Clustering of Luminous Red Galaxies and Applications to Cosmology

Clustering of Luminous Red Galaxies and

Applications to CosmologyNicRoss (Penn State)

Research Progress MeetingLBNL

8th November 2007

Ross et al., 2007, MNRAS, 381, 573Ross et al., 2007, MNRAS submitted,

astro-ph/0704.3739Cannon et al., 2006, MNRAS, 372, 425

Outline of talkOutline of talk•Motivation•The 2dF-SDSS LRG And QSO (2SLAQ)

Survey•Clustering techniques•2SLAQ LRG Clustering results and z-

space distortions•The AAOmega LRG Pilot Survey•Future BAO Surveys, The B.O.S.S.

Outline of talkOutline of talk•Motivation•The 2dF-SDSS LRG And QSO (2SLAQ)

Survey•Clustering techniques•2SLAQ LRG Clustering results and z-

space distortions•The AAOmega LRG Pilot Survey•Future BAO Surveys, The B.O.S.S.

~70%

~26%

4%

What is the Universe made What is the Universe made of?of? Baryonic matter

Dark matter

“Dark Energy”

Evidence from: SNeIa, CMB, LSS, (Clusters)

MotivationMotivation• Luminous Red Galaxies (LRGs) provide a very

good observational sample to test models of galaxy formation and evolution.

• Excellent tracers of Large Scale Structure (LSS).

• Semi-Analytic Model predictions - lines, LRG Observations - stars; z=0.24 (L), z=0.50 (R)

• Almedia et al. 2007 (astro-ph/0710.3557)

Motivation, observations vs. Motivation, observations vs. modelsmodels

MotivationMotivation• Luminous Red Galaxies (LRGs) provide a very

good observational sample to test models of galaxy formation and evolution.

• Excellent tracers of Large Scale Structure (LSS).

• 2 Point Correlation Function (2PCF) simple but powerful statistic.

Eisenstein et al. 2005 (ApJ, 633, 560)

• `Bump’ in the 2 Point Correlation Function at ~100 h-1 Mpc

• Due to ``baryon acoustic oscillations’’

• Can be used as a Standard Ruler, determine geometry of the Universe

MotivationMotivation

MotivationMotivation• Luminous Red Galaxies (LRGs) provide a very

good observational sample to test models of galaxy formation and evolution.

• Excellent tracers of Large Scale Structure.• 2 Point CF simple but powerful statistic.• Redshift-space distortions can provide

cosmological parameter constraints via Alcock-Paczyncski and clustering evolution (explained in due course)

• 2SLAQ extending SDSS LRGs to 0.4<z<0.8. • Extended redshift arm led to photo-z calibration• Proof of concept for future LRG studies e.g. BOSS

Outline of talkOutline of talk•Motivation•The 2dF-SDSS LRG And QSO (2SLAQ)

Survey•Clustering techniques•2SLAQ LRG Clustering results and z-

space distortions•The AAOmega LRG Pilot Survey•Future BAO Surveys, The B.O.S.S.

SDSS DR6: 8417 deg2, 1,271,680 spectra, 790,860 gal

LRG Photometric Selection, LRG Photometric Selection, grigri--bandsbands• Method: Use

SDSS photometry, gri-bands, to select intrinsically luminous (L > 3L*) red galaxies from z0.0 to 0.8.

• Bruzual and Charlot (2003) evolutionary model tracks superimposed on the SDSS data.

• Star/galaxy separation from SDSS images. Some red M-type stars remain.

20’’

SDSS LRG vs. 2SLAQ LRG SDSS LRG vs. 2SLAQ LRG N(z)N(z)

• SDSS LRG sky density is 12 deg-2

• 2SLAQ LRG sky density is 53 deg-2

• Same populations, different redshifts

• 2SLAQ LRG Survey: 13,121 LRGs 17.5 < i < 19.8 80 fields giving

total area 180 degs2

92% spectroscopic completeness

(P. Weilbacher)

2dFGRSSDSS LRG2SLAQ

Outline of talkOutline of talk•Motivation•The 2dF-SDSS LRG And QSO (2SLAQ)

Survey•Clustering techniques•2SLAQ LRG Clustering results and z-

space distortions•The AAOmega LRG Pilot Survey•Future BAO Surveys, The B.O.S.S.

