Title: Omar L
1Cluster Identification and Galaxy
Populations Bernards Cosmic Stories Valencia,
Spain, 2006
Omar López-Cruz Instituto Nacional de
Astrofísica, Optica y Electrónica (INAOE) Sta.
María Tonantzintla, Puebla, México omarlx_at_inaoep.m
x
2Galaxies are just a tiny fraction of the
clusters mass so, Who cares?
Jerry Ostriker, CITA Seminar, ca.
1995.
Galaxies are fair tracers of the dark
matter... Stefano Borgani, Kona 2005, Craig
Sarazin, GH2005, Roy Gal, GH2005
3Summary
- How we go about finding cluster (biased towards
optical searches, X-ays (Ettori) and SZE
(Sunyaev)) - General Properties of Clusters Richness,
Morphology, and Density Profiles. - 2-D Surface Brightness modeling, B/T, andGalaxy
Morphology. - The Color-Magnitude Relation, The Size-Magnitude
Relation and the more Complete Fundamental Plane.
- The Luminosity Function of Galaxies in Clusters
and - The Effects of the Environment on Cluster
Galaxies. - Conclusions.
4Cluster Finding Techniques
- Density enhancements above the field galaxy
counts - Based on the intrinsic properties of clusters
5Finding Overdensities in the Optical
- Counting galaxies Abell 58, Zwicky et al. 68,
ACO (Abell et al. 89), APM (Dalton et al. 92),
EDSCC (Lumsden et al. 92), PDCS (Postman et al.
96), EIS (Olsen et al. 99), LCDCS (surface
brightness fluctuations Gonzalez et al. 01)
NoSOCS (Gal et al.2004), Voronoi Tessalation
(Kim et al. 2002, Lopes et al. 2004) - Pros easy to implement
- Cons - fairly large contamination, mostly for
relatively poor and distant systems ? ill
calibrated selection function. - - loose correlation between richness
and cluster mass.
6Using Galaxy Properties
Adding color infomation (color-magnitude
relation) reduces the effects of contamination
RCS (Gladders Yee 2005), SDSS (e.g. Miller et
al. 2005, Wilson 2005 (Spitzer)) Pros - cheap!
Easy to cover large area with modern large CCD
frames on large FOV telescopes - color
information suppress contamination and false
detections efficient cluster detection out to z
? 1. - acceptable correlation with
cluster mass requires accurate photometry and
photometric redshifts (free) Cons calibration
of the selection function difficult from first
principles requires calibration with simulations
(Montecarlo or N-body)
7A Desirable calibration M-Lopt
Popesso et al. 05 Lopt from i-band SDSS
data SDSS Mdyn open squares MX from ASCA data
filled circles. Lopt a is a better mass proxy
than richness.
8Intrisic Properties of Clusters
X-ray identification (Ettoris talk)
Pre-ROSAT HEAO-1, EMSS (Gioia et al. 90), Jones
Forman (1999) RASS XBACS (Ebeling et al.
97), BCS (Ebeling et al. 01), REFLEX
(Boehringer et al. 04), NORAS (Boehringer et al.
00), NEP (Gioia et al. 03), MACS (Ebeling et
al. 01) ROSAT deep pointings RDCS (Rosati
et al. 02), 160sq.deg. (Mullis et al. 04),
SHARC (Burke et al. 03), WARPS (Perlman et al.
02), BMW (Moretti et al. 04) Several
ongoing XMM-Newton and Chandra surveys. Pros -
Calibration of the selection function
possible Cons X-ray flux sensitive to details of
the gas distribution
??connection to mass requires external
calibration or follow-up observations (e.g., T,
?v, Compton-y, lensing)
9Intrinsic Cluster Properties
- SZ identification (Sunyaev) - next to come SZA,
ACT, SPT, APEX, Planck, BOLOCAM, OCRA,GTM - Pros - No redshift dimming clusters identified
virtually at any redshift - - Selection criterion essentially
equivalent to a mass-selection one. - Cons - Contamination from radio sources (apply
multi-frequency observations) - - Contamination from fore/back-ground
structures.
