ACS Observations of Distant Clusters of Galaxies Protoclusters
ACS Observations of Distant Clusters of Galaxies & Protoclusters: New Constraints on Cluster and Galaxy Formation & Evolution Marc Postman STSc. I March 17, 2004 STSc. I Colloquium 1
Talk Outline • Overview - what we know and don’t know (about cluster & cluster galaxy formation & evolution) • Sample definition • Results at z ~ 1 – – Constraining the SF history of early type cluster galaxies Tracking the evolution of cluster galaxy morphologies Brightest Cluster Galaxy assembly and evolution Galaxies, x-rays, and gravitational lensing: looking for evidence of cluster assembly • High-z RGs - Tracers of Protocluster sites? – TN 1138 - an intriguing system at z = 4. 1 - evidence in support of an associated protocluster • Epilogue - an update to the structure formation story March 17, 2004 STSc. I Colloquium 2
My ACS IDT Collaborators • Cluster Origins – – – John Blakeslee (JHU) Frank Bartko (Colorado) Nicholas Cross (JHU) Ricardo de. Marco (ESO/JHU) Holland Ford (JHU) Marijn Franx (Leiden) Brad Holden (UCSC) Nicole Homeier (JHU) Garth Illingworth (UCSC) Marco Lombardi (ESO) Felipe Manenteau (JHU) Piero Rosati (ESO) March 17, 2004 • Weak Lensing – Myoungkook Jee (JHU) – Rick White (STSc. I) – Narciso Benitez (JHU) • Hi-z Radio Galaxies – George Miley (Leiden) – Roderik Overzier (Leiden) – Andrew Zirm (Leiden) • ACS IDT Archive STSc. I Colloquium – D. Golimowski, K. Anderson – T. Allen, W. J. Mc. Cann – A. Framarini, G. Meurer 3
Cluster Formation Scenario • • • Primordial Intra-cluster medium (ICM) shocks at the intersection of matter streams and begins to emit x-rays (epoch? ? Probably z > 2). Since z = 1, there is little or no evolution in thermodynamics & metallicity of the ICM. Massive ellipticals assemble in range 2 < z < 5 (How does assembly proceed? Direct observations? ? ); they are the first galaxies to reach dynamic equilibrium with the cluster potential. (Global cluster effects? ) Most cluster galaxy SF activity is quenched by z=0. 5 (Dressler et al. 1997). (Trigger? ) Infall of spirals results in morphological and color gradients within the cluster (When are gradients first detectable? ). This process continues into the present epoch. S 0 and d. E populations develop within the cluster core (epoch? ? Timescales? ). HSB spirals S 0; LSB spirals d. E (Moore et al 1998) (Verify. Find progenitors at high-z? ? ) March 17, 2004 STSc. I Colloquium Simulation by J. Colberg (Starts at z = 20) 4
ACS GTO Cluster Survey • Cluster Galaxy Formation & Evolution – Goals: Study, with best precision to date, the star formation rates, internal structures, & fundamental relationships in cluster galaxies at z ~ 1. Study the origins of the most massive cluster galaxies and their link to possible progenitors. – 8 rich galaxy clusters in the range 0. 8 < z < 1. 3 – 4 high-z radio galaxies: Protoclusters at 2 < z < 5? – ACS imaging in at least 2 bands, NIR & x-ray imaging, plus extensive optical spectroscopy. March 17, 2004 STSc. I Colloquium 5
ACS GTO Cluster Survey • The ACS Advantage – High sensitivity makes moderately wide coverage in at least 2 passbands affordable for a modest size sample. – Higher angular resolution provides superior constraints on galaxy structure (e. g. , morphology, half-light radii) and lensing. – The combination yields superior photometric and morphological data that enables study of mass distributions and galaxy properties in z~1 clusters in unprecedented detail. March 17, 2004 STSc. I Colloquium 6
