Missing Photons that Count:

1
Missing Photons that Count Galaxy Evolution via
Absorbing Gas
(and a little bit of fundamental physics to boot)
Chris Churchill
(Penn State)
2
Quasars physics laboratories in the early
universe
quasar
To Earth
Lyb
Lya
CIV
SiIV
CII
SiII
SiII
Lyman limit
Lyaem
Lybem
NVem
Lya forest
CIVem
SiIVem
3
Categories by Neutral Hydrogen

• Damped Lyman-a Absorbers (DLAs)
  N(HI) gt 2x1020 cm-2

Compact Star forming objects Galaxy centers Metal
lines, low ionization dominate

• Lyman Limit Systems (LLSs)
  N(HI) gt 2x1017 cm-2

Proto-galaxy structures Galaxy outskirts
(extended halos, disks) Metal lines, low,
intermediate, and high ionization

• Lyman-a Forest
  N(HI) lt 6x1016 cm-2

Cosmic Web Sheets and Filaments Metal lines, weak
to non-existent
4
Categories by Metal Lines in the Optical
Historical - Optical
C IV systems 1.8ltzlt4.9
Mg II C IV
Mg II systems 0.3ltzlt2.2
Mg II associated with LLS C IV associated with
sub-LLS
5
The Lyman Alpha Forest

• Piercing the Cosmic Web
• Tracing Structure Growth
• Constraining Ionization Evolution

N(HI) lt 1016 cm-2
Lya forest
6
Great Insights are gained from simulations of
structure growth, but these simulations are
starved for hard data to constrain the physics
(courtesy M. Haehnelt)
Note structure growth is rapid at for zgt5 (a
short cosmological time frame), and then
evolution is slower, especially from zlt1
(majority of time)
7
The Power of Simply Counting Lines
The redshift path density, dN/dz, places
constraints on simulations of structure growth as
a function of redshift
(Dave etal 1999)
8
The Power of Simply Counting Lines
The redshift path density, dN/dz, places
constraints on simulations of structure growth as
a function of redshift
(Dave etal 1999)
9
The Power of Simply Counting Lines
The redshift path density, dN/dz, places
constraints on simulations of structure growth as
a function of redshift
(Dave etal 1999)
10
The Power of Simply Counting Lines
The redshift path density, dN/dz, places
constraints on simulations of structure growth as
a function of redshift
or
(Dave etal 1999)
11
The Power of Simply Counting Lines
(Weymann etal 1999)
(Dave etal 1999)
12
C IV Systems

• Proto-galactic clumps
• Tracing Pre-galactic Structure Growth
• Constraining Kinematic/Dynamic Evolution

N(HI) 2 x 1017 cm-2
Metal Lines
13
QSO Absorption Lines Anatonomy of a Simulation
(courtesy M. Steinmetz)
Efforts have been made to include ionization
feedback, both in terms of spectral energy
distributions, photon transport, and mechanical
stirring of the gas
14
QSO Absorption Lines Anatonomy of a Simulation
Ly-a
C IV
Velocity
(courtesy M. Steinmetz)
Technology and innovation is quickly outpacing
observational data
15
The Power of Simply Counting Lines
Mg II shows no evolution (co-moving), but nothing
in known above z2.2 Lyman Limit systems (LLS)
show no evolution, measured from continuum
break at 916 A in the rest frame, N(HI)gt1017.3
cm-2 C IV systems evolve rapidly! They increase
with cosmic time until z1.5 and then show no
evolution
Structure, Ionization, or Chemical Evolution?

• Evolution measures product of
• number
• size
• ionization fraction

Is this an increase in number, in ionization
level, or in the chemical abundance of
carbon? We need low ionization data. Mg II.
16
Motivations and Astrophysical Context
Mg II arises in environments ranging over five
decades of N(H I)

• Damped Lyman-a Absorbers (DLAs)
  N(HI) gt 2x1020 cm-2

eg. Biosse etal (1998) Rao Turnshek (2000)
Churchill etal (2000b)

• Lyman Limit Systems (LLSs)
  N(HI) gt 2x1017 cm-2

eg. Steidel Sargent (1992) Churchill etal
(2000a)

