The Dark Side of

doutfanaticalΜηχανική

14 Νοε 2013 (πριν από 3 χρόνια και 11 μήνες)

64 εμφανίσεις

The Dark Side


of

Galaxy Formation

Ariyeh Maller

Jason Prochaska

Rachel Somerville

Joel Primack

MNRAS 2003, XXX, XXX

MNRAS 2001, 326, 1475

The Light Side

Maller, McIntosh, Katz and Weinberg 2003

The Dark Side

Outline

Absorbtion Systems and what they tell us about galaxy
formation

Galaxy Formation Models

Damped Lyman alpha (DLA) systems N > 2 x 10
20

cm
-
2



Low
-
ion kinematics

High
-
ion kinematics

Future observations of absorbtion systems and what we
may learn about galaxy formation

Sub
-
DLA systems N > 10
19

cm
-
2


Lyman Limit (LL) Systems N > 2 x 10
17

cm
-
2



Galaxy Formation Models

I.

Semi
-

analytic methods (SAMs)


I.

Strengths

I.

Computationally fast (Monte
-

Carlo)

II.

No limit to resolution

III.

Conceptually simple

II.
Weaknesses

I.

Not from first principals

II.

Simplifing assumptions

III.
What you put in is what you get out.

Ingrediants

Merger Tree

Extended Press Scheter

Cooling Recipe

All gas in the cooling radius becomes cold gas

Star Formation Recipe

Star formation proportional to total amount


of cold gas.

Feedback Recipe

Feedback proportional to star formation


rate and more efficient in smaller halos

d
m
star
d
t
=
m
col d
É
star
Merging Tree

Cooling
Radius

Described in:
Somerville and Primack

1999


Somerville, Primack and Faber
2001

Star formation History


Tully
-
Fisher Relation

z =3 Luminosity Function

z = 0 B band Luminosity Function

Damped Lyman
a

Systems



log(N
HI
)
≥>
20.3 cm
2


Column density N
HI


can be dertermined

from the damping

wings of the profile

These column densities

are compareable to what

is seen in local spiral

galaxies.

Low ion Kinematics

Different ions trace one
another well.

There is a large spread in
velocity width 30
-
300 km/s.

The profiles are asymmetric



One explination is massive
rotating disks, but this is in
conflict with CDM structure
formation.



Another is that the kinematics
result from multiple component
systems.

Haenelt et al

1998

Prochaska and Wolfe

1997,1998

Standard models

In standar models of disk
formation the size of gassous
disks are determined by
conservation of angular
momentum.

However, this leads to disks
that are too small to create
DLA systems by multiple
components.

Models that work

For a multiple component

model to explain the DLA

kinematics, the cold

neutral gas must be very

extended.

The model reproduces the

kinematic properties of the

low ions.


This occurs because many

lines of sight interesect

multiple gas disks.

Number density

f
N
=
?
n
?
N
?
X
=
n
N
,
N
ƒ
-
N
-
N
-
X
Only the last 3 points

are measurements because

column densities can

only be determined

for damped systems.


The first 3 points are

just the extrapolation

of the yellow to the

total number of Lyman

limit systems.

Z = 3

Dependance on halo mass

Haehnelt et al 2000 find

a slope of 2.5 is needed to

produce the observed low
-

ion kinematics which are

model also produces.


Recent results from

hydrodynamical simulations

find a similar value.

Nagamine et al
2003

High ion Kinematics

The high ions sometimes but
not always follow one another.

The high ion profiles
do not

follow the low ion profiles

The high ion profiles are
always broader then the low
ion profiles.

There is correlation between
the two but also differences.

The thick disk model can not explain the

high
-
ion kinematics!

Cold gas disk

Hot gas halo

Agreement between the

CIV kinematics and their

relationship to the low

ion kinematics with our

model.


The results follow

because the hot gas is

spherical and therefore

some lines of sight pass

through hot but not cold

gas.


On the other hand all

sight lines though cold

gas also pass through

hot gas.


Cross correlation of
high
-
and
low
-
ion profiles

Models a
-
d agree

with the data for

velocity differences

greater than 50 km/s

but deviate below this.


If we add a CIV

component that

rotates with the cold

gas the agreement

improves dramatically.

The distribtuion of CIV column densities in the data and

in the model. The hot gas metalicity is set to be

-

1.5 dex

Galaxy formation and DLA systems in
SAMs

We assume that the low ion kinematics arise because
each halo contains many sub halos that have been
accreted in its merger history. Thus each halo contains
many galaxies, like the LMC in our galaxies halo.

We associate the cold gas in the sams with the low ions
and the hot gas in the sams with the high ions.

The sams give us the amount of cold gas but not its
spatial distribution. We find that the cold gas must have
an extended distribution.

We take the hot gas to be distributed like the dark matter
and clumped into sub halos. Much of it must cool into
warm clouds that produce CIV absorbtion.

N
R
=
N
t
R
t
R
?
1
R

R
t
Sub
-
DLA

Any multi
-
component

model of the DLA systems

will produce lower column

density systems with

simplier kinematics
.

Low ion

High ion

Generally refered to as

Systems with N > 10
19

cm
-
2

A more physical description

is systems that are cold,

mostly neutral gas. In our

models therefore

N > 4 x 10
19
cm
-
2

The hot gas can

only account for

some of the

lower column

density Lyman

limit systems.

Lyman Limit Systems

Mini
-
halos

It has been suggested that

mini
-
halos (V < 35 km/s)

may be the source of LL

systems at z=3.


Abel & Mo

1998


However realistic distributions

of the gas in these systems do

not have enough cross section

to be important sources of LL

systems.


Therefore our models would

suggest most LL systems

arise outside of virialized halos.

Conclusions

The properties of damped Lyman a absorbtion systems
can be reproduces in galaxy formation models

To do this requires:

The cold gas is very extended


The hot gas is clumped like the dark matter

Some of the hot gas is cooled into warm clouds

These are not the assumptions used in todays SAMs and
they can be tested against the results of hydrodyamical
simulations.

Conclusions

The kinematics of sub
-
DLA systems can test the validitiy
of the multi
-
compontent model.

Mini
-
halos do not produce many LL systems

The majority of LL systems (at z=3) may come from gas
outside of virialized halos.

Future Work

Optical Counterparts and Star formation in DLA
systems.

Compare star formation rate from the SAM to that
measured by
Wolfe et al

2003

Determine how DLA systems cluster in comparisson
with Lyman Break Galaxies.

Determine the optical properties of the galaxies that
host the DLA system.

Thanks

Metalicity

Metalicities range from
-

0.3


to
-

2.5 dex

Average is about
-

1.5 dex

No clear trend with


redshift

F(N) in the model

Metalicity vrs. N in the model