A ring galaxy at z=1 lensed by the cluster Abell 370

5 Pages·0.28 MB·English

by  G. Soucail

A&A manuscript no. ASTRONOMY
(will be inserted by hand later) AND
Your thesaurus codes are: ASTROPHYSICS
1 (11.09.1 LRG J0239–0134 ; 11.05.2; 11.09.2; 11.19.1; 12.03.3; 12.07.1)
z ⋆
A ring galaxy at = 1 lensed by the cluster Abell 370
G. Soucail1, J.P. Kneib1, J. B´ezecourt1,2, L. Metcalfe3, B. Altieri3, and J.F. Le Borgne1
1 Observatoire Midi-Pyr´en´ees, Laboratoire d’Astrophysique,UMR 5572, 14 AvenueE. Belin, F-31400 Toulouse, France
2 Kapteyn Institute,Postbus 800, 9700 AV Groningen, The Netherlands
3 ISODataCentre,AstrophysicsDivision,SpaceScienceDepartmentofESA,VillafrancadelCastillo, POBox50727, E-28080
9 Madrid, Spain
9
9 Received 12 January 1999 / Accepted 26 January 1999
1

n
a Abstract. We present a study of a very peculiar object as natural “gravitational telescopes” to address several
J foundinthefieldofthecluster-lensAbell370.Thisobject astrophysical problems related to the properties and na-

6 displays,inHSTimaging,aspectacularmorphologycom- ture of high redshift galaxies. Morphological properties
2
parabletonearbyring-galaxies.Fromspectroscopicobser- of distant lensed sources were first addressed by Smail

vationsattheCFHT,wemeasuredaredshiftofz =1.062 et al. (1996) who estimated the intrinsic linear sizes of
2
v basedontheidentificationof[Oii]3727˚Aand[Nev]3426˚A the galaxies and showed them to be compatible with
3 emissionlines.Theseemissionlinesaretypicalofstarburst a significant size evolution with redshift. More recently,
2 galaxies hosting a central active nucleus and are in good several detailed morphological analysis of high-z arcs
1
agreement with the assumption that this object is a ring were proposed using lens modeling. Most of them ap-
1
galaxy. This object is also detected with ISO in the LW2 pear knotty with a more complex morphology than their
0
9 and LW3 filters, and the mid Infra-Red (MIR) flux ratio local counterparts (Colley et al. 1996; Franx et al. 1997;
9 favors a Seyfert 1 type. The shape of the ring is gravita- Pell´o et al. 1999) although there may be some biases due
/ tionally distortedby the cluster-lens,andinparticularby to observations in the UV restframe.
h
p a nearby cluster elliptical galaxy. Using our cluster mass Inthisletter,westudyapeculiarobjectdetectedinthe
- model,wecancomputeitsintrinsicshape.Requiringthat fieldofthegalaxyclusterA370.OntheHST/WFPC2 im-
o
the outer ring follows an ellipse we constrainthe M/L ra- ages,it displaysanunusualmorphologysimilar to nearby
r
t tio ofthe nearbygalaxyandderivea magnificationfactor ring-galaxies, although gravitationally distorted by the
s
a of 2.5 ± 0.2. The absolute luminosities of the source are cluster.Sect.2summarisesthe variousobservationsrelat-
v: then LB = 1.3 1012LB⊙ and ν Lν ≃ 4.1010 L⊙ in the ingtothissource:HSTimaging,spectroscopicandphoto-
i mid-IR. metric data including mid-IR photometry obtained with
X
ESA’sISOspacecraft(Kessler et al. 1996).Thesourcere-
r Key words: Galaxies: individual: LRG J0239–0134 – construction of the ring and a morphological analysis is
a
Galaxies: evolution – Galaxies: interactions – Galaxies: presented in Sect. 3 with a robust estimate of the lens
Seyfert – Cosmology:observations – gravitationallensing amplification. Sect. 4 discusses the spectral energy distri-
bution (SED) of the object. Discussion and conclusions
are given in the last section.
