Clustering and dusty high-z galaxies

Emanuele DaddiESO-Garching

(NOAO-Tucson)

• Properties of K-selected z=2 galaxies (K20/GOODS/other surveys) dusty SF galaxies• Efficient selection technique• Clustering

How, when did massive galaxies form ? Studying clustered high-z populations

Near-IR bright z~2 galaxies

Cimatti et al. 2002, K20 survey

K<20 at z~2 10^11 Msunstellar masses

Too many massive galaxiesat z>1.4 as compared towhat LCDM models hadpredicted

SFR~50—500 Msun/yr, Lya emission often absent,red UV continua, strong absorption lines, Irregular/merging morphologiesDaddi et al. 2004 ApJ 600 L127

Passive 1—2 Gyr galaxies, early-type morphologiesCimatti et al. 2004, Nature, 430, 184

BzK selection technique

Daddi et al. 2004 ApJ in pressastro-ph/0409041

BzK=(z-K)-(B-z) (AB mags)

BzK>-0.2 SF at z>1.4BzK<-0.2 & z-K>2.5 Old

94% of spec redshifts; K20 survey Technique calibrated with ~50 z>1.4 redshifts

Reddening independent selection of SF galaxiesAlso because of K-selection

Calzetti et al. 2000 extinction law

Would miss <10Myr bursts

Median E(B-V)~0.4 (from CSF models)

UnGRs colors of K<20 z>1.4 sample

BM/BX samples include K<20sources (Shapley et al. 2004)but miss most reddened SF(~20--30% of total SFRD)and passive galaxies

For K<20 sourcesDensity is ~ 10^-3 h^3 Mpc^-3comparable to local L>L* galsmuch smaller than LBGs

X-ray stacking

Lx~ 8x10^41 cgsHR < -0.54 (2-sigma)

star-forming galaxies <SFR> ~ 200 Msun/yr

Average 20cm radio emission also give 2-300 Msun/yr(Courtesy of K.Kellermann, J.Kelly; VLA data)

(Szokoly et al 2004; Gilli et al 2004)

SCUBA galaxies, or their remnants

Bridge to submm/mm populations ?

More time allocated SCUBA/IRAM (but slow going…)Need Spitzer and/or future facilities to detect dust emission, even more for dusty K>20 sources

6 objects:Photometric follow-up with SCUBA(with Rottgering, Cimatti, Kurk)

One detection 4.5mJy (850mu)Positive flux from the others(about 1mJy, 2.5 sigma)(preliminary numbers)

BzK colors to identify counterparts ?

See POSTER by Dannerbauer et al., this conference

But week correlation between K and submm flux, why ?• week SFR/Mass correlation ?• redshift effects ?• Large/variable obscuration even at K ?

Forming progenitors of massiveearly-type galaxies detected ?

• Large masses, near-IR bright (K<20@z~2)• Large SFRs (SFR>100 Msun/yr)• Metal enriched (Solar or more)• Density comparable to z=0 L>L* galaxies• Assembling by merging, perhaps some

Are these strongly clustered ?

Evidence for clustering

Enhanced clustering ofz~2—3 K-sel galaxies from FIRES, Daddi et al. 2003

Redshift pairings in the (SMALL)spectroscopic samples: e.g.Daddi et al. 2004; van Dokkum et al 2003

Monte Carlo simulationsof clustered samples:r_o > 7 h^-1 Mpc (95% c.l.)

Needs much larger sampleto measure clustering

ESO—Japan collaboration:ESO: Renzini, Cimatti, Mignoli, Pozzetti, Zamorani, Brusa, ED

Subaru: Arimoto, Ikuta, Kong,Onodera, Ohta, etc

450 z>1.4 candidates BzK-sel 350 arcmin^2 field to K=20 (Kong et al. in preparation)

~ 300 spectra taken with VIMOS

GMASS: ESO LP on UDF/K20/GOODS region

PI: A. Cimatti145 hours FORS2ongoing

Collaboration with GOODS

Selection:IRAC 4.5mu<23.5 ABzphot>1.4, BzK50 arcmin^2 region

raw+reduced data will be publicly released

GOODS/CDFS

BzK selection to K=22 Vega(incomlete for reddest objs)

VIMOS public observations: ~1000 BzKs (plus ~1000 UV-sel objects)PI: C. Cesarsky

Catalogue by L.Moustakas, B.Mobasher, M.Dickinson, et al.

COSMOS BzKs: 2 square degree, ~20000 BzKs

Submitted ESO LP (>500n)PI: S.Lilly

10000 BzKs to be observed+ 20000 UV-sel objects

Catalogue Subaru+KPNOB.Mobasher, Y.Taniguchi,N.Scoville(PI) H.Aussel, P.Capak

Angular clustering

About 450 K<20 z>1.4 candidates; ESO/Japan obs of EIS Deep3a field

r_o ~ 8+-2 h^-1 Mpc (fixed delta=0.8)from Limber equation (main uncertainty N(z))for BzK-SF

BzK-Old seem to have higher angular clustering

N(z)K20survey

Preliminary COSMOS results

A(1deg) ~0.008; S/N ~ 10 (fixed delta=0.8)

r_0 ~ 10+-2 h^-1 Mpc (main uncertainty N(z))

Evolution: M>10^11 Msun (L>~L*) galaxies

Tracks: evolution of DM halos fromMo & White 2002

dusty z~2 K-bright galaxies consistent to evolve into EROs and local gEs clustering comparable to SCUBA objs (Blain et al. 2004), higher density LBGs have r_o~4 h^-1 Mpc (Adelberger et al 2005)

GIF Models (Jenkins et al 1998; Kauffmann et al. 1999)

Correctly predict that at z>1.4objects with 10^11 Msun stellar mass are either old/deador very active starbursts

AND classify the active starburstsas early-type galaxies (bulgedominated, red)Agree with our conclusion thatnear-IR bright z>1.4 objectsevolve into massive E/S0s

(but predict x10 too few objects)

Conclusions

• Population of massive dusty SF galaxies at z~2: bridge between submm/mm and LBGs, forming L>~L* early-type galaxies • BzK selection is very efficient for 1.4<z<2.5 reddening

independent selection of SF galaxies• Clustering is an important measure for dusty high-z galaxies (link to evolution, DM halos properties, etc). Requires statistics (N_objs), wide/many fields, redshifts for high significance

measurements

Passive BzKs in UDFDaddi et al 2005 (in preparation)

ACS+HST spectra from GRAPES project (PI Malhotra)

ACSNICMOS imaging

Average UV spectra: de Mello et al. 2004 ApJ 608 L29

NGC 6090: Merger, dusty, LIRG, metal rich