1. Luminous Red Galaxies in the SDSS Daniel Eisenstein ( University of Arizona) with Blanton, Hogg, Nichol, Tegmark, Wake, Zehavi, Zheng, and the rest of the SDSS team.

2. Outline of the Talk I.SDSS and Galaxies II.The Luminous Red Galaxy Sample III.Stacking Spectra IV.Clustering V.Clustering with Cross-Correlations

3. The Sloan Digital Sky Survey  Data Release 3 is today! 5282 square degrees of imaging (ugriz) to r~22.2. 5282 square degrees of imaging (ugriz) to r~22.2. 500,000 spectra over 4188 sq. deg. R~1800. 500,000 spectra over 4188 sq. deg. R~1800. 375k galaxies375k galaxies 50k quasars50k quasars 70k stars70k stars http://www.sdss.org/DR3

4. I. Local Galaxy Properties  SDSS has made high-precision measurements of the properties of local galaxies.  This sets the baseline for quantification of evolution.

5. Multivariate Distributions  With massive statistics, we can explore high- dimensional spaces of galaxy parameters. Magnitude Color(s) Surface Brightness SB Profiles Velocity Dispersion Environment Emission Lines  Make these distributions yourself with the NYU Value-Added Galaxy Catalog. http://wassup.physics.nyu.edu/vagc http://wassup.physics.nyu.edu/vagc http://wassup.physics.nyu.edu/vagc Blanton et al. (2003)

6. Color Bimodality Blanton et al. (2003) Red Sequence Massive Galaxies

7. II. Introduction to SDSS LRGs  SDSS uses color to target luminous, early-type galaxies at 0.2<z<0.5. Fainter than MAIN (r<19.5) Fainter than MAIN (r<19.5) About 15/sq deg About 15/sq deg Excellent redshift success rate Excellent redshift success rate  The sample is close to mass-limited at z<0.38. Number density ~ 10 -4 h 3 Mpc -3  Co-conspirators: Annis, Connolly, Gunn, Nichol, Szalay  Science Goals: Clustering on largest scales Galaxy clusters to z~0.5 Evolution of massive galaxies

8. 55,000 Spectra

9. A Volume-Limited Sample

10. Massive Galaxies Evolve Slowly

11. Luminosity Function of LRGs  Using data from the SDSS-2dF LRG project. Using SDSS imaging to select LRGs at 0.5<z<0.7 for spectroscopy at 2dF. Using SDSS imaging to select LRGs at 0.5<z<0.7 for spectroscopy at 2dF. 5500 redshifts at z>0.4. 5500 redshifts at z>0.4. Preliminary from Wake et al, in prep

12. III. Spectral Analyses of LRGs  Stack spectra in bins of luminosity & environment. This is like studying the ensemble of stars. This is like studying the ensemble of stars.  Environment and luminosity form a 1-dimensional set of stacked spectra. 90% of all spectral variation is in first PCA component. 90% of all spectral variation is in first PCA component.  “Environment is Age; Luminosity is Metallicity” does not match data. Eisenstein, Hogg, et al. (2003)

13. IV. Clustering of Galaxies: Luminosity Dependence  Clustering is seen to depend strongly on luminosity, esp. above L *.  This is expected if more luminous galaxies live in more massive halos. Tegmark et al. (SDSS, 2003)

14. Luminosity and Color Dependences  The mean environment around galaxies is a strong function of luminosity and color.  Sharp upturn for the most massive red galaxies. Hogg, Blanton, DJE et al. (2002)

15. Zoom in on the Massive End  We’re going to look at the high- luminosity end with much more precision.  Goal: Probe the halo occupation of these massive galaxies and the role of environment in building these system. Tegmark et al. (SDSS, 2003)

16. LRG Correlation Functions   8 = 1.80±0.03 up to 2.06±0.06, r 0 = 9.8h -1 up to 11.2h -1 Mpc  Obvious deviations from power-laws! Zehavi, DJE, et al. (2004)

17. Halo Occupation Modeling  The distribution of dark matter halo masses for the galaxies determines their clustering.  Generically predict an inflection in  (r). Zehavi et al. (2004); Zheng et al, in prep.

18. V. Cross-Correlation Analysis  Angular correlations of objects with known redshift with objects of unknown redshift.  Only physical correlations (save weak lensing) must be at the known redshift. Therefore, we can map angles to transverse distance and flux to luminosity. Therefore, we can map angles to transverse distance and flux to luminosity. E.g., can take a fixed metric aperture and track galaxies of a constant luminosity bin. E.g., can take a fixed metric aperture and track galaxies of a constant luminosity bin.  Background subtraction: yields projection of correlation function: w p = Int  ([R 2 +z 2 ] 0.5 ) dz.

19. A New Method  Statistical isotropy permits deprojection.  However, this usually requires dw p /dr.  Instead, consider an integral of the correlation function.  This reduces to a weighted sums over pairs.  Very flexible! Eisenstein (2002)

20. Application to LRG Sample  30,000 spectroscopic LRGs.  16 million galaxies from SDSS imaging.  At each LRG redshift, select imaging galaxies in range M*-0.5 to M*+1.0.  Measure overdensity of L* galaxies around LRGs at radii from 200 kpc to 6.4 Mpc, as a function of LRG luminosity. Eisenstein et al. (2004)

21. Environment (200 kpc) vs. Luminosity

22. What are these numbers?  For  0  =10, we have:  0 V  = 0.92 galaxy near the LRG, weighting the count by W(r).  0 V  = 0.92 galaxy near the LRG, weighting the count by W(r).  (200 kpc) = 500.  (200 kpc) = 500.  0  =10 gal h 3 Mpc -3 approx. mean density 200 kpc from LRG.  0  =10 gal h 3 Mpc -3 approx. mean density 200 kpc from LRG.  If  scales as b LRG b *, then b LRG varies by 4!

23. Environment (200 kpc) vs. Luminosity

24. Environment (400 kpc) vs. Luminosity

25. Environment (800 kpc) vs. Luminosity

26. Environment (1.6 Mpc) vs. Luminosity

27. Environment (3.2 Mpc) vs. Luminosity

28. Environment (6.4 Mpc) vs. Luminosity

29. Clustering on 5 Mpc scales

30. Smaller Scales, in comparison  Can passive evolution from z>1 do this? Or must we invoke environmental effects?

31. Deviations from Power Law  Opportunity for detailed modeling of halo occupation.

32. Conclusions  The SDSS is providing enormous amounts of data on massive galaxies at low redshift (z<0.5). Quantitative baseline for evolution comparison. Quantitative baseline for evolution comparison.  Stacked spectra refute “Environment = Age; Luminosity = Metallicity” hypothesis.  High-precision measurement of clustering of massive galaxies and L* galaxies. Small-scale bias is a very steep function of luminosity, a factor of 4 from 2L* to 8L*. Small-scale bias is a very steep function of luminosity, a factor of 4 from 2L* to 8L*. Luminosity dependence is steeper on small scales. Luminosity dependence is steeper on small scales.  Future work: Halo occupation. Evolution of LRGs.

33. Take Our Data. Please!

35. Scale & Luminosity Dependent Bias

36. Red Fractions vs. Scale & Luminosity

37. Red Fractions versus Time