Throughoutthe paper,we considera Hubble constant
1. Introduction
of H0 =50 h50 km s−1 Mpc−1, with Λ=0 and Ω=1.
A useful property of gravitational lensing is the magni-
fication of background objects: the gain in spatial reso-
lution allows the morphological properties of distant and 2. Observations
resolved objects to be probed in greater detail and the
2.1. HST imaging
gain in apparent flux allows fainter sources than would
otherwise either be detected or studied to be probed sta- InthedeepF675WWFPC2 imagedescribedinB´ezecourt
tistically. Massive clusters of galaxies can then be used et al. (1999a, hereafter Paper I), a spectacular distorted
ring is detected close to a bright cluster elliptical (#32
Send offprint requests to: G. Soucail, [email protected]
in the numbering scheme of Mellier et al. (1988) with
⋆ Based on observations with the NASA/ESA Hubble Space
z = 0.370). The object displays a clearly resolved central
Telescope obtained from the data archive at the Space Tele-
bulgesurroundedbya∼4.8′′-diameterdistortedringanda
scope European Coordinating Facility, with ISO; an ESA
project with instruments funded byESA Member States with secondary1.5′′-diameterinnerring(Figure1).Thisobject
the participation of ISAS and NASA, and with the Canada- is very similar in aspect to the Cartwheel galaxy (Struck
France-Hawaii Telescope at Mauna Kea, Hawaii, USA. et al.1996),with similar ratiobetweenthe tworing radii.
2 G. Soucail et al.: A ring galaxy lensed by A370
Fig.1. Multi-colour images of the ring galaxy and its surroundings: Left: HSTWFPC2 F675W image (with a 12′′
size) – Center: isocontours of the HST F336W (U-band) image overplotted on the F675W image – Right: ISO LW3
contours overplotted on the F675W image. Galaxy # 32 has been subtracted for clarity. Note the accurate matching
of the ISO source with the nucleus. North is top, East is left.
We used the APM catalogue to measure the astrometry
of the field. The absolute coordinates of the lensed ring
galaxy are: αJ2000 = 2h 39m 56.51s, δJ2000 = −1◦ 34′
25.66′′ (witha0.2′′ rmsaccuracy).Thereforewereference
hereafter this object as: LRG J0239–0134.
A photometric analysis was performed on the F675W
HSTimage.First,the nearbygalaxy#32wassubtracted
after a radial fit of the isophotes with the “ELLIPSE”
package in the IRAF/STSDAS environment. The total
integrated flux for LRG J0239–0134 gives a magnitude
R675W = 20.7±0.1 that can be separated into the emis-
sion of the central part R675W =21.5±0.1 and the outer
ringcontributionR675W =21.4±0.2(lessaccuratedueto
larger uncertainties in the local sky background). In the
HST/WFPC2 U-Bandimage(seeB´ezecourtetal.1999b)
the ring-like object is also detected and appears less cen-
trally concentrated than in F675W. However the lower
signal-to-noise prevents a detailed morphologicalanalysis
oftheextendedemission(Figure1)althoughtheremaybe
some inhomogeneities in the ring due to hot spots similar
to that seen in the Cartwheel. Fig.2. The spectrum ofthe ring galaxysuperimposedon
asyntheticspectrumofalate-typespiralredshiftedtoz =
1.062.Thedataweresmoothedto fitthe resolutionofthe
2.2. Spectroscopic and photometric observations
spectrographand the most prominentlines are identified.
SpectroscopicdatawereobtainedduringaCFHTrunwith Above is the original 2D spectrum after sky subtraction,
the OSIS-V instrument (Le F`evre et al. 1994) in August showing in particular the spatial extent of the emission
1997.A 1′′-wide long-slitwas positionedthroughthe cen- line at 7685˚A.
tralbulgeandthering(slitorientationwasEast-Westlim-
iting the contamination by the envelope of galaxy #32).