10Optical Measurements
- Optical observations are extremely efficient
- At low-moderate redshift, ground based telescopes
sufficient - Current surveys to z0.3-0.5 DPOSS, APM, SDSS
- Future surveys to z1 RCS2, LSST, Pan-STARRS
- Detection measurement of basic properties does
not require deep data - Two filters already good for rough photo-zs and
CMDs - Can detect poor systems groups where most
galaxies reside
11Richness
- Simple galaxy counting
- -in what radius
- -what mag limits?
- -color cuts?
- Observationally computationally inexpensive -
but can it be a proxy for mass, which is what we
want? - Abell (1958) - of galaxies with m3ltmltm32
- within radius of .83 h180-1Mpc
1.5h100- - -Poorly correlated with modern measurements
Gal et al. 2003
12Richness
- ?cl - equivalent number of L galaxies within
some radius in a cluster - Ltot ?cl x L
- Correlates luminosity richness
- Used by Kim et al., Kepner et al. on SDSS data
- Ngal from Annis et al. BCG technique - number of
galaxies within 2s of E/S0 CMR brighter than L1 - -gives smaller numbers due to color limitation
- -may vary with cluster pops
- Measures correlated but noisy
13Richness
- Bgc - amplitude of galaxy - cluster correlation
?(r)Bgc r? - Taken from radio studies (Longair Seldner 1979)
- Yee López-Cruz (1999) measured Bgc for 47 Abell
clusters, previous work by Prestage
Peacock(1988,1989) - Robust against magnitude cuts and radial
coverage. - RAbell is overstimated for clusters at zgt0.1!
- A655 (z0.18) the only R5 cluster is not that
Rich! - Bgc requires knowledge of the LF and its
evolution, and assumes spherical symmetry for
deprojection. ?-1.77
Expected Bgc vs. R
Vel. Disps require 10 zs.
14A few Words on Galaxy and Cluster Classifications
- There is a mask of theory over the phase of
Nature - Schemes should be based on a quantifiable
variable property. Useful schemes strike a
fundamental property that is related to physical
processes. - Categories first kind (purely descriptive),
second kind (quantifiable varying properties, but
no physical mechanism), third kind (quantifiable
varying property rooted on physical processes,
e.g. MK classification of stars), Fourth kind
(rooted on a fundamental physical process, e.g.,
the Periodic Table of the elements) - For galaxies and clusters our schemes are only
second kind!!! - Our cosmic inventory is not complete. And we do
not have a compelling theory for galaxy
formation, yet.
15Classification of GalaxiesWhat is useful and
what is not...
16What is a cD galaxy?
- cD are supergiant galaxies up to 4 mags. brighter
than M. They concentrate almost half of the
total cluster light ( in the R-band LcD1013
h50-2 Lsun ). - cD galaxies are only found in clusters,
independent of cluster richness. - They can have blue cores and multiple nuclei
- cD are often powerful radio-galaxies (WAT), in
fact the term cD galaxy was introduced in a study
of optical counterparts of luminous
radio-galaxies (Matthews,Morgan, Schmidt,
1964). The first 10 cD galaxies discovered A389,
A401, A754, A787, A1775, A1795, A1904, A2029,
A2199, A2670
17Classification of Clusters of Galaxies
Regular
Irregular
All the proposed schemes underline a sequence
from irregular to regular
The Rood-Sastry Classification Scheme
18The Hercules Cluster an example of an irregular
cluster
19Irregular
Regular
20Coma, entire cluster
21Two Classification Schemes
- Rood-Sastry(1971, Struble Rood 1982)
- cD cluster that contain a cD (A401)
- B (binary) two BCG of similar brightness
(A1656) - L (line) three or more galaxies line up (A426)
- C- (core-halo) the 4 BCG located near the center
(A2065) - F (flat) galaxies in flattened configuration
(A397) - I (irregular) (A400) , Is (smooth), Ic (clumpy)
- Bautz-Morgan Types (1970)
- I -clusters containing a centrally located cD
galaxy (A2199, A2029) - I-II -Intermediate
- II -Between cD and Virgo gEs (A2197)
- II-III Intermediate (A426, A400)
- III No dominant galaxy (Virgo, A2065)
- III-E (with ellipticals) III-S (with spirals)
22(No Transcript)
23A Simplified Scheme
- From X-ray observations it seems that cooling
core cluster have different properties from those
without them, as seen by their morphological
structure, temperature structure and metallicity
(De Grandi et al. 2004) - It has been recently recognized that
cluster-cluster merger are frequent. (see S.
