Title: Clusters of galaxies The ICM, mass measurements and statistical measures of clustering
1Clusters of galaxiesThe ICM, mass measurements
and statistical measures of clustering
2Plan of this class
- The intracluster medium, its origin, dynamics and
general properties - Evidence of Dark Matter in clusters
- Masses derived by the virial theorem, x-rays and
gravitational lensing - Results from studies of gravitational lensing in
clusters - Statistical measures of clustering
3The intracluster medium
4Clusters are among the most luminous X-ray
sources in the sky. This X-ray emission comes
from hot intracluster gas.
5For comparison,
- Cataclismic variables Lx 1032 1038
erg/s - Milky Way, M31 Lx 1039 erg/s
- Clusters of galaxies Lx 1043 1045
erg/s - Only Seyferts, QSOs, and other AGN rival clusters
in X-ray output - Clusters may emit nearly as much energy at X-ray
wavelengths as visible - L(optical) 100 L galaxies 1045 erg/s
6The Lx s correlation
7What is the origin of cluster X-ray emission?
- Answer hot (107 108 K) low-density (10-3
cm-3) gas, mostly hydrogen and helium, that fills
space between galaxies. At these high
temperatures the gas is fully ionized. - Two emission mechanisms
- 1) Thermal bremsstrahlung (important for T gt 4
x 107 K) - free electrons may be rapidly accelerated
by the attractive force of atomic nuclei,
resulting in photon emission - because the emission is due to Coulomb
collisions, X-ray luminosity is a function of gas
density and temperature - Lx nelectron nion T1/2
rho_gas2 T_gas1/2 - 2) Recombination of electrons with ions
(important T lt 4 x 107 K) -
8Dynamics of the intracluster gas
- The intracluster gas can be treated as
- An ideal fluid
- In hydrostatic equilibrium
- At a uniform temperature
9(No Transcript)
10(No Transcript)
11(No Transcript)
12(No Transcript)
13(No Transcript)
14X-ray spectra
- Spectroscopy of the intracluster gas provides
information on its temperature and composition - Observed spectra show exponential decrease at
high-frequencies that is characteristic of
bremsstrahlung.
Coma Cluster Hughes et al. 93
15- Emission lines due to Fe, Ni and other heavy
elements are seen. This suggests that much of the
intracluster gas must have been processed through
stars. - Chemical abundance of the intracluster gas can be
measured from the equivalent widths of these
emission lines. It is found to be about 30-40 of
solar abundance - If the galaxies and gas are both in thermal
equilibrium in the cluster potential well, then
one expects - m v(gal)2 3 kbTgas
- Tgas proportional to v(gal) 2
16What is the origin of the intracluster gas? Two
possibilities
- The intracluster gas once resided in galaxies and
was later removed. - - this would explain
high metallicity of gas - - galaxies in the cores
of rich clusters are - observed to be
deficient in HI gas, which - suggests that
stripping has occurred. - The gas is primordial, originating at the time of
cluster - formation.
- - but since Mgas gtgt Mgal
it is difficult to - understand how so much
material could - have been stripped
from galaxies -
17How much gas is there in clusters?
18Cluster Mass estimates X-ray gas
19The total gas mass in clusters exceeds the total
galaxy mass. Gas contributes as much as 10-20 of
the total cluster mass.
David, Jones and Forman 95
20Evidence of Dark Matter (DM) in clusters
21Dark Matter in ClustersA more accurate name for
clusters of galaxies would be clusters of dark
matterObservational evidence suggests that
80-90 of the mass in clusters is in an invisible
form1) What evidence is there for dark
matter?2) How much dark matter is there?3) What
is the distribution within clusters?
