1. Galaxy Formation: What are we missing? C. Steidel (Caltech) Quenching  : How to Move from the Blue Cloud  Through the Green Valley  and to the Red Sequence  in 9 Easy Steps in the Face of Downsizing 

2. Some Questions: How efficient is galaxy formation as a function of time, environment, mass? Do we understand gas cooling and star formation on galaxy scales? Feedback in general: is it needed, is it observed, how does it modify the process? –What process ends major star formation in galaxies? AGN: when do they happen, how do they influence the life of a (massive) galaxy? Where or how do “massive galaxies’’ fit in with other galaxies? –What did they look like when they were forming most of their stars?

3. Using optical (rest-UV) and near-IR (rest optical) to quantify physical properties of z~2-3 galaxies Optical spectra: IMF stellar photospheric abundances ISM metallicity ISM kinematics Near-IR spectra: Kinematics/dynamical masses Ionized gas metallicity SFR estimate (cold) gas mass estimates precise redshifts!

4. Bolometric Luminosity vs. Stellar Mass Reddy et al 2006; see also Papovich et al 2006 Note that only ~20% of “red” z~2 galaxies are consistent with being passive. The rest are dusty LIRG-ULIRGs.

5. Typical mean velocities are ~200-400 km/s with respect to nebular line (systemic) redshifts Typical line widths are ~650 km/sec (barely resolved in typical spectra) Difference between Ly alpha emission peak and interstellar absorption centroid is ~650 km/s at both z~3 and z~2; full velocity fields of +/- 500- 1000 km/sec Close to unity covering fraction is common Lines of all accessible ionization stages seen (e.g., OI to OVI) Galactic Scale Outflows are Ubiquitous @ high redshift Ly  emission relative to H  emission relative to H  emission CS et al 2006

7. Blue: M bar <5x10 10 M  Red: M bar >5x10 10 M  (M bar =M * +M gas ) V max ~800 km s -1 Sample of ~100 galaxies with accurate (H  ) systemic redshifts, good UV spectra CS et al 2006

8. OSIRIS+LGSAO: Spatially Resolved Emission Line Maps of z~2-3 Galaxies Collapsed H  line flux maps for galaxies at z ~ 2 which show no coherent velocity shear and  ~ 80-100 km/s. Collapsed [O III] line flux map for a galaxy at z ~ 3 which shows velocity shear of ~ 150 km/s over 4 kpc. Shear was previously undetected with long-slit spectrographs. 0.05” Resolution Law et al 2006  v abs ~1000 km s -1

9. Stellar Mass/Metallicity Relation at z~2.2 Erb et al 2006 No Lum-Metallicity Relation Based on H-alpha (K-band) Spectroscopy of ~90 z~2-2.6 galaxies

10. Mass Comparisons, z~2 star forming galaxies (masses on physical scales of r~5 kpc) Erb et al 2006b Dynamical vs. M*Dynamical vs. (M gas +M*)

11. Gas Fraction vs. Stellar Mass, z~2-2.5 star forming galaxies (H  sample, rest-UV selected) Erb et al 2006b Gas fraction inferred from H  surface brightness, inverting Kennicutt/Schmidt Law

12. “Specific Star Formation Rate” vs. Stellar Mass @z~2 SFR/M* Reddy et al 2006 Blue/yellow Blue/yellow dots: rest-UV selected Red boxes: DRGs triangles: BzK triangles: BzK

13. Simple Chemical Evolution Models, z~2 Galaxies (Cold) Gas Fraction (  ) Metallicity Erb et al 2006a Curves: Black: Closed box (no outflow) Blue: outflow=1xSFR Green: outflow=2xSFR Red : outflow=4xSFR Yield of Salpeter IMF

14. What Does All of this Mean? Almost looks like galaxies receive their full gas supply at the beginning of an extended starburst episode, which is then “processed”, much of which (~80%?) leaves the galaxy and ~20% of which becomes stars. –a typical galaxy in SF samples is 50% cold gas and 50% stars in the central few kpc @z=2-3 (accounts for ~30% of total baryons given halo masses of ~10 12 M  ). –low stellar mass objects in current samples are not low mass objects, they are just almost entirely gas and very young. Range of properties among star forming galaxies is dictated by where we catch them in working their way through their gas supply Because high stellar mass galaxies have low gas fractions, both the ISM metallicity and the opacity increase very rapidly, hence they are extremely dusty just prior to running out of gas. –clear evidence for increasing prominence (or duty cycle) of AGN in high stellar mass, low gas fraction objects (~25% compared to ~3-4% in lower stellar mass objects) –most of the gas is already gone by the time this occurs (though SFR can still be fairly high).

