Morphology of z~1 galaxies from deep K-band AO imaging in the COSMOS field

1
Morphology of z1 galaxies from deep K-band AO
imaging in the COSMOS field

• M. Huertas-Company (LESIA), D. Rouan (LESIA), G.
  Soucail (LAT), L. Tasca (LAM), O. Le Fèvre (LAM)

A century of cosmology, Venice, August 2007
2
Goals

• Understand the building up of the Hubble sequence
• Evolution of the SFR
• Evolution of sizes
• Link between morphology and environment

3
Observing procedure
Cosmological Surveys (COSMOS, GOODS, HDF..)
Thousands of galaxies
Redshifts
Masses
Morphological Classification
Sizes
SFR
4
Difficulties

• Angular resolution
• Wavelength (K-correction)

5
Adaptive Optics in the NIR

• High resolution
• The telescope diffraction limit can be reached
• NIR
• Probed stellar populations closely related to the
  underlying mass (Complement to HST)

6
The Data

• 7 fields of 11 in the Ks band (2.2µm) largest
  AO survey (79 detected objects)
• 3h exposure time per field
• Pixel scale 54mas (undersampling)
• Mean FWHM 0.1
• Complete up to K(vega)22
• Mean photo-z 0.8

7
?
?
HST/ACS - I band
CFHT/Megacam - I band
30
VLT/NACO - K band
8
Classical methods

• 2 types
• Parametric Analytical model fitting (GIM2D,
  GALFIT)
• Non-parametric Measure parameters on the galaxy
  image

Model
Real galaxy

PSF
I
C
S
A
E
Real galaxy
A
C
9
Parametric Analysis Recovering the PSF

• Particular shape
• Undersampled
• Non constant

Galaxy Model (Sersic Exponential Profile)

PSF
Real Galaxy
10
Comparing with CFHT
Huertas-Company et al. 2007, AA, 468, 937
11
and with HST
Simard et al. 2002
20
Huertas-Company et al. 2007
12
Non-parametric analysis
1. Calibration

• Parameters depend on
• Physical properties
• Instrumental effects

Use a calibration sample close to the real sample
2. Boundaries

• Classically
• Linear boundaries

Optimal boundaries?
2. Number of parameters
More than 3?

• Classically
• lt 3 parameters

13
2
3
1
4
14
Z
MAG
15
Morphological parameters

• Morphology
• Concentration
  (Abraham et al., 94, 96, Bershady et al. 00)
• Asymmetry
  (Conselice et al., 00)
• S
• Gini
• M20
  (Lotz et al., 04)
• Shape ellipticity, S/G
• Size area, petrosian radius
• Luminosity magnitude, surface brightness
• Distance redshift

(Conselice et al., 00)
(Abraham et al. 03, Lotz et al., 04)
12-D space
16
Support vector machines

• Particular type of learning machine (Vapnik,
  1995)
• Finds the optimal boundary between distributions

• No linear boundary and non separable data

17
Results
18
Summary and conclusions

• Promising results
• Parametric morphology comparable to HST up to
  K19
• Non-parametric morphology up to K22
• But need more objects! Whats next?
• LARGE PROGRAMME ESO
• ...but rejected 2 times