Transcript Clusters as laboratories for the study of galaxy evolution

Clusters as Laboratories for the
Study of Galaxy Evolution
Alan Dressler
Carnegie Observatories
Ringberg 2005
“Distant Clusters of Galaxies”
“Clusters…are laboratories for the study of galaxy evolution and
may become as useful as star clusters are in the study of stellar
evolution” -- Dressler, 1984 Annual Reviews
Back in 1984, (and still?) the prevalent view was that clusters were
themselves agents that influenced the evolution of galaxies, so that the
processes that produced different morphological types, for example,
could be observed at work in (today’s) rich clusters.
This notion is very different than the case of star clusters, where we
exploit the coeval populations of stars to understand how stars evolve,
but do not expect that the evolution of these stars has been influenced
by the cluster environment. Given this difference, my “laboratories”
statement would have been an imperfect analogy, at best.
In this talk I will review evidence that clusters of galaxies may indeed be
a good place to study galaxy evolution, but, not primarily because of the
present-day cluster environment – just the like star clusters.
Topics to be covered in this talk:
1.
Investigate the evidence that nearly all elliptical galaxies (and
some S0 galaxies) in today’s rich clusters were formed very
early and their properties are basically independent of the
cluster environment. Look at some counter-evidence, too.
2. How is the evolution of disk galaxies affected by the cluster
environment, or is that too a second-order effect? Star
formation is more enhanced in cluster galaxies at z>0.3
compared to z=0, but it is always less than in the field at the
same-epoch. A morphology-density relation holds for the
z > 0.3 clusters, and starbursts are quite common. What
mechanisms are responsible for these effects – do they
operate in rich clusters or in the infall to rich clusters?
3. Describe an instrument well suited for studying the outskirts
of galaxy clusters: IMACS on the Magellan-Baade Telescope,
and introduce the IMACS Cluster-Building Survey)
Present-epoch morphology/density
relation (Dressler 1980)
At the Nearly Normal Galaxies
II meeting, Sandy Faber
showed this morphologydensity relation and remarked
that it influenced all of us to
take seriously the idea that
elliptical galaxies are formed
by mergers (following Toomre
& Toomre)
I was very surprised – we all
know that ellipticals in today’s
cluster cores are in
environments that suppress
mergers (high ΔV’s). The
alternate point of view is that
cluster environment has
nothing to do with it, at least
not today’s environment, but
that the prevalence of E’s in
rich clusters tells us about the
environment in which these
galaxies were born, at z > 2.
Let’s look at the evidence that suggests that cluster elliptical
galaxies are very old, zf > 2. If this is true, E’s formed in a
very different environment from the cores of today’s clusters,
in a less-deep potential well (lower velocity dispersion), without
a pervasive hot ICM and without strong tidal effects from the
cluster core.
from Gladders et al.
Color-magnitude relations (CMR) as a tool for determining
the stellar ages of elliptical galaxies
z=0
*Bower R. G., Lucey J. R., Ellis R., 1992, MNRAS, 254, 589
were the first to do this well, in Virgo and Coma, concluding
that the small scatter of the red sequence implied formation
ages z > 2.
Extending this to higher redshift increases the “leverage” on
the formation age of the red galaxies:
“The Homogeneity of Spheroidal Populations in Distant
Clusters” – Ellis et al. (Morphs), 1997 ApJ, 483, 582.
“The evolution of early-type galaxies in distant clusters” -Stanford, Eisenhardt, & Dickson, 1998 ApJ, 492, 461.
“The Slope of the Cluster Elliptical Red Sequence: A Probe of
Cluster Evolution” – Gladders et al, 1998 ApJ, 501, 571.
z=1.2
“Advanced Camera for Surveys Photometry of the Cluster
RDCS 1252.9-2927: The Color-Magnitude Relation at z = 1.24”
-- Blakeslee et al., 2003 ApJ 596, L143
“Simulations using the latest stellar
population models indicate an age
scatter
for the ellipticals
about
Deep HST/ACS
imaging –ofgalaxies
34%,
with abymean
age τL>~2.6 Gyr
selected
morphological
(corresponding
zL>~2.7,
classification,to
(3.5
mag ofand
LF)the
last
star formation
occuring
at
including
distinguishing
E’s from
zend
>~1.5.
elliptical
galaxies
S0’s.
52 …Thus,
galaxies:
σ = 0.029
mag!
were
already
well
in Xor 0.024
mag
forestablished
30 ellipticals
ray-luminous clusters when the
universe was a third of its present
age.”
Slope (top) and scatter (bottom) of the rest-frame (U-B) CM
relation as a function of redshift. The filled circle is from the
present work; the open circles show results from, in order of
increasing redshift, Bower, Lucey, & Ellis (1992); van Dokkum et
al. (1998); Ellis et al. (1997); van Dokkum et al. (2000); and van
Dokkum et al. (2001). The dotted lines indicate the average
values.
No dependence on zero-point or slope of CMR with
regard to richness and x-ray luminosity
“The Environmental Dependence of Galaxy Colors in Intermediate-Redshift
X-Ray-selected Clusters” – Wake et al., 2005 ApJ 627, 186
“We find that the “red” galaxies form a tight color-magnitude relation
(CMR) and that neither the slope nor zero point of this relation changes
significantly over the factor of 100 in X-ray luminosity covered by our
sample.” – neither does the blue fraction, but both “depend significantly
with cluster radius…It thus appears that the global cluster environment, in
the form of cluster mass (Lx), has little influence on the properties of the
bright cluster galaxies, whereas the local environment, in the form of galaxy
density (radius), has a strong effect.
