Fan Fiction

The Massive Hosts of Radio Galaxies Across Cosmic Time

The Massive Hosts of Radio Galaxies Across Cosmic Time
of 66
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
    a  r   X   i  v  :  a  s   t  r  o  -  p   h   /   0   7   0   3   2   2   4  v   1   9   M  a  r   2   0   0   7 2007 March 9 th The Massive Hosts of Radio Galaxies Across Cosmic Time Nick Seymour 1 , Daniel Stern 2 , Carlos De Breuck 3 , Joel Vernet 3 , Alessandro Rettura 4 ,Mark Dickinson 5 , Arjun Dey 6 , Peter Eisenhardt 2 , Robert Fosbury 3 , Mark Lacy 1 , PatMcCarthy 6 , George Miley 7 , Brigitte Rocca-Volmerange 8 , Huub R¨ottgering 9 , S. AdamStanford 10 , 11 , Harry Teplitz 1 , Wil van Breugel 10 , 11 & Andrew Zirm 4 ABSTRACT We present the results of a comprehensive Spitzer  survey of 69 radio galax-ies across 1 < z < 5 . 2. Using IRAC (3 . 6 − 8 . 0 µ m), IRS (16 µ m) and MIPS(24 − 160 µ m) imaging, we decompose the rest-frame optical to infrared spec-tral energy distributions into stellar, AGN, and dust components and determinethe contribution of host galaxy stellar emission at rest-frame H  − band. Stel-lar masses derived from rest-frame near-IR data, where AGN and young starcontributions are minimized, are significantly more reliable than those derivedfrom rest-frame optical and UV data. We find that the fraction of emitted lightat rest-frame H  − band from stars is > 60% for ∼ 75% the high redshift radio 1 Spitzer  Science Center, California Institute of Technology, 1200 East California Boulevard, Pasadena,CA 91125, USA. [email: ] 2 Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA. 3 European Southern Observatory, Karl Schwarzschild Straße, D-85748 Garching, Germany. 4 Johns Hopkins University, 3400 N. Charles Street, Baltimore, MD 21218, USA. 5 National Optical Astronomy Observatory, Tucson, AZ 85719, USA. 6 Carnegie Observatories, 813 Santa Barbara Street, Pasadena, CA 91101, USA. 7 Leiden Observatory, University of Leiden, PO Box 9513, 2300 RA Leiden, Netherlands. 8 Institute d’Astrophysique de Paris, 98bis Bd Arago, 75014 Paris, France. 9 University of California, Davis, CA 95616, USA. 10 Institute of Geophysics and Planetary Physics, Lawrence Livermore National Laboratory, Livermore,CA 94551, USA. 11 University of California, Merced, PO Box 2039, Merced, CA 95344, USA.  – 2 –galaxies. As expected from unified models of AGN, the stellar fraction of the rest-frame H  − band luminosity has no correlation with redshift, radio luminosity, orrest-frame mid-IR (5 µ m) luminosity. Additionally, while the stellar  H  − bandluminosity does not vary with stellar fraction, the total  H  − band luminosityanti-correlates with the stellar fraction as would be expected if the underlyinghosts of these radio galaxies comprise a homogeneous population. The resultantstellar luminosities imply stellar masses of 10 11 − 11 . 5 M  ⊙ even at the highest red-shifts. Powerful radio galaxies tend to lie in a similar region of mid-IR color-colorspace as unobscured AGN, despite the stellar contribution to their mid-IR SEDsat shorter-wavelengths. The mid-IR luminosities alone classify most HzRGs asLIRGs or ULIRGs with even higher total-IR luminosities. As expected, theseexceptionally high mid-IR luminosities are consistent with an obscured, highly-accreting AGN. We find a weak correlation of stellar mass with radio luminosity. Subject headings: galaxies: active — galaxies: high-redshift — galaxies: evolu-tion 1. Introduction Luminous radio galaxies were the first class of obscured, or type 2  , quasars to be dis-covered and characterized. They have accreting super-massive black holes whose continuumemission at UV, optical, and soft X-ray energies is absorbed by dust, thus allowing a clear  view of the host galaxy. More recently, hard X-ray and mid-IR surveys have identified theradio-quiet cousins to luminous radio galaxies ( e.g., Norman et al. 2002;Stern et al. 2002; Mart´ınez-Sansigre et al. 2005;Polletta et al. 2006;Lacy et al. 2006). The main evidence that radio galaxies host super-massive black holes comes from the high luminosities of their radiolobes, which are fed by radio jets srcinating at the host galactic nuclei. The lobe spatialextents (up to a few Mpc,Saripalli et al. 2005) and luminosities ( L 1 . 