• represents the excess probability of finding a PAIR of galaxies compared with a random distribution:

• Power Law behaviour:

• Measure the redshift-space CF which include peculiar velocities due to cluster infall and random motions leading to “redshift-space distortions”.

• Can measure in two dimensions, with , perpendicular and , parallel, to line-of-sight where,

and effects of z-space distortions seen

The 2 Point Correlation The 2 Point Correlation FunctionFunction

€

dP12 = n2 1+ ξ (r)( ) dV1 dV2

€

(r) = rr0

⎛ ⎝ ⎜

⎞ ⎠ ⎟−γ

)(s

222 +=

€

(σ ,π )

, sloper0, correlation length

€

(r)

Hawkins et al. (2003, MNRAS, 346, 78)

a=0 kms-1

=0

a=500 kms-1

=0

a=0 kms-1

=0.4

a=500 kms-1

=0.4

The 2 Point Correlation The 2 Point Correlation FunctionFunction

(perpendicular to the l.o.s.)

(a

long

the

l.o.s

.)

• Redshift-space, (s), and Real-space, (r), CFs related by (Kaiser 1987):

• Also, and M related (Peebles 1980; Lahav1991; Peacock+ 2001; Hawkins+ 2003; Zehavi+ 2002)

where b is the `linear bias parameter’, which is the ratio of galaxy to (underlying) mass fluctuations; g = b2 m

• This linear bias, b, important because it reduces the fractional error due to shot noise, i.e. b , no. of objects needed (e.g. Blake&Glazebrook, 2003; Tegmark 2006)

The 2 Point Correlation The 2 Point Correlation FunctionFunction

⎟⎠⎞⎜⎝

⎛ ++= 2

51

321)()( ββξξ rs

bzM )(6.0

≅

Outline of talkOutline of talk•Motivation•The 2dF-SDSS LRG And QSO (2SLAQ)

Survey•Clustering techniques•2SLAQ LRG Clustering results and z-

space distortions•The AAOmega LRG Pilot Survey•Future BAO Surveys, The B.O.S.S.

2SLAQ LRG 2SLAQ LRG RedshiftRedshift-space 2PCF-space 2PCF• 2SLAQ LRGs at

redshift z = 0.55, r0=7.45 +/- 0.35 h-

1 Mpc• SDSS LRGs

(Zehavi et al. 2005) at z=0.28, r0=9.80+/-0.20 h-1 Mpc

• 2dFGRS Luminous Early-Type (Hawkins et al. 2003; Norberg et al. 2002), at redshift z≈0.1, r0≈6 h-1 Mpc

2SLAQ 2-D Correlation Function, 2SLAQ 2-D Correlation Function, ((,,))

(perpendicular to the l.o.s.)

(a

long

the

l.o.s

.)

• Ratio of observed radial size to angular size, varies with cosmology:

• Have an intrinsically isotropic structure (e.g. the clustering of galaxies) and observe the radial/angular ratio (A&P, 1979; Ballinger, Peacock, Heavens ’96; Popowski+’98; Hoyle+’02; da Angela+’05; Kim&Croft+’07)

• Pros: Geometric test to determine ; no messy galaxy evolution; v. complimentary to BAOs.

• Cons: Peculiar velocities also affect isotropic structure.

Alcock-Paczynski Test Alcock-Paczynski Test

δθδz

z

• Compare data to range of test cosmologies

• Degeneracy along M-, recall

• Additional info: e.g. (z=0), 2dFGRS (Hawkins+ 2003)

= 0.45

M =0.25

M

(0)

(z)

1.0

00 1.0

Cosmological ConstraintsCosmological Constraints

05.0+03.0−

+0.10- 0.15

bzM )(6.0

≅

Outline of talkOutline of talk•Motivation•The 2dF-SDSS LRG And QSO (2SLAQ)

Survey•Clustering techniques•2SLAQ LRG Clustering results and z-

space distortions•The AAOmega LRG Pilot Survey•Future BAO Surveys, The B.O.S.S.