Also, “Homogeneity of early-type galaxies across clusters” – Andreon, 2003
A&A, 409, 37. SDSS investigated – no dependence of red sequence on group
or cluster richness.
z=0
“Hubble Space Telescope Photometry and Keck Spectroscopy
of the Rich Cluster MS 1054-03: Morphologies, ButcherOemler Effect, and the Color-Magnitude Relation at Z = 0.83”
– van Dokkum et al., 2000 ApJ, 541, 95.
van Dokkum and collaborators have argued that this tight
CMR red-sequence at z~1 is an illusion.
Progenitor bias – you only see on the red sequence galaxies
where star formation has ceased, and this could be recent,
and the progenitors of the next bunch of red-sequence
galaxies can be actively forming stars and blue only 1-2 Gyr
earlier.
z=1.2
The quantitative effects of progenitor bias:
“The elliptical galaxy colour magnitude relation as a discriminant
between the monolithic and merger paradigms” – Kaviraj et al., 2005
MNRAS, 360, 60.
“We use a cold dark matter hierarchical merger model of galaxy formation to
investigate the existence and redshift evolution of the elliptical galaxy CMR in the
merger paradigm. We show that the SFH of cluster ellipticals predicted by the
model is quasi-monolithic, with only 10 per cent of the total stellar mass forming
after z = 1…[The CMR] is not a meaningful tool for constraining the SFH in the
merger paradigm, since a progressively larger fraction of the progenitor set of
present-day cluster ellipticals is contained in late-type star-forming systems at
higher redshift, which cannot be ignored when deriving the SFHs.
“Spectroscopic Confirmation of Multiple Red GalaxyGalaxy Mergers at z < 1” -- Tran et al. 2005 ApJ627,L25
The apparent mergers are real…R<14 kpc, ΔV < 165
km s-1. “…these bound pairs must evolve into E/S0
members by z = 0.7…most if not all, of its earlytype members evolved from (passive) galaxy-galaxy
mergers at z < 1”
Identified as a special
epoch in the cluster’s
life, or significant
subcluster merging
(relative velocities
still low), before
virialization
Can progenitor bias “save the day” for CDM hierarchical manufacture of
massive elliptical galaxies?:
Problems to overcome:
(1) The scatter in the CMR does not seem to increase with z, as predicted
(2) The luminosity function of clusters does not shift as predicted.
(3) The fundamental plane is not populated differently at z~1
(4) The morphology-density relation at z~1 shows the same fraction of
luminous ellipticals at a given density – it should instead contain the
progenitors, which would include late-types.
(5) The high Mg/Fe ratios of luminous ellipticals seem to argue for prompt
enrichment, i.e., early star formation, with later (Pop II supernoave)
suppressed (Trager et al. 2000 AJ, 119, 1645)
Kaviraj et al.
“Evolution of Cluster Elliptical Galaxies” – Barrientos & Lilly, 2003
ApJ, 596, 129
A study of 8 clusters at z ~ 0.45
“We find that the luminosity function for red galaxies with -25.6 < MV < -17.6…
are well described by a Schechter function for all clusters. In spite of the
fact that the clusters span a wide range in richness and X-ray
luminosities, all the cluster luminosity functions are consistent with a single
value for their characteristic magnitude (M*). The value of M* at z = 0.45
appears brighter than that observed in local clusters ( M = -0.94 ± 0.41), in
agreement with models of passive evolution and with studies of surface
brightness and fundamental plane at these redshifts.”
“Ks-band luminosity function of the z = 1.237 cluster of galaxies
RDCS J1252.9-2927”, Toft et al., 2004 A&A, 422, 29.
“The brightening of the characteristic magnitude, and lack of evolution in the shape
DeepisVLT/ISAAC
K-band
(plusformation
of the bright end of the LF to redshift z=1.237
consistent with
a simple
optical
forthe
photo-z’s)
scenario in which the massive elliptical galaxies
thatimaging
dominate
bright end of
a very distant
cluster
the K-band LF are passively evolving systemsof
assembled
at high
redshift…also in
agreement with the observed properties of the CM-relation of elliptical galaxies in
RDCS J1252.9-2927, … consistent with old populations of stars formed at 2.7 < zf <
3.6... From the evolution of the CM relationship alone it is not possible to distinguish
between formation scenarios where the old stars are formed in monolithic
collapse…[or] in the disks of less massive late-type galaxies which later merge to
form the ellipticals, as long as the merging does not trigger significant star
formation. From the lack of evolution in the shape of the bright end of the K-band LF
we can however deduce that if the massive ellipticals in clusters formed through
merging, it took place at higher redshifts (z >> 1 ) than is predicted by current semianalytical models.”
p.s. van Dokkum and Stanford, 2003 ApJ, 585, 78, have measured velocity
“…bright
endfor
of the
was
already
in
dispersions
3 ofLF
the
bright
galaxies
in this cluster, which confirm the large mass
place
at z=1.237,
while the
M > 10^11
solar masses
forflattening
two of these. The third is a factor of two less massive
of
thestrong
faint end
slope
suggests
that
with
Balmer
lines
in its spectrum.
