Observing the Assembly of Galaxies Hans-Walter Rix Max-Planck-Institute for Astronomy Heidelberg HWR Princeton, 2005 Overview I.

Download Report

Transcript Observing the Assembly of Galaxies Hans-Walter Rix Max-Planck-Institute for Astronomy Heidelberg HWR Princeton, 2005 Overview I.

Observing the Assembly of Galaxies
Hans-Walter Rix
Max-Planck-Institute for Astronomy
Heidelberg
HWR
Princeton, 2005
Overview
I. The Build-Up of the Stellar Mass in Galaxies
II. The Formation and Evolution of Massive Galaxies
Thursday May 5, 2:00PM
III. The Evolution of (Internal) Galaxy Structure
Wednesday May 11, 2:00PM
IV. Archeo-Cosmology in the Local Group
Friday, May 13, 2:00PM
HWR
Princeton, 2005
I. The Build-Up of Stellar Mass
1. Casting the problem into specific questions
2. Diagnostic Tools
3. A brief survey of surveys
4. Estimating the star-formation rate = f(z)
5. Estimating the stellar mass density = f(z)
6. Results
HWR
Princeton, 2005
1. Re-phrasing “the build-up of stellar mass”
•
What is <SFR(z)> and
<r*(z) >?
• What epoch encloses the formation of most stars?
• How to best measure <SFR (z)r
> and <r*(z) > ?
• How much important are mergers in triggering SF and in
setting the present-day mass function?
• What are the expectations from models?
HWR
Princeton, 2005
2. Diagnostic tools for
star-formation rates and stellar masses
• Star formation rate estimates are
based on UV luminosity produced by
hot, massive, short-lived stars
– Observe the UV
– Observe Ha
– Observe absorbed UV flux, reradiated by dust in thermal IR
! LIR(re-radiated) >> LUV(escaped) !
– Mtot estimate is based on stars >10Mo,
which are small fraction of Mtot
HWR
Princeton, 2005
Kroupa 2002
Starlight and
Re-processed Starlight
Single-age, dust-free stellar population
Devriend et al 2000
HWR
Princeton, 2005
ground
Herschel
(2007)
Spitzer
SED of an ageing stellar
population of solar
metalicity with dust
HWR
Princeton, 2005
f(24mm) vs Lbol
Papovich and Bell 2003
Given that Spitzer
can only observe well
at 24 mm, what are
the bolometric
corrections?
HWR
Princeton, 2005
Mass measurements in cosmologically distant
galaxies
• Dynamics:
– OK to z~1, but quite expensive.
– Very limited spatial resolution conceptually problematic
– Currently not feasible for most galaxies z>1.5
• Clustering:
– Measures halo mass, not stellar mass
• M* = L x (M/L)* with M/L from SEDs
HWR
Princeton, 2005
Stellar Masses from Spectral Energy Distributions
Near-degeneracy of age,
metallicity and dust
Source of despair or
opportunity?
tstars =
[Gyrs]
B
K
Bell and de Jong 2001
Optical/near-IR spectra of galaxies
are a nearly 1D sequence
HWR
Princeton, 2005
• Mapping one or few integrated galaxy colors to
– age
– dust extinction
– metallicity
is poor!
• Mapping (optical -- across 4000A break)
color to M/L should be robust!
HWR
Princeton, 2005
M/L from Colors? Compare to Mdyn!