We used the 2K×2K Loral thinned CCD and the grism 2). Several features are detected: a strong emission line
R150 which gives a dispersion of 6˚A/pixel, with a resolu- at λ = 7685±3˚A, a weaker one at λ = 7064±9˚A and
tion of 18˚A across the wavelength range 5000 to 9000˚A. an absorption line at λ = 5773±3˚A, all visible on the
Two 1.8 ksec exposures were obtained just before morn- 2D spectrum. We unambiguously identify those lines as:
ing twilight. The data were reduced with standard pro- [Oii] 3727˚A,[Nev] 3426˚A and Mgii 2800˚A giving a red-
cedures for flat-fielding, wavelength and flux calibration. shift of z =1.062±0.001 for the source. Moreover,a ten-
Sky subtraction was performed on the 2D image (Figure tative identification at this redshift suggests weak emis-
3
Table 1. Multi-wavelength fluxes of the ring galaxy (see
text for more details). These fluxes are not corrected for
the gravitationalmagnification of the source.
Filter λ (µm) Magnitude Flux (µJy) Resolution
U336W 0.336 21.00 ±0.2 4.45±0.9 0.1 ′′
Bj 0.450 21.81 ±0.2 6.36±1.2 1.1 ′′
R 0.646 20.38 ±0.4 21.3±9.4 1.1 ′′
R675W 0.673 20.66 ±0.1 15.8±1.5 0.1 ′′
I 0.813 19.31 ±0.2 45±9 0.7 ′′
J 1.237 17.82 ±0.2 118±24 1.2 ′′
K′ 2.103 16.34 ±0.2 212±42 1.2 ′′
LW2 6.7 750±225 6 ′′
LW3 14.3 1800±540 6 ′′
sion from the [Neiii] 3869˚A line though the spectrum is
Fig.3.ContourplotoftheestimatorEforthegalaxy#32
contaminatedby sky residuals and Hδ and Hε absorption
when fitting the outer ring to be projected as an ellipse
lines. Redder lines fall outside the visible range and un-
inthe sourceplane.The dottedline correpondstomodels
luckily, Hα lies at 1.353µm, a wavelength domain hardly
detectablefromtheground(outsidetheJ-band).Further- with constant M(<a32)/L.
more, the fit of the continuum with a typical late-type
spiral galaxy is satisfactory.Finally, the [Oii] line is more
suretimeof16.1ksecinthedeepestpartoftheimage.The
spatiallyextended thanthe underlyingstellar continuum,
datawerereducedfollowingtwosubstantiallyindependent
and a contribution from the ring itself is likely, which is
methods: a Multi-resolution Median Transform method
notthecaseforthe[Nev]line.Unfortunately,noinforma-
(PRETI, Starck et al. 1998), and the Vilspa method de-
tion on any velocity field can be provided with this low scribedinAltierietal.(1998).ThecenteringofISOmaps
S/N ratio spectrum.
on optical data was obtained by correlating the positions
Multi-bandphotometryofthisgalaxyisavailablefrom of 7 sources, giving a final astrometric accuracy of 1′′.
alargesetofdataexistingontheclusterAbell370.Inad- The correspondancebetweenthe ringgalaxyandthe ISO
dition to the HST U and R images, we used B, R and I source is quite secure (Figure 1). The photometric accu-
magnitudes from deep CFHT images (Kneib et al. 1994). racyachievableforsuchfaintmid-IRsources,allowingfor
Near-IR data were obtained with the Redeye camera at alltheuncertaintiesinthedatareduction,isaround30%.
