J.A.,etal.2009a, ApJ, A.Heck(Dordrecht:Kluwer),109
698,956 Guedes,J.,Madau,P.,Mayer,L.,&Callegari,S.2011,ApJ,729,
Comerford,J.M.,Griffith,R.L.,Gerke,B.F.,etal.2009b,ApJ, 125
702,L82 Ho,L.C.2008,ARA&A,46,475
Comerford,J.M.,Pooley,D.,Gerke,B.F.,&Madejski,G.M. Ho,L.C.2009,ApJ,699,626
2011, ApJ,737,L19 Ho,L.C.,&Ulvestad,J.S.2001,ApJS,133,77
Comerford,J.M.,Gerke,B.F.,Stern,D.,etal.2012,ApJ,753,42 Hobbs,G.,Archibald,A.,Arzoumanian,Z.,etal.2010,Class.
Comerford,J.M.,Schluns,K.,Greene,J.E.,&Cool,R.J.2013, Quantum Grav.,27,084013
ApJ,inpress,arXiv:1309.2284v1 Hoffman,L.&Loeb,A.2007,MNRAS,377,957
Condon,J.J.1992, ARA&A,30,575 Juneau,S.,Dickinson,M.,Alexander,D.M.,&Salim,S.2011,
Condon,J.J.,Kellermann,K.I.,Kimball,A.E.,&Perley,R.A. ApJ,736,104
2013, ApJ,768,37 Juneau,S.,Dickinson,M.,Bournaud,F.,etal.2013,ApJ,764,
deGrijp,M.H.K.,Keel,W.C.,Miley,G.K.,Goudfrooij,&Lub,J. 176
1992, A&AS,96,389
6 Wrobel, Comerford & Middelberg
Kartaltepe,J.S.,Sanders,D.B.,LeFloc’h,E.,etal.2010,ApJ, Perley,R.A.,&Butler,B.J.2013,ApJS,204,19
721,98 Ranalli,P.,Comastri,A.,&Setti,G.2003,˚a,399,39
Kim,M.,Ho,L.C.,&Im,M.2006,ApJ,642,702 Rosario,D.J.,Shields,G.A.,Taylor,G.B.,etal.2010,ApJ,716,
Komassa,S.,Burwitz,V.,Hasinger,G.,etal.2003,ApJ,582,L15 131
Koss,M.,Mushotzky, R.,Teiester,etal.2012, ApJ,746,L22 Schinnerer,E.,Smolcic,V.,Carilli,C.,etal.2007, ApJS,172,46
Kukula,M.,Pedlar,A.,Baum,S.A.,&O’Dea,C.P.1995, Schinnerer,E.,Sargent,M.T.,Bondi,M.,etal.2010, ApJS,188,
MNRAS,276,1262 384
Kulkarni,G.&Loeb,A.2012, MNRAS,422,1306 Scoville,N.,Abraham,R.G.,Aussel,H.,etal.2007,ApJS,172,38
Lal,D.V.,&Ho,L.C.2010, AJ,139,1089 Shen,Y.,Liu,X.,Greene,J.E.,&Strauss,M.A.2011,ApJ,735,
Lilly,S.J.,LeFevre,O.,Renzini,A.,etal.2007,ApJS,172,70 48
Liu,X.,Shen, Y.,Strauss,M.A.,&Greene,J.E.2010,ApJ,708, Smith,K.L.,Shields,G.A.,Bonning,E.W.,etal.2010,ApJ,716,
427 866
Liu,X.,Shen, Y.,&Strauss,M.A.2012,ApJ,745,94 Smolcic,V.,Schinnerer,E.,Scodeggion, M.,etal.2008,ApJS,
McMullin,J.P.,Waters,B.,Schiebel,D.,Young, W.,&Golap, 177,14
K.2007, AstronomicalDataAnalysisSoftwareandSystems Terashima,Y.,&Wilson,A.S.2003,ApJ,583,145
XVI(ASPConf.Ser.376),ed.R.A.Shaw,F.Hill,&D.J.Bell Trump,J.R.,Impey,C.D.,Kelly,B.C.,etal.2009,ApJ,700,49
(SanFrancisco,CA:ASP),127 Trump,J.R.,Impey,C.D.,Kelly,B.C.,etal.2011,ApJ,733,60
Middelberg,E.,Deller,A.T.,Norris,R.P.,etal.2013,˚a,551,A97 Trump,J.R.,Impey,C.D.,Kelly,B.C.,etal.2013,erratum
Mundell,C.G.,Wrobel,J.M.,Pedlar,A.,&Gallimore,J.F.2003, Ulvestad,J.S.,&Ho,L.C.2001,ApJ,558,561
ApJ,583,192 VanWassenhove, S.,Volonteri,M.,Mayer,L.,etal.2012,ApJ,
Murphy,E.J.,Condon,J.J.,Schinnerer,E.,etal.2011, ApJ,737, 748,L7
67 Wang,J.,Chen,Y.,Hu,C.,Mao,W.,&Bian,W.2009,ApJ,
Napier,P.J.,Bagri,D.S.,Clark,B.G.,etal.1994, Proc.IEEE, 705,L76
82.658 Wang,J.,Fabbiano,G.,Elvis,M.,etal.2011a,ApJ,736,62
Owen,F.N.,O’Dea,C.P.,Inoue, M.,&Eilek,J.A.1985, ApJ, Wang,J.,Fabbiano,G.,Elvis,M.,etal.2011b,ApJ,742,23
294,L85 Wright,E.L.2006,PASP,118,1711
Perley,R.A.,Chandler,C.J.,Butler,B.J.,&Wrobel,J.M. Zakamska,N.L.,Strauss,M.A.,Heckman, T.M.,Ivezic,Z.,&
2011, ApJ,739,L1 Krolick,J.H.2004,AJ,128,1002