Van der Wel, Franx, can Dokkum and Rix, 2004
at z~1
HWR
Princeton, 2005
Look-back Galaxy Surveys: Desiderata
•
•
Select SFR surveys by SFR, and mass surveys by stellar mass
–
SFR: assure most of the intense star-burst are not missing due to
dust
–
Stellar mass: select galaxies lobs > (1+z) 4000A break
Number of galaxies as a function of
–
–
Epoch  redshift (few %)
Luminosity/stellar mass
–
Color/stellar age
1,000 – 10,000 galaxies
•
Measure galaxy sizes/internal structure ~0.3” resolution
•
Either Nfield >> 1 or qfield > 2xcorrelation length ~10’
HWR
Princeton, 2005
A Survey Survey
Name
Nfield
Field
size
HST
imaging
# of
bands
Depth
Nredshift
HDFs/UDF
3
2.5’
+
7
R=29
700
GOODS
2
12’
+
10
i=27.5
400
3000 (5%)
FIRES
2
5’
+
10
KAB=26
600
(5%)
COMBO-17
GEMS
3
30’
+
22
R=24
30,000
(1%)
MUNICS
3
30’
-
7
K=19.5
20.000 (5%)
GDDS/LCIRS
2
30’
-
7
H=21.5
500(2000)
SUBARU
RestUV
Steidel et al
RestUV
HWR
Princeton, 2005
HWR
Princeton, 2005
COMBO-17
Wolf, Meisenheimer, Rix et al. 01/03
Heidelberg, Oxford,Potsdam,Edinburgh
•
•
•
3 fields @ 30’x30’
17 filters to mr~23.6
~10.000 redshifts (1.5%)+ SEDs per field
Z
Wavelength
HWR
Princeton, 2005
[nm]
MB
Comparison of COMBO-17 with VIMOS Spectra
(data from Le Fevre et al 2004)
HWR
Princeton, 2005
A quick Tour through Redshift Space
GEMS(CDFS)
HWR
Princeton, 2005
Abell 901
S11 (random)
HWR
Princeton, 2005
Stellar Masses from the COMBO-17 Survey
Borch, Rix, Meisenheimer et al 2005
• Stellar masses to z~1 can
be estimated for 10.000s
of galaxies
• Flux limit (R-band) is
VERY different from
mass limits.
0.65<z<0.75
HWR
Princeton, 2005
FIRES: FaintInfra-Red-Extragalactic-Survey
ultra-deep VLT survey
*HDF-south
100 hours in JHK
FWHM=0.45”
*MS1054:
5xlarger area
25 hours in JHK
per pointing
Franx, Rix, Rudnick, Labbe, van
Dokkum, Foerster-Schreiber,
Trujillo, Moorwood, et al.
2001-2005
Selecting and studying
galaxies z>2 in their
rest-frame optical bands
HWR
Princeton, 2005
Not a Ly-break!!
Just a red SED
HWR
Princeton, 2005
What kind of galaxies are found in such a search?
• Galaxies without many (really) young stars won’t be found by
their Ly-break or their sub-mm dust emission.
• Ditto for galaxies with significant dust extinction that are
not powerful enough for a sub-mm detection.
• Remember: both UV searches (dust) and sub-mm searches
(fainter galaxies) have ~10 corrections to get total SFR
HWR
Princeton, 2005
SED fits for DRGs
Near-IR selected
UV selected
HWR
Princeton, 2005
Förster-Schreiber, Franx, Rix et al; FIRES
Improving Mass, SFR and Av Estimates at z~2.5
through IRAC (3.6mm-8mm) data
Labbe, Franx, Rix et al 2005
Förster-Schreiber, Rix et al 2005; FIRES
HWR
Princeton, 2005
Comparing dynamical (?) with SED masses
HWR
Princeton, 2005
Van Dokkum, Franx, Rix, et al. 2004
Results I: Cosmic Star-Formation Rate
HWR
Princeton, 2005
SFR’s from thermal-IR flux 0<z<1
Zheng, Rix, Rieke, Bell et al 2004
Stacking galaxy classes (z,L) from
COMBO-17 and measuring the 24mm flux
HWR
Princeton, 2005
SFR’s from thermal-IR flux 0<z<1
Zheng, Rix, Rieke, Bell et al 2004
• Through stacking,
(single source) confusion limit
LIR/LSpitzer’s
UV = f(SFR) @ all z,Lopt
can be beat by >10 to <10mJy
• IR flux dominates in all galaxies (to 3% of L*) to z~1.2;
–
large majority of UV photons absorbed.
