Ultra-Steep Spectrum Radio Galaxies at Hy Redshifts
a r
X
i
v
:a
s t
r o
-
p
h
/
9
9
1
3
1
1v
1
1
8
O
c
t
1
9
9
9
Ultra–Steep Spectrum Radio Galaxies at Hy Redshifts Wil van Breugel,Carlos De Breuck 1&Adam Stanford 2Inst.of Geophysics &Planetary Physics,LLNL,Livermore,CA 94550Huub R¨o ttgering &George Miley Leiden Observatory,Leiden,The Netherlands Daniel Stern 3Astronomy Department,University of California,Berkeley,CA 94720Dante Minniti P.Universidad Catolica,Santiago 22,Chile Chris Carilli National Radio Astronomy Observatory,VLA,Soccoro,NM 87801Abstract.Radio sources have traditionally provided convenient bea-cons for probing the early Universe.Hy Spinrad was among the ?rst of the tenacious breed of observers who would attempt to obtain optical identi?cations and spectra of the faintest possible ‘radio galaxies’to in-vestigate the formation and evolution of galaxies at hy redshift.Modern telescopes and instruments have made these tasks much simpler,although not easy,and here we summarize the current status of our hunts for hy redshift radio galaxies (HyZRGs)using radio spectral and near–IR selec-tion.
1.Hy Z Radio Galaxies:Why?
The ?rst optical identi?cations of (bright)radio sources with (faint)galaxies were made when Hy Spinrad was still a teenager (Windhorst 1999;these proceedings).After that it was soon realized that the ‘invisible’universe of radio sources provided convenient beacons to locate very distant galaxies and thus might be used to study their formation and evolution.As so eloquently described by
several of Hy’s colleagues,collaborators and ex–students in these proceedings, he became an early key player in these distant galaxy hunts.
For most radio astronomers in those days the Universe stopped at the POSS limits.Surely,many radio sources could not be identi?ed,but so what?It just con?rmed that the Universe was bigger than the biggest optical telescope, but...not bigger than the biggest radio telescope!Occasionally Hy would write letters to Leiden Observatory radio astronomers with requests of radio maps and accurate positions.When provided he would spend many hours using one of the world’s then?nest telescopes at Lick Observatory,to obtain optical identi?cations and spectra of these HyZRG candidates(presumably squeezed in between observations of standard stars for Jim Liebert;these proceedings).How foolish this seemed,to some of us(WvB).
Since then extraordinary progress in the development of optical and near–IR detectors,larger telescopes,and better selection techniques have resulted in discoveries of radio galaxies at increasingly hyer redshifts.Paradoxically this task was eased by the discovery,?rst by Hy and his collaborators,that the Lyαemission line was very strong in radio galaxies and could easily be detected in <1hr integrations with3m–class telescopes(Spinrad et al.1985),provided that the redshifts would be hy enough(z>1.6)so that Lyαwould enter the observable optical window.
It has now become clear that HyZRGs are both a boon and a curse for students of galaxy evolution.A boon because the near–IR Hubble K?z relation for radio sources appears well represented by the predicted‘passive’evolution of massive(5?10L?)galaxies with hy formation redshifts(Lilly&Longair 1984;Eales et al.1997;Best et al.1998;van Breugel et al.1999),despite the e?ects of k–correction and morphological evolution(van Breugel et al.1998). No matter the reason for this relationship,it suggests that radio sources might be used to?nd massive galaxies and their likely progenitors out to very hy redshift.This method was?rst successfully used by Lilly(1988)to identify the HyZRG B20902+34at z=3.395.Deep spectroscopic observations of a few relatively weak radio sources(but still of the powerful double FRII class !),where the AGN do not a?ect the rest–frame UV as much(see below),have shown directly examples of radio galaxies at z～1.5with old(～>3.5?4.5Gyr) stellar populations with implied formation redshifts z f>10(e.g.,Spinrad et al.1997).
For galaxy formation studies HyZRGs are also cursed because their struc-tures are aligned with their associated radio sources,suggesting that the col-limated out?ow and ionizing radiation from their AGN profoundly a?ect their parent galaxy host appearances at UV,blue and green wavelengths(McCarthy 1999,and Dey1999;these proceedings).HyZRGs are very well suited for study-ing the e?ects of powerful AGN on ambient dense gas,including induced star formation(e.g.,Bicknell et al.1999),and may even be used as searchlights to in-vestigate the properties of proto–galactic material in the early Universe(Cimatti et al.1997;Villar–Mart′?n et al.1997).