Baryon Acoustic OscillationsBaryon Acoustic Oscillations• In the tightly coupled

baryon-photon fluid prior to Recombination, acoustic waves create a characteristic scale – the sound horizon, RS.

• At Recombination, vs 0, wave stalls and the imprint of RS, (BAO), is frozen (but still evolves gravitationally) in the matter, and later, galaxy correlation functions.

• Rs(z=1089) can be determined accurately with CMB (148 Mpc), so BAO become a very promising standard ruler, DA(z) and H(z).

• Wayne Hu, http://background.uchicago.edu/~whu/• Martin White, http://cdm.berkeley.edu/doku.php?id=baopages

Future LSS BAO LRG Future LSS BAO LRG ProjectsProjects• What you need (minimum):

Large volume of Universe (>1 Gpc3) Large number of objects (105 - 106)

• Our Idea: ~300,000 LRG redshift surveyover ~3,500 sq. degs. (95 deg-2) with redshift 0.5 < z < 1.0

• Requirements: Imaging from SDSS and VST-ATLAS LRGs down to i < 20.5 in ~1 hour 4m-class telescope, reasonable site Multi-object spectrograph (Australian)

AAOmeAAOmegaga• AAOmega,

392 fibre MOS, on 4m AAT

• Blue and Red arms

• 5600-8800Å 4000 Å @ z = 0.4 - 1.2 (~0.9).

•`Large’ ~200 night proposal•Use LRGs to measure BAO at <z>=0.7•Pilot Program 03 Mar 2006 – 07 Mar 2006

LRGLRG riz riz-selection-selection

QuickTime™ and aTIFF (Uncompressed) decompressor

are needed to see this picture.

• Again, use SDSS imaging

• Select in riz-bands, down to i<20.5 (cf. 2SLAQ i<19.8)

• High stellar contamination, can do better

• Panels show selection areas and model tracks

AAOmega LRG Pilot RunAAOmega LRG Pilot Run• Mean

redshift, <z>=0.68

• Exposure times to get absorption line redshifts varied from 1 to 3 hours, (typically 1.67hrs)

AAOmega LRG Correlation AAOmega LRG Correlation function function

• wp() projection of 3-D (s) • AAOmega LRG r0 = 9.03+/-0.93• AAOmega LRGs sample now comparable to SDSS LRGs for LSS studies• However…

Outline of talkOutline of talk•Motivation•The 2dF-SDSS LRG And QSO (2SLAQ)

Survey•Clustering techniques•2SLAQ LRG Clustering results and z-

space distortions•The AAOmega LRG Pilot Survey•Future BAO Surveys, The B.O.S.S.

Baryon Oscillation Spectroscopic Baryon Oscillation Spectroscopic SurveySurvey• PI: David Schlegel, part of “SDSS-III”

Use existing 2.5m telescope, upgrade optics and spectrographs

1.5x106 LRGs, 0.2<z<0.8 over 10,000 deg2

160,000 Ly Forests from QSO sightlines, 2.3<z<2.8, 8,000 deg2

• dA to 1, 1.1 and 1.5% at z~0.35, 0.6, 2.5• Due to start imaging in latter half

2008, spectroscopy 2009

ConclusionsConclusions• 2SLAQ LRG Survey complete, >13,000 LRG

spectroscopic redshifts, 0.4 < z < 0.8.• 2SLAQ LRGs have r0=7.45+/-0.35 h-1 Mpc.• Using dynamical (peculiar velocity) and

geometric (Alcock-Paczynski) information find: M=0.25+0.10

-0.15 and (z=0.55) = 0.45+0.05-0.03

• Alcock-Paczynski test, v. complimentary to BAOs

• SDSS riz-band selection pushes <zLRG>=0.7• Future Projects, e.g. BOSS (LRG & Ly) Survey

CreditsCredits• Tom Shanks, Jose da Angela, Phil Outram,

Alastair Edge, David Wake (Durham)• Bob Nichol (Portsmouth)• Russell Cannon, Scott Croom, Rob Sharp

(AAO)• Michael Drinkwater, Isaac Roseboom, Kevin

Pimbblet (UQ)• Daniel Eisenstein (Arizona)• John Peacock (Edinburgh) • And Nikhil P. and Martin W. for inviting me.