Cluster Survey Targets Cluster Redshift Velocity Dispersion X-ray Lum. (1044 erg/s) ACS Filters Total Orbits MS 1054 0. 831 (143) 1112 23. 3 V, i, z 24 CL 0152 0. 837 (72) 1283 7. 8 r, i, z 24 CL 1604 +4304 0. 897 (22) 1226 2. 0 V, I 4 CL 1604 +4321 0. 924 (44) 935 <1. 2 V, I 4 CL 0910 1. 101 (10) N/A 1. 5 i, z 8 CL 1252 1. 235 (34) 755 2. 5 i, z 32 CL 0848 -A, B 1. 265 (~40) 640 (A) 1. 5 (A), ~1 (B) i, z 24 March 17, 2004 STSc. I Colloquium 7
CL 0152 & MS 1054 @ z = 0. 83 CL 0152 (z=0. 837) March 17, 2004 MS 1054 (z=0. 831) STSc. I Colloquium 8
CL 1604+43 @ z ~ 0. 91 CL 1604+4304 (z=0. 897) March 17, 2004 CL 1604+4321 (z=0. 924) STSc. I Colloquium 9
CL 0910 + CL 1252 @ z > 1 CL 0910+5422 (z=1. 101) March 17, 2004 CL 1252 -2927 (z=1. 235) STSc. I Colloquium 10
Constraining the ages of early-type cluster galaxies from the Color-Magnitude Relation • “Red” Sequence in clusters – a galaxy population, largely of early-type morphology, with a relatively narrow range in color but spanning a modest range in luminosity. It is a common feature in most rich clusters to z < 1. • The CM relation is most likely due to an underlying mass-metallicity relation (e. g. , Kodama et al. 1997). • Slope of and scatter about the CM relation provide constraints on the formation epoch and star formation history of galaxies. March 17, 2004 STSc. I Colloquium 11
r-i vs i V-I vs I �Elliptical S 0 Spiral V-I vs I March 17, 2004 i-z vs z STSc. I Colloquium 12
Estimating Epoch of Last Major Star Formation Event z. F > 1. 6 March 17, 2004 STSc. I Colloquium 13
Additional constraints from the observed evolution of scatter about & slope of the CM relation Blakeslee et al. 2003 Coma Cluster @ z =0. 02 CL 1252 @ z=1. 24 INT = 0. 023± 0. 007 mag (15 Egal spec-confirmed members) INT = 0. 026 mag (31 Egals, sigma-clipped result) INT = 0. 029 mag (All 52 Egals with z mag < 24. 5) March 17, 2004 STSc. I Colloquium 14
Modeling the scatter Model 1: Galaxies form stars in single bursts at random times between t 0 (z~1000) and t. END (z > 1. 24). Model 2: Galaxies form stars at constant rates in selected times (t 1, t 2) where t 0 < t 1< t 2 < t. END In both model scenarios, we vary t. END and at each step compute the colors for 10, 000 galaxies by interpolation and integration of the BC (2003) solar metallicity models. Then find t. END that yields the best match to the observed intrinsic scatter of a given early type population (E or S 0) March 17, 2004 STSc. I Colloquium 15
Modeling the scatter Model 1: Minimum age of last epoch of SF in CL 1252 ellipticals is 1. 6 Gyr prior to z=1. 24 lookback time (z. END = 1. 9) Mean Luminosity weighted age is 3. 3 Gyr (z = 3. 6) with a scatter of ~30% For S 0’s: z. END = 1. 5 with age scatter ~44% Model 2: Minimum age of last epoch of SF in CL 1252 ellipticals is 0. 53 Gyr prior to z=1. 24 lookback time (z. END = 1. 4) Mean Luminosity weighted age is 2. 6 Gyr (z = 2. 7) with a scatter of 38% Even though epoch of last star formation activity is “recent” the mean age of the ensemble is still high. For S 0’s: z. END = 1. 3 with age scatter ~47% Both models give observed colors that match the observations. March 17, 2004 STSc. I Colloquium 16