4GHz ≥ 10 25 WHz − 1 )clearly rule out a stellar srcin for their energetics ( e.g., Rees 1978). In terms of the orien-tation unification scheme for AGN ( e.g., Barthel 1989;Antonucci 1993;Urry & Padovani 1995), radio galaxies are radio-loud quasars seen from an angle where an optically-thick torusobscures emission from the region closest to the central engine.Due to their large radio luminosities, radio galaxies were the predominant way to probethe distant universe until the advent of 8 − 10m class telescopes and the Lyman-break tech-nique in the last decade. In fact, radio galaxies were the first galaxies to be found aboveredshifts one, two, three and four (seeStern & Spinrad 1999, and references therein). Acrosscosmic time, the host galaxies of powerful radio sources appear to be uniquely robust in-  – 3 –dicators of the most massive galaxies in the universe. At low redshift this result has beenknown since the first optical identifications of extra-galactic radio sources showed them tobe associated with massive, giant elliptical (gE and cD) galaxies (Matthews et al. 1964). Inthe more distant universe, indirect evidence that this association remains intact comes fromthe detection of host galaxies with r 1 / 4 law light profiles in Hubble Space Telescope ( HST  )observations of high-redshift radio galaxies (HzRGs) at 1 ∼ <z ∼ < 2 (Pentericci et al. 2001;Waddington et al. 2002;Zirm et al. 2003;Bunker et al. 2007); the tendency for HzRGs to reside in moderately rich (proto-)cluster environments (Le F`evre et al. 1996;Pascarelle et al. 1996;Venemans et al. 2002,2003,2004,2005,2006;Stern et al. 2003); the spectacular ( > 100 kpc) luminous Ly α haloes seen around several sources, implying large gas reservoirs(Reuland et al. 2003;Villar-Mart´ın et al. 2003); sub-mm detections of HzRGs, implying vi- olent star formation activity up to ∼ 1000 M  ⊙ yr − 1 (Archibald et al. 2001;Reuland et al. 2004); and a few direct kinematic measurements of HzRGs (Dey et al. 1996;Dey & Spinrad 1996;Nesvadba et al. 2006). The most compelling evidence of this association of HzRGs with the most massive systems, however, is the tight correlation of the observed near-infrared Hub-ble, or K  − z , diagram for powerful radio sources (Fig.1;Lilly & Longair 1984;Best et al. 1998;Eales et al. 1997;van Breugel et al. 1998;Jarvis et al. 2001;De Breuck et al. 2002; Willott et al. 2003;Rocca-Volmerange et al. 2004;Brookes et al. 2006): HzRGs form a nar- row redshift sequence which traces the envelope of radio-quiet galaxies and is well modeledby the evolution of a stellar population formed at high redshift from an accumulation of up to10 12 M  ⊙ of pre-galactic baryonic material. The large-scale, double-lobed radio morphologiesand enormous radio luminosities suggested early on that HzRGs must have spinning super-massive black holes powering relativistic jets in their centers (Rees 1978;Blandford & Payne 1982). With the more recent discovery that the stellar bulge and central black hole massesof galaxies are closely correlated ( e.g., Tremaine et al. 2002), it is no longer a surprise thatthe parent galaxies of the most powerful radio sources occupy the upper end of the galaxymass function.Despite two decades of study since the initial discovery of the HzRG K  − z rela-tion (Lilly & Longair 1984), the nature and tightness of the relation remains mysterious.The scatter in the relation is surprisingly low out to z ∼ 1, and only increases mod-estly at higher redshifts as the observed K  -band samples rest-frame optical emission (e.g.,Fig.1;De Breuck et al. 2002). Interpretations have generally relied on these observed near-IR observations probing the stellar populations of the distant radio galaxies ( e.g., Rocca-Volmerange et al. 2004). However, detailed interpretations have been complicatedby (i) uncertain contributions of AGN-related light which are most important in the rest-frame UV ( e.g., Vernet et al. 2001), (ii) evolution of the host galaxy stellar population(s),and (iii) band-shifting, meaning that observed 2 . 2 µ m samples very different emitting wave-  – 4 –lengths for different redshift sources, reaching into the rest-frame ultraviolet for the mostdistant HzRGs.In this paper, we present observations of a large sample of HzRGs obtained with the Spitzer Space Telescope (Werner et al. 2004). By observing the same rest-frame near- tomid-IR spectral range for sources over a large redshift range (1 < z < 5 . 2), we removemany of the complications which plagued previous studies. In particular, the 1 . 6 µ m peakof the stellar emission provides a reasonably robust measure of the stellar mass for stellarpopulations with ages ∼ > 1Gyr (Sawicki 2001). While UV and optical emission has strongcontributions from the youngest and hottest stars in a galaxy, near-IR emission primarilyderives from the low-mass stars which dominate the stellar mass of a galaxy. Thus, whilerest-frame UV and optical studies of galaxies are well-suited to probe galaxy star-formationrates, rest-frame near-IR studies are well-suited to probe galaxy stellar masses. We discussthis approach and other recent results that might effect it (e.g.Maraston 2005) in detail in § 6.Fig.2illustrates the advantage of rest-frame near-IR derivations of stellar