All three fall into the same “extremely red
clusters
at z~1 “…our
containresults
relatively
object class.”
show that it is hazardess to use simple models for
smaller
fractions
of low
converting
luminosity
tomass
massgalaxies
for these objects.” See also Holden et al. 2005 ApJ,
than
in the
local
620,clusters
L83. “Thus,
even
in universe.”
the centers of massive clusters, there appears to have been
significant star formation in some massive (> 10^11 Msun) galaxies at z ~ 1.5”
The Buildup of Red Sequence in Galaxy Clusters since z
~ 0.8, De Lucia et al, ApJ, 610, L77.
Uses 4 clusters from EDisCS (VLT)
project, z=0.7-0.8 to study the
evolution of the red sequence in
earlier epoch clusters. Comparing
to the Coma cluster, they find that
the luminous L/L* > 0.4 galaxies are
“in place,” and old, passively evolving
population. In contrast, the less
luminous galaxies 0.1 > L/L* > 0.4 are
under-represented. They conclude
that many of these galaxies are
arrived more recently, and that star
formation continued for them after
z = 0.8.
From Toft et al., a summary of K-band cluster photometry – the evolution of K* - and its comparison with models of Kodama & Arimoto (1997)
“On the evolutionary status of early-type galaxies in clusters at z ~ 0.2
I. The Fundamental Plane” – Fritz et al. 2005 MNRAS, 358, 233
Photometry, morphology, and spectra for galaxies in A2390 & A2218
“For the total sample of 34 E+S0 cluster galaxies which enter the FP we deduce only a
mild evolution with a zero-point offset [from Coma] of 0.10 ± 0.06, corresponding to a
brightening of 0.31 ± 0.18 mag. Elliptical and lenticular galaxies are uniformly distributed
along the FP with a similar scatter of 0.1 dex. Within our sample we find little evidence
for differences between the populations of elliptical and S0 galaxies… Our results can be
reconciled with a passive evolution of the stellar populations and a high formation
redshift for the bulk of the stars in early-type galaxies. However, our findings are also
consistent with the hierarchical formation picture for rich clusters, if ellipticals in
clusters had their last major merger at high redshift.”
Two projections of the fundamental plane:
“The Detailed Fundamental Plane of Two High-Redshift Clusters: MS 205304 at z=0.58 and MS 1054-03 at z=0.83” – Wuyts et al., 2004 ApJ, 605, 677
“To avoid a bias induced by the magnitude limit of our sample,
we focus on the high-mass end, where selection effects are less
relevant. Applying a mass cut at M > 10^11 M to all four
considered clusters below z = 0.6 and at M > 10^11.5 Msun to
the two higher redshift clusters at z = 0.83 and 1.27 increases
the best-fitting formation redshift from zform = 2.26 to 2.95.”
“The Fundamental Plane of Cluster Elliptical Galaxies at z=1.25”
- Holden et al., 2005 ApJ 620, 83
-
Velocity dispersion of 4 cluster members, plus 3
members of another cluster at z=1.27, gives
characteristic age of 3.0 Gyr, or z* = 3.4 +/- 0.5.
Including progenitor bias lowers these numbers to
2.1 Gyr and z* = 2.3 – still quite old!
High-z cluster
members are red dots
and green crosses;
Coma cluster
ellipticals are “stars”
“Two lower mass galaxies in our z=1.25 sample have much lower M/L
values, implying significant star formation close to the epoch of
observation. Thus, even in the centers of massive clusters, there
appears to have been significant star formation in some massive,
M~10^11 Msun, galaxies at z~1.5.”
Kuntschner & Davies “The Ages and Metalicities of early-type
galaxies in the Fornax Cluster (1998 MNRAS, 295, L33)
“…we find that the Fornax ellipticals follow the locus of fixed
age in Worthey’s models…the lenticular [S0] galaxies, however,
exhibit a substantial spread to younger luminosity-weighted ages,
indicating a more extended star formation history.
cluster
field
unpublished work of Trager, Faber, and Dressler
Result so far: elliptical galaxies in clusters 5 Gyr ago look
younger, but old even then: ages of > 4 Gyr 5 Gyr ago
Black points – Coma; Red points A851 (z = 0.41)
Squares = ellipticals, triangles are S0’s
Before we leave these kinds of studies, we can briefly ask
the question, are field ellipticals really all that different
(in age) from cluster E’s?
“Evolved Galaxies at z>1.5 from the Gemini Deep Deep Survey: The Formation
Epoch of Massive Stellar Systems”, McCarthy et al. (GDDS), 2004 ApJ, 614, L9
“Conservative age estimates for 20 galaxies with z > 1.3…give a median age of 1.2 Gyr and
zf = 2.4. One-quarter of the galaxies have inferred zf > 4. Models restricted to [Fe/H] ~0
give median ages and zf of 2.3 Gyr and 3.3, respectively. These galaxies are among the most
massive and contribute 50% of the stellar mass density at 1 < z < 2. …Our results point
toward early and rapid formation for a significant fraction of present-day massive galaxies.”
Can feedback with hierarchical CDM explain? See Cirasuolo et al, 2005 ApJ, 629, 816.