Mean Spectral Features in CL 1252 (z=1. 24) Local Elliptical Local Spiral Significant H in absorption seen in co-added spectrum from 10 brightest early type cluster members Most of the early type galaxies contain “post-starburst” stellar population Consistent with a formation redshift of z. F ~ 3 Rosati et al. 2003 de. Marco et al. 2004, in prep March 17, 2004 STSc. I Colloquium 17
Constraints from the Evolution of the Mass-to-Light Ratio Results for CL 1252 are preliminary! (good to within ~20%) M 2 R L R 2 M/L 2/ R Jorgenson, Franx & Kjaegaard 1995 Kelson et al 2000 Holden, van der Wel et al 2004 E+A Galaxy Van Dokkum & Stanford 2003 The M/L evolution is ~1. 1 magnitudes in the rest-frame B band between z=0. 33 and z~1. 25. This amount of evolution is consistent with a solar metallicity BC 03 SSP model with an initial star formation epoch at z=3. March 17, 2004 STSc. I Colloquium 18
Within central 1. 5 Mpc region of distant (z ~ 0. 8 - 0. 9) clusters, the fraction of galaxies with OII EQW > 15 Å is ~45% (Postman, Lubin, Oke 2001), substantially higher than the 10 to 20% active fraction seen in the centers of 0. 2 < z < 0. 55 clusters (Balogh et al. 1997) de. Marco et al. 2004 March 17, 2004 SFRs in z~1 clusters are also substantially higher than in z~0 clusters. Within central 1 Mpc, evidence STSc. I Colloquium 19 for suppressed SFRs relative to those at larger radii.
Ages of Early Type Cluster Galaxies • Ellipticals are well established in x-ray luminous clusters when universe was ~1/3 its present age • There is no significant evolution in the CMR (other than expected from passive evolution) • Photometric and spectroscopic data suggest a mean formation era of z ~ 3 with a spread in formation times of ~34 (± 15) % (~0. 7 Gyr) • Star formation is higher and more frequent in z>0. 8 cluster S 0 s and spirals than what is seen today March 17, 2004 STSc. I Colloquium 20
Are the morphologies of cluster galaxies evolving? • What we know: – The relative fraction of galaxy morphologies depends on density (Dressler 1980, Postman & Geller 1984) and/or on clustocentric radius (Whitmore & Gilmore 1993) Goto et al 2003 – Physical processes exist that can alter galaxy morphology on timescales much less than the current age of the universe: ram pressure, tidal disruption, mergers – Some data suggests there is Quilis, Moore, & Bower 2000: Ram detectable evolution in the pressure induced evolution of morphological composition clusters gaseousdisk moving thru hot ICMof - all over theis past ~5 in. Gyrs (z~0. 5 to the HI gas stripped 100 Myr present epoch) March 17, 2004 STSc. I Colloquium Poggianti 2000 21
Classifying Galaxies • Visually – Over 3500 galaxies with i, z < 24 classified. No selection in color or position. Classifications done in bandpass that is closest to the rest-frame B-band. – Methodology: ~3 team members (MP+) classify galaxies using a common reference set. Agreement in E, S 0, Sp, Irr typically 75 - 85%. No systematic errors between classifiers detected. – Morphological “k-correction” has been shown to be a small effect (e. g. , Bunker et al. 2000; Windhorst et al. 2002) • Machine-generated Parameters – Compactness, Asymmetry, BPZ Spectral template type show good correlations with visually determined type March 17, 2004 STSc. I Colloquium 22
