4.Results

1
Preliminary Results on Stellar Populations of
LAEs at z4.8
S. Yuma, K. Ohta, K. Yabe (Kyoto U.), K.
Shimasaku, M. Yoshida (U. Tokyo), I. Iwata (OAO),
M. Ouchi (Carnegie), M. Sawicki (St. Marys U.)
ABSTRACT
We present the preliminary Spectral Energy
Distribution (SED) fitting results of Lyman Alpha
Emitters (LAEs) at z4.8 in GOODS-N and its
flanking fields. Observing through the NB711
narrow-band filter Å, FWHM 73 Å
attached on Suprime-Cam, Subaru telescope, we
obtained 33 Lyman Alpha Emitters (LAEs). 8 of
them are detected in mid-IR data obtained by IRAC
camera on Spitzer Space Telescope (SST) toward
GOODS-N and flanking field. The rest-frame UV to
optical SEDs of 8 LAEs at z4.8 are constructed.
After fitting the observed SEDs with stellar
population synthesis models by Bruzual
Charlot(2003), various physical properties of
LAEs are derived. In this work, we adopt the
constant star formation history with fixed 0.2Z?
metallicity. By taking into account the H?
emission line, the best fitted stellar mass and
age are in the ranges of M?
and 1-550 Myr, respectively. These values are
broadly consistent with other studies on LAEs at
other redshifts (e.g. Gawiser et al.20062007,
Lai et al.2007, Nilsson et al.2007, Finkelstein
et al.2007, and Pirzkal et al.2007). However, our
derived star formation rates (SFRs) and dust
extinctions are significantly larger than those
of LAEs by other studies. Comparing the derived
mass and age with those of LBGs at z5, it is
found that our LAE candidates are younger and
less massive than LBGs.
1.Data and LAE Selection
Broadband photometry of 8 objects are used to
construct Spectral Energy Distributions (SEDs).
The SEDs consist of photometry from 4 bandpasses
which are Ic and z bands from Subaru data and
IRAC channels 1 and 2. The observed SEDs are then
fitted with population synthesis models by
BruzualCharlot (2003) as shown in the diagram
below

• Data From Ground-base to space telescope
• The optical data are obtained by Suprime-Cam on
  Subaru Telescope using NB711 narrow band and
  BVRIcz broadbands centering at Hubble Deep
  Field-North RA(2000) ,
  Dec(2000) . The observations
  are operated in 2005 April.
• Mid-IR data
• GOODS-N field There are publicly available data
  taken under SST Legacy Science Program by IRAC
  camera on Spitzer Space Telescope (SST) in 4
  bandpasses. http//ssc.spitzer.caltech.edu/legacy
  /
• GOODS-FF The flanking regions are also observed
  by using the same instrument (i.e. IRAC on SST)
  in 2005 December and 2006 June. Note that the
  data observed by us are 1 mag shallower than
  those in GOODS-N field.

Parameters used to build models

• BC03 Padova evolutionary track(1994)
• Salpeter IMF 0.1-100 M?
• 0.2Z? metallicity
• Calzetti dust extinction law (2000) with E(B-V)
  0-1.0 (0.02 step)
• Time runs from 0 to 1.2 Gyr (Age of the universe
  at z4.8) with equal logscale time step of 0.1
• Constant Star Formation History (CSF)
• Adding H? line to 3.6?m bandpass

• Selection Criteria for LAEs at z4.8
• 1) Strong detection in narrow band NB711 lt 26.1
  (3? at 2.5 ?)
• 2) Large Ly? equivalent width RI-NB711 gt 0.9 mag
• RI continuum brightness at Ly? wavelength,
  (RI)/2
• This criterion corresponds to the observed EW
  more than 109 Å
• 3) Non-detection in B and V bands
• B gt 28.81 and V gt 28.15, 2? limiting magnitude at
  2.5 diameter aperture)
• 4) The continuum-break criteria as same as for
  LBGs at z5 by Iwata et al.(2007) to reject any
  possible low-redshift interlopers Note that
  these criteria of continuum break are applied to
  objects, Ic and z photometry of which are
  brighter than 3? limiting magnitude at 2.5?,
  26.58 and 25.77 respectively.
• V-Ic gt 1.55, and V-Ic gt 7.0(Ic-z)0.15

4.Results
Figure 1. Plots of the observed SEDs of LAEs at
z4.8 with the best fitted model spectra. The
vertical error bars show errors in photometry,
while bandwidths of each filter are illustrated
by horizontal ones.
Because the observed fields of GOODS-N and its
flanking fields are smaller than those observed
by Subaru Telescope, among 33 LAE candidates,
there are 23 objects in the GOODS-N and its
flanking fields, 10 in GOODS-N and 13 objects in
GOODS-FF. Requiring 2? detection in IRAC
photometry reduces the number of LAEs, which can
be used to fit with models, to 8 objects 5 in
GOODS-N and 3 objects in GOODS-FF.
AB Magnitude
2.Photometry

• Subaru Data Use mag_auto output from SExtractor
• IRAC data Use aperture photometry with a 2.4
  diameter aperture and correct to
• total magnitudes

Observed Wavelength (Å)
An example of Stamp pictures of LAEs In all
available bands
5.Comparison to LAEs at other z
6.Comparison to LBGs at z5
(a)
(a)
(b)
Histograms illustrate the distribution of derived
masses, ages, dust extinction, and SFR of our LAE
candidates and those of LBGs at z5. The good
comparison to our work seems to be the stellar
populations derived by Yabe et al., using exactly
the same models as ours. Figure (a) shows that
the stellar masses of LAEs at z4.8 are mostly
distribute in the low mass ranges compared to the
distribution of LBGs masses. In figure (b), Ages
of LAEs seem to be younger than those of LBGs
however, they are comparable to those by Verma et
al.. Figure (c) indicates that the amount of
E(B-V) of LAEs is less than those of LBGs if
compared to those by Yabe et al. Our derived SFRs
are smaller than those of LBGs from both studies
as seen in figure (d).
Figure (a) indicates that ranges of the stellar
masses of our sample are comparable to those of
LAEs at other redshifts except ones at z5 by
Pirzkal et al.(2007). The difference in masses
between our LAEs and those by Pirzkal et al. may
be explicable by the difference in rest-frame
optical brightness between these two samples.
Pirzkal et al. used the upper limits for IRAC
photometry, thus they selected objects from the
fainter population of LAEs at the redshift. The
fainter the rest-frame optical photometry, the
less massive the stellar mass, as seen in figure
3. It is seen in figure (b) that our derived ages
are broadly consistent with those of LAEs at
other redshifts. Most of our LAEs are young with
ages in the order of few Myr which agree well
with those by Pirzkal et al. On the other hand,
they are higher than the ages of LAEs at z3.1.
It is however difficult to make a strong
constraint on age comparison due to large
uncertainties in derived ages. According to
table 2, dust extinctions of LAEs at z4.8 agree
well with those of LAEs at z4.4 and 5.7, whereas
the discrepancy can be seen if comparing them
with those at z3 and z5. SFRs derived in our
work are hence different from those at the
redshifts.
(b)
(d)
(c)