Download The first billion years of galaxy formation and evolution

Galaxy Formation and Evolution from the Epoch
of Reionization to z=4
Thomas R. Greve
Max-Planck Institute for Astronomy
Purple Mountain Observatory, Nanjing, April 3rd 2009
Outline of this talk
1) Cosmic history: the Universe beyond z > 4
- Identifying outstanding problems in galaxy formation and evolution and
key science drivers for the next decade
2) How do we find galaxies at z > 4?
- Dust obscured star formation at z > 4
- All-sky optical/near-IR surveys: hunting for z > 4 QSOs
- Pristine galaxies at z > 4: Lyman-α Emitters
3) Understanding the interstellar medium in z > 4 galaxies?
- How interstellar medium studies can help solve the key problems in galaxy
formation and evolution
4) Summary
1) Cosmic history: the Universe beyond z > 4
Identifying outstanding problems in galaxy formation and evolution
and key science drivers for the next decade
Cosmic History
Galaxy (z=6.4)
Old Stars
Young star/ionized gas
Molecular gas
z=?
The new
cosmic
frontier
z=4
Galaxy (z=2.5)
Galaxy (z=0)
The new cosmic frontier: the epoch of reionization
Key questions to be addressed in the
coming decade:
-When did the EoR start?
-How and when did the first
galaxies form?
z=?
The new
cosmic
frontier
z=4
-How and when did the first
supermassive black holes form?
-What was the inter-relationship
between supermassive black hole
and host galaxy growth at these
early epochs?
This requires large, robust samples of z >
4 galaxies!
2) How do we best observe the first galaxies at z > 4?
Dust obscured star formation at z > 4
The dust-obscured Universe
HDF-N
Hughes et al. (1998)
OB stars
UV
IR/FIR
dust
Far-IR
luminosity
(Obscured) star
formation rate
850μm
LIR = 1x1013L
SCUBA
JCMT, Hawaii
The submm probes
the reionization
epoch!
The submm Universe
~1 sq. degree of sky has been surveyed at submm
wavelengths to date resulting in the detection of
more than ~400 bright SMGs (>3mJy)
~20-30% of the (sub)mm background has been
resolved by blank-field surveys. ~80% by galaxy
cluster surveys but poor number statistics
Submm/FIR
Optical/UV
Submillimetre/Millimetre Surveys
Greve et al. (2008)
Borys et al. (2005)
HDF-N
Hughes et al. (1998)
Submm surveys suffer from poor
resolution (FWHM=11-15”)
Radio inteferometry, however,
offers <1” resolution
Optical spectroscopy
of 90 radio-ID
submm galaxies
The radio-FIR correlation
?
Condon (1992)
Chapman et al. (2005)
A significant population of z > 4 SMGs?
?
Model prediction of the volume
density of SCUBA galaxies
A significant population of z > 4 SMGs?
Discovery: a z=4.76 submm-selected source not associated with a QSO
870μm APEX/LABOCA Survey
Extended Chandra Deep Field South
Weiss et al. (2009)
SMMJ033229.5 (z=4.76 from optical spectrum)
Coppin et al. (2009)
A significant population of z > 4 SMGs?
z=4
Student project!
A multi-wavelength ‘hunt’ for
submm-selected galaxies at z > 4
Quantify their abundance and
intrinsic properties
Model prediction of the volume
density of SCUBA galaxies
The next submillimetre revolution
SCUBA-2 (first light 2009)
SCUBA-2 will deliver thousands of submm-selected sources
Sub-arcsecond submm/mm interferometry with ALMA:
- immediate identification (no need for radio identification)
A census of the z > 4 submm population
ALMA (first light 2012)
2) How do we best observe the first galaxies at z > 4?