“Spectra of evolved GDDS galaxies with z > 1.3. The
SDSS LRG composite has been overlaid on each
spectrum, and an offset has been applied to each, in
steps of 10^-18 ergs cm-2 Å-1”
“Both composite spectra show strong Mg II 2800, Mg I 2852
absorption and broad spectral features due primarily to Fe II
absorption. Overlaid in red is a single-burst Bruzual & Charlot
spectral synthesis model with an age of 2 Gyr, solar abundances,
and a Salpeter IMF cutoff at 120 Msun”
The DEEP Groth Strip Survey. IX. Evolution of the Fundamental Plane of
Field Galaxies – Gebhardt et al., 2003 ApJ 537, 239.
Velocity dispersions and optical photometry for 21
“early-type” galaxies in the Groth strip.
“The difference in the degree of evolution between our field sample and
published cluster galaxies suggests a more recent formation epoch, around z=1.5
for field galaxies compared to z>2.0 for cluster galaxies. The magnitude
difference implies that the field early-type galaxies are about 2 Gyr younger
than the cluster ellipticals using standard single-burst models. However, the
same models imply a significant change in the rest-frame U-B color from then to
the present, which is not seen in our sample. Continuous low-level star formation,
however, would serve to explain the constant colors over this large magnitude
change. A consistent model has 7% of the stellar mass created after the initial
burst, using an exponentially decaying star formation rate with an e-folding time
of 5 Gyr.”
Put this together with the claim that evolution in the cluster
galaxies at z ~ 1 is luminosity dependent, and you find another
related result:
“Mass-to-Light Ratios of Field Early-Type Galaxies at z ~ 1 from Ultradeep
Spectroscopy: Evidence for Mass-dependent Evolution”
– Van der Wel et al., 2005 ApJ, 145, 162
27 early-type field galaxies 0.6 < z < 1.15, VLT spectra, HST photometry
“For galaxies with masses M > 2 × 10^11
Msun , we find no significant difference
between the evolution of field and cluster
galaxies: ln(M/LB) = (-1.20 ± 0.18)z for
field galaxies and ln(M/LB) = (-1.12 ±
0.06)z for cluster galaxies.”
“corrected” for cluster galaxy evolution
Wilmer et al. 2005, astro-ph 506041, The DEEP2 Redshift Survey: The Galaxy
Luminosity Function to z ~ 1
Faber et al. 2005 astro-ph 506044,
Galaxy Luminosity Functions to z~1:
DEEP2 vs. COMBO-17 and Implications
for Red Galaxy Formation
Quenching star formation in spheroids?
Alternatively, are we seeing the end of star
formation in the spiral disks that become
S0’s? Faber says no, because these are
the brightest early-types, which are
usually E’s.
Summing up part 1:
1) A good deal of evidence points to an early formation epoch, z > 2,
for elliptical galaxies in rich clusters, with no significant star
formation after z = 1.5.
2) Progenitor bias could mean that some much younger systems
became ellipticals and that they experienced ongoing star
formation to z < 1, but there is no clear evidence that this actually
happened often and contributed significantly to the population of
ellipticals in rich clusters. Various tests, LF, FP, σ(CMR) seem to
suggest progenitor bias may not be a big effect.
3) More sensitive measurements of age and metallicity in stellar
populations at 0.5 < z < 1.3 should decisive in distinguishing between
these two possibilities in the near future.
Implications of part 1:
1) For those ellipticals that formed at z=2 or earlier, mergers are a
likely formation mechanisms, as hierarchical clustering suggests, and
they are likely to be mergers of gas rich systems. However, the
systems that are merging would not have been very much like the
giant spiral galaxies we see today, which we believe formed over a
much longer timescale.
2) The places where this happened were destined to be the cores of
rich clusters, but they too were unlike the rich clusters we see
today. A better picture is that of a modern group environment, but
with a higher density of galaxies than we see in today’s groups and
very gas rich systems.
3) Most elliptical galaxies are not influenced by the environment of
today’s rich clusters.
4) Questions: How different was the origin of most field ellipticals?
What becomes of the famous examples of disk-disk mergers we see
today? S0’s outnumber field E’s by 2:1.
Present-epoch morphology/density
relation (Dressler, 1980 ApJ, 236, 1)
Commonly cited as evidence
for ram pressure stripping
in the conversion of spirals
to S0’s
…but not by me. I
suggested that a
relationship like this, such
a slow function of density,
must have arisen at early
times when the density
contrast was much smaller.
And, most S0 galaxies are
in the field!
In Dressler et al (1997) the Morphs claimed
that there is a striking deficiency of S0
galaxies in intermediate-redshift compared to
environments in present-epoch. If confirmed,
this would be critical for understanding the
fate of the infalling, star-forming galaxies.
Present-epoch
morphology/density relation
…Morphs clusters at z ~ 0.5.
Comparing at the same galaxy density
(which is important), ellipticals represent an
equal (or even greater) percentage of the
cluster population. The big difference in the
2 diagrams is the paucity of S0 galaxies
compared to today’s clusters.
