Morphological Diversity among the highz Galaxies in the

1
Morphological Diversity among the high-z Galaxies
in the
Great Observatories Origins Deep Survey
(GOODS)
Swara Ravindranath
Inter-University Center for Astronomy
Astrophysics (IUCAA, Pune)
Collaborators Mauro Giavalisco, Henry C.
Ferguson, Chris Conselice, Martin Weinberg, Neal
Katz, Mark Dickinson, Jennifer Lotz, Michael
Fall, Bahram Mobasher, Casey Papovich
To appear in ApJ (October 20, 2006)
astro-ph/0606696
At the Edge of the Universe - Latest results from
Deepest Astronomical Surveys, Sintra, Portugal
October 13, 2006
2
Observed evolution in morphology with redshift..

• evidence for spirals and ellipticals by z1,
  with only mild evolution at z lt1 from HST
  surveys.
• (Lilly et al. 1998, Im et al. 1999, Simard et al.
  1999, Stanford et al. 2004, Ravindranath et al.
  2004, Barden et al. 2005)

• At z gt1, regular Hubble type morphology can be
  identified to z 1.5.
• Hubble sequence not observed beyond z 1.5
  mergers appear dominant (Conselice et al. 2005
  Papovich et al. 2005)

3
GOODS HST/ACS Treasury Program
P.I. Mauro Giavalisco
GOODS 320 arcmin2 UDF 10 arcmin2
HST/ACS F435w, F606w, F775w, F850lp
courtesy F. Summers
4
GOODS

• Large sample for morphological analysis

U-dropouts (z3.1) 1526 B-dropouts (z3.8)
2463 V-dropouts (z4.9) 878

• 292 starburst galaxies at z1.2 (low-z sample)

• Study rest-frame UV ( 1400-1600Å) morphology
  consistently for z3,4, and 5

• High spatial resolution 700-800pc for LBGS

PSF FWHM 0.1? and pixel sampling 0.03?/pixel
5
Morphological k-correction
HDF galaxies generally look very similar in
rest-frame UV and rest-frame optical. (Dickinson
2000) Quantitative measurements of internal
color dispersion show k-correction can be
significant for the lower redshifts, but not so
for the LBGs. (Papovich et al. 2003) Rest-frame
UV morphology gives fair estimate of the dominant
morphology.
6
U-dropouts LBGs at z3
lt-----------gt
3"
7
B - dropouts LBGs at z 4
8
V-dropouts LBGs at z 5
9
Morphological Analysis

• - Quantitative morphologies for LBGs is
  challenging, and mapping to the galaxy
  morphologies seen locally is not easy (faint,
  small, and different stellar population).
• Various approach to derive quantitative
  morphology CAS (Conselice 2003), Gini coeffs and
  M20 (Lotz et al. 2004)
• Our attempt is to classify broadly as smooth
  disks, mergers, and objects with high central
  concentration.

• Surface brightness profile shapes (Giavalisco et
  al. 1996)
• -Axial ratio distributions (Odewahn, Burstein,
  Windhorst 1997)

10
Surface brightness profile Analysis
- 2-D modelling using a single Sérsic function
Exponential disks n 1 R1/4 spheroids n 4
GALFIT Software (Peng et al. 2002)

• allows convolution by the point spread function
• better handle on flux in the galaxy wings where
  S/N drops at low surface brightness levels
• Measurement biases minimized

Quality control low chi2, small errors on
parameters, mfit mauto0.5
11
B-dropout with n 5 (centrally
concentrated)
?
3"
?
? 3"?
100 x 100 pixels
12
B-dropout with n gt 3.0 (spheroid-like)
13
B-dropout with n 0.8 (disk-like)
14
B-dropout with nlt0.5 (mergers, multiple
cores)
15
Hubble Ultra Deep Field
U-dropouts 32 B-dropouts 85 V-dropouts 26
Small statistics, but allows to evaluate the
measurement errors on the shape parameters.
Surface brightness of 1-? isophote GOODS 25.46
AB magnitudes/arcsec2 HUDF 27.60 AB
magnitudes/arcsec2
Beckwith et al. in prep
16
Comparison of structural parameters from GOODS
HUDF
17
Distribution of Profile shapes for LBGs and z1.2
starbursts
Morphological types LBGs at z gt 2.5 40
exponential disks 30 spheroid-like 30
mergers, multiple cores
Star - forming galaxies at z 1.2 26
exponential disks 16 spheroid-like 58
mergers, irregulars?
18
Probing the Intrinsic shapes through ellipticity
distribution
Observed peak in the ? (1- b/a) , and skewed
distribution ? Not population of
spheroids and circular disks seen at random
orientations
Likely Prolate or triaxial
Peak ? shifts lower at lower z
19
Ellipticity distribution for different LBG
profile types.
20
Results
? Contrary to previous results, profile shapes
suggest LBGs are predominantly disks mergers.
Similar results also from gini coeffs analysis
Lotz et al. (2005). But fraction of spheroids
increases at high-z ? Striking contrast to low
redshift galaxy population in ellipticity
distribution, with skew towards high ?.
Elongated morphologies dominate LBGs !
21
Implications for galaxy formation models
Possible explanations for the excess of
Elongated morphologies among LBGs ! ?
Rotation-dominated disks? Edge-on projections and
selection effects ? Star forming clumps along
gas-rich filaments of cold gas infall in DM
halos ? High-z bars at early epochs of galaxy
formation?
22
Rotation curves at z2
GOODS large area and angular resolution enabled
the first systematic kinematical studies of
galaxies at z2
Erb et al. 2004 Steidel et al. 2004
13 GOODS-N galaxies morphologically selected
23
Rotation curves at z2
Only 3 out of 13 galaxies show evidence of
coherent velocity field/shear
What is the relationship between morphology and
kinematics?
Erb et al. 2004 Steidel et al. 2004
HDF-BX1397
24
Galaxy kinematics at z2
Elongated galaxies appear to have lower velocity
dispersions
Erb et al. 2004 Steidel et al. 2004
25
Hydro-dynamical (SPH) sims (Keres et al. 2005)
cold accretion dominates for Mgal 1010.3
Msun Mhalo ?1011.4 Msun
26
Observing the first gas-rich bars among LBGs at z
gt 2.5?
Classic bar morphology in the first few billion
years!
Bar encompasses the whole galaxy 2-3 kpc
scalelength
27
Summary

