Transcript SXDS Highlights : Subaru / FOCAS Spectroscopy

Subaru
Adaptive Optics (AO) Rest-frame V-band Imaging of
Galaxies at z~3 :
High Surface Density Disk-like Galaxies ?
Masayuki Akiyama (Subaru Telescope, NAOJ)
Kouji Ohta (DoA, Kyoto Univ.)
Yosuke Minowa (Mitaka, NAOJ)
Naoto Kobayashi (IoA, Univ. of Tokyo)
Ikuru Iwata (OAO, NAOJ)
ApJS accepted, arXiv.0709.2714
Subaru User’s Meeting 20080130
Naive motivation
Subaru

The morphology of galaxies at z~1 still follows Hubble sequence seen
in the nearby universe. How about galaxies further away ?
3col images of z=1 galaxies in GOODS
Rest-frame optical morphology is important
Subaru

Rest-frame optical morphology of galaxies reflects the stellar mass distribution
of galaxies, and provides important information on the dynamical structure of
galaxies.
Two spiral galaxies at z~1
Shorter than 4000A
break:
Longer than
4000A break:
Distribution of
young stars =
distribution of star
forming regions
Distribution of red
and long-lived
stars = distribution
of stellar mass
K-band = 5600A @ z=3
Adaptive Optics
0.1-0.2” = 0.8-1.5kpc
Targets for Observations
Subaru



Our main targets are U-band dropout Lyman Break Galaxies (LBGs)
• Steidel et al. 2003 is the largest sample of spectroscopicalyconfirmed z~3 galaxies selected by U-dropout Lyman Break method.
• Select a sample not affected by the redshift uncertainty with LBG
An radio galaxy (4C28.58 at z=2.891)
We also examined morphologies of serendipitously observed Distant
Red Galaxies (DRGs) in our FoVs. DRG criterion of J-K>2.3 also selects
red galaxies at similar redshifts to U-dropout LBGs.
Observation: Subaru Telescope Intensive Program
Subaru
Subaru 8.2m
+ AO36 system:
Low-order correction
with low-noise
Shack-Hartmann
wavefront sensor
= Good for extragalactic studies !





Natural guide star AO system on Subaru telescope with IRCS.
154 hours of observation in total.
13 FoVs with 36 LBGs , 1 RadioG., and 7 DRGs are observed.
Typical on-source effective integration is 5 hours.
Typical PSF size at the target position is FWHM=0.2” (~1.5kpc@z=3)
Observation: Subaru Telescope Intensive Program
Subaru
An example of an FoV
with 6.2h integration
PSF-reference(20”)
FWHM=0.20”
PSF-reference (15”)
FWHM=0.18”
LBG@z=3.261
AO Guide Star
LBG@z=3.088





Natural guide star AO system on Subaru telescope with IRCS.
154 hours of observation in total.
13 FoVs with 36 LBGs, 1 RadioG., and 7 DRGs are observed.
Typical on-source effective integration is 5 hours.
Typical PSF size at the target position is FWHM=0.2” (~1.5kpc@z=3)
Images of LBGs in order of K-band magnitudes
Subaru
Kvega<21.5
Kvega<22.5
No detection
36 LBGs are observed, 31 are detected
 3.5”x3.5” ~ 30kpc x 30kpc

Luminosity vs. J-K color of the LBGs
Subaru


The observed sample covers a wide range of the rest-frame optical absolute magnitude
(between Mv*-0.5 and Mv*+3.0)
The LBG-selected galaxies cover not only the less-massive bluer galaxies (U-V~-0.3) but
also the massive redder galaxies (U-V~0.5) similar to DRGs.
Offset between optical and K-band Images
Subaru

Bright LBGs show significant offsets between K-band (rest-frame optical) and
seeing-limited optical (rest-frame UV) images. This indicates optical and UV
morphologies are different.
One component Sersic profile fitting for bright (~Mv*) LBGs
Subaru
Kvega<21.5
Kvega<22.5
No detection

36 LBGs are observed, 31 are detected
Examples of Sersic profile fittings for LBGs with Kvega<21.5

Subaru
LBGs are described better with n=1 Sersic profile (similar to disk
galaxies, less concentrated; green) than n=4 Sersic profiles (similar to
spheroidal galaxies, more concetrated; blue).
Summary of Sersic fittings for Kvega<21.5 LBGs (+DRGs)
Subaru

Most of the LBGs (+an RadioG +DRGs) are fitted well with Sersic profiles with
n<2.
Summary of Sersic fittings for Kvega<21.5 LBGs (+DRGs)
Subaru