“Evolution since z = 1 of the Morphology-Density Relation for
Galaxies” -- Smith et al., 2005 ApJ, 620, 78
“The Morphology-Density Relation in z ~ 1 Clusters” -- Postman et
al., 2005 ApJ, 623, 721
“The evolution in the MDR is confined to
densities > 40 galaxies Mpc^-2 and
appears to be primarily due to a deficit of
S0 galaxies and an excess of Sp+Irr
galaxies relative to the local galaxy
population. The f(E)-density relation
exhibits no significant evolution between
z = 1 and 0. We find mild evidence to
suggest that the MDR is dependent on
the bolometric X-ray luminosity of the
intracluster medium.”
Environmental influences on galaxies
• (Galaxy-galaxy collisions stripping, a la Baade & Spitzer)
• Ram pressure stripping – Gott & Gunn, Abadi, Moore, & Bower…
ram pressure induced star formation? -- Dressler & Gunn,
Bekki & Couch 2003
• Evaporative gas removal – Cowie & Songaila, Nulsen
• Starvation (removing gas-rich halos) – Larson, Tinsley, & Caldwell
Bekki et al.
• Mergers, accretions, tidal interactions – Toomre, Mihos
(driving secular evolution? – see Zhang 1999 ApJ 518, 613)
• Harassment (tidal, galaxy & cluster) – Moore et al.
In the 1970’s, global processes like ram pressure stripping were nearly
exclusively cited for environmental changes such as the rarity of spirals in
clusters. There is now good evidence that stripping does occur, but that
it is not the primary mechanism of S0 production. In fact, these
mechanisms associated with the dense cluster core may not be the most
important for turning spiral galaxies into S0’s.
“A Wide-Field Hubble Space Telescope Study of the Cluster Cl 0024+16 at z =
0.4. I. Morphological Distributions to 5 Mpc Radius” – Treu et al., 2003 ApJ,
591, 53
Discusses the efficacy of different environmental processes and concludes
lower-density “long-range” processes are dominant, not cluster-core processes.
“The Environmental Dependence of Galaxy Colors in Intermediate-Redshift Xray Selected Clusters” – Wake et al., 2005 ApJ, 627, 186.
“It thus appears that the global cluster environment, in the form of cluster mass (LX),
has little influence on the properties of the bright cluster galaxies, whereas the local
environment, in the form of galaxy density (radius), has a strong effect. The range of
~100 in LX corresponds to a factor of ~40 in ram pressure efficiency, thus suggesting
that ram pressure stripping or other mechanisms that depend on cluster mass, like tidal
stripping or harassment, are unlikely to be solely responsible for changing the galaxy
population from the “blue” star-forming galaxies, which dominate low-density
environments, to the “red” passive galaxies, which dominate cluster cores.”
Courtesty of Andrey Kravstov: Accretion,
Mergers Shocks, & Turbulence when subclusters
of galaxies coalesce
gas density
gas
gasentropy
entropy(slice)
Norman & Bryan 1998
Nagai, Kravtsov & Kosowsky 2003
Sunyaev, Norman & Bryan 2003
Also thanks to Jaqueline
van Gorkum
10 Mpc
Courtesy of A.Vikhlinin: Chandra images of z = 0 and z = 0.6 clusters
nearby clusters
distant clusters (z ~ 0.6)
“The Evolution
of of
Population
Gradients
in Galaxy Clusters: The
Spectroscopic
version
morph-density
relation:
Butcher-Oemler
Effectcan
andbe
Cluster
Infall”
– Ellingson, Lin, Yee, &
“It appears
that the cluster
modeled
as the
2001 ApJ,
547, 609 component of
sum ofCarlberg,
two components:
a virialized
older galaxies, and a younger component which, while
Spectroscopy
of CNOC1 sample
of cluster
x-ray
itCFHT
is probably
in quasi-equilibrium
with the
luminous clusters
< z < recently
0.55. PCA
classification
potential…has
fallen0.18
in more
and
which may
of spectra
good stepto
forward.
eventually
beaexpected
transform both spectrally
and dynamically to blend with the older population.
Confirm Butcher-Oemler effect but note that, for
Galaxies in the midst of this transformation (e.g.,
this x-ray luminous sample, considering only galaxies
K+A galaxies, or the "PSF" PCA component) appear
within 0.5 R200 would result in much smaller evolution.
to inhabit an intermediate
spatial distribution.”
“The main difference in these curves is then a simple
vertical shift, and these results can thus be
interpreted as a straightforward decline in the infall
rate of new galaxies into the cluster.
“…implying that about 6% of the stellar mass in the
cluster has recently fallen into the cluster at an
observed epoch of z ~ 0.45. Thus, the ButcherOemler effect illustrates the sensitivity of galaxy
populations in tracing even a very small fraction of
infall into the cluster.
…further suggests that the infall rate has declined
further by a factor of 3, by z ~ 0.2.”
“The
Color-Magnitude
Relation
CL 1358+62
at Zcluster
= 0.33: center:
Evidence for
Gradient
in C-M scatter
withindistance
from
Significant
Evolution
Population”
– van
1998
“The intrinsic
scatterininthe
theS0
rest-frame
B-V
CMDokkum
relationetofal.,
the
ApJ,
500,galaxies
714.
elliptical
is very small, ~0.022 mag. The CM relation of the
ellipticals does not depend significantly on the distance from the
This study uses spectroscopically-confirmed cluster members, and
cluster
center. In contrast, the CM relation for the S0 galaxies does
HST morphologies, to select a sample of E’s and S0’s. With HST
depend on radius: the S0's in the core follow a CM relation similar to
multi-color photometry, the scatter in the color properties are
that of the ellipticals, but at large radii [R > 0.5 h Mpc] the S0's are
compared, and also studied as a function of radius: 4.6’  ~1 Mpc –
systematically bluer and the scatter in the CM relation approximately
not very far, but still…
doubles, to ~0.043 mag.”
