1. Bob Fosbury ST-ECF; Leiden Nov. 2002 UV/optical properties of z~2.5 RG Massive galaxy formation during the “Quasar Epoch” Bob Fosbury (ST-ECF) Joël Vernet, Sperello di Serego Alighieri (Arcetri) Marshall Cohen (Caltech), Laura Pentericci (MPIfA) Montse Villar-Martín, Andrew Humphrey (U Hertfordshire) Based on: Keck LRISp and VLT ISAAC observations

2. Bob Fosbury ST-ECF; Leiden Nov. 2002 Radio quasars and radio galaxies have different orientations The galaxies exhibit a ‘natural coronograph’ Why radio galaxies?

3. Bob Fosbury ST-ECF; Leiden Nov. 2002 Why redshift ~ 2.5? High star formation rate Peak of quasar activity Epoch of elliptical assembly Rapid chemical evolution Groundbased access to UV and optical restframe spectrum Courtesy Blain, Cambridge

4. Bob Fosbury ST-ECF; Leiden Nov. 2002 Earth forms 3C radio galaxies z2p5 radio galaxies Most distant QSO reionization The Quasar epoch Now

5. Bob Fosbury ST-ECF; Leiden Nov. 2002 Main results The interstellar medium of the galaxy, ionized by the quasar, tells the story of early chemical evolution in massive galaxies Dust absorption, scattering and re-radiation processes are very important for the overall SED — from the restframe UV through the FIR/sub-mm

6. Bob Fosbury ST-ECF; Leiden Nov. 2002 Expectation Evidence for the hidden quasar Actually hard to see anything else! An ISM excited by the AGN Chemical composition Kinematics Starlight Red/blue ratio and 4000Å break => evolutionary state of stellar population Absorption lines Stellar and interstellar

7. Bob Fosbury ST-ECF; Leiden Nov. 2002 Strategy Hi-res images in optical and NIR with HST (WFPC2 & NICMOS) Optical spectropolarimetry of the restframe UV from Ly  to ~2500Å -> resonance emission and absorption lines, dust signatures, continua from young stars and from the scattered (hidden) AGN -> separate the stellar from the AGN-related processes IR spectroscopy of the restframe optical: [OII] -> J [OIII] -> H H  -> K (constrains z-range) -> forbidden lines and evolved stellar ctm. Understand the K Hubble diagram (K–z)

8. Bob Fosbury ST-ECF; Leiden Nov. 2002 What is unique to this study? 3 to 8 hrs of Keck LRISp integration for each of 12 objects => P(continuum) to ±1 or 2 % and high s/n spectrophotometry Use of the first publicly available 8m IR spectrograph (ISAAC) to see the restframe optical continuum The Keck and VLT samples partially overlap which gives us ~continuous spectral coverage from Ly  to H 

9. Bob Fosbury ST-ECF; Leiden Nov. 2002 The complete spectral range

10. Bob Fosbury ST-ECF; Leiden Nov. 2002 A note on sample selection Optical sample: Radio galaxies from the ultra-steep spectrum selected sample (Röttgering et al. 1995) with z>2 accessible to Keck IR sample: Overlapping sample but with 2.2 < z < 2.6 to ensure the major emission lines fall in the J, H and K windows. LRISp; LRISp+ISAAC; ISAAC Objectz 4C+03.243.570 MRC0943-2422.922 MRC2025-218 2.63 MRC0529-549 2.575 USS0828+1932.572 4C-00.62 2.527 4C+23.562.479 MRC0406-244 2.44 B30731+4382.429 4C-00.542.360 4C+48.482.343 TXS0211-122 2.340 MRC0349-211 2.329 4C+40.362.265 MRC1138-262 2.156

11. Bob Fosbury ST-ECF; Leiden Nov. 2002 K-band Hubble diagram (PMC)

12. Bob Fosbury ST-ECF; Leiden Nov. 2002 The Keck II LRISp data 3,900–9,000Å, R~400 dual beam polarimeter One cycle -> 4 x 30min at standard HWP angles Reduced to I, Q and U Stokes spectra -> unbiassed estimates of P and  Error estimates from Monte-Carlo simulations Slit aligned with radio axis Fluxes scaled to HST magnitudes Corrected for Galactic extinction

13. Bob Fosbury ST-ECF; Leiden Nov. 2002

16. 4C+48.48 This is the RG with SCUBA sub-mm measurements — used later for scaling the dust scattering/ emission model