Recent cosmological theories are providing additional incentives to use radio galaxies as probes to study the early Universe.Within standard Cold Dark Matter scenarios the formation of galaxies is a hierarchical and biased process. Large galaxies are thought to grow through the merging of smaller systems,
Figure1.The Hubble K?z diagram for HyZRGs.Filled triangles
are Keck measurements of HzRGs from van Breugel et al.(1998),the
large triangle is TN J0924?2201at z=5.19,and all other photometry
is from Eales et al.(1997).Two stellar evolution models from Bruzual
&Charlot(1999),normalized at z<0.1,are plotted,assuming param-
eters as shown.
and the most massive objects form in over–dense regions,which will eventually evolve into the clusters of galaxies seen today(e.g.,White1997).It has also been suggested that the?rst massive black holes may grow in similar hierarchical fashion together with their parent galaxies(e.g.,Kau?mann and Haehnelt1999) or,because of time scale constraints,may precede galaxy formation and be primordial(e.g.,Loeb1993).To confront these theories it is therefore of great interest to?nd the progenitors of the most massive galaxies and their AGN (active massive black holes)at the hyest possible redshifts and to study their properties and cosmological evolution.
While optical,‘color–dropout’techniques have been successfully used to?nd large numbers of’normal’young galaxies(without dominant AGN)at redshifts even surpassing those of quasars and radio galaxies(Weymann et al.1998;Hu et al.1999),the radio and near–infrared selection technique has the additional advantage that it is unbiased with respect to the amount of dust extinction. HyZRGs are therefore also important laboratories for studying the large amounts of dust(e.g.,Ivison et al.1998)and molecular gas(Papadopoulos et al.1999), which are observed to accompany the formation of the?rst forming massive galaxies.Indeed,a signi?cant part of the scienti?c rationale for building future large mm-arrays is based on the expectation that to understand galaxy formation will ultimately require understanding of their cold gas and dusty environments.
Table1.Radio Surveys
WENSS TEXAS MRC
Frequency(MHz)325365408
Sky regionδ>+29??35.?7<δ<+71.?5?85?<δ<+18.?5
#of sources229,57667,55112,141 Resolution54′′×54′′cosecδ10′′2.′62×2.′86sec(δ?35.?5) Position uncertainty1.′′50.′′5—1′′8′′
(strong sources)
RMS noise～4mJy20mJy70mJy
Flux density limit18mJy150mJy670mJy
NVSS FIRST a PMN
Frequency(MHz)140014004850
Sky regionδ>?40?b>45?87.?5<δ<+10?
#of sources1,814,748550,00050,814 Resolution45′′×45′′5′′×5′′ 4.′2
Position uncertainty1′′0.′′1～45′′(strong sources)
RMS noise0.5mJy0.15mJy～8mJy
Flux density limit 2.5mJy1mJy20mJy
Figure 2.Logarithmic spec-
tral index distribu-
tion for WENSS–NVSS (full
line),Texas–NVSS (dot–dash
line)and MRC–PMN (dotted
line).The vertical line indi-
cates the –1.3cuto?used in
our spectral index
selection.Figure 3.Identi?cation frac-tion on the POSS as a function of spectral index for the com-bined WN and TN sample.
(Gri?th et al.1993)surveys it is now possible for the ?rst time to de?ne a large sample of USS sources with extremely steep radio continuum spectra (α≤?1.3,Fig.2),and using 10–100times lower ?ux density limits than has been possible before (Chambers et al.1996;R¨o ttgering et al.1994;Blundell et al.1998).
Using these surveys we constructed 3sub–samples,covering di?erent regions of the radio sky using the deepest low and high frequency surveys available in each area (Table 2).Our largest,and most complete sample is based on the WENSS survey at δ>29?,together with the NVSS and FIRST surveys.In the remaining area covered by the northern hemisphere radio telescopes,we used the Texas survey at low frequencies,which produces a similar sample but at a higher ?ux density and is less complete.We also used two southern surveys to construct the ?rst USS sample in the deep southern sky.More details about the samples are given in De Breuck et al.(2000a ).