The authors go on to
model the continuing star
formation in the S0’s in
contrast to the E’s and
conclude that “…~15% of
the population in presentday early-type galaxy
population in rich clusters
was accreted between
z=0.33 and z=0.”
“MS 2053 is a classic Butcher-Oemler cluster: 24% of its
“Infall, themembers
Butcher-Oemler
of Blue (44%) are
are blue Effect,
galaxies,and
andthe
an Descendants
even higher fraction
Cluster Galaxies
at z~0.6”
– Tran et
al., 2005
ApJ,of619,
star-forming.
However,
more
than half
the134.
blue/star-forming
members belong to the infalling structure (MS 2053-B). Unlike
Extensiveprevious
HST imaging
andthat
Keckhave
spectroscopy
confirmed
studies
found only(149
indirect
evidence for the
cluster members!)
of MS
a massive,
X-ray
luminous
cluster
link between
the2053-04,
B-O effect
and galaxy
infall,
our unique
data set
at z = 0.59.
enables us to show conclusively that this is the case in MS 2053.”
from the abstract: “MS 2053's deficit of S0 galaxies combined
“We
find implies
two pronounced
redshift
with its overabundance of blue
spirals
that many
of these
at z = 0.5840
(MS 2053-A)
late-type galaxies will evolve peaks,
into S0one
members…Our
observations
a smaller
at z = 0.5983
(MS
show that most of MS 2053'sand
blue
cluster one
members,
and
2053-B)”
– ΔV ~originate
2700 km/s
ultimately most of its low-mass
S0 galaxies,
in the
field.”
“More than half of the blue/late-type
cluster members are in MS 2053-B.”
What seems to be
emerging from all these
studies of the radial
dependence of cluster
populations is that the
effects of environment
on disk galaxies begin at
surprisingly low galaxy
densities.
This is actually an old
result. Notice how, in
the morphology-density
relation the population
gradients in this relation
begin at very low
densities, orders-ofmagnitude lower than
cluster cores.
Postman & Geller 1984,
ApJ, 281, 95
Extended the
morphology-density
relation outside of rich
clusters of galaxies out
into low-to-moderate
density groups. The
trends were seen in
these environments,
which are very unlike
the cores of rich
clusters.
Pisces-Perseus Supercluster
Giovanelli, Haynes, & Chincarini 1984 ApJ, 300, 72
Distribution of all
morphological types
E,S0,S0a
Sa,Sab, Sb,Sbc
Sc, S
Later than Sc
Sloan Digital Sky Survey result
75th
SFR-density
Morphological fraction
Critical Density: 1 galaxy per h2 Mpc-2
Sc
Morph-radius
25th
C4 Clusters
median
Surface density (10th NN) in a redshift
slice (pseudo 3D density) Mpc-2
SDSS sees a dramatic
decrease in the strongly starforming Sc galaxies: in-fall
regions of clusters
There is universal agreement among researchers that dense
environments “quench” star formation, and the evidence I’ve just
shown suggests it doesn’t take much of a rise in density to see the
effect. In most studies, [O II] is the sole indicator of star
formation activity.
It requires higher resolution, and higher-S/N spectra to see the
Balmer absorption lines, which record the history of starbursts. My
colleagues and I believe that starbursts, which are much more
prevalent in earlier-epoch clusters than today, are an important clue
as to what mechanisms are actually at work in the “quenching,” and
the accompanying conversion of spirals to S0’s
…a few slides to demonstrate the prevalence of
starbursts at z > 0.3
“Studying the Star Formation Histories of Galaxies in Clusters from Composite
Spectra” – Dressler et al., 2004 ApJ, 617, 867.
The Morphs have co-added, cluster by cluster, the 30-60 spectra in
each of 8 clusters. These composite spectra, high S/N, enable
accurate measurements of line strengths.
1. K-type spectral features (metal lines) show the fractional
representation of “old” (τ > 3 Gyr) stars
2. [O II] strength (or better, Hα emission – if you have it) measures
the rate of ongoing star formation (timescale: τ ~ 10-100 Myr)
3. Balmer absorption, specifically Hδ absorption, reveals the presence
of a starburst – the fraction of intermediate age stars (A-F stars),
enhanced over normal levels compared to galaxies with continous
(quiescent) star formation (timescale: τ ~ 1 Gyr)
Composite [O
spectra
for 12 present-epoch
II]
Hδ clusters and
8 intermediate-redshift clusters
– typically 30-60 galaxies coadded per cluster
Spectra are normalized, then luminosity-weighted and summed
5 clusters (8 total)
z ~ 0.5
Morphs sample
5 clusters (12 total)
z~0
Dressler-Shectman
sample
Coadded
by Hδ
strength
...a remarkable change in the amount and character of
star formation in the recent past.