masses. First,the impact of dust extinction falls sharply with wavelength, making quantities derived fromrest-frame near-IR observations more than an order of magnitude less susceptible to uncertaindust extinction corrections relative to optical observations. Second, since the main sequencelifetimes of low-mass stars exceed the Hubble time, galaxy masses derived from rest-framenear-IR observations are relatively insensitive to the age of stellar populations or the starformation history. In particular, secondary bursts of star formation will affect UV and opticalmagnitudes significantly more than near-IR magnitudes.We here use mid-IR observations of a large sample of radio galaxies to probe theirhost galaxy stellar masses. By observing consistently on both sides of the 1 . 6 µ m peakof the stellar emission, we avoid complicated k − correction effects and derive a reasonablyrobust measure of the stellar mass for stellar populations with ages ∼ > 1Gyr. Observationsat longer wavelengths allow us to determine the contribution of warm, AGN-heated dustemission to the rest-frame near-IR emission. Our paper is organized as follows. Section 2describes the Spitzer  HzRG sample. Section 3 presents the Spitzer  mid- to far-IR dataand their reduction. Section 4 describes our χ 2 fitting of the spectral energy distributions(SEDs) and § 5 presents the results of this fitting. Section 6 describes how the derivedrest-frame H  − band luminosities are converted into stellar masses. Section 7 discusses andconcludes this analysis. We present notes on individual sources as an Appendix. This paperpresents our entire Spitzer  data set and analyzes the bulk properties of the sample. Spitzer  observations of individual sources from this program have been the subject of detailed studiesbyVillar-Mart´ın et al.(2006, MRC 2104 − 242 at z = 2 . 49),Stern et al.(2006, LBDS 53W091  – 5 –at z = 1 . 55),Broderick et al,(2007, PKS at z = 2 . 156) andDe Breuck et al.(2007, 4C 23.56 at z = 2 . 48). Throughout we assume a concordance model 1 of universe expansion, Ω M  =1 − Ω Λ = 0 . 3, Ω 0 = 1, and H  0 = 70 kms − 1 Mpc − 1 . Inferred luminosities presented in thispaper are of the form νL ν  /L ⊙ , where L ⊙ = 3 . 9 × 10 23 W. 2. The Spitzer  High-Redshift Radio Galaxy Sample Our HzRG sample is drawn from radio galaxy surveys executed during the last 45 years(starting with the 3CR;Bennett 1961). We have searched both flux-limited surveys suchas the 3C (Spinrad et al. 1985), 6CE (Eales et al. 1997), 7C (Lacy et al. 1999;Willott et al. 2001), MG (Bennett et al. 1986;Lawrence et al. 1986), and MRC (McCarthy et al. 1996), as well as surveys filtered by their ultra-steep radio spectra (USS; e.g., Chambers et al. 1996;R¨ottgering et al. 1997;De Breuck et al. 2000). The former provide a sample unbiased in their radio properties, but lack the high redshift ( z > 2) sources mainly found in the USSsamples. Because the USS samples also probe fainter radio flux densities, including suchsources allows us to break the strong luminosity − redshift degeneracy of flux density-limitedsurveys.For the purposes of this work we define a HzRG as a radio galaxy above a redshift of one with a rest-frame 3GHz luminosity greater than 10 26 WHz − 1 . This choice is to ensurethat we include only those objects with very powerful obscured AGNs. This value is similarto that used by other authors to separate AGN based upon their radio luminosities. Theclassical luminosity that separatesFanaroff & Riley(1974) type 1 and type 2 sources (which are morphologically distinct), converted from 178MHz to 3GHz, is 10 25 . 4 (10 24 . 8 )WHz − 1 assuming a steep (ultra-steep) radio spectral index of  α = − 0 . 8 ( − 1 . 3). A more recentradio-loud/radio-quiet division is L 3GHz = 10 26 . 3 (10 26 . 4 )WHz − 1 (Miller et al. 1990) and anupper limit to radio-quiet QSOs of  L 3GHz = 10 25 . 2 (10 25 . 1 )WHz − 1 was found byGregg et al.(1996), both converted to 3GHz luminosity assuming the same spectral index values.To examine how Spitzer  -derived quantities depend on redshift and radio luminosity, weselected a representative subset of 69 targets from the parent sample of 263 known z > 1HzRGs (circa 2003; Fig.3). Table1lists the selected subset, which spans the full range of  1 These common values of the “concordance” cosmology underestimate luminosities (and hence stellarmasses) by ∼ 2 − 3% at z = 3 − 4 when compared to the latest Wilkinson Microwave Anisotropy Probe mea-surements, Ω M  = 1 − Ω Λ = 0 . 25, Ω 0 = 1, and H  0 = 73 kms − 1 Mpc − 1 (Spergel et al. 2006). Lower-redshiftHzRGs show an even smaller systematic adjustment, implying that the assumed cosmological parametersare inconsequential to our analysis.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!