Transcript margarita2007

Galaxy Formation in a Λ Cold Dark
Matter Universe
Mario G. Abadi
Instituto de Astronomía Teórica y Experimental
Observatorio Astronómico de la Universidad Nacional de Córdoba
Consejo de Investigaciones Científicas y Técnicas, Argentina
In collaboration with
Julio Navarro, Matthias Steinmetz, Vince Eke
Amina Helmi, Laura Sales and Andres Meza
XII Reunion Regional Latinoamericana de Astronomía de la IAU
Isla de Margarita, Venezuela
October 22-26 2007
Milky Way Galaxy
Surviving Satellites
Duncan Forbes 2007
Merging Satellites
•
•
•
•
•
Sagittarius stream: warp around the
entire MW nearly perpendicular to
the midplane (Ibata et al. 2001).
Progenitor=Sagittarius dwarf
Monoceros ring: a proposed ring of
stars around the galactic plane
(Yanny et al 2003 & Ibata et al 2003).
Progenitor=Canis Major dwarf?
Orphan stream: a tidal stream that
extends over 50° in the north
galactic cap (Belokurov et al 2007).
Progenitor=Ursa Major II??
Andromeda stream: a giant stream
of stars uncovered in the halo of
M31 (Ibata et al. 2001)
External galaxy: a dwarf satellite
galaxy in the process of being torn
apart by gravitational tidal forces
(Forbes et al 2003)
Merged Satellites
• Substructure in the
galactic halo (Helmi et al.
1999)
• Arcturus stream (Eggen
1971, Navarro et al. 2004)
• Debris from omega Cen
parent galaxy in the solar
neighborhood (Meza et al.
2005)
• Substructure in the
galactic disk (Helmi et al.
2005)
Orbital Classification of Satellites
Distance
Surviving:
Rgalaxy<Rsatellite<Rvirial
Merged:
0<Rsatellite<Rgalaxy
Escaping
Rvirial<Rsatellite<∞
Time
R-band Simulated Galaxy
40 kpc
Stellar halo and Satellites
600 kpc
Dark Matter Halo
600 kpc
Photometric Properties

I-band surface brightness
Surface Brightness
profile of the simulated galaxy
seen projected face-on

The surface density profile
agrees very well with an
observed Sab galaxy (UGC615)

Also colours are in agreement
with UGC615
Distance
Photometric Properties

I-band surface brightness
Surface Brightness
profile of the simulated galaxy
seen projected face-on

The surface density profile
agrees very well with an
observed Sab galaxy (UGC615)

Also colours are in agreement
with UGC615
Distance
Kinematics Properties
Velocity
• There is a poor agreement
between the spheroiddominated, declining rotation
curve of the simulated galaxy
and the observed galaxy.
Distance
Kinematics Properties
Velocity
• There is a poor agreement
between the spheroiddominated, declining rotation
curve of the simulated galaxy
and the observed galaxy.
Distance
I-band Absolute Magnitude
Tully Fisher Relation
Rotation Speed
• Angular momentum problem:
Simulated disk galaxies only
marginally agree with the
observed TF relation for latetype spirals. They fit rather well
in the relation appropriated for
S0 due to the net angular
momentum transfer from the
gas to the dark matter halo
during the galaxy formation
process.
• Solutions: involve some sort of
feedback
processes
that
regulate the gas supply to the
galaxy. But, Governato et al
ascribed the problem to poor
numerical resolution.
Feedback Solution
The shapes of the CDM halo
mass function and the
luminosity function are very
different (Baugh 2006)
Simulations with no or week feedback
tend to overpredict the cosmic star
formation history substantially
(Springel 2006)
Feedback Solution
Weak
Strong
• Feedback effects regulate the
delicate balance between gas
supply and morphology
• Identical initial conditions with
different feedback recipe and
star formation produce different
galaxies
• Strong (weak) feedback
produce disk (bulge)-like
galaxies
Zavala Okamoto & Frenk 2007
Dynamical Components

Distribution function of
circularity
Counter-rotating stars

Spheroid: no net rotation

Thick disk: slow rotation

Thin disk: fast rotation
Co-rotating stars
Star formation rate (solar masses per year)
Age Distribution

Inefficient feedback leads to
high rates of star formation at
high-redshift.