In rich clusters of galaxies,
but even in “the field,” the
mode of star formation
appears to shift to a much
greater fraction which are
starbursts (20-30%)
The disappearance of
starbursts appears to be
cosmic evolution rather than
cluster specific (no evidence
for variation cluster-tocluster).
Recipe for a cluster at z = 0.5
(from the Morphs cookbook)
Starbursts!
About 30%
of the
population!
It’s important to emphasize that, unlike young
starbursts – i.e., those associated with AGN or HII
galaxies – these spectra arise from elevated star
formation rates for long periods, τ > 108 yr, which is
comparable to the dynamical time in the galaxy. ‘A’
stars diffuse out of the star formation regions and
are visible over extensive regions. In some cases, it
appears that the starburst has been a global
phenomenon in the galaxy, not just confined to the
center as many simulations suggest.
Las Campanas
Redshift
Survey: 1 in 10
DresslerShectman
nearby clusters
Liu & Kennicutt
nearby mergers
2+ σ
limits
A851 (z=0.41)
cluster members
field galaxies
0.35 < z < 0.55
Conclusion: Starbursts are a common mode of star formation
at z ~ 0.5, both in and out of clusters
According to these early
data, both in clusters and
the field, intermediateredshift galaxies seem to
have a much stronger
propensity for strong
starbursts than today.
Is this an extension to
higher luminosities what is
seen in some present-epoch
dwarf galaxies?
Given the timescale for
this phenomenon, and the
significant percentage of
the population (10-30%)
that seems to be involved,
one might speculate that
earlier than z = 1 we will
find no continuously-starforming spirals like the
Milky Way.
Gallart, Aparicio, & Freedman
Strong Balmer line galaxies
are mostly disk galaxies. A
significant amount of ongoing
star formation could be
hidden (Glenn Morrison’s VLA
work on radio continuum
detections)
k+a (and a+k)
– post starburst
Because the Balmer line galaxies
are mostly disk galaxies, it’s
reasonable to believe that – in
addition to quenching – the
starburst activity that is so
common in intermediateredshift clusters is part of the
process that turns spiral
galaxies into S0 galaxies out of
spiral since z ~ 1.
e(a) – with weak
emission: dusty
starbursts!
What might trigger these starbursts?
Exploring the outskirts of Abell 851 (z = 0.405)
Image by
Kodama & Smail
“il tripico” – three
Abell 851 is a
members of a small
massive cluster
infalling group, two with
undergoing a
strong Balmer absorption,
major subcluster
one with very strong
merger.
emission (starburst)
ROSAT HRI
image of A851
(Schlinder &
collaborators)
ΔV = 2214 km s-1
σ = 280 km s-1
HST + X-ray contours from Rosat
3 megaparsecs
Spectral types
of 113 cluster
members in
A851
A low yield out
in the
filaments –
only 9/58
spectra (with
redshifts) are
members of
A851!
sum spectra…
filaments
HST
WFPC2
starbursts
The infalling group or filament, at the redshift of
A851, cold (σ ~ 200 km s-1), far away from the
cluster core (~2.5 Mpc) and full of starbursts.
KPNO groundbased image
from Morrison
The suburbs of
A851: 7/9 are
starbursts!
only 2 k-type
and no e(c)’s!
Hypothesis: As gas-rich disk
6 are e(a),
galaxies fall into these rich
1 k+a
clusters, either individually
or in small groups, they are
severely jostled by tidal
interactions with each other
and their new neighbors.
This seems to give rise to
starbursts and effectively
ends their lives as spiral
galaxies, leaving most as
burned-out S0 galaxies.
Alternate hypothesis:
shocks from intracluster
medium drive the
starbursts.
If the data for Abell 851 are representative, they suggest that the
major evolutionary effect that is going on in clusters since z = 1 is
the infall of spiral galaxies, most in small groups, in which merging
and interactions, and/or a strong interaction with a dynamic ICM,
drive starbursts that push these galaxies along the path to S0s. For
many starbursts, we see evidence of distortions and multiplicity
indicative of mergers and/or tidal effects, yet in others we see little
of this – yet a starburst has triggered. Could this be due to shocks in
the ICM driven by cluster mergers?
Either way, such observations suggest that cluster-core processes
are not the first to influence the evolution of infalling galaxies, and
they may be less important.
A new wide-field imager
and spectrograph for
Magellan: IMACS
The Magellan Consortium: Two 6.5-m telescopes shared
by Carnegie, Harvard, Arizona, MIT, and Michigan
IMACS – Inamori-Magellan Areal Camera & Spectrograph
A 5-year, $6M project of the Carnegie Observatories commissioned on the
Baade (Magellan I) telescope in August 2003
IMACS is an imaging multislit spectrograph at the Nasymth focus.
The challenge: move 1,000 lbs of optics with
6,000 lbs of steel and keep deformations to a
few thousandths of an inch (for mirrors and
gratings in the pupil, only a few microns!)
IMACS optical layout
Magellan telescope
5 feet
collimator
f/2.5 camera
1/2º field
f/4 camera
1/4º field
field
lens
Reflecting grating
for spectroscopy
spectrum
64 million pixel
CCD camera
IMACS will record up
to 500 spectra of
stars and/or galaxies
at one time.
IMACS is ideal for imaging and spectroscopy of galaxy clusters.