22. 2 mag S 0/a E S 0/a Sb/Sc Sa or later S 0/E Sb/Sc E S 0/E E S 0/a S 0/E Sb or later E S 0/a Sa or later E Sc/Sd S 0/E Sb/Sc S 0/a E/S 0 E March 17, 2004 S 0/E Sa E/S 0 E S 0/E S 0/a Sc/Sd Sd S 0/a E E E/S 0 S 0/E E/S 0 Sb or later STSc. I Colloquium E E S 0/a Sb or later Sb/Sc Sa or later S 0/a Sb/Sc 23. 0 mag 23
Photometric Redshifts 0. 33 ≤ z. B≤ 0. 53 0. 93 ≤ z. B≤ 1. 13 March 17, 2004 0. 53 ≤ z. B≤ 0. 73 1. 13 ≤ z. B≤ 1. 33 STSc. I Colloquium 0. 73 ≤ z. B≤ 0. 93 1. 33 ≤ z. B≤ 1. 53 25
Morphology-Density Relation @ z = 0. 83 CL 0152 + MS 1054: BPZ Sample 10 100 Spectroscopic Sample 1000 10 1000 Density (Galaxies Mpc-2) Postman et al 2004, in prep March 17, 2004 STSc. I Colloquium 26
Morphology-Density Relation @ z = 1. 24 Postman et al 2004, in prep March 17, 2004 STSc. I Colloquium 27
Morphology-Radius Relation Whitmore & Gilmore 1993 March 17, 2004 STSc. I Colloquium 28
All Galaxies (photo-z sample) CL 1252 -2927 CL 0152 -1357 March 17, 2004 STSc. I Colloquium 29
Spiral Galaxies (photo-z sample) CL 1252 -2927 CL 0152 -1357 March 17, 2004 STSc. I Colloquium 30
S 0 Galaxies (photo-z sample) CL 1252 -2927 CL 0152 -1357 March 17, 2004 STSc. I Colloquium 31
Elliptical Galaxies (photo-z sample) CL 1252 -2927 CL 0152 -1357 March 17, 2004 STSc. I Colloquium 32
Evolution of the Early Type Population Fraction CL 0152 MS 1054 CL 1252 Lubin, Oke, Postman 2002 1. 24 March 17, 2004 STSc. I Colloquium 33
Postman et al 2004, in prep No significant evolution of morphological population between z=1. 24 (8. 5 Gyr ago) and z=0. 4 (4. 3 Gyr ago) Poggianti 2000 March 17, 2004 0. 83 STSc. I Colloquium 1. 24 34
Evolution of Morphological Composition in Clusters • • • Morphological segregation is in-place by z=1. 24, at least in clusters with x-ray luminosities greater than 2. 5 x 1044 erg/s. M-D and M-R relations are qualitatively consistent with what is seen locally - spiral galaxies are not common in dense environments. Given the irregularity** of the galaxy distributions, the presence of a strong M-D relation suggests that density, rather than clustocentric radius, may be the more fundamental variable. There is no strong evolution in the early type fraction in x-ray selected clusters between z ~ 1. 24 and z ~ 0. 5 - but, as always, a larger sample is needed. This may suggest that morphological fractions in current epoch clusters are the result of more recent environmental interactions (a la MORPHs) ACS / WFC enables relatively easy and robust early vs late type classification down to i = 23 mag & down to i = 24 mag with care. March 17, 2004 STSc. I Colloquium 35
Brightest Cluster Galaxies • BCGs could exhibit significant photometric & morphological evolution between z~1 and now. • What fraction of z~1 BCGs appear to be in process of merging? • How does the rest-frame Bband BCG luminosity in our z~1 clusters compare with current epoch BCGs? BCG Formation Simulation by J. Dubinski (1998) 10 Gyr in 15 seconds March 17, 2004 STSc. I Colloquium 36
Brightest Cluster Galaxies CL 0152 -13 Elliptical S 0/a CL 1604+4321 March 17, 2004 MS 1054 Elliptical CL 1604+4304 Sb/Sc S 0/a Elliptical CL 0910+54 CL 1252 -29 STSc. I Colloquium 37
BCG Mergers? • MS 1054 has asymmetric outer isophotes. • CL 1252 BCG “pair” leave significant residuals when best-fit elliptical isophotes subtracted - suggestive of tidally stripped stars. This pair will likely merge within a few Gyr. • CL 0152 BCG is well-fit by concentric elliptical isophotes. March 17, 2004 STSc. I Colloquium 38