• sub-LLSs (low redshift forest!)
  N(HI) lt 6x1016 cm-2

eg. Churchill Le Brun (1998) Churchill etal
(1999) Rigby etal (2001)
Mg II a-process ion Type II SNe
enrichment from first stars (lt1 Myr)
Fe II iron-group ion Type Ia SNe late
stellar evolution (gtfew Gyr)
Mg II selection probes a wide range of
astrophysical sites where star formation has
enriched gas these sites can be traced from
redshift 0 to 5
17
Present Day Coverage and Astrophysical Context
2796
2803
18
Simple Kinematic Models of Absorbing Gas from
Galaxies
(Charlton Churchill 1998)
Absorption kinematics is symmetric about the
galaxys systemic velocity
Absorption kinematics is offset in the direction
of stellar rotation compared to the galaxys
systemic velocity
Halo/infall Rotating/disk produces both
signatures in single profile
19
Mg II 2796 Absorption Profiles from HIRES/Keck
(Churchill 2001)
Galaxy redshifts can be matched to the absorbers
20
Mg II 2796 Absorption Profiles from HIRES/Keck
(Churchill Vogt 2001)
21
Mg II 2796 Absorption Profiles from HIRES/Keck
Each Mg II system has several Fe II transitions
and Mg I (neutral)
The clouds are modeled using Voigt profile
decomposition
Obtain number of clouds, temperatures, column
densities, ionization conditions (from modeling)
22
Build the Database and the Simulations will Follow
Ultimately, the simulations need to be driven by
the data as we have seen the great successes in
this arena for the Lya forest to z5, and are
seeing the new successes for metal enriched
diffuse objects to z5.
(courtesy M. Haehnelt)
We will begin to see the successes of galaxy
evolution in more detail, including structure
evolution, kinematics, metallicity, and
ionization. The data are lacking. Wholesale
inventory of Mg II absorbers is the best approach.
23
Q0827243
Q1038064
Q1148387
(Steidel etal 2002)
24
Kinematics Stellar, Mg II 2796, and C IV 1548,
1551
Mg II traces stellar kinematics yet is difficult
to explain as extended disk rotation (at 72 kpc
impact parameter!). C IV traces Mg II
kinematics but has strongest component at
galaxys systemic velocity, as highlighted in
l1551. What physical entity is giving rise to
this C IV component?
(Churchill 2003 Churchill etal, in prep)
25
Equivalent Width Distribution
Using HIRES/Keck, we discovered that the EW
distribution followed a power law, with no
observable cut off down to W0.02 A. - these are
high metallicity forest clouds. 5 papers over
10 years predicted that none of these weak
systems existed! They outnumber galaxies by
1106.
Differential Number Density Distribution
As the lower EW cutoff of the sample, Wmin, is
increased, the number of systems per unit
redshift decreases (As Wmin increases, the mean
redshift increases ) differential redshift
evolution
Redshift Path Density
Comoving redshift path density is consistent with
no structure/ionization evolution for Wmin0.02 A
(red) and Wmin0.3 A (blue).
dN/dz ns(1z)g .
26
Evolution of Strongest Systems
Scenario of kinematic evolution of gas
As Wmin increased evolution is stronger
R E D S H I F T
dN/dz N0(1z)g
What is the nature of the evolution??? Is it
related to high velocity clouds, presence of
supperbubbles, or superwinds???
27
Present Day Coverage and Astrophysical Context
The epochs of greatest evolution are un-probed
(Based upon Pei etal 1999)
28
Constraints on Global Galaxy Evolution Models
No coverage for Mg II for zgt2.2 No high
resolution coverage for Mg II for zgt1.4
W(stars)
W(gas)
W(IGM metals)
Mg II provides metalicity for high-z forest in
lower ionization gas- heretofore un-probed
W(baryons)
W(gas flow)
(Pei etal 1999)
29
Population of Weak Systems Where do they arise?

• Their equivalent width distribution follows a
  power law down to 0.02 A
• Arise in optically thin H I (Lya clouds)
• 25-100 of all Lya forest clouds with
  column densities 1015.5ltN(HI)lt1016.5 cm-2 at
  0.4ltzlt1.4
• almost all have zgt0.1 solar metallicity
• Many are iron rich, suggesting later stages of
  star formation
• 90 cannot be associated with galaxies (within 70
  kpc)

(Churchill etal 1999 Rigby etal 2002)
30
Population of Weak Systems Where do they arise?

• 1990-1992 Yanny York used narrow band imaging
  to find OII emission at Mg II absorber redshifts
• They found several emission line objects within
  200-300 kpc of the QSOs substantialy further out
  than the big, normal galaxy picture
• The technique is prime for searching wide fields
  for OII emission in the weak systems do they
  exhibit indicators of star formation? Either
  way, what are the astrophyiscal implications?

• Fabry-Perot at APO is perfect for this job.
• weak systems
• revisit strong systems

31
Some Future Plans

• High Resolution optical spectra of QSOs to get
  Mg II kinematics to cover 1.4ltzlt2.2
• High Resolution HST ultraviolet spectra of
  higher ionization gas
• Low Resolution infrared spectra of QSOs to get
  Mg II statistics for 2.2ltzlt4.0
• Moderated Resolution HST ultraviolet spectra of
  higher ionization gas
• High Resolution infrared spectra of QSOs to get
  Mg II kinematics for 2.2ltzlt4.0

• Leading international collaboration Keck,
  Subaru, VLT, HET, LBT
• Student opportunities include observing, echelle
  data reduction, data analysis
• --- VP decomposition, statistics, distribution
  function (DF) evolution

• Collaborating with N. Kobayashi (Subaru/IRCS),
  future VLT
• Student opportunities include observing, UV and
  IR data reduction, data analysis
• --- visibility function, sample completeness,
  statistics, DF evolution

• This is 5-10 years futureVLT, LBT

32
And now for something completely different.
33
Evidence For Cosmological Evolution of the Fine
Structure Constant?
a e2/hc
Da (az-a0)/a0
34
Procedure
1. Compare heavy (Z30) and light (Zlt10) atoms,
OR 2. Compare s p and d p
transitions in heavy atoms. Shifts can be of
opposite sign. Illustrative formula
Ez0 is the laboratory frequency. 2nd term is
non-zero only if a has changed. q is derived
from relativistic many-body calculations.
Relativistic shift of the central line in the
multiplet
K is the spin-orbit splitting parameter.
Numerical examples Z26 (s p) FeII
2383A w0 38458.987(2) 1449x Z12 (s p)
MgII 2796A w0 35669.298(2) 120x Z24 (d
p) CrII 2066A w0 48398.666(2) - 1267x
where x (az/a0)2 - 1
MgII anchor
35
High-z
Low-z
36
Uncorrected Quoted Results