Star-Forming Galaxies with VIRUS/HETDEX

1
Star-Forming Galaxies withVIRUS/HETDEX

• Caryl Gronwall (Penn State)

2
Searches for Primeval Galaxies

• Primeval galaxy in the act of formation
• Expect strong star-formation (hence strong
  emission-lines), little dust metals, high
  redshift
• Search primarily consists of looking for Ly?
  emission-line galaxies
• Understanding the properties of such galaxies can
  help with our picture of galaxy formation

3
Searches for Primeval Galaxies

• Search techniques
• Slitless spectroscopy (high sky background)
• Long-slit spectroscopy (small effective area)
• Narrow-band imaging (small volume)
• Lots of telescope time devoted to these searches,
  as of 1994 this pervasive population has so
  far eluded detection, Pritchet, PASP review
• Problems galaxies fainter than expected,
  detectors were too small

4
Searches for Primeval Galaxies

• By late 1990s, detecting z3 galaxies via Lyman
  break (Steidel et al.) was easy! Why bother
  looking for Ly? galaxies anymore?
• But Keck long-slit narrow-band surveys in 1998
  (Cowie Hu, Hu et al) found 15 Ly? emitters in a
  30 sq. arcmin region
• Since then, narrow-band surveys using mostly 4-m
  class telescopes have been extremely successful
  Stiavelli et al. 2001 (z2.4) Arnaboldi et al.
  2002, Ciardullo et al. 2002 (z3.1) Fujita et
  al. 2003 (z3.7) Rhoads et al. 2000 (z4.5)
  Ouchi et al 2003 (z4.9), Rhoads Malhotra 2001,
  Ajiki et al. 2003 (z5.7) Kodaira et al 2003,
  Rhoads et al 2004 (z6.5)

5
Our Observations

• Narrow-band O III?5007 imaging using Mosaic
  camera on the CTIO 4-m telescope
• Region the Extended Chandra Deep Field South
• 35 x 35 arcmin region
• 20-hour(!) exposure through the 50 Å FWHM filter
• z 3.09 - 3.13
• Off-band derived from deep B V images
• (Gronwall, Ciardullo, Gawiser, Feldmeier, Hickey,
  et al.)
• Detect galaxies via
• Difference image
• On-band Off-band color

6
On Off Difference
7
(No Transcript)
8
35
9
Results

• Detect 150 Ly? emitters brighter than 1.52 x
  10-17 ergs/cm-2 /s. This sample is
  significantly (5 times) larger than existing
  samples at z3 and higher redshift.
• Will allow us to study this population in detail
  and study evolution in number of Ly? emitters.
• Provides good estimate of number of galaxies that
  will be seen by HETDEX

10
Luminosity Function
11
Luminosity Function
12
Luminosity Function

• -1.35 /- 0.12
• n(lt 1.5 x 10-17 ) 0.00146 /- 0.00014 ) Mpc-3

13
Rest Frame Equivalent Width
14
Implications for HETDEX

• Expect 4.84 LAEs per square arcmin per unit
  redshift interval (to 1.5 x 10-17erg/cm2/s).
• HETDEX will detect 977,000 galaxies in 30 sq.
  degrees with 1.8 lt z lt 3.7 to a flux limit of 1.5
  x 10-17erg/cm2/s. If flux limit is 2.0 x
  10-17erg/cm2/s, number decreases by 25.
• Primary possible contaminant will be O
  II??3727,3729 from z0 to 0.53. For zlt0.17, H?
  will be visible.
• Will need to rely on lack of continuum blueward
  of the emission line.
• Multicolor imaging would be helpful!

15
Science with HETDEX

• 1 million LAEs from 1.8 to 3.7 will enable some
  very interesting science versus redshift
• Evolution of luminosity function, and star
  formation history. Note that LAEs dominate the
  faint end of the LF at high redshift.
• Determine fraction of total star formation
  contained in LAEs.
• Search for presence of HeII ?1640 line, both in
  individual and combined spectra -- indicates
  presence of Population III stars.
• LAEs properties vs. galaxy environment.

16
Science with HETDEX -- O II

• HETDEX will also detect a similarly large number
  of O II galaxies from z0 to 0.53.
• Estimates using existing LFs from Gallego et al
  (2002) z0 and Hogg et al. (1998) z0.3 to
  1.5, median z0.6 imply 3.22 O II galaxies per
  square arcmin to VIRUS redshift limit assuming
  conservative flux limit of 3.0 x 10-17erg/cm2/s.
  Courtesy of Niv Drory
• In 30 square degrees, HETDEX will detect at least
  348,000 O II emitters, probably substantially
  more!

17
From Gallego et al (2002)
18
Science with HETDEX -- O II

• Locally (z lt 0.095), SDSS has found a SFR-density
  relation, SFRs are lower in dense environment.
• HETDEX will be able to explore the evolution of
  such relations with redshift.
• Using the low resolution red grating on VIRUS
  would allow the extension of such work to z
  0.8.

19
From Gomez et al (2003)