BCG Luminosity Evolution z ~ 1 BCG exhibit a similar dispersion in their rest-frame Bband luminosities (~32%) as their z=0 counterparts M 2 - M 1 in all but one of these z~1 clusters is smaller (<0. 13 mag) than that in ~90% of the z~0 rich Abell clusters [Exception is MS 1054 which has 0. 36 mag contrast] We expect some of these BCGs to undergo a doubling in mass by z~0. 5 (e. g. , CL 1252) March 17, 2004 STSc. I Colloquium 39
Mapping the Cluster Mass Distribution • Key questions: – Is the distribution of luminous matter an accurate tracer of the total mass? Are substructures seen in x-rays and galaxies reflected in mass as well? – Is the mass estimate from gravitational lensing consistent with other (x-ray, kinematic) mass estimators? March 17, 2004 STSc. I Colloquium 40
Chandra imaging: Evidence for On -going Cluster Accretion CL 0152 -1357 MS 1054 -0321 CL 1252 -2927 Rosati et al. 2002 March 17, 2004 STSc. I Colloquium 41
Cluster Mass Distribution: CL 0152 Luminosity Density for cluster members March 17, 2004 M. Jee et al. 2003 STSc. I Colloquium 42
Cluster Mass Distribution: MS 1054 Luminosity Density for cluster members March 17, 2004 M. Jee et al. 2003 STSc. I Colloquium 43
Cluster Mass Distribution: CL 1252 ------ NIS model ------ NFW model ------ X-ray, T=[5. 6, 6, 6. 4] ke. V Contours show the surface mass distribution as derived from the co-added i+z band image. The detection shown is significant at the 6 level. March 17, 2004 STSc. I Colloquium Luminosity Density for probable cluster members Lombardi, Rosati et al. 2003 44
Gravitational arcs discovered in CL 1252 -2927 @ z=1. 24 Galaxy @ z= 3. 36 Arc A Arc B March 17, 2004 STSc. I Colloquium 45
Cluster Mass Distribution • Strong lensing is detected in at least two clusters at z=0. 83 and z=1. 24. Weak lensing detected in all z>0. 8 clusters studied so far. • Galaxies, ICM, and mass exhibit qualitatively similar distributions. The frequency of significant substructure or elongated cluster galaxy distributions tends to increase with redshift (cf. Lubin & Postman 1996) - consistent with our ACS observations and current theoretical models. • Lensing maps reveal, however, that highest galaxy overdensities don’t always correspond to highest mass peaks, suggesting variations in M/L within a cluster. • Lensing mass profile in good agreement with x-ray temperature and galaxy kinematics. March 17, 2004 STSc. I Colloquium 46
Finding Protoclusters at z > 1. 3 • Powerful High-z Radio Galaxies: Forming BCGs? – Amongst the brightest galaxies at every z (e. g. De Breuck et al. 2000) – Merging of multiple L* clumps (e. g. Pentericci et al. 1998) – ~100 kpc envelopes of emission-line gas (e. g. v. Ojik et al. 1996) – LFIR > 1013 Lo implying SFR > 2000 Mo/yr – Highly clustered (e. g. Blake & Wall 2003, Overzier et al. 2003) – SM blackholes powering AGN – Associated with GALAXY OVERDENSITIES! WFPC 2 (F 606 W) image of PKS 1138 -262 at z=2. 2 with VLA radio contours (Pentericci et al. 1998) March 17, 2004 STSc. I Colloquium 47