17. Bob Fosbury ST-ECF; Leiden Nov. 2002 Scattering model A simple dust scattering model is borne out by analytical and Monte-Carlo simulations of transfer through a dusty, clumpy medium (eg, Varosi & Dwek, 1999; Witt & Gordon 1999) The scattering is approximately grey (from Ly  to H  ) but with dust signatures A ‘luminosity weighting’ process ensures that most of the light we see comes from  ~ 1 F rg ~ F qso  scat exp(–  ext )

18. Bob Fosbury ST-ECF; Leiden Nov. 2002

19. ISAAC K-band spectra

20. Bob Fosbury ST-ECF; Leiden Nov. 2002 TXS0211-122 z = 2.340

21. Bob Fosbury ST-ECF; Leiden Nov. 2002

24. MRC1138-262 — ISAAC

25. Bob Fosbury ST-ECF; Leiden Nov. 2002 Results: the UV-optical continuum Dominated in the UV by scattered light from the hidden quasar. The evidence is: The polarization The continuum shape and intensity The presence of (polarized) broad lines with ~the expected EW The nebular continuum (computed from the recombination lines) is a minor contributor In low P objects there is some evidence for starburst light — population is constrained by the continuum colour In the optical, the continuum can comprise 3 components: evolved stars, scattered quasar, direct (reddened) quasar

26. Bob Fosbury ST-ECF; Leiden Nov. 2002 Results: The FIR continuum The dust scattering model can be used to calculate the FIR emission from the quasar-heated dust within the ionization cones This does not include the AGN torus emission (from hot dust in the MIR), nor does it include any contribution from dusty star-formation Figure based on 4C+48.48

27. Bob Fosbury ST-ECF; Leiden Nov. 2002

28. Results: the emission lines Two main contributors to the emission lines Scattered light from the quasar — characterised by polarization; both broad and (weak) narrow components Fluorescent emission from the ISM which is ionized predominantly by the AGN — seen directly and thus unpolarized BOTH of these components are spatially extended In some objects, we also see direct (reddened) quasar light at longer wavelengths (eg. H  )

29. Bob Fosbury ST-ECF; Leiden Nov. 2002 Ly  /CIV & NV/CIV vs P(%) correlations blue: sources with similar data from literature

30. Bob Fosbury ST-ECF; Leiden Nov. 2002 What does NV/CIV vs. P imply? Using the modelling, we can rule out ionization, density or depletion explanations The simplest explanation is a variation of metallicity with nitrogen changing quadratically wrt C/H or O/H => secondary nitrogen production As the enrichment proceeds, dust is produced and dispersed — leading to increasing obscuration and scattering. AGN-powered ULIRG are the end-point of this process

31. Bob Fosbury ST-ECF; Leiden Nov. 2002 Comparison of the kpc-scale ISM data from the RG with the BLR data discussed by Hamann & Ferland Quasar BLR

32. Bob Fosbury ST-ECF; Leiden Nov. 2002 Illustrative enrichment model from Hamann & Ferland (1999). The gE exhausts its gas after ~ 1Gyr followed by passive evolution. O/H

33. Bob Fosbury ST-ECF; Leiden Nov. 2002 Metallicity sequence The top two diagrams illustrate the narrow range in ionization The lower right confirms the metallicity sequence

34. Bob Fosbury ST-ECF; Leiden Nov. 2002 Spectral sequence From low- polarization, metal-poor radio galaxies to High- polarization, metal-rich ULIRG

35. Bob Fosbury ST-ECF; Leiden Nov. 2002 Comparison with Ly-break galaxy Pettini et al. 2000 Note dramatic difference in interstellar absorption line spectra …there are interesting correlations between the behaviour of the low ionization interstellar lines and the continuum polarization…

36. Bob Fosbury ST-ECF; Leiden Nov. 2002 SiII +OI CI I 0

37. Bob Fosbury ST-ECF; Leiden Nov. 2002 Summary Radio sources mark the sites of massive galaxy and cluster formation Radio galaxies have a built-in coronograph UV spectra are dominated by AGN-related processes: dust scattering and line fluorescence Re-radiaton in the FIR accounts for only a small fraction of the observed sub-mm flux (=> dusty SF?) Emission lines measure the physical and chemical and kinematic properties of the ISM Evidence for chemical evolution in the host galaxies during the “epoch of the quasars” Optical spectra -> stellar population and more detailed picture of chemical composition

38. Bob Fosbury ST-ECF; Leiden Nov. 2002 Conclusions Quasar-ionized ISM reveals a phase of rapid chemical evolution during the assembly of massive galaxies Can construct a consistent model for the transfer of quasar radiation in a dusty galaxy which accounts for both the UV/optical and a fraction of the FIR continuum

39. Bob Fosbury ST-ECF; Leiden Nov. 2002