22Evidence of Dark Matter in clusters
- Virial mass estimates
- If a cluster is in virial equilibrium then
its mass can be estimated from Mvirial Rltv2gt/G - Observations indicate that the total
cluster mass exceeds the combined masses of all
galaxies by factors of 10-20. - Example the Coma Cluster
- Mvirial 1 x 1015 h-1 solar masses
- Ltot 4 x 1012 h-2 solar
luminosities - Assuming a typical galaxy with M/L 10
- Then Mvirial/Mgalaxies 25
-
23Typical mass to light ratios
- Globular clusters 1-2 M/L
- Elliptical galaxies 5-10 h M/L
- Groups of galaxies 100-300 h M/L
- Rich clusters 300-500 h M/L
-
24Mass to light ratio of Coma
25 Galaxy Dynamics
Mass estimate using the Virial theorem
26X-ray mass estimates
- If the intracluster gas is in hydrostatic
euilibrium in the cluster potential, then the
cluster mass can be determined from -
27Gravitational lensing studies provide another
independent evidence for DM in clusters
28Gravitational Lensing some history
- 1913 Einstein predicted that the gravitational
field of massive objects can deflect
light rays. - 1919 Eddington measured the deflection of
starlight by the Sun, confirming Einsteins
prediction. - 1937 Zwicky suggested that galaxy clusters may
produce observable lensing. - 1987 First evidence of strong gravitational
lensing by clusters was found (Lynds/Petrosian,
Soucail et al.) - 1990 Weak gravitational lensing by clusters
was discovered (Tyson et a. 1990). - Today Evidence of lensing has been found for
several dozen clusters. New examples are being
discovered all the time. -
-
29(No Transcript)
30STRONG LENSING
- 1986 Lynds Petrossian discover the first
gravitational arcs in clusters of galaxies
1987 Soucail et al. determine the distance to
the arc twice the distance to the cluster that
contains it.
31Gravitational lensing the basic ideas
32(No Transcript)
33Galaxy cluster
Background galaxy
Observer
Strong lens
Weak lens
34- Strong lensing occurs when
- Long arcs and multiple images are produced.
- Weak lensing occurs when
- Small arclets and distortions are produced.
-
35(No Transcript)
36(No Transcript)
37Strong Lensing
A 1451
z 0,199
?
38(No Transcript)
39A 1451
z 0,199
?
40Weak Gravitational Lensing
Mellier 99
41Why Weak Lensing ?
42Measuring Faint Galaxy Shapes
Cypriano et al. 2005
43Mass ? Light
A2029
In 77 of the cases the center of light and mass
distributions are consistent with each other...
Light
Mass
44Mass ? Light
...but there are exceptions
45Mass ? Light
A4010
Light
Mass
46Mass ? Light
There is a strong alignment between the BGC and
the dark mater main axis
47Comparison with X-Rays
A1451
A2163
A2744
48Comparison with the Velocity Dispertion
49The dynamical state of the clusters
50The dynamical state of the clusters
Interpretation There are two structures along
the line of sight
Chandra observations confirms fusion along the
line of sight (Kempner David 2004)
51Which method is the best one ?
Weak Lensing
? Independent of the dynamical state
? Reconstruct the 2-D potencial
? Needs good seeing
? Cannot separate components along the line of
sight.
52Which method is the best one ?
X-Rays
? Depend of thermal/dynamical state of the ICM
? Cannot separate components along the line of
sight.
? All Sky Surveys (e.g. ROSAT) can provide large
and homogeneous samples
53Which method is the best one ?
Dynamics of galaxies
? Depend on the dynamical state of the cluster
galaxies (galaxies relaxes later than the ICM)
? Reliable results depends on a large number of
galaxy velocities over a large area (e.g. Czoske
et al. 2002)
? Can separate structures along the line of
sight
No single method is perfect !
54What can we learn from gravitational lensing?
- Gravitational lensing can be used to determine
the amount and distribution of dark matter in
clusters. - Unlike virial or X-ray mass determinations,
lensing requires no assumptions about the
dynamical state of the cluster! - The arc thickness is related to the cluster mass
distribution. More concentrated mass
distributions produce thinner arcs. - Modelling the positions and shapes of arcs and
arclets allows the cluster potential to be
mapped. Lensing models have become so good that
in can predict the locations of faint additional
arcs. - Gravitational lensing causes images to be
magnified. Clusters of galaxies can be used as
natural telescopesto study extremely distant
galaxies that would be otherwise too faint to
see. - Lensing can also be used to place cosmological
constraints, because distances (Dos, Dol, Dls)
depend on omega, Ho and lambda. -
-
55z 5.6
Ellis, Santos, Kneib Kuijken (2001)
56What have we learned so far from gravitational
lensing?
- Samples of strong and weak gravitational lensing
have been found in several dozen clusters. - Lensing mass estimates indicate large quantities
of dark matter in clusters - Lensing mass estimates agree with virial and
X-ray masses (with a few exceptions). - The exceptions are probably clusters which are
not in equilibrium. - Hot clusters tend to present dynamical activity
(major concern for experiments designed to
constrain cosmological parameters). - Mass follows light in most cases.
