Download R. Somerville

How are galaxies influenced
by their environment?
Predictions & insights from hierarchical models
rachel somerville
STScI
with thanks to
Eric Bell
the COMBO/GEMS team
Risa Wechsler
Andrey Kravtsov
Sandy Faber
what are observations telling us?
old wisdom: familiar correlations between
galaxy observables
color-morphology
color-magnitude
luminosity-metallicity
new wisdom: correlation between
intrinsic/physical properties , e.g.
stellar mass and star formation rate
stellar mass & metallicity
new wisdom: many fundamental galaxy
properties have bimodal distributions, divided
by a critical stellar mass
color
SDSS: color and magnitude
luminosity
Blanton et al. 2003
Baldry et al. 2003
~stellar age
stellar mass and age…
SDSS
stellar mass
Kauffmann et al. 2003, 2004
relative star formation rate
stellar mass and relative star
formation rate
stellar mass
Brinchmann et al. 2004
what imposes these relationships
on galaxies (internal/external)?
the old wisdom:
morphology-density relation: early type fraction
increases with density
Butcher-Oemler effect: early/blue fraction
decreases with cosmic time
the new wisdom:
(Hogg et al., Blanton et al., Balogh et al., Kauffmann et al.):
structural properties have weak dependence on
environment
spectro-photometric properties have a stronger
dependence on environment -- critical density?
morphology-density as a
function of redshift
high
density
low
density
projected density
lookback time
Smith et al. 2004
structure and density
increasing density
Kauffmann et al. 2004
age and relative star formation
and local density
increasing N
increasing N
Kauffmann et al. 2004
luminosity
has a strong
dependence
on local density
color has a weaker
dependence on
local density
Hogg et al.
2003
Balogh et al. 2004
fraction of red galaxies
increases with density
but the mean color of the red and blue
distributions changes little with density
the color magnitude
relation is in place at
z~1 and evolution is
consistent with passive
RDCS1252 z=1.24
magnitude
Blakeslee et al. 2003 (ACS GTO team)
…in the field as
well as clusters
(COMBO-17/
GEMS)
red dots:
early type
blue dots:
late type
age =
8.4Gyr
age =
5.5 Gyr
rest V magnitude (luminosity)
Bell et al. 2003
•do hierarchical models predict
this behaviour?
•can they give us any insight into
what is going on?
time
cluster
halo
‘Milky Way’
halo
Wechsler et al.
hierarchical simulations
show a clear correlation
between
color/morphology
and density, in
qualitative
agreement with
observations
Kauffmann et al. 1999
VIRGO/GIF simulations
see also
Benson et al. 2001;
Springel et al. 2001
dependence of mean color and
morphological fraction on halo mass
color
log halo mass
fraction of bulge/disk galaxies
Diaferio et al. 2001
inflation
primordial power spectrum
merger tree
collisional heating
radiative cooling
star formation
stellar feedback
chemical enrichment
stellar populations
dust absorption & emission
galaxy
observables
specific model ingredients
SFR 
*
 Vc

 *   t 

200km/s 
mcold
0
* dyn
*
 rh
 Vc 
mÝrh  SFR   

200km/s 
0
SN

reheated gas
ejected if
Vc>150 km/s
•major mergers (>4:1) trigger bursts of star formation
•Bruzual & Charlot 2003 multi-metallicity stellar
population models
SDSS & 2MASS luminosity functions
u-band
g-band
r-band
i-band
z-band
K-band
magnitude
(observed LF from Bell et al.
luminosity functions
by morphology
disk dominated
bulge dominated
gas fraction
SDSS Bell et al.
log stellar mass
gas fraction distributions
increasing stellar mass -->
color-magnitude relation
r-band magnitude
bright-->
color histograms
faint
g-r color
color of a passively
evolving burst formed
at z=5
u-r
g-r
Z=2xsolar
Z=solar
Global star formation history
stellar mass
assembly
history
new observational
estimates from
COMBO-17 and
GOODS
rss et al. 2004
stellar mass
assembly
history
estimates from
Glazebrook et al. (GDDS)
Rudnick et al. (FIRES)
Dickinson et al. (HDFN)
Fontana et al. (K20)
rss et al. 2004
color-magnitude and morphology at high redshift
red: n>2
blue: n<2
red: B/T>0.5
blue: B/T<0.5
missing EROs
13.5
5.8
3.2
1.0
0.5
0.1
GOODS
KAB<22
rss et al. 2004 GOODS ApJL
status: low redshift
hierarchical models can be made to
reproduce global luminosity/stellar mass
distributions at low redshift but don’t
produce enough luminous red galaxies
color magnitude relation has correct
slope (well, sort of) but distributions do
not match data and are not bimodal
status: high redshift
hierarchical models produce enough
massive galaxies to z~2
but, do not produce enough red
galaxies
the mean stellar ages of the massive
galaxies are old enough -- color
problem is caused by ‘frosting’ of young
stars
what makes red galaxies red?
need a process that quenches star
formation in the most massive galaxies
without drastically altering the mass
assembly/star formation history
environment: ram pressure or tidal
stripping, harassment?
internal: SN or AGN driven wind? global
instability?
neighbor counts (R=2 Mpc)
L=114 Mpc mp=3x108 M_sun r_force=1.5 kpc
number of
neighbors in 2
Mpc spheres
cyan: 0-1
blue: 2-3
green: 4-6
rust: 7-11
red: >11
number of neighbors vs. halo mass
log
halo mass
log(1+N)
color-magnitude by density
g-r
r-band mag
N=0-1
N>17
Kauffmann et al. 2004
N=0-1
increasing density-->
N=2-3 N=4-6 N=7-11
N>11
-22.5
-21.5
-20.5
-19.5
-18.5
increasing density-->
decreasing luminosity-->
u-r
Balogh et al. 2004
relative SFR vs. mass
log stellar mass
star formation/age as function of local density
increasing N
increasing N
Kauffmann et al. 2004
where does this leave us?
where do galaxies become red?
simulations: massive halos
real universe: all environments (though more often
in dense ones)
how do galaxies become red?
simulations: whole distribution shifts
real universe: galaxies ‘hop’ from one distribution
to the other
why do galaxies become red?
simulations: strangulation
real universe: ???
internal processes
all gas driven out of galaxy after a
merger by SN or AGN winds?
has almost no effect on colors because of
continuous infall of fresh gas
star formation ‘turned off’ when a bulge
(BH?) has grown beyond a critical
mass?
star formation shut
off when m_bulge
> 2 x 1010 M_sun
g-r color
better!
r-band magnitude
color distributions
bimodal!
bright
faint
g-r color