Transcript Friday03

MNRAS, submitted
group catalogue makers
Background
Cluster Galaxies
• Segregation of
luminosities,
morphologies, and
emission line fraction
well-known
• Early-types consistent
with passive evolution
since z>2
• Small fraction of
actively star-forming
galaxies
Nature or Nurture?
• Nature? Elliptical galaxies only form in
protoclusters at high redshift. Rest of
population is due to infall.
• or Nurture? Galaxy evolution proceeds along
a different path within dense environments.
Butcher-Oemler effect (for T.S.)
• Concentrated clusters at
high redshift have more blue
galaxies than concentrated
clusters at low redshift
• some of those blue galaxies
show signs of recent
changes (e.g. Couch & Sharples
1987)
• an argument for nurture??
Butcher & Oemler (1984)
(Why are we still talking about)
the Butcher-Oemler effect
• Magnitude, colour, and
radius cuts have strong
effect and need to be
theoretically motivated
to some extent
• A lot of scatter
– appears to be mostly due
to correlation with
cluster richness
• still room to worry
about cluster selection?
Margoniner et al. (2001)
But: Field galaxy evolution
• But field population also
evolves strongly (Lilly et al.
1996)
• Post-starburst galaxies
equally abundant in the field
(Zabludoff et al. 1996; Goto et al.
2003)
• So: does BO effect really
point to cluster-specific
physics, or just the evolving
field and infall rate (Ellingson
et al. 2001)?
Steidel et al. (1999)
If it is “nurture”…
Groups
Clusters
• Could cosmic SFR
evolution be a
consequence of
environment?
• Only if star formation
rates are low in groups
and low-density
environments as well as
clusters
Bower & Balogh (2003)
Theoretical expectations (?)
• Isolated galaxies have (invisible) halo of hot gas that
can cool and replenish the disk
– allows star formation to continue for longer
• Cluster galaxies lose this gas, so their SFR declines
more quickly. Also cluster galaxies form earlier.
– therefore SFRs should be lower in clusters
• no ram-pressure stripping, harassment needed to
achieve reasonable match to observed clusters (Diaferio
et al. 2001; Okamoto et al. 2003).
Observations from the SDSS
and 2dFGRS
First steps: nearby clusters
• Analysis of 2dFGRS
– Ha equivalent widths,
within 20 Mpc of known
clusters
– dependence of mean
SFR on local density
– “critical density”?
• Gradient is consistent
with the strangulation
hypothesis (Balogh et al.
2000)
Lewis, Balogh et al. (2002)
•
New study based on combination of SDSS and
2dF galaxy redshift surveys
•
Volume-limited sample of 24,968 galaxies at


0.05<z<0.1
Mr<-20.6 (SDSS); Mb<-19.5 (2dFGRS)
3 measures of environment:
1.
2.
3.
“traditional” projected distance to 5th nearest neighbour
3-dimensional density on 1 and 5 Mpc scales
velocity dispersion of embedding cluster or group