Maurogordatos talk) - RS B- clusters are clusters are merging clusters
(Tremaine 1989). - cooling core clusters are cD clusters (BM I,
I-II). - Three classes cD, non-cD and Mergers (RS
B-clusters, presence radio relics, and halos
(Feretti 2006))
24Morphology
- Modern methods
- PA, ellipticity (Binggeli 1982) - alignments
along filaments - Moments of galaxy distribution (Rhee et al. 1989,
Plionis et al. 1991, Basilakos et al. 2000, de
Theije et al.1995) - Flatness - inverse of ellipticity or elongation
(Struble Ftaclas 1994) - Fitting ?-models (Strazzullo et al. 1995)
25Morphology
- Radial profiles
- Ideally, we would like 3-d mass distributions
- X-ray temperature, surface brightness profiles
can be used for comparison to simulations (Loken
et al. 2002, Arnaud et al. 2002, Markevitch et
al. 1999) -
- Optical data requires lots of spectroscopy
- (such as CNOC, SDSS)
- Carlberg et al. 1997 derive
- SN(R) - projected number density profile
- sp(R) -projected velocity dispersion profile
- Fit to projection of Hernquist profile
- Could also use NFW
26Morphology
- Radial profiles in cluster cores
- -CDM models make specific prediction of
universal mass profile - -Lensing (strong weak) can be used to test
mass profile, compare with light, x-rays - -Need for multiple radial tangential arcs to
distinguish NFW vs. isothermal (Gavazzi et al.
2003) - -Arcs useful for accessing central density
profiles - NFW r-1, Moore r-1.5 or other
(Molikawa Hattori 2001) - -Inner slope may be as low as 0.5
- (Sand et al. 2004) suggesting complex
mass-light - relationship in cluster centers
- -NFW profiles are only for collisionless CDM
- particles baryons can behave differently
- - need to add to simulations (See Session 7)
- cD/BCG galaxies may have an appreciable effect
-
A383 mass model Sand et al. 2004
27Morphology
- Substructure
- Merging clusters, infalling groups
- Rate predicted by CDM, related to Om (Buote 1998)
- Detailed studies in optical are recent (Geller
Beers 1982,West Bothun 1990) - Lensing contraints (Natarajan Springel 2005)
- Evolution with time - dynamical times comparable
to tHubble - More substructure at high z (Jeltema 2004)
- 2dF results show high rate of substructure in
poor clusters (Burgett et al.2004) - -supports long relaxation times
- Alignment with filaments stronger for dynamically
active clusters (Plionis Basilakos 2002) - X-ray and optical substructure are correlated
(Kolokotronis et al. 2001, Rosati et al. 2002) - Different measures but need to compare
observations simulations - Wavelets (2-d, Girardi et al. 1999), Lee
statistic (Fitchett 1988), skewness/kurtosis
(Bird Beers 1993), subclumps via ?-test
(Dressler Schechtman 1988), etc.
28The Magnitude Zoo
- Aperture Magnitudes
- Isophotal Magnitudes
- Petrosian radius
- ? (mag)2.5log(5d log r/d? mag)
- Total or asymptotic magnitudes (parametric or
non-parametric) - Vega base magnitudes (based on the SED of Vega)
- AB magnitudes (same zero point for all filters. A
source with a flat SED will have color0)
slope of the growth curve
29NGC 3377
Surface Brightness Profile
Growth Curve
Petrosian Radius
Image from Sandage Perelmuter 1990
30Surface Brightness Profilesand Curve of Growth
- The surface brightness profile and the growth
curve are related, e.g., for the de
Vaucouleurs profile - I(R)Ioexp-7.67((R/Re)1/4-1
- Ltot7.22?Re2Ie.