During the course of our optical and near–IR imaging and optical spec-troscopy programs we have ?ne–tuned our selection technique.Previously it had been found that the identi?cation fraction of radio galaxies decreases with spectral index (Tielens et al.1979;R¨o ttgering et al.1995),which provided the rationale for using the USS source selection technique.With our ‘hyper-steep’radio spectrum selection (α≤?1.3)most sources remain unidenti?ed,at least on the POSS (R ～<20).Only ～15%of the sources can be identi?ed,usually with bright galaxy clusters,as indicated by the frequent overdensity of galaxies around them,and by X-ray detections (De Breuck et al.2000a ).This identi?-cation fraction appears to be independent of spectral index (Fig.3),in support of the idea that these are mostly foreground objects.
This also explained why our initial optical imaging campaign on3m–4m–class telescopes(R～<24)was not very succesful in?nding R?band identi?ca-tions.Furthermore,for the typically expected R?K～4values of HyZRGs,it would even be a challenge to detect most HyZRGs in the near–IR at Lick Obser-vatory.We therefore decided to entirely skip the optical identi?cation program and go straight to near–IR imaging at the Keck I telescope.This has produced, to date,a100%identi?cation rate with good photometric magnitudes to select HyZRG candidates using the Hubble K?z diagram(Fig.1).
We have now spectroscopically observed30faint USS HyZRG candidates with the following results.Only5of the sources have z<2,7have2 https://www.sodocs.net/doc/009685197.html,S samples Sample Density Spectral Index Flux Limit#of Sources sr?1mJy WN151α1400 325≤?1.30S1400>10343 TN48aα1400 365≤?1.30S1400>10268 MP26α4800 408 ≤?1.20S408>700;S4850>3558 D E C L I N A T I O N (J 2000)RIGHT ASCENSION (J2000)13 38 26.526.426.326.226.126.025.925.8 -19 42 28 30 32 34 36 Figure 4. 4.85GHz VLA radio contours overlaid on a Keck K ?band image of TN J1338?1942.Figure 5.VLT spectrum of TN J1338?1942at z =4.11. that the central region is obscured by dust.Such asymmetric radio sources are not uncommon,even in the local Universe,and are usuallly thought to be due to strong interaction of one of its radio lobes with very dense gas (e.g.,McCarthy et al.1991;Feinstein et al.1999). A high signal–to–noise spectrum was also obtained with the VLT Antu telescope (De Breuck et al.1999b ).The spectrum is dominated by the bright Ly αline (W rest Ly α=210?A )which shows deep and broad (～1400km s ?1)blue–ward absorption.The latter is probably due to resonant scattering by cold HI gas in a turbulent halo surrounding the radio galaxy and has also been seen several other HyZRGs (van Ojik et al.1996;Dey 1999).The continuum is relatively bright (F 1400～2μJy)and if all due to young O–B stars this would imply a total SFR of several hundred M ⊙/yr,resembling 4C41.17,and suggesting that TN J1338?1942may be another HyZRG in which induced star formation might occur (c.f.,Bicknell et al.1999). In Table 3we have listed the Ly αproperties of the known 7most distant radio galaxies for which high quality optical slit spectroscopy data taken with Keck or the VLT are available.We have assumed H 0=65km s ?1Mpc ?1,q 0=0.15,and Λ=0.The Ly α?uxes are as measured i.e.,uncorrected for blue–ward absorption.TN J1338?1942is the most luminous Ly αgalaxy and,after 4C41.17,also the brightest (in similar apertures).In all cases the brightest Ly αemission occurs on scale sizes of 1′′?2′′,comparable to those of the brightest radio structures.4C41.17is known to have a very extended halo (Chambers et al.1990)and the total size quoted is a lower limit,based on the deep (9hrs)Keck spectropolarimetry data from Dey et al.