✔Dust obscured star formation at z > 4
All-sky optical/near-IR surveys: hunting for z > 4 QSOs
Pristine galaxies at z > 4: Lyman-α Emitters
All-sky optical/near-IR surveys: hunting for z>4 QSOs
All-sky surveys such as the SLOAN have found
numerous, extremely luminous z > 4 QSOs by
means of drop-out techniques in the optical
They represent massive, extremely rare,
overdensities in the primordial density
distribution.
z=4
Gunn-Peterson trough
Becker et al. (2006)
All-sky optical/near-IR surveys: hunting for z>4 QSOs
The extreme luminosities of z > 4 QSOs make them ideal laboratories to study galaxy
formation and black hole growth at the high-mass-end
Submm/mm photometry:
1/3 of optically selected QSOs are IR hyper-luminous (LIR ≥ 1013L)
The most distant QSO known
SDSSJ1148+5152 (z=6.42)
Bertoldi et al. (2003)
mm-emission/near-IR
1 arcmin
5 arcsec
Walter et al. (2003)
Wang et al. (2007)
CO (3-2)
All-sky optical/near-IR surveys: hunting for z>4 QSOs
The extreme luminosities of z > 4 QSOs make them ideal laboratories to study galaxy
formation and black hole growth at the high-mass-end
Submm/mm photometry:
1/3 of optically selected QSOs are IR hyper-luminous (LIR ≥ 1013L)
The most distant QSO known
SDSSJ1148+5152 (z=6.42)
Bertoldi et al. (2003)
mm-emission/near-IR
1 arcmin
Wang et al. (2007)
Extreme galaxy in place <1Gyr after the Big Bang!
LFIR ≈ 1013L
Mgas ≈ 7 x 1010M
Mdust ≈ 109M
Future large samples of distant QSOs
Full UKIDSS Large Area Survey (4000 deg2, Y<19):
# 8.0 > z > 5.8 QSOs: 17
Full Pan-STARRS Survey (10,000 deg2, Y<20.5):
# 8.0 > z > 5.8 QSOs: 73
These samples of QSOs will be prime targets for multi-line
molecular/atomic follow-up observations!
1) How do we best observe the first galaxies at z > 4?
✔Dust obscured star formation at z > 4
✔All-sky optical/near-IR surveys: hunting for z > 4 QSOs
Pristine galaxies at z > 4: Lyman-α Emitters
Pristine galaxies at z > 4: Lyman-α Emitters
In the absence of dust and strong optical
continuum, the easiest way to find the first
galaxies is via the Lyα recombination line: the
strongest emission line produced by the
hydrogen atom (Partridge & Peebles 1967)
i’
z’
NB
z=6.541
z=4
i’
z’
NB
z=6.578
Kodaira et al. (2003)
Pristine galaxies at z > 4: Lyman-α Emitters
Lyman-α Emitters (LAEs) are likely to be pure starbursts – and
representing the first building-blocks of galaxies
The large number of z > 6 LAEs (30 per
0.25 sq. deg) implies that they could
play a dominant role in reionizing the
Universe
Low stellar masses (<109M) and
star formation rates (<30M/yr).
No dust (very metal-poor)
Small linear scales (<1kpc)
Gawiser et al. (2007)
Future samples of distant Lyman-α emitters
There are currently several hundreds known LAEs at z > 4
JWST+ELT will be able to detect the smallest and most distant galaxies (z > 7),
increasing the number of LAEs by order of magnitude
Extremely Large Telescope
30m optical/near-IR ground-based telescope
James Webb space telescope
6.5m optical/near-IR/mid-IR telescope in space
1) How do we best observe the first galaxies at z > 4?
✔Dust obscured star formation at z > 4
✔All-sky optical/near-IR surveys: hunting for z > 4 QSOs
✔Pristine galaxies at z > 4: Lyman-α Emitters
What is the most effective way of studying these first galaxies
in order to maximize constraints on formation and evolution
models?
The role of gas in galaxy formation and evolution
The gravitational hierarchical build-up of dark matter structures provides the framework
for galaxy formation and evolution
The interstellar medium (gas and dust) is a key ingredient in galaxy formation and
evolution as it provides the ‘fuel’ for star formation and supermassive black hole accretion
…so understanding the physical properties of the interstellar medium (ISM) in distant
galaxies is fundamental to our picture of galaxy formation and evolution
Galaxy
Dark matter
z = 6.4 (t = 0.9 Gyr)
Springel et al. (2006), Nature
Galaxy
Dark matter
z = 2.5 (t = 4.0 Gyr)
Galaxy
Dark matter
z = 0 (t = 13.6 Gyr)
Observing the interstellar medium
C
O
J=2-1 (ν = 230GHz)
...
