1. Measuring the clustering of galaxies in COSMOS Measuring the clustering of galaxies in COSMOS Olivier Le Fèvre, LAM Why ? Why ? How ? correlation function How ? correlation function What has been done from other datasets ? What has been done from other datasets ? Toward a strategy for COSMOS Toward a strategy for COSMOS –Use the unique parameter space –Capitalize on morphology measurements

2. Clustering measurements: a unique tool Characterize mass assembly over time Characterize mass assembly over time Clustering is progressing following the underlying cosmology frame Clustering is progressing following the underlying cosmology frame Probing scales from ~10kpc to ~10-20Mpc Probing scales from ~10kpc to ~10-20Mpc Need large areas Need large areas Measure: Measure: –Correlation length r 0 –  =  0.6 /b –pair-wise velocity dispersion Caveat: tracing clustering with luminous matter Caveat: tracing clustering with luminous matter

3. Standard method Compute 2-point correlation function  (r p,  ) and projection w(r p ) Compute 2-point correlation function  (r p,  ) and projection w(r p ) Higher order statistics: a powerful tool in COSMOS field Higher order statistics: a powerful tool in COSMOS field 2DFGRS COSMOS-Subaru, Guzzo et al.

4. Constraints on COSMOS: what has been done before ? COSMOS do not need to repeat previous experiments COSMOS do not need to repeat previous experiments Many other datasets with deep multi- wavelength and/or HST morphology information exist Many other datasets with deep multi- wavelength and/or HST morphology information exist –Photo-z –Spectro-z –Wide fields / deep fields No other single daset with same wide field, depth AND HST morphology No other single daset with same wide field, depth AND HST morphology

5. Current measurements: VVDS VVDS-Deep VVDS-Deep 9000 spectroscopic redshifts, IAB=24 9000 spectroscopic redshifts, IAB=24 0.7x0.7 deg², up to z~2 0.7x0.7 deg², up to z~2

6. VVDS-0226-04 cone: Galaxy density field, 6217 redshifts I AB  24 (C. Marinoni et al.) 2DFGRS/SDSS stop here z=0.5 z=0.6 z=0.7 z=1.3 z=0.8 z=0.9 z=1 z=1.1 z=1.2 160Mpc 30Mpc COSMOS will be 3x larger scales

7. VVDS: measuring clustering evolution from z~2 Le Fèvre, Guzzo, et al., A&A, in press

8. Evolution of galaxy bias Marinoni, et al., A&A, submitted Clustering length Full population Clustering length per galaxy type

9. Strategy for clustering measurement with COSMOS Unique to COSMOS: bringing together depth, morphology, and wavelength selection, over a large contiguous field Unique to COSMOS: bringing together depth, morphology, and wavelength selection, over a large contiguous field Compute  (r,M( ),type,z) Compute  (r,M( ),type,z) –Photo-z –Spectro-z Probe the evolution of different types of galaxy populations Probe the evolution of different types of galaxy populations –Early types: when are they already in place in the densest environments ? –Star forming galaxies –Try to link populations in evolution scenario

10. key parameter at high-z: morphology Select large morphology-selected sub-samples Select large morphology-selected sub-samples –assume 500-1000 galaxies for a proper calculation of wp ACS-based ~3500A rest z~0.8: ACS-based ~3500A rest z~0.8: –~800,000 images: 1600 bins (z,type,M) NICMOS-based, ~1  m rest: NICMOS-based, ~1  m rest: –~20,000 images: 40 bins Need to train machine on eye classification Need to train machine on eye classification –Reference sample need to be defined for COSMOS

11. H-band NICMOS images: a very important dataset H band improves photo-z for z>1 H band improves photo-z for z>1 Compare ACS and NICMOS morphology measurements Compare ACS and NICMOS morphology measurements –Understand differences On-going work: Lidia Tasca et al. Paper for ApJ 1st release NICMOS tile originalGIM2D Model MaskResidual

12. Clustering analysis of COSMOS data for 1st ApJ release Evolution of the correlation function per morphological type: Evolution of the correlation function per morphological type: –I band selected to z~1.2 –H band selected to z~1.5 –Photometric redshifts Measurements at z~3-5 ? Measurements at z~3-5 ? –Need iteration on photometric catalog On-going work: Baptiste Meneux et al. Paper for 1st ApJ release

13. Select by ACS morphology early type late type irregulars Select by NICMOS morphology early type late type irregulars Select by apparent magnitude I band B band H band NUV (Galex) Spitzer 3.6-8 microns Select by absolute magnitude / volume limited: UV to 1  m rest Two-point correlation function I. Photo-z: down to I~26 all galaxies selected from z~0 to z~1.2 selected populations: LBGs z~3-5, EROs II. Spectro-z: down to IAB~22.5 all galaxies from z~0 to z~1.2 (zCOSMOS-bright) all galaxies z~1.4-2.5 (zCOSMOS-deep)

14. Photo-z: comparing different methods On-going work: O. Ilbert, L. Tasca, S. Arnouts Mobasher photo-z code Arnouts / Ilbert photo-z code