Subsequent mergers form a
massive spheroid

Galactic disks date from the
time of the last major merger
and form from the inside out
Old
Stellar Age
Young
Mass Fraction
Natives and Immigrants
The spheroid is old. There are no
young stars in this component
The thin disk contains a significant
number of old stars (15% are older
than 10 Gyrs)
Fraction of Natives
More than 90% of this old thin disk
is a result of satellite accretion
events
The thick disk is old
The old thick disk is not a former
thin disk thickened by a minor
merger but actually the debris from
satellite accretion events
Old
Stellar Age
Young
Density Profiles
Number density Profile
• Luminous satellites are
resilient to disruption by
tides and they can survive
as self-bound entities
closer to the primary,
where substructure in
dark-matter only
simulations may quickly
disrupt (White & Rees
1978)
• Stellar halos are much
more concentrated than
dark matter halos
Distance to the centre
Cumulative Profiles

Simulated and observed Milky
Cumulative number fraction
Way and Andromeda satellites
have very similar profiles
Distance to the centre
Cumulative Profiles
Cumulative number fraction
Simulated satellites trace
very well the underlying
dark matter distribution
and are the best dark
matter tracers in the outer
parts of the galaxy
• Milky Way globular
clusters and simulated
stellar halo have very
similar distributions which
are much more
concentrated.
Distance to the centre
Anisotropy Velocity Dispersion
Fraction of Natives
• The simulated stellar halo
consists almost exclusively of
accreted stars
=1-T2/(r2)
• Satellites are only slightly more
radially anisotropic than the
dark matter β~0.4
• The anisotropy of the stellar
halo is much more pronounced
β~0.8
• The fact that the stellar halo is
made of disrupted satellites
suggests that there is an
intrinsic difference between
merged and surviving satellites
Radial
Isotropic
Distance to the centre
Accretion Mass Distribution
Number
Any satellite more
massive than ~10% Mhost
is not able to survive.
Large mass satellites
are very likely to merge
with the host
Light
Satellite Mass
Heavy
Fraction
Accretion Redshift Distribution
Accretion Redshift
• Merged satellites are more
massive and have been
accreted earlier than surviving
one
• Surviving satellites are
predominantly low-mass
systems and have been
accreted recently
• The building blocks of the
stellar halo were on average
more massive and were
accreted (and disrupted) earlier
than de population of satellites
that survive until the present
• These results may help to
explain the difference between
the abundance patterns of halo
star in the solar neighborhood
and in galactic dwarfs
Conclusions
1)Photometric properties of simulated galaxies agree with observations
2)Simulated galaxies marginally agree with Tully Fisher relation, but
stronger feedback alleviate this problem
3)Spheroid by mergers, disk by gas accretion
4)Satellite density profile similar to dark matter halo but much more
extended that the stellar halo
5)Satellite kinematics: also similar to dark matter and MW but different
from stellar halo
6)Satellite Orbits: more massive satellites are accreted earlier and
merge. Less massive satellite are accreted later and survive
Papers








Dynamical and Photometric properties
Fine Structure of galactic disks
Tidal torques
Omega Cen
Stellar halo
Substructure in the galactic disk
Surviving and merging satellite
Satellite on extreme orbits
Bibliografia

2002 Freeman K. and Bland-Hawthorn J.
The New
Galaxy:Signatures of its Formation

2003 Kauffmann G.
The Formation and
Evolution of Galaxies

2006 Baugh C.M.
A prime on
hierarchical Galaxy Formation: the Semi-Analytical Approach

2006 Avila-Reese V.
Understanding Galaxy
Formation and Evolution

2007 Cecil G. and Rose J.A.
Structure and Evolution from the Light of Nearby Systems
Constraints on Galaxy
Escenario Cosmologico
Ωtotal=100$
Ωlambda=76$
ΩmateriaOscura=20$
Ωgas=$3 con 80cts
Ωgalaxias=20cts
Challenges to CDM on Galactic
Scales









Too much dark matter in halo centers?
Halo substructure issues
Halo and galaxy merging history?
Halo occupation statistics ok?
Angular momentum issues
Does CDM correctly predict galaxy
number density (luminosity function)?
morphology, kinematics, and colors?
formation and evolution?
Galaxy Centers



Problem first recognized by Flores and me, and
Moore, but HI errors were underestimated
(Swaters).. Small galaxies are mainly dark matter
so the complications of baryonic physics are
minimized. The only case where Blitz group see
no radial motions is consistent with CDM r-γ with
γ≈1.
The non-circular motions could becaused by
nonspherical halos (Navarro).
Too Many Halos





“missing satellites”? Squelching
(Somerville) and properly identifying halos
hosting satellites
(Gnedin, Klypin, Kravtsov, Zentner) seem to agree
with
data. Enough subhalos at small radii to explain
anomalous
flux ratios in radio gravitational lenses?
Halo Occupation Distribution




HALO OCCUPATION DISTRIBUTION Agree
with
observations (as a function of mass, redshift)?
Predicted inner steep part of correlation function
seen at
high redshift (SUBARU).
Merging History





Are there too many galaxy mergers to account for the
numbers of (classical) bulges seen in disk galaxies
(Kormendy)?
How important are major and minor
mergers in forming stars, spheroids, and AGN? In
accounting for the brightest optical and IR galaxies at z >
2?
Do predicted mergers agree with numbers of peculiar
galaxies and galaxy pairs seen (as a function of galaxy
luminosity, redshift, environment, …)?
Angular Momentum Issues

Catastrophic loss of angular momentum due to overcooling hydrodynamic simulations (Navarro, Steinmetz).