It is a direct-imaging spectrograph, not a fiber spectrograph
Grism dispersers:
1000 < R < 5000
27 arcmin
Grating dispersers:
2000 < R < 12000
Also echellette mode
R =25000
15’
Durham
IFU
7
7”
5”
60”
f/2.5 camera
f/4 camera
0.2” per pixel 0.2 sq deg
0.11” per pixel 0.06 sq deg
Spectroscopy for up to ~700
objects per mask
Similar to DEIMOS, VMOS
~ 200 objects per mask
2 x 1000
fibers
0.2”
resolution
First light, NGC 253
August 18, 2003
the “cosmic web”
ICBS
IMACS Cluster
Building Survey
Alan Dressler, Gus Oemler, Mike Gladders, and Bianca Poggianti
are studying the R < 5 Mpc spheres around 10 clusters at z =
0.4, 0.8, and (we hope) 1.1 The goal is to pick out, among the
thousands of field galaxies, the few hundred infalling cluster
galaxies that are building the cluster. The IMACS Cluster
Building Survey is a study of the infall regions of distant
clusters, relating galaxy evolution and the building of rich
clusters, and exploring how present-day environment/population
relationships came about.
The first observations were taken in January ’04.
In addition to studying what’s going on 2-5 Mpc from
the centers of rich clusters, we want to connect
different epochs of cluster building. This means an
“apples to apples” comparison over epochs. To do
this, we define equal volumes at
0.35 < z < 0.50
0.75 < z < 0.85
1.05 < z < 1.20
Using the Red-sequence Cluster Survey (Yee &
Gladders), and supplemented by the SDSS, we choose
the richest clusters in each of these three volumes,
which insures that they will be “ancestors.” Our
initial goal is to study 5 clusters z ~ 0.4 and 5
clusters z ~ 0.8.
These are N-body dark-body
simulations made by Gladders
from the Hydra Consortium
code, with a mass resolution
of (643 particles) 1011 Mo and
a box size of 62 h-1 Mpc.
These 3 simulated clusters
are well assembled by z = 0.2,
but were in pieces at z ~ 1
and even at z ~ 0.5 that are
spread over many megaparsecs. Note how both
“field” and grouped galaxies
are incorporated. The dashed
circle shows the f/2 camera
field of IMACS, which
reaches to r ~ 5 Mpc for
z > 0.4.
z = 0.3
z = 0.6
z = 0.9
z = 1.2
12’
IMACS field
is twice this
diameter!
862 redshifts in the field of RCS0221-0346
RCS0221 field -- z distribution
70
Open: R > 2 Mpc
Shaded: R < 2 Mpc
60
50
redshift
40
Frequency
30
20
10
79
0.
76
0.
7
73
0.
number= 1050 km s-1
σ0(R<2 members)
0.
67
0.
64
0.
61
0.
58
55
0.
52
0.
0.
49
0.
46
0.
4
43
0.
0.
37
0.
34
0.
31
0.
28
0.
25
22
0.
19
0.
0.
16
0.
13
0.
0.
1
0
Every small dot is a
redshift, 862 in this
field. Large dots are
138 cluster and
supercluster
members.
HST/ACS
653 redshifts in the field of RCS1102
RCS1102 all redshifts
70
Open: R > 2 Mpc
Shaded: R < 2 Mpc
60
50
number
40
Frequency
30
20
10
redshift
σ0(R<2 members)
= 815 km s-1
8
0.
77
0.
74
0.
71
0.
68
0.
65
0.
62
0.
59
0.
56
0.
53
0.
5
0.
47
0.
44
0.
41
0.
38
0.
35
0.
32
0.
29
0.
26
0.
23
0.
19
0.
16
0.
13
0.
0.
1
0
653 redshifts in
field, 149 cluster
and supercluster
galaxies
Very flattened!
Hδ strengths in
composite spectra of
RCS0221
R < 1.2 Mpc
all member spectra
R > 1.2 Mpc
One of several in our sample, a starburst so strong
that the Balmer lines down in Hη are in emission.
The suburbs of
A851: 7/9 are
starbursts!
only 2 k-type
and no e(c)’s!
6 are e(a),
1 k+a
These early results of
two outer fields in the
ICBS suggest a very
rapidly starforming
population, with a
significant fraction of
active starbursts
EW[O II] < -40 ang
RCS0221 R>1.2 Mpc
RCS1102 R>1.2 Mpc
However, our early
measurements of the
“field” at this redshift
shows the same
distribution of EW[O II] –
is the cluster environment
playing a role or not?
Summing up: evolution of the disk galaxies in clusters
1.
Many cluster S0 galaxies were made – from spirals – since
z = 1.
2. Star formation is suppressed (quenched) in regions of high
galaxy density, but the influence of environment is evident
far from cluster cores.
3. Starbursts appear to be much a more common mode of star
formation at z > 0.3 than today.
4. Merging and tidal effects in the “group phase” may be the
dominant reason for 2 & 3, but the processes of the dense
core, including the dynamical process of cluster mergers,
may be final step in the conversion.
5. The IMACS Cluster Building Survey will extend studies of
intermediate-redshift clusters to the outskirts of these
clusters, with an extent and completeness not done before.
The “laboratory” may be just that
– a laboratory, not a petri dish.