Local relation
• Mean LIR/LUV drops with galaxy luminosity
 faint galaxies contribute hardly to SF integral
• “Correction” seems to be a function of (absolute) SFR only
– Insensitive to stellar luminosity, redshift
HWR
Princeton, 2005
State of Affairs: Star-fomration rate
Borch, Rix, Meisenheimer et al 2005
HWR
Princeton, 2005
Why the drop of the SFR since z~1?
or
In what type of galaxies did stars form back then?
HWR
Princeton, 2005
Whence the UV flux at z~0.7?
j280 (z~0.7) ~ 4 x j280nm (now)
[not necessarily true in massive,
old systems]
Explore “morphology” of
galaxies that give rise to
these photons
Subjective – use 6 eyes
[Morphologies from GEMS, see
Thursday]
HWR
Princeton, 2005
UV luminous
“blue”
Pick f(2800A) as a proxy
for young stars (t<tdyn)
UV-to-optical flux (M280nm – V)
Wolf, Bell, Rix et al 2004
0.65<z<0.75
UV-light contribution by
•
At MV>-19 and z~0.75
– ½ the flux comes from
seemingly normal spirals
– 20% from visibly interacting
systems
• only minority of UV flux from
manifestly interacting
systems at z~0.75
• drop in (major) merger
rate not cause of SFR drop
HWR
Princeton, 2005
Galaxy type at z~0.75
z~0.75
Normal
spirals
Results II:
Evolution of the Stellar Mass Density with Redshift
HWR
Princeton, 2005
The Evolution of the Stellar Mass Function
over the Last 7 Gyrs
COMBO-17 survey; 30,000 galaxies
Mean stellar mass Build-Up
Present-day
stellar mass
function
HWR
Princeton, 2005
Borch, Meisenheimer, Rix, Bell et al 2005, COMBO-17
Where is the stellar mass at z=2-3.5?
DRGs (“distant red galaxies”) vs Ly-Break Galaxies
Distant red galaxies
likely dominate the
mass budget
HWR
Princeton, 2005
<r*(z)>: State of Affairs
Borch, Meisenheimer, Rix et al 2005
HWR
Princeton, 2005
…half the mass since z~1.5…
Borch et al 2005
HWR
Princeton, 2005
Putting it together
Borch, Meisenheimer, Rix et al 2005
HWR
Princeton, 2005
Summary
• Waning SFR not a consequence of waning major mergers
– Waning cold gas supply
• SED-based stellar mass estimates now available for 1000’s
of galaxies to z~3
– Need to observe at least to lrest>4000A
– Available testing against dynamics OK
• “Distant red galaxies”, between Ly-break and sub-mm
galaxies, may contain the bulk of stellar mass 2<z<3.5
– Found through near-IR surveys
– Quite frequent objects with SFR x tSFR ~1010-11M
• <r*(z) > can be traced from z~3.5 to 0
– enclosing ~90% of all stars formed
• Integral over SFR estimate agrees with <r*(z)
– Assuming diet-Salpeter IMF (e.g. Kroupa 2002)
– Leaves not much room for overlooked SFR
HWR
Princeton, 2005
> to < 2
Where do we go from here?
• Role of merging in the build-up of the galaxy mass
function is observationally barely constrained
• Comprehensive linkeage of SED-based and
dynamical masses
• Beat field-to-field variations at z>2
• Relate stellar masses at different z to halo
masses
– Lensing, clustering
HWR
Princeton, 2005
HWR
Princeton, 2005
HWR
Princeton, 2005
HWR
Princeton, 2005
HWR
Princeton, 2005
HWR
Princeton, 2005
HWR
Princeton, 2005
HWR
Princeton, 2005
HWR
Princeton, 2005
HWR
Princeton, 2005
Improving Mass, SFR and Av Estimates at z~2.5
through IRAC (3.6mm-8mm) data
Labbe, Franx, Rix et al 2005
Förster-Schreiber, Rix et al 2005; FIRES
HWR
Princeton, 2005
SED Fitting of FIRES Galaxies
HWR
Princeton, 2005
Where is the stellar masses at z=2-3.5
DRGs (“distant red galaxies”) vs Ly-Break Galaxies
HWR
Princeton, 2005
Förster-Schreiber, Franx, Rix et al 2005