Spiral galaxies would be hard to form if ordinary matter has the same specific angular momentum distribution as dark matter (Bullock).

How do the disk baryons get the right angular momentum?

Mergers give halos angular momentum – too little for halos that host disks, too much for halos that host spheroids?

Role of AGN and other energy inputs? Role of cold inflows (Birnboim, Dekel, Katz, Weinberg …)?

Can simulated disks agree with observed Tully-Fisher relation and Luminosity Function at all redshifts?
Luminosity Function
.
Bimodality
SUMMARY

We now know the cosmic recipe. Most of the universe is invisible stuff called
“nonbaryonic dark matter” (25%) and “dark energy” (70%).

Everything that we can see makes up only about 1/2% of the cosmic density, and invisible
atoms about 4%. The earth and its inhabitants are made of the rarest stuff of all: heavy
elements (0.01%).

The ΛCDM Cold Dark Matter Double Dark theory based on this appears to be able to
account for all the large scale features of the observable universe, including the details of
the heat radiation of the Big Bang and the large scale distribution of galaxies.

Constantly improving data are repeatedly testing this theory. The main ingredients have
been checked several different ways. There exist no convincing disagreements, as far as I
can see. Possible problems may be due to the poorly understood physics of gas, stars,and
massive black holes.
Fundamental Observational
Properties of Galaxies
1) Galaxy contribution to the total density of the Universe: Why star formation is
such an inefficient process?
2) Luminosity Function : Why there is a characteristic mass for galaxies?
3) Bimodality: Why are there 2 distinct populations or a bimodality in properties
such as colour?
4) Spatial Distribution (Correlation Function, Morphology Density Relation):
What role does the environment play in galaxy formation?
5) Scaling Laws in size, luminosity and velocity: Why are there remarkably tight
correlations between certain galaxy properties?
6) High redshift
Satellite Luminosity Function
Satellite Luminosity Function
Density Profile
• Dark matter halo follows an “NFW”
•
•
•
•
•
•
•
density profile ρ~rˉ¹(1+r)ˉ²
Number density profile of satellites,
after rescaling their positions to the
virial radius of each host and stacking
all 8 simulations
There is little difference in the shape
of the dark matter and satellite
profiles.
Half mass radius very similar: R½=0.3
for dark matter particles and R½=0.37
for satellites
Simulated satellites differ from
substructure halos, whose density
profile is known to be significantly
shallower than the dark matter
(Ghigna et al. 1998,2000 Diemand
Moore & Stadel 2004 Gao et al 2004).
Luminous satellites are resilient to
disruption by tides and they can
survive as self-bound entities closer to
the primary, where substructure in
dark-matter only simulations may
quickly disrupt (White & Rees 1978)
The stellar halo is much more centrally
concentrated R½=0.05
Density Profile
• Dark matter halo follows an “NFW”
•
•
•
•
•
•
•
density profile ρ~rˉ¹(1+r)ˉ²
Number density profile of satellites,
after rescaling their positions to the
virial radius of each host and stacking
all 8 simulations
There is little difference in the shape
of the dark matter and satellite
profiles.
Half mass radius very similar: R½=0.3
for dark matter particles and R½=0.37
for satellites
Simulated satellites differ from
substructure halos, whose density
profile is known to be significantly
shallower than the dark matter
(Ghigna et al. 1998,2000 Diemand
Moore & Stadel 2004 Gao et al 2004).
Luminous satellites are resilient to
disruption by tides and they can
survive as self-bound entities closer to
the primary, where substructure in
dark-matter only simulations may
quickly disrupt (White & Rees 1978)
The stellar halo is much more centrally
concentrated R½=0.05
Density Profile
• Dark matter halo follows an “NFW”
•
•
•
•
•
•
•
density profile ρ~rˉ¹(1+r)ˉ²
Number density profile of satellites,
after rescaling their positions to the
virial radius of each host and stacking
all 8 simulations
There is little difference in the shape
of the dark matter and satellite
profiles.
Half mass radius very similar: R½=0.3
for dark matter particles and R½=0.37
for satellites
Simulated satellites differ from
substructure halos, whose density
profile is known to be significantly
shallower than the dark matter
(Ghigna et al. 1998,2000 Diemand
Moore & Stadel 2004 Gao et al 2004).
Luminous satellites are resilient to
disruption by tides and they can
survive as self-bound entities closer to
the primary, where substructure in
dark-matter only simulations may
quickly disrupt (White & Rees 1978)
The stellar halo is much more centrally
concentrated R½=0.05
Satellite Orbit
Stellar Halo
Satellite Velocities
ΛCDM Cosmological Model
Large Scales: Concordance model with parameters fitted to reproduce
many observational results
76% Dark Energy, 20% Dark Matter and 4% Baryons (Spergel et al 2007)
Galactic Scales: Many observational challenges (Baugh 2006)
1) Galaxy contribution to the total density of the Universe
2) Luminosity Function
3) Bimodality
4) Spatial Distribution
5) Scaling Laws
6) Evolution
Galaxy Formation