catalogues of Nichol, Miller et al. and Eke et al.
Bimodality
• SDSS colours show two
distinct populations
• Red population may be
the result of major
mergers at high redshift,
followed by passive
evolution
(u-r)0
Baldry et al. (2003)
Bimodality
• Same is seen in Ha
distribution: SFR is
not continuous
• galaxies do not have
arbitrarily low SFR
• So mean/median do
not necessarily trace a
change in SFR
The star-forming population
• Amongst the starforming population,
there is no trend in
mean SFR with
density!
• Hard to explain with
simple, slow-decay
models (e.g. Balogh et al.
2000)
Ha in Rich Clusters at z~0.3
(Field)
• Number of emission
lines galaxies is low in
all clusters
• However, shape of
luminosity function
similar to field:
– consistent with shift in
normalisation; not in
Ha luminosity
Couch et al. (2001)
Balogh et al. (2002)
Correlation with density
• The fraction of
star-forming
galaxies varies
strongly with
density
2dFGRS
• Correlation at all
densities; still a
flattening near the
critical value
Isolated Galaxies
• Fraction of SF galaxies in
lowest density
environments is not much
larger than the average
Average value
in full sample
2dFGRS
– So strong evolution in
global average cannot be
due only to a change in
densities
Isolated Galaxies
All galaxies
Bright galaxies
• Selection of isolated
galaxies:
– non-group members,
with low densities on
1 and 5.5 Mpc scales
• ~30% of isolated
galaxies show
negligible SF
– challenge for models?
– environment must not
be only driver of
evolution.
Group catalogues
• 2dFGRS (Eke et al.)
– Based on friends-of-friends linking algorithm
– calibrated with simulations. Reproduces mean characteristics (e.g.
velocity dispersion) of parent dark matter haloes
– is highly complete, at expense of having unphysical contamination,
esp. at low masses
– selected subsample with at least 10 members above our luminosity
limit.
• SDSS (Nichol, Miller et al.)
– Search for clustering in spatial and colour space; also calibrated with
simulations
– Selected subsample with Gaussian velocity dispersions
– is a highly pure sample, at expense of being incomplete
2dF groups
SDSS groups
circle size is
proportional to
virial radius
(vel. dispersion)
Large scale structure
• Little dependence
on cluster velocity
dispersion
• SFR depends
mostly on galaxy
density, not
embedding halo
mass.
Large scale structure
● s > 600 km/s
● 200 < s < 400
• Measured 3-d
density on 1.1 and
5.5 Mpc scales
• groups are wellseparated in this
plane, by velocity
dispersion
Large scale structure
r5.5 (Mpc-3)
0.050
0.010
0.005
• Emission-line
fraction appears to
depend on 1 Mpc
scales and on 5.5
Mpc scales.
Increasing fraction of Ha emitters
Interpretation?
Nature vs. Nurture
• Nature:
1. Dense regions just form a little earlier?
• would expect to see lower SFR among active population in
high-z clusters: not observed
2. Early-type population formed at high redshift?
• would have to be a substantial fraction of today’s cluster
population: so why does the fraction of SF galaxies evolve?
(or does it?)
Nature vs. Nurture
• Nurture: long
timescale processes?
z~0.3
z~0.1
– Difficult to
accommodate
strangulation model
– would need offsetting
effect (starbursts?
infall?) to keep HaLF
the same.
Nature vs. Nurture
Passive spirals in SDSS
•
Nurture: short timescale
processes?
1. trends at low densities and large
scales rule out ram-pressure
stripping as dominant effect
2. interactions: known to affect
SFR on short timescale (e.g.
3.
Goto et al. (2003)
Lambas et al. 2002)
Passive spirals (e.g. van den Bergh
1976; Poggianti et al. 1999; Balogh et
al. 2002; McIntosh et al. 2002)?
 More common in clusters
(Goto et al. 2003)
 mechanism?
Most likely scenario (for
bright galaxies)?
• Brightest ellipticals likely result of initial
conditions
• Galaxy-galaxy interactions:
–
–
–
–
more common in dense regions
change SFR on short timescale
effective over wide range of environment
evolve strongly with redshift (Patton et al. 2002; Conselice et
al. 2003)
– only environment known to effectively transform
SFR of a galaxy (e.g. Lambas et al. 2002)
Conclusions
• Distribution of star formation rates is
bimodal, not continuous (unlike
morphology?)
• SFR distribution among active population is
independent of environment
• Fraction of SF galaxies depends on local and
large-scale densities (?)
• Galaxy-galaxy interactions are the most likely
cause of observed segregation
True/observed emission line fraction
Projection Effects?
projected population
at field density
projected population 10
times more dense than
field
• Is star-forming
population all
projected??
• No: at high density,
contrast is high, and
area is small
– at low density, trend is
weak, so signal not
diluted by projection
Balogh et al. (2003)
Nature vs. Nurture
Blue galaxies only: (g-r)<0.7
• Nurture: clusters
directly affect SFR?
– short timescale?
• few (<0.1 %) E+As
• normal SFR for colour
• however, these don’t
provide strong
constraints: it is possible
to generate entire nonSF population in this
way