Aspen Summer Session 5/30/03 Zaritsky EDisCS ESO Distant Cluster Survey Las Campanas Distant Cluster Survey --surface brightness enhancements instead of galaxies 130 sq. deg, half lost to masking 1073 cluster candidates success rate ~70% applied to SDSS this technique would give several x10^5 clusters 15000 with 5x10^14 Msun 30 LCDCS clusters fed to EDisCS chosen on surf brtness (mass), plus coord cuts, and visual inspection 13 of sample at z=0.5, 17 at 0.7-0.8 snapshot VLT imaging for confirmation left 20 clusters (10 and 10) low-z sample has BVIKs imaging hi-z VRIJKs 20 NTT nights for IR imaging ->photoz via Hyperz , and Rudnick use to reject non-cluster members leaves ~50% of field, 90% of cluster members deep spectra with 22 nts FORS2 for redshifts to I~23, line indices to I~22.5, int. veldisp to I~21.5, rot. curves to I~22 expected stats z=0.5 z=0.7 spectra 1000 1300 redshifts 900 1200 cluster memb 400 420 line ind. 100/150 150/250 cluster/field Sig. internal 75/70 110/70 rot curves 50/100 80/130 +80 orbits HST for morpho at hiz, +WFI in 3 colors prelim results -LF fit by 2dF local cluster LF in shape but L* is brighter by 1.05mag at z=0.5 by 1.2mag at z=0.75 in rest frame B -brighter than expected from FP but includes more blue galaxies -CM relations: large variation in red sequence strength; some atz=0.8 do not have enhancements in RS -large # of disks in red sequence - red sequence scatter depends on redshift AND vel. disp Yee: RS dispersion is f(veldisp); not just increased bg contrib for poorer clusters -emission line fraction as f(veldisp) decreases from 0.8 to 0.2 from 400 to 1200 km/s using fixed METRIC aperture, not r_200 or is it field contribution? Mohr-emline gals @low z clusters evenly distributed so may just be at right z, not in cluster -mass estimates sig_veldisp sig_weaklens 477 503 799 678 830 742 933 980 -structure coadd hi-z clusters, trace to 1.5Mpc radius, maps of substructure -Halpha imaging (Jband): global SF rates 20-70Msun/yr h^-2 No radial dependence in general, 1/3 centrally concentrated, fraction of Ha galaxies rises with decreasing sigma ====================================================================================== ====================================================================================== Yee Using cluster surveys for cosmology N(z,m) -> Omega_m, sigma_8, w constraints in w-Omega_m orthogonal to SN measurements need ~1000 sq. deg cluster survey to z=1 two main hurdles -how to get 1000 sq. deg. -how to get mass for all those objects (10^4 clusters) RCS approach 100 sq. deg in 2 colors R,z' @ CFHT 12k & CTIO 8k R<25.4, z'<23.5 at z=1, assuming 1 mag of evolution, get to L*+1 at z=1 use CMD to pick clusters from red sequence take cuts in CMD color based on Kodama models, make map, find cluster, also gives photo-z based on area of RCS done so far, 5-6000 clusters in whole catalog, 5% contam, complete to Abell R=1 Move on to 1000 sqdeg survey CFHT Megacam 40 2kx4.5k CCDs -> 1 sq. deg./pointing don't need to go so deep for w meas. 830 sq. deg. of own data - 3 filters g (4 min, 25.3), r (8 min,24.8), z (6 min,22.5) if delivers 0.5" images ->20 min/sq. deg. ==> 43 nights, time already received!, will take 4 semesters. Use Legacy survey of 170 sq. deg. (france/canada) deeper than RCS2 data, good for calibration of lensing mass How to get the mass? need cheap & fast mass proxy richness or light gets good M_est seen in CNOC survey; get sigma to 15%, but these are all EMSS clusters get M_200 to 30%, T_x to 30% but what the hell is mass? How to connect to cosmology/sims? use WL, Nbody models to connect light to mass Hank Hoekstra - measure r_Einstein for 1800 RCS1 clusters, bin in richness & stack for WL signal richness measure B_gc vs r_E correlates amazing well. !!Optical richness and WL mass BOTH measure projected distribution in cylinder, minus background!! so they should correlate well. Needs to be a redshift dependent relation At z<0.5 can calibrate with RCS2. At z~0.6-0.7 can use Legacy survey to get WL mass. Will need HST or something else at z>0.7 Gus: Mimicking selection effects in simulations maybe a problem? A: color distributions in field & clusters, comparison of detection by mass vs galaxies in sims can give a handle Voit: Would LCDCS poor red-sequence guys be missed Zaritsky: cosmo from RCS1? A: from 10 sq. deg. , results agree with LCDM. But poor calibration of mass, small area, etc. Mohr: Mass fn. behavior depends on mass definition. How does Bgc or WL defined mass affect MF behavior? A: initial results (mass in tube) seem ok Mohr: but the evolution of this MF should be more complicated. Need N-body sims for all cosmologies to compare! No simple Jenkins-like mass fn. ====================================================================================== ====================================================================================== Pat Henry Cosmology from