1. Formation and Clustering of High-redshift Galaxies 3. Galaxy Clustering
Eric Gawiser
Rutgers University
MUSYC E-HDFS UBR composite

2. Protogalaxies at z=3: TLAs
LBG=Lyman Break Galaxy
selected via Lyman break, blue continuum (starburst)
LAE=Lyman Alpha Emitter
selected via strong emission line (early stage of star formation)
DRG=Distant Red Galaxy
selected via Balmer break in observed NIR
SMG=Sub-Millimeter Galaxy
selected in sub-mm, use radio to get position
DLA=Damped Lyman  Absorption system
selected in absorption, N(HI)>1020 cm-2

3. Images from HST-ACS: irregular morphology at z=3
AGN
z=3.60
R=22.4
LBG
z=3.37
R=24.3
LBG
z=3.24
R=23.8
From 10kpc resolution to 1 kpc resolution
Need Spitzer to see real morphology of most of stellar mass, not just rest-frame UV,
LAE
z=3.10
R=26.1

4. ECDFS RJK
Does anyone know why stars look blue? Because falling spectra from J to K as peak of graybody is between them.

5. NIR selects rest-frame Balmer/4000Å break at 2<z<4
Balmer vs A break
Reddy et al 2005

6. Distant Red Galaxies (DRG)
SED fitting results
Don’t be fooled by nice template spectrum - only photometry is observed!
van Dokkum et al 2005, in prep.
MUSYC
van Dokkum et al 2006

7. Sub-Millimeter Galaxy contribution to Star Formation Rate Density
Chapman et al 2005

8. LBGs and LAEs in MUSYC-ECDFS
Which do you think has strongest clustering? Vote…
Angle -sure. LBGs almost poisson but small-scales more clumps and voids
Which clusters more strongly in 3-D? How can we determine that? Vote…
1240 LBGs
162 LAEs

9. Spatial and angular cross-correlation functions:
dP(r) = 12[1 + 12(r)] dV1 dV2
dP() = 12[1 + w12()] d1 d2
Projection of 12(r )=(r/r0)- into w12() = dz1 dz2 p1(z1)p2(z2) 12(r(z1,z2, )) Need redshifts to determine selection functions pi(z) for inversion of w12() to determine 12(r)
For those who had been lamenting the lack of equations in this talk…
If 1=2, auto-correlation
Xi is not quite a power-law, just a convenient approximation
Limber is widely used, but be careful for cross-correlation (p1*p2) or very thin delta_z
For autocorrelation and acceptable geometry,
Limber approximation 
w() = 1- r0 (1/2, (-1)/2) p2(z) (1+z)1- DA(z)1- H(z)/c dz

10. LBG and LAE redshift distributions
Point out much narrower z range for LAEs - re-vote!
2.8<z<3.7 expected for UVR
Dark curve shows selection function: narrow-band filter response convolved with EW distribution

11. Measuring angular auto-correlation
(): Excess probability of finding pairs separated by angular distance  over uniform distribution
Landy-Szalay estimator uses counts of pairs separated by 
DD: between data pairs
DR: data-random pairs
RR: random-random pairs
The random pairs “subtract” the excess probability that is due to the geometry of the survey
Usually assumed that () = A-

12. LBGs and LAEs in MUSYC-ECDFS

13. LBGs in MUSYC-ECDFS
Clustering analysis by
Harold Francke
1240 LBGs

14. Clustering analysis by H. Francke
LAEs in MUSYC-ECDFS
Much stronger w(theta) caused by narrow slice of LSS I.e. narrow redshift distribution - actually implies
Less spatial clustering than LBGs
Clustering analysis by H. Francke
162 LAEs

