Transcript - Lorentz Center

Open Questions
Satellite galaxies II
(and other stuff)
Data rich subject
• Lots of data out there:
– Observers have
quantified how satellite
properties depend on
• stellar mass
• halo mass
• redshift
• But what do these
trends mean?
Take-away message from
yesterday:
• Stellar mass dominates
the properties of galaxies
•
Or is it “velocity dispersion”?
• Halo mass determines
fraction of galaxies that
are passive
• But star forming satellites
look just like centrals
A working hypothesis
• DM halo grows by accretion of
smaller haloes
• Smaller haloes orbit
– Outer halo is stripped along with
diffuse gas reservoir
– This breaks cycle of gas ejection
and recapture
– Star formation declines
– Sufficiently massive galaxies
spiral to center and merge
• What’s wrong with this picture?
• How does this stack up to the
data?
Strangulation = suffocation
= starvation
Hot gas
reservoir
SNe winds
quickly
exhaust
disk gas
Data rich subject
• Stellar mass
dominates the
properties of
galaxies
•
AGN become
effective
M*
Mhalo
0.5
Galaxy
Mergers
Or is it “velocity
dispersion”?
• Halo mass
determines fraction
of galaxies that are
passive
• But star forming
satellites look just
like centrals
Mhalo 2
Supernova
feedback
Mhalo
Open questions
• Why do AGN become effective at a
particular halo mass scale?
• Why are supernovae more effective in
low mass haloes?
Data rich subject
• Stellar mass
dominates the
properties of
galaxies
•
Or is it “velocity
dispersion”?
• Halo mass
determines fraction
of galaxies that are
passive
• But star forming
satellites look just
like centrals
AGN become
effective
M*
Mhalo
0.5
Lots of scatter
– correlated
with
formation
history?
Mhalo 2
Galaxy
Mergers
Supernova
feedback
Mhalo
Open questions (?!):
• Why do AGN become effective at a
particular halo mass scale?
• Why are supernovae more effective in
low mass haloes?
How to shape the mass function….
….including recapture in semi-analytic models (Bower et al 2012)
Varying wind mass
loading
Varying wind speed
…allows us to understand how to shape
the mass function…. But this doesn’t help
match evolution of the mass function any
better!
Data rich subject
• Stellar mass
dominates the
properties of
galaxies
•
Or is it “velocity
dispersion”?
• Halo mass
determines fraction
of galaxies that are
passive
• But star forming
satellites look just
like centrals
Satellites of massive
haloes have been
satellites in smaller
haloes
Open Question:
does the IG(r)M
also play a role –
denser in larger
haloes? (group haloes
are well below the cosmic
baryon fraction)
The need for a holistic approach
•
Blowing the gas out of the halo (NEW)
– Don’t just blow the ISM out of the
galaxy…
• Blow the diffuse gas out of
the halo
– Needs a lot of energy!
• But there is lots.
– Is the comparison to
group binding energy
just a coincidence?
– For most semi-analytic models the
lack of observable halo gas is a
problem
• For this model it is a
prediction
• Comparison to the observed
gas fractions and MW
luminosities
Bower et al 2008
SAM
McCarthy et al. (2010), OWLS hydro-sim
Data from
M. Sun et al.
(2009)
Data rich subject
• Stellar mass
dominates the
properties of
galaxies
•
Modify semi-analytic model to
allow satellites haloes to keep
their own halo gas (and thus allow
them to continue star formation)
until “strangulation-pressure” is
exceeded
Or is it “velocity
dispersion”?
• Halo mass
determines fraction
of galaxies that are
passive
• But star forming
satellites look just
like centrals
Font et al 2008
Passive/star forming fractions
are very encouraging… but…
But why are centrals and satellites so
similar?
Weinmann et al. (2010)
• Stellar mass
dominates the
properties of
galaxies
•
Or is it “velocity
dispersion”?
• Halo mass
determines fraction
of galaxies that are
passive
• But star forming
satellites look just
like centrals
In the models. passive/star forming fractions
are very encouraging… but… colours of
satellites are distinct from centrals unless
truncation is abrupt
• …but if truncation is abrupt why is passive
fraction low??
See Sean
• Need a delay, then sudden decline??
McGee’s
• Open Question: … its even worse… just
talk
cutting halo growth makes satellite
colours redder than centrals!
What has happened to the
Butcher-Oemler effect?
•
•
? ?
Haynes et al 2008
log(Mstar) ~ 10.5
ICBS, zCOSMOS, GEEC/GEEC2, GCLASS
Woo et al. (2013)
•
•
MB: Quiescent fractions are high
out to z=1
SF galaxies are much more SF – in
both environments!