CFHT in August 1994. The final near-IR images corre- In both ISO bandpasses, the ring galaxy is the bright-
spondtodeepexposureswithintegrationtimesof3.4ksec est extragalactic source in the field covered by ISOCAM,
in J and 7.2 ksec in K′ in good observing conditions. The apart from the giant arc.
details of this photometric setup will be published else-
where. The photometry of the ring galaxy is summarised
3. Source reconstruction of the lensed ring galaxy
in Table 1. From this multi-band photometry, a photo-
metric redshift of 0.95±0.1 was found (R. Pell´o, private This ring galaxy is magnified and distorted by the grav-
communication) confirming the spectroscopic identifica- itational shear induced by the cluster and the nearby el-
tion through a fit to the continuum shape. In addition, liptical cluster galaxy #32. This is particularly visible in
the fainter object close to the ring (labelled A in Figure the distorted shape of the outer ring. In order to recon-
1) has a photometric redshift zphot = 0.4 ± 0.1 and is structits true intrinsic size,shape anddetermine its total
therefore not related to the ring galaxy. amplification, we traced the rays back through the lens
into the source plane, using the model proposed in Pa-
2.3. ISO data perI.Thismodelwasoptimisedfromthe identificationof
several multiple images detected on the HST images and
The ring galaxy was detected with the ISO camera (ISO- takesintoaccountthemassivehalosofthebrightestgalax-
CAM, C´esarsky et al., 1996), as part of a programme of iesusingstandardscalinglaws(e.g.Kneib et al. 1996)re-
imaging through gravitational lensing clusters (Metcalfe latedtotheirluminositieswithscalingdependancesimilar
et al. 1999). A370 was deeply imaged on a wide 7′ ×7′ to the properties of the Fundamental Plane:
field, in micro-scanning mode, with the 3′′ pixel-field-of-
view, and using two broad-band, high-sensitivity filters: L 14 L 0.8 L 12
LW2 (5–8.5µm) and LW3 (12–18µm) with a total expo- σ0 =σ0∗(cid:18)L∗(cid:19) ; rt =rt∗(cid:18)L∗(cid:19) ; r0 =r0∗(cid:18)L∗(cid:19) .
4 G. Soucail et al.: A ring galaxy lensed by A370
Fig.5.Thespectralenergydistributionoftheringgalaxy.
Thedatacorrespondtotheobservedphotometryfrom0.3
Fig.4. Image-reconstruction of the ring galaxy in the µm to 15 µm. The curve shows the synthetic SED of an
sourceplaneusingthebestfitmodelparameters.Thebot- Sc galaxy from Fioc and Rocca-Volmerange (1996). The
tom part of the ring was not included in the fitting. This mid-IR excess is clearly non-stellar.
reconstruction suggests that this object does not belong
to the outer ring.
the aperture a32. Fixing rt=42.5 h−501kpc, the best value
for the velocity dispersion is σ0 = 220 ±20 km/s, which
In particular for the elliptical galaxy #32 (R675W = corresponds to a correction factor of 1.4 for the aperture
18.47) this means that LR = 2.2 1011 h−502 L⊙ which mass leading to M(<a32)/L = 8.0 ±1.5 (M/L)⊙.
translates to σ0 = 173 km/s, rt = 42.5 h−501 kpc (or 6.8′′ The magnification factor of the source has a mean
at the cluster redshift) and a total M/L ratio of 6.4 h50 value of 2.5±0.2 but ranges from 3.6 to 2.1 depending on
(M/L)⊙ for this galaxy and its halo. Within the aperture thedistancetogalaxy#32(Figure4).Theintrinsicradius
a32 = 26.8 h−501 kpc defined by the distance from #32 to for the outer ring found is 7.7±0.4h−501 kpc, a value com-
the ring galaxy nucleus we have M(< a32)/L = 5.7 h50 parable with the characteristicsofnearby similar objects.
(M/L)⊙. The axis ratio of the ellipse is cosi = b/a = 0.76 that
We then examined the influence of the parameters translates in a inclination angle of i=40 degree assuming
(σ0,rt) in the source reconstruction. In order to quantify anintrinsiccircularring.TheB-bandabsoluteluminosity
this we identified in the image plane 20 points (xi,yi) de- of the source, corrected from the magnification and mea-
scribingtheouterring.Wethentunedthetwoparameters sureddirectlyfromtheImagnitude,isLB =1.31012LB⊙,
(σ0,rt) to check how close to anellipse the corresponding about 10 times brighter than the Cartwheel galaxy (Ap-
source points (xsi,ysi) are. For this purpose we defined pleton & Marston 1997).