- The growth curve is
- F(R)LTot1-exp(-z)1?n1...7 (zn/n!),
- where z7.67(R/Re)1/4
31(No Transcript)
32Distance Modulus
Re d s h i ft
R14
R23
33 A few questions
- Can the CMR be seen in every nearby cluster?
- Is the CMR affected by the environment ( i.e.,
cooling flows), temperature gradients, AGNs
(radio, X-ray) ? - When did ETG form?
- When are the effects of the environment
important? - Are the optical and Cluster X-ray properties
related?
34 Pointed Observations of low-z Clusters.
- Over 160 000 photometrical measurements
(galaxies, stars, and garbage). 63 350 galaxies
at the 5s. The completeness limit of R21.5
mag, 0.9m T2KA (23.2 arcmin X 23.2 arcmin FOV)
- 9 clusters 0.02ltzlt0.04 clusters were observed
with KPNO 0.9m telescope MOSA (1 deg X 1 deg
FOV) - 1 Mpc lt ? lt 4 Mpc), with a resolution of
0.68''/pixel. - X-ray selected (Jones Forman 1999) Abell
Clusters (ARC 0) and 7 control fields in R and
B. - Star/galaxy classifications and photometry using
PPP (Picture Processing Program, Yee (1991), Yee
et al. (1996) - López-Cruz, Barkhouse, Yee (2004)
35The CMR was found for every cluster in the
sample. The CMR extends down to 8 mag. No
breaks in the CMR were observed.
36The CMR are fitted using an a robust scheme
based on the biweight, the errors are derived by
bootstraping. ????mag
37Galaxies with B/T gt 0.7 for a sample of 28
clusters of galaxies with varying richness from
Barrientos et al. 2004. If you classify the
galaxies your CMR are cleaner.
38Galaxy Morphology forGalaxies in the Coma Cluster
GALFIT (Peng et al. 2002) Sérsic Bulge Exp.
Disk 0.0 B/Tlt0.4---Spirals 0.4 B/Tlt0.6
---S0 0.6 B/T1.0 ---E
Gutiérrez et al. (2004).
-1 Spirals, 0 S0 ,
1 E
39CMR by galaxy types for 11 Abell Clusters zlt0.05
How do S0s form?
Christopher Añorve (GH2005, poster)
Añorve 2006, M.Sc.Thesis, INAOE
40Galaxy Morphology
Distribution of the Sérsic Index similar to
Blanton et al. (2003)
41Sérsic Index vs. Luminosity
42(No Transcript)
43(No Transcript)
44CLJ1251, z1.235, inside 1 M pc (2 arcmin)
Coma
filled circles E filled squares S0
Blakelessle et al. 2003
45An economical redshift indicator
Dispersion about the fit 0.010
46A cluster finding tool
This is the basis for the Mike Gladders
RSCS. Background contamination important for
substructures studies and LF estimation.
X-rays
Isopleths
47Improved Cluster Finding using DTFE
A690, DTFE map generated by Pablo Arayn da.
Technique due to Bernardeau van de Weygaert
1996, Shaap van de Weygaert 2000) See posters
by Platen et al.
Background cluster at z2.3
48The Size-Magnitude Relation
Salperter IMF
Schade, Barrientos, López-Cruz (1997)
49Kormendy Relation
µe(3.5 /-0.17)log(Re) (19.4/-0.11) Añorve
(2006) µe(3.5 /-0.2)log(Re)
(19.4/-0.4) Coenda et al.(2005)
50It is universal for cluster galaxies
Changes can be explained assuming passive
evolution. Data from Jorgensen et al. (1999)
51Fundamental Plane for Coma Galaxies using
Sérsics Law
logRe1.29(log(s)0.29ltµgte)-5.8 s taken from
Jorgensen, Franx, Kjaergaart (1995).