Halo assembly is much more rapid
Does this make sense? Do models
have predictive power?
Field
Massive clusters (>1014)
Large Groups (0.5-1 x 1014)
Small groups (<5x1013)
Two more issues
Pasquali et
al. 2010
• Metal abundance of
satellites vs centrals
– Why is the offset so small
if satellites are closed
boxes?
• Satellite conformity
– Why are satellites of
moderate mass passive
galaxies more likely to be
passive?
Kauffmann
et al 2013
Open Questions
•
Why are there so many blue satellites?
– Does this mean a long delay time?
– Or is this the result of IG(r)M ?
•
Why is the environment independent of galaxy mass?
– Is it just that processes balance out, or is there more to it?
•
Why are star forming satellites so like star forming centrals?
– Does this mean a short and sudden death?
•
We have assumed that “quenching” is a one-way trip…
– Is “refueling” a possibility?
•
What creates “conformity”
– Is it connected to the spread in halo mass at given stellar mass?
– Galaxies being blasted by AGN feedback in the central?
An advert for the EAGLE project
– “stop screwing around with semi-analytic models
and do it properly. We just need a hydrodynamic
simulation that matches the present-day galaxy
mass function”
• Ian McCarthy, 2008
EAGLE
(Evolution and Assembly of GaLaxies and their
Environments)
constructors: Booth, RGB, Crain, Dalla Vecchia,
Frenk, Furlong, Jenkins, Rosas-Guevara,
Schaye, Theuns, McCarthy, Schaller, Springel,
+++
A hierarchy of galaxy formation
simulations
ISM simulation
Sub-grid
physics
Galaxy Population
EAGLE
(Hydro-simulation)
emulation
large scale structure
N-body + GALFORM
(semi-analytic)
Blackhole accretion
1 pc
100 pc
100 Mpc
The big uncertainties in galaxy formation simulations
• Star formation
• The ISM equations of state
• Supernova driven winds
• Black hole accretion and feedback
100 Gpc
Eagle’s Heritage
OWLS (schaye et al 2010)
• Suite of 50+ simulations
• Varying physics
• Gas particle resolution
~108MO (to z=0)
~106MO (to z=2.5)
• Boxes of 25 –
100h-1Mpc
GIMIC (Crain et al 2010)
• 4 simulations to redshift 0
• One physics model
• High resolution ~106MO (resolves
the Jeans instability in galactic disks)
• Radius of ~20h-1Mpc
But Gimic did
not include AGN
feedback
Sub-grid physics in EAGLE
•
Star formation law from Dalla
Vecchia & Schaye 2008
•
•
•
Implicitly uses Kennicut-Schmidt
relation
Metalicity dependent star
formation threshold
Effective EOS treatment of the ISM
Material heated by
feedback
Cooling accretion
flows
• Thermal feedback in EAGLE
• No decoupling!!!
• Stochastically heat particles to
target temperature (Creasey et al
ISM: imposed
equation of sta
2010; Dalla Vecchia & Schaye 2012)
• Heating to fixed T ensures
particles are hot enough not to
cool immediately
• Probability of heating ensures
that correct energy is input
Primoridal gas
Eagle Ecology
• A key goal of the Eagle
project has been to
match the stellar mass
function
• This provides a strong
basis for exploring the
effect of environment
• Look at how complex
it is…
– Filaments penetrate
halo “pursuing”
satellites
– Widespread impact
of AGN feedback
Movie shows formation
of 1013 Mo group
Part II
The small satellites of “normal”
galaxies
Why so different?
• Rapidly increasing efficiency
of star formation in the
central galaxy.
– Much larger gap between
central and satellites
– Efficiency trend required to
match mass function
Picture from Quo?
(satellites of other
galaxies?)
• Do the same physics apply?
– Extrapolation of
cluster/group trends to
lower halo mass
– Much weaker gas halo?
Only 3% of MW mass galaxies have
SMC/LMC companions Robotham et al 212
“too big to fail”?
• Dwarf satellites may
yet prove a problem
for L-CDM
– Predicted central
velosity dispersions
are too large
– Is this an indication
of residual thermal
velocities in the
DM?
– Or a consequence of
baryonic feedback
Summary – open questions
• Accounting for the properties of satellites
– Broad picture exists suggesting:
• Trends with halo mass
• Trends with redshift
• …but why do satellites look so like centrals?
– important to match the “details”
• Why the abrupt decline?
• Why the separability?
• MW satellites
– Continuation of group/cluster trends?
– Or something completely different?