(1997). 3.2.TN J0924?2201at z =5.19 TN J0924?2201is one of the steepest spectrum sources in our USS sample (α1.4GHz 365MHz =?1.63)and therefore was one of our primary targets for near–IR identi?cation.A deep K–band image at Keck showed indeed a very faint (K = Figure6.Keck/NIRC K-band image of TN J0924?2201,with radio contours superposed.Figure7.Keck spectra of TN J0924?2201at z=5.19on two di?erent nights. 21.3±0.3),multi–component object at the position of the small(1.′′2)radio source(Fig.6).The expected redshift on the basis of the K?z diagram was z>5,and spectroscopic observations at Keck showed that this was indeed the case,based on a single emission line atλ～7530?A which we identi?ed as Lyαat z=5.19(van Breugel et al.1999;none of the z>5galaxies have more than one line detection). Among all radio selected HyZRGs TN J0924?2201is fairly typical in ra-dio luminosity,equivalent width and velocity width(Table2).It does have the steepest radio spectrum,consistent with theα?z relationship for powerful ra-dio galaxies(e.g.,R¨o ttgering et al.1997),and also has the smallest linear size. The latter may be evidence of its‘inevitable youthfulness’or a dense con?n-ing environment,neither of which would be surprising because of its extreme redshift(Blundell&Rawlings1999;van Ojik et al.1997).Among the radio se-lected HyZRGs TN J0924?2201appears underluminous in Lyα,together with 8C1435+63,which might be caused by absorption in an exceptionally dense cold and dusty medium.Evidence for cold gas and dust in several of the most distant HyZRGs has been found from sub–mm continuum and CO–line obser-vations(e.g.,Ivison et al.1998;Papadopoulos et al.1999). The second hyest redshift radio galaxy currently known listed in Table3 is VLA J123642+621331at z=4.42(Waddington et al.1999).This source was not USS selected and therefore provides an interesting alternate view on the possible selection e?ects of our method of?nding HyZRGs.The source is an asymmetric double and although its radio luminosity is about a factor1000 times lower than that of its much more luminous brothers at similar redshifts,it still quali?es as a FRII–type,though with a radio luminosity close to the FRI/ FRII break at408MHz(P408～3.2×1026W Hz?1).Its radio spectrum is steep (α1.4GHz 8.4GHz ～?1.0,using the?ux densities given by Waddington et al.),but not as steep as our USS selected HyZRGs,and the Lyαluminosity is a factor5–10 times less.Apart from the luminosity these properties are not hugely di?erent from expected on the basis of radio selection and indicate that less extreme steep spectrum selected samples(α Table3.Physical Parameters of the Highest HyZRGs Name z L Lyαa L365aα1400 365W rest Lyα ?Lyαa Size a Ref.b TN J0924?2201 5.19 1.37.5?1.63>11515008WvB99 VLA J1236+6213 4.420.20.0035?0.96>50440Wad99 6C0140+326 4.4116 1.3?1.157********DeB00 8C1435+63 4.25 3.211?1.31670:180028Spin95 TN J1338?1942 4.1125 2.3?1.31200100037DeB99 4C41.17 3.79812 3.3?1.25100140099Dey97 4C60.07 3.7916 4.1?1.481502900:65R¨o t97 question how these can form so shortly after the putative Big Bang may prove even more challenging then that of the formation of galaxies(e.g.,Loeb1993; Silk&Rees1998). Acknowledgments.WvB is grateful for the many interesting conversa-tions,advice,and wonderful collaborations he has had with Hy Spinrad