CO 1-0
2-1 3-2
… 5-4
Temperature
J=1-0 (ν = 115GHz)
Density
Molecular hydrogen (H2) is by far the main component of
the ISM – but its lack of a permanent dipole moment
makes it virtually impossible to observe directly
Instead the rotational lines of CO are mainly used to
study the ISM
Other important molecular gas tracers: HCN and HCO+
Atomic fine-structure lines: [CI] and [CII] (ν = 490-1900GHz)
Atmospheric transmission vs. frequency
The CO J=1-0 line from a local galaxy falls within
the 3mm atmospheric window,
…as does the (redshifted) CO J=5-4 line from a
galaxy at z=4 (νobs = 575GHz/(1+z) = 115GHz)
Observational status
This excitation-bias prevents a meaningful comparison between the molecular gas
properties of local and high-z galaxies
High-J CO lines
Dense, warm gas
CO 3-2 in SDSSJ1148+5152 (z=6.42)
Walter et al. (2003)
Low-J CO lines
Diffuse gas
z>4
Greve (2009)
The highest CO detection to date
Universe was 1/16 of its current age
A new golden era in ISM astronomy
The next five years will see a quantum leap in our ability to study the ISM in galaxies
across the Cosmos - one that will take us from an epoch of merely detecting
molecular lines at high-z to multi-line surveys capable of fully characterizing the ISM
Herschel, launch
2009
ALMA, first light 2012
EVLA, first light 2012
z=0
A new golden era in ISM astronomy
A full understanding of galaxy formation and evolution at z > 4…
Requires:
An exhaustive inventory of the microscopic properties (e.g. chemistry, density,
thermal balance) of the ISM in z > 4 galaxies, and their effect on macroscopic
properties (e.g. star formation, luminosity, morphology)
Key Questions:
-When did the EoR start?
Method:
- Sampling of the spectral line distributions of CO,
HCN and HCO+, [CII] 158μm and [CI] 369μm
- Spatially and kinematically resolved dust and
molecular line observations
- For large samples of z > 4 objects (QSOs, SMGs,
and LAEs)
-How and when did the first
galaxies form?
-How and when did the first
supermassive black holes form?
-What was the inter-relationship
between supermassive black hole
and host galaxy growth at these
early epochs?
High-z ISM studies at sub-kpc scales
Imaging galaxy formation
Submm galaxy at (z=2.49)
High-resolution observations of the dust
and molecular gas provide a direct image
of the formation morphology, and can
distinguish between several scenarios
i)A major merger between two gas-rich
components (‘wet’ merger)
ii)Many minor bursts distributed within an
extended potential and interspersed with
periods of no star formation
Tacconi et al. (2008)
SDSSJ1148+5152 (z=6.42)
Walter et al. (2003)
iii)A single monolithic collapse
In addition, one obtains accurate
dynamical masses, merger fractions etc.
CO(3-2)
High-z ISM studies at sub-kpc scales
Black hole and galaxy host growth at z > 4
An unusually tight relation between the
mass of the supermassive black hole
The local MBH-Mbulge relation (Magorrian et al. 1997)
and that of its host spheriod has been
established in the local Universe.
This relation connects a phenomenon
ocuring on spatial scales of ≈10-5pc
(black hole accretion) to the spheriod
which is 8 orders of magnitude larger
(≈103pc )
This suggest a deep, co-evolutionary link
between the supermassive black hole
and the galaxy spheriod.
What is the underlying physics?
How does the relation evolve with
redshift?
Mbulge=0.002MBH
scatter < 0.30dex
Häring & Rix (2004)
High-z ISM studies at sub-kpc scales
Black hole and galaxy host growth at z > 4
High-resolution CO studies can uniquely probe the MBH-Mbulge relation at high-z
QSOs
Did the black holes start to
grow first?
SDSSJ1148+5152 (z=6.42)
Local relation
Walter et al. (2003)
Local relation
CO(3-2)
High-z ISM studies at sub-kpc scales
Black hole and galaxy host growth at z > 4
High-resolution CO studies can uniquely probe the MBH-Mbulge relation at high-z
QSOs
Or did the bulges grow first?