mostly X-ray L,T functions Theory gives N(M,z,cosmology) and power spectrum P(k) also fn. of z and cosmo due to small numbers and non-contiguous obs. usually measure mass fn. Universal mass fn n(m) for all cosmologies Jenkins et al. 2001, Evrard et al. 2002, Hu & Kravtsov 2003 (N-body) Sheth & Torman 1999 (analytic) but measuring mass is hard so measure L_x, T_x theoretical and measured relations, but still hard to measure mass! So hard to make empirical relation. Allen et al. astro-ph/0208394 -most round clusters,17 with Chandra data -direct decomposition of gas density & temp, plus WL masses -total mess, high scatter, weak correlation, but all are at different redshifts -so already need a theory to get M(z) -need normalization of M(T) reln.; this is the best we can do now. B_TM*kT = f(z,cosmo)M^(2/3) also written as T/T* = f"""" get B_TM = 1.42/T* B_TM from 0.5 to 1.5 (T* from 2.8 to 0.95), theory tends to give higher B_TM by factor of ~50% ; empirical: 0.72 Theory: 1.03 Omega_M(z=0) sigma_8 WMAP 0.27pm0.04 0.84pm0.04 Rosati et al 03 0.26pm0.06 0.78pm0.08 Luminosity evoln Ikebe 02 0.26pm0.09 0.94pm0.13 Local Temp Pierpaoli 0.30 0.77pm0.05 Local Temp Pierpaoli 0.30 0.79pm0.06 Local Lum Schuecker 0.29 0.70 N(z)for z<0.3,+P(k) So, good agreement! And independently measured from WMAP, WMAP indep of X-ray Field has converged to (an) answer; much better than few yrs ago Cluster Temp Fn. Evolution Distant sample is unique - complete, has T's 19 clusters from EMSS with kT>3, 0.3most influenced by heating cooling mechanism Rising gas frac. profile predicted by simulations, even w/0 heating/cooling Does group f_gas catch up to clusters within R_vir Entropy conserved in adiabatic process must rise monotonically with radius for convective stability get slope for S(T) of 0.65, flatter than expected from preheating, agrees with simple cooling model S(R) ; preheating gives S(R) ~ R^1.1 within shock region + flat entropy core see this in data, get R^1.1. consistent with accreting gas cooler systems have greater entropy excess but see no large flat entropy core optical luminosity scales wth mass Voit: for groups M/L may be lower, would affect optical mass estimators Mohr: yes, see slope, but not well measured Yee: see M/L change from clusters to groups Clowe: WL shows constant M/L but mass-light distribution is different; need to go to large radii Evrard: confidence in f_Baryon changing within R-500 as f(M,T, etc.) ====================================================================================== ====================================================================================== Zabludoff Galaxy evolution in groups & clusters Pillars of gal ev. in dense environs: Morphology-density relation (MD) Butcher-Oemler effect (BO) ascribed to cluster effects -ram pressure Gunn & Gott -galaxy harassment Moore et al. -cluster tides Observations may not support ANY of these @low-z 1) early type fraction f_e in some poor groups (those with xray halos) is comparable to theat in clusters -see increase in f_early with vel. disp from (0.1,100 to 0.6,400); saturates at cluster value by 400 km/s -how do small groups (xray only 1kev) get such high f_e? -relation would predict >100% f_e for rich clusters, not seen -suggestion: saturation in f_e at veldisp=400km/s corresponds to system in which L* galaxy has ahd time to undergo merger 2)E+A post-starburst galaxies now also found in poor groups and in the field HST imaging shows dramatic tidal features for all E+As Must have comparable galaxy densities in groups and clusters or else MD relation fails BO effect may be more group accretion in clusters at higher z; also suggested by CNOC So, is it galaxy-galaxy interactions in groups driving galaxy evolution? dwarf/giant ratios in groups consistent with rich clusters, at least to L*+3 fraction of early types with SF also comparable BCGs in clusters do not sit at cluster center, but in center of a subclump gEs in groups always seen bang in group center. -->suggests groups are formation sites of BCGs Also suggestive: veldisp of BCG always < 400 km/s, like a group! At this veldisp, mergers in a group get less likely. SDSS/2dF results show break in SFR vs. projected local density relation @ density of few gals/Mpc^2, which is group density Zabludoff sees no differences in M/veldisp/Lx/Tx relations between groups and clusters down to 1kev What is needed to get this right? -control samples: --isolated groups and those that are falling into clusters --would allow comparison of SFR,HI,L,morphology ; if same, then cluster environment unimportant! -groups at hi-z (>0.5) to see evolution within groups --how to find? QSO lensing galaxies tend to be in groups, so survey strong QSO lenses -so far 5/11 are groups -need simulations that have resolution at lower masses i.e. 10^13 Msun Evrard: is red sequence of groups different from clusters? A: not enough good data yet