Initially dark matter and gas are almost uniformly distributed in an
expanding universe

Small density fluctuations are present with a power spectrum given
by the ΛCDM cosmological model in both dark matter and gas
components

Perturbations take part of the universal expansion until they
eventually stop, collapse due to they own gravitational attraction and
relax to form dark matter halos

Gas is dragged by dark matter, shock heated to the halo virial
temperature, acquires angular momentum due tidal torques, cools
due radiative cooling and form stars (White & Rees 1978)
Numerical Simulations
• Initial conditions (positions and velocities) given by the ΛCDM model
• Initially only dark matter and gas particles
• Astrophysics: gravitation, hydrodynamics, radiative cooling, star
formation, feedback (SN or AGNs) and metals
• Gas particles transformed into star particles according to their
density and local divergence
• Properties computed using stellar evolution models
• (Navarro & Steinmetz 2000, Tacker & Couchmann 2001, Sommer
Larsen et al. 2002, Abadi et al. 2003, Governato et al. 2004, Robertson
et al. 2004, Okamoto et al. 2005, Governato et al. 2007, etc.)
Simulated Galaxy System
Dark matter
600kpc
Satellites
600kpc
Stellar halo
600 kpc
Galaxy
40kpc
How Galaxies form?

How Milky Way and other galaxies formed?

Origin of different components: spheroid, thick disk, thin disk, stellar
halo, satellites, dark matter halo, dark matter subhalos

Monolithic collapse (Eggen Lynden-Bell & Sandage 1962) or
successive mergers (Searle & Zinn 1978)

How many progenitors? Properties.

Natives and Immigrants stars: Where do they live now?
Normalized radial velocity
MW Satellite Velocities
Normalized Distance
• Curves delineating the top and
bottom boundaries show the
escape velocity given by NFW
profile
• Open squares correspond to the
11 brightest Milky Way satellite
(van den Bergh 1999)
• Velocities and positions of Milky
Way satellites are consistent
with the simulated satellite
population
• Leo I is located at the virial
radius and moving outward at
the escape velocity similar to
one of our simulated satellites
• Similar results are found for
Andromeda (Sales et al. 2007)
Density Profiles
Number density Profile
• Simulated satellites differ from
substructure halos, whose
density profile is known to be
significantly shallower than the
dark matter
• (Ghigna et al. 1998,2000
Diemand Moore & Stadel 2004
Gao et al 2004).
• Luminous satellites are resilient
to disruption by tides and they
can survive as self-bound
entities closer to the primary,
where substructure in darkmatter only simulations may
quickly disrupt (White & Rees
1978)
• Stellar halos are much more
concentrated than dark matter
halos
Distance to the centre
Eventos de Acreccion en el Disco de
la Via Lactea
Cen
La distribucion de Jz estrellas
pobres en metal en la
vecindad solar sugieren la
presencia de distintos grupos
cinematicos
Meza et al. 2005
Orbital Classification of Satellites
Distance
Surviving:
Rgalaxia<Rsatelite<Rvirial
Merged:
0<Rsatelite<Rgalaxia
Escaping:
Rvirial<Rsatelite
Time
Normalized radial velocity
MW Satellite Velocities
Normalized Distance
• Curves delineating the top and
bottom boundaries show the
escape velocity given by NFW
profile
• Open squares correspond to the
11 brightest Milky Way satellite
(van den Bergh 1999)
• Velocities and positions of Milky
Way satellites are consistent
with the simulated satellite
population
• Leo I is located at the virial
radius and moving outward at
the escape velocity similar to
one of our simulated satellites
• Similar results are found for
Andromeda (Sales et al. 2007)
Escaping Satellites
Distance
• Escaping satellites are accreted
in pairs of unequal mass.
• The heavier satellite follows a
conventional orbit for a
“merging” satellite.
• The light member of the pair is
ejected from the system as a
result of the gravitational
interaction between the pair and
the host during the first
approach
Time