High-resolution CO studies of submm galaxies
Student project:
Local relation
SMGs
Spatially resolve (<1” FWHM) the
gas-kinematics in a large sample of
z>4 QSOs and SMGs in order to
study the MBH-Mbulge relation in the
early Universe
CO-detected SMGs (Alexander et al. 2007)
Tacconi et al. (2008)
A new golden era in ISM astronomy
A full understanding of galaxy formation and evolution at z > 4…
Requires:
An exhaustive inventory of the microscopic properties (e.g. chemistry, density,
thermal balance) of the ISM in z > 4 galaxies, and their effect on macroscopic
properties (e.g. star formation, luminosity, morphology)
Key Questions:
-When did the EoR start?
Method:
- Sampling of the spectral line distributions of CO,
HCN and HCO+, [CII] 158μm and [CI] 369μm
- Spatially and kinematically resolved dust and
molecular line observations
- For large samples of z > 4 objects (QSOs, SMGs,
and LAEs)
-How and when did the first
galaxies form?
-How and when did the first
supermassive black holes form?
-What was the inter-relationship
between supermassive black hole
and host galaxy growth at these
early epochs?
The ISM conditions at z > 4: the density structure of the gas
The dense gas fraction of the ISM in a galaxy may govern its star formation
efficiency and hence its evolutionary path.
Determining the density structure of the ISM requires a very complete
sampling of the CO rotational ladder
Weiss et al. (2006)
Is the ISM in QSOs more
excited than in submmselected galaxies?
APM0827 (z=3.9) Weiss et al. (2006)
CO(4-3)
CO(6-5)
CO(10-9)
CO(11-10)
CO(9-8)
The ISM conditions at z > 4: the density structure of the gas
Determining the density structure of the ISM requires a very complete
sampling of the CO rotational ladder. As well as of dense gas tracers such
as HCN and HCO+
HCO+(1-0) in the Cloverleaf (z=2.6)
Weiss et al. (2006)
Riechers et al. (2006)
APM0827 (z=3.9) Weiss et al. (2006)
CO(4-3)
CO(6-5)
CO(10-9)
CO(11-10)
CO(9-8)
HCN(5-4)
The ISM conditions at z > 4: gas cooling
The [CII] 158μm line is the main cooling line in our
Galaxy and in typical local starburst (L[CII]/LIR ≈ 5x10-3)
The first fully sampled CO spectrum (up to
J=6-5) of a local IR-luminous galaxy
(Papadopoulos et al. 2007)
However, [CII] cools 10x less efficiently in the most
IR-luminous galaxies (at low- and high-z)
Normal local galaxies
Local ultra IRluminous
galaxies
High-z
SDSSJ1148+5152 (z=6.42)
[CII]
CO(6-5)
Hailey-Dunsheath (2008)
Walter et al. (2009)
The ISM conditions at z > 4: gas cooling
The [CII] 158μm line is the main cooling line in our Galaxy
and in typical local starburst (L[CII]/LIR ≈ 5x10-3)
Student project
However, [CII] cools 10x less efficiently in the most
IR-luminous galaxies (at low- and high-z)
In metal-poor systems, however, we
can have L[CII]/LIR ≈ 0.5-1x10-2
Normal local galaxies
Local ultra IRluminous
galaxies
High-z
An z=7 LAE with LIR ≈ 2x1011L
(SFR=30M/yr) will be detectable
with ALMA!
SDSSJ1148+5152 (z=6.42)
[CII]
CO(6-5)
Hailey-Dunsheath (2008)
Maiolino et al. (2005)
Detecting the first objects at z > 7 with ALMA
CO(8-7)
[CII]
The [CII] 158μm line may be the line of
choice for z > 7 objects with ALMA
Weiss et al. (2006)
CO J > 8 no highly excited
[CII]
[CII] is 5x
brighter than
CO(6-5)
CO(6-5)
Maiolino et al. (2005)
Summary
Future surveys with PanSTARRS/UKIDSS, SCUBA-2, and JWST/ELT will
drastically increase sample sizes of z > 4 galaxies
The next 5-10 years will see the advent of a number of new, groundbreaking cm/submm/far-IR facilities (e.g. ALMA, EVLA) allowing us to study
such samples effectively
For the first time it will be possible to do a detailed characterization of the
ISM in primeval galaxies during the epoch of reionization
This will revolutionize our understanding of galaxy formation and evolution
at all cosmic epochs