and his students over the past15years.The work by W.v.B.,C.D.B.and S.A.S.at IGPP/LLNL was performed under the auspices of the US Department of En-ergy under contract W-7405-ENG-48.W.v.B.also acknowledges support from NASA grants GO6608,and D.S.from IGPP/LLNL grant98–AP017. References Becker,R.H.,White,R.L.,&Helfand,D.J.1995,ApJ,450,559 Bicknell,G.,Sutherland,R.,van Breugel,W.,Dopita,M.,Dey,A.,Miley,G. 1999,ApJ,in press,astro-ph/9909218 Blundell,K.M.,et al.1998,MNRAS,295,265 Blundell,K.M.,&Rawlings,S.1999,Nature,399,330 Bruzual,A.G.,&Charlot,S.1999,personal communication Chambers,K.C.,Miley,G.K.,&van Breugel,W.J.M.1990,ApJ,363,21 Chambers,K.C.,Miley,G.K.,van Breugel,W.J.M.,&Huang,J.-S.1996, ApJS,106,215 Cimatti,A.,Dey,A.,van Breugel,W.,Hurt,T.&Antonucci,R.1997,ApJ, 476,677 Condon,J.,et al.1998,AJ,115,1693 De Breuck,C.,van Breugel,W.,R¨o ttgering,H.,Miley,G.&Carilli,C.1999a,in ’Looking Deep in the Southern Sky’Edited by R.Morganti&W.Couch (Springer),p.246 De Breuck,C.,van Breugel,W.,Minniti,D.,Miley,G.,R¨o ttgering,H.,Stanford, S.,&Carilli,C.1999b,A&A,in press(astro-ph/9909178) De Breuck,C.,van Breugel,W.,R¨o ttgering,H.,&Miley,G.2000a,A&AS, submitted De Breuck,C.,van Breugel,W.,R¨o ttgering,H.,Miley,G.,&Stern,D.2000b, in preparation Dey,A.,van Breugel,W.,Vacca,W.D.,&Antonucci,R.1997,ApJ,449,698 Dey,A.,Spinrad,H.,Stern,D.,Graham,J.,&Cha?ee,F.1998,ApJ,498,L93 Dey,A.1999,in’The Most Distant Radio Galaxies’,ed.H.R¨o ttgering,P.Best &M.Lehnert(Amsterdam:KNAW),p.19 Douglas,J.N.,et al.1996,AJ,111,1945 Dunlop,J.,&Peacock,J.1990,MNRAS,247,19 Eales,S.,et al.1997,MNRAS,291,593 Fan,X.et al.1999,apjl,in press,astro-ph/9910001 Feinstein,C.,et al.1999,astro-ph/9906059 Gri?th,M.&Wright,A.E.1993,AJ,105,1666 Haehnelt,M.,Natarajan,P.&Rees.M.J.1998,MNRAS,300,817 Hu,E.,McMahon,R.,&Cowie,L.1999,ApJ,522,9 Ivison,R.J.et al.1998,ApJ,494,211 Kaplan,D.L.,et al.1998,ApJS,119,75 Kau?mann,G.,&Haehnelt,M.1999,MNRAS,in press,astro-ph/9906493 Large,M.I.,Mills,B.Y.,Little,A.G.,Crawford,D.F.,&Sutton,J.M.1981, MNRAS,194,693 Lilly,S.J.&Longair,M.,1984,MNRAS,211,833 Lilly,S.1988,ApJ,333,161 Loeb,A.1993,ApJ,403,542 Martin,C.,Halpern,J.,&Schiminovich,D.1998ApJ,494,211 McCarthy,P.,van Breugel,W.,&Kapahi,V.K.1991,ApJ,371,478 McCarthy,P.1999,in’The Most Distant Radio Galaxies’,ed.H.R¨o ttgering, P.Best&M.Lehnert(Amsterdam:KNAW),p.5 Papadopoulos,P.P.,et al.1999,ApJ,in press Pentericci,L.et al.1998,A&A,341,329 Rengelink,R.,et al.1997,A&A,124,259 R¨o ttgering,H.J.A.,Miley,G.K.,Chambers,K.C.&Macchetto,F.1995,A&AS, 114,51 R¨o ttgering,H.,van Ojik,R.,Chambers,K.,van Breugel,W.,&de Ko?,S. 1997,A&A,326,505 Shaver,P.,Wall,J,Kellerman,K.,Jackson,C.,&Hawkins,M.1996,Nature, 384,439 Silk,J.&Rees,M.1998,A&A,331,L1 Spinrad,H.,Filippenko,A.V.,Wycko?,S.,Stocke,J.T.,Wagner,R.M.,& Lawrie,D.G.1985,ApJ,299,L7. Spinrad,H.,Dey,A.&Graham,J.1995,ApJ,438,L51. Spinrad,H.,Dey,A.,Stern,D.,Dunlop,J.,Peacock,J.,Jimenez,R.&Wind-horst,R.1997,ApJ,484,581 Spinrad,H.et al.1998,AJ,116,2617 Tielens,A.G.G.M.,Miley,G.K.,&Willis,A.G.1979,A&AS,35,153 van Breugel,W.,Stanford,S.A.,Spinrad,H.,Stern,D.,&Graham,J.R.1998, ApJ,502,614 van Breugel,W.et al.1999,in’The Most Distant Radio Galaxies’,ed.H. R¨o ttgering,P.Best&M.Lehnert(Amsterdam:KNAW),p.49 van Ojik,R.,R¨o ttgering,H.J.A.,Miley,G.K.,&Hunstead,R.W.1997,A&A, 317,358 Villar-Mart′?n,M.,Tadhunter,C.,&Clark,N.1997,A&A,323,21 Weymann,R.et al.1998,ApJ,505,L95 White,S.1997,ESO-VLT Workshop”Galaxy Scaling Relations:Origins,Evo-lution and Applications”,astro-ph/9702214