the following estimator:
1 4. Spectral Energy Distribution
E(σ0,rt)= NΣNi=1|fσ0,rt(i)|
Joining together all the multi-wavelength photometric
with measurements, we can construct the SED of the ring
[(xs −x )cosθ−(ys −y )cosθ]2 galaxy over a large wavelength range (Fig. 5). This range
i c i c
fσ0,rt(i) = a2 covers both the rest-frame stellar emission and the IR
[(xs −x )sinθ+(ys −y )sinθ]2 flux emitted by the warm component of the interstellar
+ i c i c −1 medium. The significant excess of light emitted in the
b2
mid-IR with respect to the stellar contribution can orig-
where(x ,y ,a,b,θ)arethe parametersofthe ellipsethat inate from three possible sources: either a central active
c c
minimise,foreachsetof(σ0,rt),theestimatorE.Foraset nucleus heats the dust torus around it, or the warm dust
of points belonging to an ellipse, E is zero. Furthermore, is heated by a violent starburst induced by the merger
by definition, E is scale-invariantand does not depend on that produced the ring, or we are witnessing a combina-
theintrinsicsizeofthering.Figure4showsiso-contoursof tion of the two phenomena. A strong argument in favour
theestimatorE(σ0,rt).Theoriginalpointofthe modelof of an active nucleus is the detection in the optical spec-
paperIisindicatedasacross.Notsurprinsingly,theshape trum of the [Nev] emission line typical of Seyfert galax-
of the ring does depend strongly on the M/L ratio within ies. Moreover, the bulk of the optical emission is clearly
5
stellar, with Balmer absorption lines and a typical stel- Acknowledgements. WewishtothankRoserPell´oforherhelp
lar continuum. This suggests that a starburst is occur- in the estimations of photometric redshifts of the objects and
ring and may dominate the optical light. For this tar- Dani`eleAlloinforfruitfuldiscussionsaboutAGNopticalspec-
get in the mid-IR it is difficult to differentiate between tra.ManythankstoIanSmail,RobIvison,AndrewBlainand
JimHigdonforusefuldiscussionsandcomments.Thisresearch
the contribution from the nuclear emission and that from
has been partly conducted under the auspices of a European
the starburst (see the discussion of the Cartwheel galaxy
TMRnetworkprogrammemadepossibleviagenerousfinancial
in Charmandaris et al. 1998). The flux ratio LW3/LW2
support from theEuropean Commission
is about 2.4, or equivalently the spectral index is about
(http://www.ast.cam.ac.uk/IoA/lensnet/).
−1.15 for LRG J0239–1034.Taking into account the fact
that we observe the rest-frame fluxes at 3.3µm and 7µm,
this can be compared easily with the mid-IR spectra References
shownbySchultz etal.(1998)fordifferenttypes ofAGN.