Results of “cloning” simulations show if there are large number of elliptical or
bulge-dominated galaxies at z~3, they should be detected, and should be fitted
well with large n-index.
Concentration vs. Size distribution of Kvega<22.5 LBGs / DRGs
Subaru

For fainter LBGs/DRGs, profile fittings with free n is not reliable, thus we
compared their concentration with those of nearby galaxies. The distribution
of LBGs/DRGs are more consistent with n<2 disk-like profile than with n>2
spheroidal-like profiles.
Surface brightness & surface stellar mass density
Subaru
z=0-1 from Barden 2005


If we assume that the LBGs/DRGs have disk-like morphology, V-band
surface brightnesses inferred from the size-luminosity relation is 2.9mag, and
1.7mag brighter than z=0 and z=1 disk galaxies, respectively.
Surface stellar mass densities inferred from the size-stellar mass relation is 36 times larger than z=0-1 disk galaxies shown with thick solid line.
Summary of the results



Subaru
K-band peaks of bright red LBGs show offsets from the optical
positions. Their inside stellar mass distributions are different from
the distributions of star forming regions.
Radial profiles of LBGs (+RadioG. +DRGs) are relatively flat, and
similar to disk-galaxies in the local universe.
Rest-frame optical surface brightnesses of the z=3 LBGs (DRGs)
are brighter than z=0-1 disk galaxies. Surface stellar mass
densities of massive LBGs are also larger than z=0-1 disk galaxies.
Naive speculation: placing the z~3 galaxies in the growth paths of
galaxies
Subaru
Basically, gas-poor dissipation-less merging produce concentrated structure similar to
elliptical galaxies. So in order to maintain the disk-like structure of the galaxies, gasrich merging process can be a key (e.g., Springel & Hernquist 2005).
New era of high-z morphology study with Laser Guide stars
Subaru
Current sample is not sufficient statistically, especially for bright (<Mv*)
galaxies …



In order to confirm the disk-like morphology of z~3 galaxies, the
distribution of ellipticities is a next important observable.
Most of the bright (Mv*) z~3 LBGs in Steidel et al. (2003) with
Natural Guide stars are observed in this program, thus in order to
extend the sample of bright LBGs, we need AO observation with
Laser Guide star.
• Gemini / Altair / NIRI observation is in the S07B ques of the
current semester, BUT ONLY 7 hours out of 16 hours (A)+8
hours(B), NOT SO CONVINCING EVEN FOR Rank A !!
Stellar dynamics is also important, but difficult.
Why LBGs to understand formation and evolution of galaxy bulges ?
Subaru



Strong spatial clustering of LBGs indicates that they reside in massive
halos and are progenitors of massive galaxies (=elliptical or bulgedominated galaxies) in the local universe (e.g. Giavalisco & Dickinson
2001).
The apparent sizes of the LBGs in the rest-frame UV-band are similar to
the sizes of the spheroids in the local universe (e.g. Steidel et al. 1996).
Therefore, LBGs are thought to be closely related to the formation of the
spheroidal (elliptical or bulge) component of galaxies.
Why Study Rest-frame Optical Morphologies of z~3
Galaxies
Subaru


HST/NICMOS H-band Observations are not sufficient ! H-band observation
only covers up to 4000A in the rest-frame, and star-forming regions can
dominate the morphology.
HST/NICMOS sample is limited to a small number of objects in Hubble Deep
Field and does not have bright (~Mv*) galaxies at z~3. The physical
properties of LBGs clearly depends on the luminosity (more luminous LBGs
have redder color, have stronger clustering, have weaker Lya emission, and
so on), thus it is still important to observe a sample covering wide luminosity
range.
Shorter than 4000A
break:
Distribution of
young stars =
distribution of
star forming
regions
Longer than 4000A
break:
Distribution of
red and longlived stars =
distribution of
stellar mass
K-band
Adaptive Optics
0.1-0.2” = 0.8-1.5kpc
“Cloning” z=3 galaxies with GOODS Data
Subaru


Compare the K-band morphologies of z=3 LBGs with z=0.4-0.6 galaxies in the
GOODSN region. K-band@z=3 corresponds to I,z- band @ z=0.4-0.6.
Covered volume @z=0.4-0.6 by GOODSN is comparable to that @z=3 by IRCS/AO
LBGs.
2PLE case
Estimated the PSFs at the target positions


Subaru
Estimate the PSF shape at the positions of the targets, using a
few stars in the FoV.
During the Sersic profile fitting, the parameters are changed
within the range shown with yellow hatch.