15. Clustering Determination
LBG, LAE, and DRG samples are large enough to use r0 to determine bias
xLBG-LBG( r ) = (r/r0)- =b2LBG xDM( r )
SMG, DLA samples are small, so study cross-correlation with numerous LBGs to determine bias
xDLA-LBG( r ) =(r/r0)- = bDLAbLBG xDM( r )
bLBG etc. determine typical dark matter halo masses of each family of protogalaxies
Method for auto-correlation from Mo & White 1996, MNRAS 282, First applied to cross-correlation by Gawiser et al 2001, ApJ 562, 628
see also Infante et al 2003, ApJ 588, 90 for QSO-LAE cross-correlation
Only do analysis at scales >1 Mpc where halo occupation effects are minimal

16. Bias  minimum DM halo mass number abundance of host halos
MUSYC
Quadri et al 2007

17. Clustering vs. halo abundance

18. What are the low-redshift descendants of z=3 galaxies?
Fresh research result!
LAE
Gawiser et al 2007, in prep

19. 5 Unsolved Problems in Galaxy Formation
What does a protogalaxy look like?
Did galaxy, stars, supermassive black holes all form simultaneously?
Possible research topics in galaxy formation…

20. 5 Unsolved Problems in Galaxy Formation
What does a protogalaxy look like?
When/how did each component form?
Thin disk: 10 Gyr - formed at z~2 but simulations have trouble making. Angular momentum coupling between DM & baryons affects bar/disk formation and bulge cuspiness.
Globular clusters: formed by Pop III stars in M halos?

21. 5 Unsolved Problems in Galaxy Formation
What does a protogalaxy look like?
When/how did each component form?
When/how did galaxy sequences evolve?
Hubble sequence not yet present at z>2
Red/blue sequences (bimodality of properties)
require “gastrophysics”

22. 5 Unsolved Problems in Galaxy Formation
What does a protogalaxy look like?
When/how did each component form?
When/how did galaxy sequences evolve?
What role did feedback play?
Feedback from AGN & supernovae regulates BH/bulge formation, cuspiness of DM halo, baryonic mass loss, IGM enrichment, minimum galaxy mass

23. 5 Unsolved Problems in Galaxy Formation
What does a protogalaxy look like?
When/how did each component form?
When/how did galaxy sequences evolve?
What role did feedback play?
When/how was the universe reionized?
Top-heavy IMF predicted at high-z due to low metallicity but exact mass range/epoch unknown and nature of “surviving” galaxies is sensitive

24. Coming Attractions Unification of galaxy formation and evolution
Needle-in-haystack techniques  evolved gals at z>2
Multiwavelength  high-z analogs at low-z
Evolutionary sequence
(e.g. DLALAELBGSMGDRG)
as part of “grand unified” model of galaxies & AGN
Spitzer is key
Shouldn’t extrapolate more than 6 years I’ve been working on GF
Spitzer is the rosetta stone to unlocking this holy grail

25. Coming Attractions Unification of galaxy formation and evolution
ISM in emission at high-redshift
CO, [CII] 158 micron with ALMA
Compare gas mass with stellar mass
Compare tips of high-redshift gas and stellar luminosity functions
Shouldn’t extrapolate more than 6 years I’ve been working on GF
The instrument formerly known as KAOS

26. Coming Attractions Unification of galaxy formation and evolution
ISM in emission at high-redshift
High-redshift galaxies constrain dark energy?
Baryon oscillations as standard rod - need z>1 point to constrain equation-of-state (w) of dark energy
A million redshifts needed? WFMOS!
Blake & Glazebrook 2003, Linder 2003, Seo & Eisenstein 2003
Shouldn’t extrapolate more than 6 years I’ve been working on GF
The instrument formerly known as KAOS

27. Coming Attractions Unification of galaxy formation and evolution
ISM in emission at high-redshift
High-redshift galaxies constrain dark energy?
More jargon
Sub-classes (sub-DLAs, bDLAs) may force FLAs!
N2 cross-correlation functions
Finally, a prediction I'm confident in!

28. MUSYC Public Data Release
June 1, 2007 at
UBVRIzK imaging of 1.2 square degrees to U,B,V,R = 26, K=22 (AB)
JHK imaging of 0.1 square degrees to K=23 (AB)
Deep Spitzer+IRAC imaging (all 4 bands) of ECDF-S
Feel free to contact me if you're thinking about using these data