Appleton P.N. & Marston A.P., 1997, AJ 113, 201
Our results clearly favour a Seyfert 1 type, in accordance Altieri B., Metcalfe L., Ott S. et al., 1998, “ISOCAM Faint
with the weak detection of the [Nev] emission line in the Sourcereport” (http://www.iso.vilspa.esa.es)
optical spectrum. B´ezecourt J., Kneib J.P., Soucail G., Ebbels, T.M.D., 1999a,
A&A,in press (astro-ph/9810199, Paper I)
B´ezecourt J., Soucail G., Ellis R.S., Kneib J.P., 1999b, in
preparation (paperII)
5. Discussion
C´esarsky C.J. et al., 1996, A&A315, L32
In this letter we analyse in details the nature of a pecu- Charmandaris V., et al., 1999, A&A 341, 69
Colley W.N., Tyson J.A., TurnerE.L., 1996, ApJ461, L83
largalaxybasedonmultiwavelengthobservations.Firstly,
Fioc M., Rocca-Volmerange B., 1996, A&A 326, 950
the outer ring of the galaxy constitutes good morpholog-
Franx M., Illingworth G.D., Kelson D.D., van Dokkum P.G.,
ical evidence for a starburst induced by a recent grav-
Tran K-V., 1997, ApJ486, L75
itational interaction, although the progenitor is not yet
Ivison R.J., Smail I., Le Borgne J.F., Blain A.W., Kneib J.P.,
identified. The presence of an extended [Oii] 3727˚Aemis- B´ezecourt J., Kerr T.H., Davies J.K., 1998, MNRAS 298,
sionline in the 2Dspectrumdemonstratesthe occurrence 583
ofstarformationinthering.Inaddition,thegalaxyhosts Kessler et al., 1996, A&A315, 27
an active nucleus in its centre, probably triggered by the Kneib J.–P., Mathez G., Fort B., Mellier Y., Soucail G., Lon-
interaction, as shown both by the detection of the [Nev] garetti P.–Y., 1994, A&A,286, 701
line in the optical and by the properties of the mid-IR Kneib J.–P., Ellis R.S., Smail I., Couch W.J., Sharples R.M.,
fluxes. This ring galaxy also corresponds to the source 1996, ApJ, 471, 643
Le F`evre O., Crampton D., Felenbok P., Monnet G., 1994,
L3 detected in the sub-millimeter domain with SCUBA
A&A,282, 325
by Smail et al. (1998). This shows that emission arises
Mellier Y., Soucail G., Fort B., Mathez G., 1988, A&A, 199,
from cold dust in the source, but without adequate spa-
13
tial information no conclusion can be drawn about the
Metcalfe L., Altieri B., McBreen, B. et al., 1999. To appear
preferred location of this cold dust (in the ring or nu-
in ESA conference Proceedings of the “The Universe as
cleus). We can also compute its intrinsic mid-IR luminos- seen by ISO”, October 1998, Unesco, Paris, France, astro-
ity, after correction for both the gravitational magnifica- ph/9901147.
tion and a k-correctionestimated from a power-law fit to Pell´o R., Kneib J.P., Le Borgne J.F., B´ezecourt J., Ebbels
the spectrum. This gives F(3.3µm)=1.71023 W/Hz and T.M.,TijeraI.,BruzualG.A.,MirallesJ.M.,SmailI.,Sou-
F(7µm) = 3.71023 W/Hz, or equivalently in solar lumi- cail G., Bridges T.J., 1999, submitted,astro-ph/9810390
nosity:ν Lν ≃4.1010 L⊙.Incomparison,thetotalmid-IR Smail I., Dressler A., Kneib J.P., Ellis R.S., Couch W.J.,
Sharples R.M., Oemler A.Jr, 1996, ApJ469, 508
luminosityoftheCartwheelatsimilarwavelengthsisonly
ν Lν ≃2.108 L⊙.Thedifferenceinluminositycomesmost SmaLil21I., Ivison R.J., Blain A.W., Kneib J.P., 1998, ApJ 507,
probably by the nucleus emission, relatively weak in the
Starck J. L. et al., 1998, in Extragalactic Astronomy in the
Cartwheel. LRG J0239–0134 belongs more specifically to
Infrared; Editions Fronti`eres, G. Mamon Ed.
the sample of high redshift galaxies for which the mid-
Schultzetal.SchultzB.,ClavelJ.,AltieriB.etal.,1999Toap-
IR and sub-mm fluxes emissions are powered by a strong pear in ESA conference Proceedings of the “The Universe
and obscured active nucleus (Ivison et al. 1998), possibly as seen by ISO”,October 1998, Unesco, Paris, France.
triggered by galaxy interactions. Struck C., Appleton P.N., Borne K.D., Lucas R.A., 1996, AJ
In conclusion, we have shown in this letter that gravi- 112, 1868
tationalmagnificationcanbeusedtogiveadetailedanal-
ysisofthenatureofafewpeculiardistantgalaxies.Clearly
cluster-lenses are and will be useful in the future to ob-
serveandstudy the nature ofthe mostdistantgalaxiesin
a wide domain of wavelengths and emission processes.