52The FP is a probe for galaxy evolution.
FP evolution 0 ltzlt0.6, K band. Pahre,
Djorgovski, de Carvalho (2005)
53Dynamical Effects in Clusters
- Cluster Mean Tidal Field
- Mergers
- Collisional Tidal Stripping
- Dynamical Friction Cannibalism
- Harassments
54dEs trace the cluster better!!
55dIr dSph are missing in the center
56Signs of Disruption
MKW 7 Tidal Debris Plume (Feldemeier et al. 2002)
CentaurusTidal Debris Plume (Cálcaneo-Roldán et
al. 2002). The plume is 8 arcmin long (gt100 Kpc)
B-V0.9, V-R0.6, V-I1.2 Stellar Colors !!!!
57Luminosity Function
- Most studies fit to Schechter (1976) function
- ?(L)dL ?(L/L)? exp(-L/L)d(L/L)
- Need to determine
- L The characteristic luminosity
- ? The faint end slope
- ? The normalization
- number per unit volume
- Do these vary among clusters,
- and between cluster field ?
- Cluster LF from Schechter 1976
58(No Transcript)
59Observers have messed up!!
Driver et al. (1994) found a steep faint-end
slope ?????1.8. Suggested a universal trend.
Trentham (1997), De Propris et al. (1997), etc,
they all got steep faint-end slopes. But Driver
et al. (1998) and Trentham (2002) have changed
their minds. Goto et al. (2002) and De
Propis et al. (2003) do not see
variations Lopez-Cruz et al. (1997), Barkhouse
et al. (2006), Paolillo et al. (2001), Mercurio
et al. (2003) Hansen et al. (2005) find LF
variations.
60We do not detect an universal pattern
61LF generated using redshifts
Christlein Zabludoff (2003)
62- We have identified a group of clusters that we
have termed flat-LF clusters. - Rich clusters
- cD galaxies (B-M I, I-II)
- Very luminous X-rays clusters, single-picked
63Variations at the bright-end of the LF
64Variations at the bright-end of the LF
The whole sample
Bgc (Yee López-Cruz 1999)
65Variations at the bright-end of the LF
cD Clusters
66(No Transcript)
67Variations at the bright-end of the LF
68Conclusions
- Clusters can be found searching for overdensities
in the optical, color information - improves the success rate. Clusters can be found
by other intrinsic properties (X-rays, SZ effect,
lensing, etc.) - Quantitative galaxy morphology possible through
2-D surface brightness modelling - The CMR is a useful property for cosmological
studies. It provides us with a galaxy formation
clock. It can be used to find clusters and get
their redshifts. - Changes in the CMR and the SMR can be explained
under passive evolution. It is very likely that
the epoch of ETG formation happened at zgt3. - The fundamental plane(universal) also indicates
passive evolution. - The LF is not universal but shows a clear
dependence with the environment. i.e., dynamical
effects are important Gus Oemler, Alan
Dressler and BST were right. - Suggestive differences between cD and non-cD
clusters. - M for non-cD clusters could be used as a
distance indicator. - Do dwarf galaxies help in the formation of cD
galaxies?
69Where is it?
Mgas as measured by Jones Forman (1999) within
1 Mpc
fgal0.19fgas? Allen (2005)
Integrated light due to galaxies in CMR within 1
Mpc
70A few details....
We use the a classifier C2 that measures the
compactness of an object, it is defined as
C2 (NA-2)-1? (mi-mi)-C0
, where NA is the adopted largest aperture mi
and mi are the instrumental magnitudes at the
ith aperture of the object and a selected
reference star, respectively, and C0 is a
normalization constant. The magnitudes that
PPP try to measure are asymptotic or total
magnitudes they are based on Growth-Curve
analysis using circular apertures. Galaxy colors
are determined using 11 h-1 Kpc apertures.
71And we do not try to fall in the temptation of
trying smart corrections. Well... just a little
one