? About 4700 LBGs identified in the GOODS images
with redshifts 2.5 lt zlt 5.0 morphological
analysis based on 1333 LBGs. ? Based on profile
type 40 disks, 30 mergers. Only 30 spheroids,
not dominant as in previous results ? Fraction of
spheroids higher by 20-25 at high-z compared to
star-forming galaxies at z1.2 ? Ellipticity
distribution skewed to high ? with a peak that
shifts to lower ? at low z, implies elongated
morphologies for most LBGs ? Possible
explanations include star formation along dense
filaments of cold accretion in DM halos, or
prevalence of high-z gas-rich bars which
encompass the entire galaxy!
28
More bar signatures among LBGs at z gt 2.5
Spiral arms from bar ends?
29
And morepossible bars among LBGs at z gt 2.5
Star formation at bar ends?
30
LBGs attractive candidates to explore the
morphology at high-z (Z gt 2)

• LBG selection is an efficient way to identify
  young (lt 2.5 Gyr) galaxies confirmed by
  spectroscopic redshifts to be at zgt2.5.
• Largest samples available among high-z galaxies
  selected by various selection methods.
• UV selection suggests relatively low dust
  obscuration
• high star formation rates, small sizes,
  clustering - attractive candidates for being
  building blocks of local spheroids in
  hierarchical galaxy formation.

31
Lyman-break galaxiescolor selection
Giavalisco et al. 2004
B-V
V-i
i-z
V-z
B-dropouts, z4
V-dropouts, z5
(V-i) gt 1.5 0.9 (I-z) ?
(B-V) ? 1.2 1.4 (V-z) ? (B-V) ? 1.2 ? V-z ? 1.2
(V-i) gt 2.0 ? (V-I) ? 1.2 ?
(i-z) ? 1.3
32
Hubble Sequence linking morphologies and global
properties to galaxy formation

• Morphologies serve as easy observables locally,
  and at cosmological distances.
• Local ellipticals, and bulges of spirals are
  described by the r1/4 law disks have
  exponential profiles.
• Star formation, gas content, increase from early
  to late types.
• Kinematics vary along with Hubble type, with
  rotation becoming dominant in disks.

33
How do galaxies evolve with look-back time?
Galaxy morphology Luminosity function Galaxy
sizes Stellar mass assembly
Observational challenge

• ? decrease in angular size ? high resolution
  images
• Surface brightness dimming ? deep images
• Band shifting at high redshifts ?
  multi-wavelength
• Cosmic variance ? large area, independent fields

34
Illustration of dropout technique for B-dropouts
B
V
i
z
35
Redshift distribution function for the LBGs
36
Star-forming galaxies at high redshift (LBGs)
routinely identified by the color signature of
the Lyman limit and Ly a breaks.GOODS is
sensitive to LBGs at zgt2.5.At z3, GOODS
reaches 0.3-0.5 mag deeper than previous large
surveys (Steidel et al. 1999).
Color Selection of Distant Galaxies Giavalisco,
ARAA, 40, 576, 2002
37
Comparison of Structural Parameters from GOODS
and HUDF
38
Lyman Break Galaxies at z gt 3Proto Spirals or
Proto Ellipticals?
C ? 5log r80/r20
B dropouts z4
disks
disks
spheroids
spheroids
- early results from GOODS 3-epoch data
Ferguson et al. 2004
39
Evolution in the morphology of star-forming
galaxies at high-z
LBG LUV gt 0.5 L ( MUV -21.02 Steidel et
al. 1999) z 1.2 sample LUV gt 0.5 L (MUV
-18.04 Wyder et al. 2005)
40
Fraction of Morphological Types among the zgt2.5
LBGs Starbursts at z1.2
41
LBGs at z3 with n lt 0.8 (multiple cores,
mergers, chains?)
42
LBGs as bursts of star-formation in filaments of
cold gas in DM halos
43
Size Evolution at high redshifts (z gt 2)

• Lyman break Galaxies at z3,4, 5 selected by
  color criteria
• Sample restricted to 0.7 Llt L lt 5L
• Half-light radii (SExtractor) and Petrosian radii

Standard ruler
RH(z)-2/3
RH(z)-1
Ferguson et al. 2004
44
(No Transcript)
45
(No Transcript)
46
(No Transcript)
47
(No Transcript)
48
LBGs at z 3 with steep n values
49
LBGs at z 4 with steep n values
50
LBGs at z4 with nlt0.8