15. Use dense sampling of galaxies, AGN to try to directly detect signatures of feedback (both radiative and hydrodynamical)  true IGM “tomography” is possible now. Galaxy/AGN/IGM Interface @High Redshift

16. z QSO Background Foreground Densely Sampling the Universe @z~1.8-3.2

17. Galaxy/CIV Cross-Correlation: 542 CIV systems, 1044 Galaxies Above N(CIV)~10 13, galaxy/absorber cross- correlation is equal to or exceeds the galaxy auto- correlation function Suggests causal connection of the strongest systems with the observed galaxies. In any case, metals (as traced by CIV) are where the observed galaxies are. Adelberger et al 2005

18. QSO sightline, 85 kpc away Galaxy BX210:  v~660 km/sec v max ~ -600 km/sec

19. 2 sightlines separated by 0.45 kpc @76 kpc distance from z=2.356 galaxy (separate gravitationally lensed images of the QSO)

20. Z=2.109 Z=2.012 Separation: 43 kpc (proper) MD80: M*=5x10 9 M sun BX513: M*=10 11 M sun High ions >> low ions, at same velocity in MD80 spectrum and in BX513 spectrum ~20 hours, Keck/LRIS-B

21. Where are the QSOs/AGN relative to the galaxies @z~2-3? Adelberger & CS 2005a,b

22. Where are the QSOs/AGN relative to the galaxies? Adelberger & CS 2005 galaxy-galaxy correlation length z~2-3 galaxies

23. Inferences about Black Hole Masses and Host Halos at z~2-3 Adelberger & Steidel 2005b

24. Environments of Very Bright QSOs (m~16-17) @z  2.5-2.8 In 12 fields of ultraluminous QSOs, only 1 bright QSO is in a substantial over- density: HS1549+19 @z=2.84

25. ~7’x5’, ~140 objects z~1.8-3.2

26. Two m~24 QSOs @ same redshift as HS1549 (2.845)

27. AGN and Star-forming galaxies at z~2-3.5: Bright and faint (optical) AGN, over ~ 10 mags of dynamic range, inhabit essentially the same dark matter halos and the same environments as the blue star-forming galaxies in the spectroscopic sample –r 0 ~5h -1 Mpc at z~2-3, independent of AGN (UV) luminosity –Significantly weaker clustering than red galaxies at similar redshifts (cf. Quadri et al 2006- see also his poster) Faint AGN have much longer duty cycle (they are “on” for a much larger fraction of the time) than bright ones Apparently, quasars cluster like ~ L* star forming galaxies (galaxies with the largest gas supplies) at all redshifts –e.g., Coil et al 2006 z~1: r 0 ~3.7h -1 Mpc

28. HS1700+64 z=2.300 Proto-cluster 186 spec. 1.7<z<3.2 8 AGN Data Include: Optical and near-IR spectroscopy (Keck) Deep J,K imaging (Palomar) Deep H  (2.17  m) (Palomar, 25 hours) Ly  imaging (in progress) HST/ACS Chandra (coming soon) Spitzer/IRAC, extremely deep MIPS

29. Inferred Properties vs. Environment Spike: =1450 Myr Field: =700 Myr spike / field =1.8 CS et al 2005

30. HS1700+64 z=2.300 Proto-cluster “Map” Black: spec. confirmed z=2.30 (rest-UV selected) Blue: H  n.b. candidate z=2.30 Red: all DRGs K<21.0 Green: Ly  n.b. candidate

31. “(Some) Ways Forward” Establish the details of the spatial relationship between galaxies and AGN/QSOs at the redshifts where big galaxies are emerging. –Lots of spectroscopy is required, both optical and near-IR Understand “feedback” processes: beyond “hand waving” –Would like in situ, not just circumstantial, evidence –Probably involves observing gas, not stars Do we really understand the implications of “down-sizing”? –Are we missing something important in our understanding of how gas cooling/SF works at high redshifts?