NARROW-BAND LYA IMAGING of POWERFUL RADIO GALAXIES has revealed z > 2 ‘PROTOCLUSTERS’ (Keel et al. 1999, Pentericci et al. 2000, Venemans et al. 2002, 2004, Kurk et al. 2001, 2003) TN J 1338 -19 at z = 4. 11: • 33 confirmed Ly emitters (Venemans et al. 2002, 2004 in prep) • Velocity dispersion ~350 km/s • Comparable to mass of Coma cluster However, Lya only detects about ~20% of Lyman Break galaxies (Steidel et al. 1999, Shapley et al. 2001) VLT/FORS March Lya 17, at 2004 z = 4. 11 STSc. I Colloquium 48
HST/ACS observations of TN J 1338 at z = 4. 11 18 orbits (5 g, 4 r, 4 i, 5 z), single pointing. Look for LBG population associated with protocluster at z~4: g-r > 1. 5, g-r > (r-i)+1. 1 r-i < 1. 0 i < 27 Miley et al. 2004; Overzier et al. in prep. March 17, 2004 STSc. I Colloquium 49
56 LBGs to i. AB=27 Cloning (R. Bouwens) of Bdropouts suggests 3 -5 sigma overdensity of LBGs compared to GOODS (Giavalisco et al. ); Concentrated towards radio galaxy. 12 Lya emitters Radio Galaxy 2 arcsec/16 kpc March 17, 2004 STSc. I Colloquium i. AB=23. 3 50
Does the structure of galaxies depend on their Ly emission? Preliminary result from Overzier et al. 2004 R H(z)-1 ( virial velocity) R H(z)-2/3 ( virial mass) No evolution TN 1338 z=4. 1 Ferguson et al. 2003 Bias? Lya = invf(L, SFR) March 17, 2004 STSc. I Colloquium 51
HST/ACS IDT protocluster program Target z Note TN 0924 5. 19 OBSERVED TN 1338 4. 11 OBSERVED 0316 3. 15 GO PROPOSED 1138 2. 16 SCHEDULED 1138 -262 at z=2. 16 Richest structure of galaxies known at z~2: Blue: EROs Black: Ly emitters Red: H emitters Green: X-ray emitters (6 confirmed) (Kurk et al. 2003, Pentericci et al. 2002) March 17, 2004 STSc. I Colloquium 52
Summary: Hi-z RG/Proto-cluster • Distant Radio Galaxies are excellent signposts for locating distant protoclusters. • TN 1338 (z=4. 1) has a significant excess of spectroscopically confirmed Ly emitters and g-band drop outs. Ground-based Lya map Ly halo morphology unusually assymetric and narrow – points away from center of overdensity effects of bouyancy? SF injects lighter gas into heavier intraprotocluster medium? March 17, 2004 STSc. I Colloquium to center of protocluster 53
Updates to the Structure Formation Story • • • Protoclusters exist at z~4 (~10% current age of universe) High-z clusters with high elongations or significant substructures are common, consistent with the idea that significant cluster accretion is occurring along filamentary structures. The bulk of the stars in early type cluster galaxies are formed over a relatively narrow timeframe (~0. 5 - 1 Gyr) in the redshift range 2. 3 < z < 4. 5. The Brightest Cluster Galaxies at z~1 are still very much “works in progress”. Many are not much brighter than their nth-ranked companions and some are up to a factor of 2 below the predicted rest-frame B-band passively evolving luminosity. The overall morphological population of clusters does not appear to change significantly until more “recent” (z ~ 0. 4 - 0. 5) epochs. If true, what triggers this recent evolution? There is plenty of ICM and time for a variety of environmental processes to have modified galaxies prior to z = 0. 4. Alternatively, one can argue that the strength of even the recent “evolution” in morphological fractions is not highly significant - perhaps the answer to the “origins” of the morphological fractions in clusters lies beyond z=1. 3. March 17, 2004 STSc. I Colloquium 54
Clusters: They’re not just for research anymore! March 17, 2004 STSc. I Colloquium 55
- Slides: 54