- Cluster dark matter has a very steep radial
distribution. - Models of the cluster potential provide strong
evidence of substructure in the dark matter
distribution. - Gravitational lensing has been seen in clusters
at zgt1 -
-
57Clusters as Tracers of Large-scale Structure
58Why use clusters to map the large-scale structure
of the universe?
- Advantages
- Clusters provide an efficient way of surveying a
large volume of space - Cluster distribution provides information about
conditions in the early universe - Clusters can be seen at great distances
- Disadvantages
- Their low space density makes clusters sparse
tracers of the large scale structure - Results may depend on the chosen cluster sample
- Redshifts of many clusters are still unmeasured
59(No Transcript)
60(No Transcript)
61(No Transcript)
62(No Transcript)
63Large scale structure 2dF
64Some history
- 1933 Shapley noticed several binary and triple
systems among the 25 clusters that he catalogued
it is possible that clusters are but nuclei or
concentrations in a very extensive canopy of
galaxies. - 1954 Shane and Wirtanens galaxy maps showed a
strong tendency for clusters to occur in groups
of two or more. - 1956 Neyman, Scott and Shanes pioneering
statistical models of galaxy clustering included
second-order clusters, I.e., superclusters. - 1957 Zwicky declared that there is no evidence
at all for any systematic clustering of clusters
clusters are distributed entirely at random. - 1958 Abell examined the distribution of
clusters in his catalogue, and concluded that
clusters of clusters of galaxies exist - Today No doubt that galaxy clusters are
clustered. Instead, debate is about the SCALE of
this clustering. -
65Statistical measures of clustering
- 1) The two-point correlation function
- 2) The power-spectrum
- 3) Cluster alignments
-
66Probability of finding objects in dV1 and dV2
separated by distance r
67(No Transcript)
68Two-point correlation function for Abell clusters
- Abell cluster correlation function has the same
power-law form as that for galaxies - ? (r) A r? 1 (r/r0) ?
? (r) 1 at r r0 - ? - 1.8
- r0 20-25 h-1 Mpc
- Richer clusters are more strongly clustered than
poorer clusters - The Abell cluster correlation function has the
same power-law form as the galaxy correlation
function, but with a 15 times greater amplitude
(r0 5 h-1 Mpc for galaxies r0 20 h-1 Mpc for
Abell clusters - Why is ? (r) different for galaxies and clusters?
Biasing! - If Abell clusters have formed from rare
high-density peaks - (? gt 3s) in the matter distribution, then
their clustering tendency will be enhanced by an
amount ?cluster ?2 ?matter (Kaiser 1984).
69Two-point correlation function for other cluster
samples
- APM and EDCC clusters show a weaker clustering
tendency than Abell clusters - r0 13-16 h-1 Mpc for both
samples - ROSAT X-ray selected clusters
- r0 14 h-1 Mpc
- Why do different cluster samples give different
results? - Three possibilities
- (a) The Abell catalogue is
unreliable - (b) Richness-dependence of the
cluster correlation function. Abell, APM and EDCC
clusters are fundamentally different types of
objects. - (c) X-ray selected samples are
flux-limited rather than volume-limited. This
means that any X-ray selected sample will contain
a mixture of nearby poor clusters and distant
rich clusters.
70Statistical measures the power spectrum
71Statistical measures the power spectrum
- Although P(k) is more complicated to measure
than the two-point correlation function it has
two big advantages - 1) it can be more directly compared with theory
- 2) it is a more robust measure
- ? (r) 1 Npairs/Nrandom Npairs/(n 4/3
p r3) - which is proportional to 1/n
- Uncertainties in n produce large
uncertainties in ? when ? ltlt 1. - For P(k), each dk is proportional to n.
Hence the shape of the power-spectrum is
unaffected.
72Statistical measures cluster alignments
- Clusters are often embedded in large-scale
filamentary features in the galaxy distribution. - Cluster major axes tend to point along these
filaments towards neighbouring clusters, over
scales of about 15 h-1 Mpc, perhaps up to 50 h-1
Mpc. - These cluster alignments may provide important
clues about cluster formation and cosmology -
73Clusters as LSS tracers
- Clusters of galaxies are efficient tracers of
the large-scale structure of the universe. - There is strong evidence of structure on scales
of over 100 h-1 Mpc in the cluster distribution. -