1. Constraints on Star Forming Galaxies at z>6.5
Samantha Hickey (University of Hertfordshire) Andrew Bunker (University of Oxford)
Matt Jarvis (University of Hertfordshire)

2. Introduction
WMAP observations indicate the universe is neutral at z>10 (Kogut et al. 2003)
Complete absorption of QSO spectra at z~6.2 (Becker et al. 2001)
6<z<10 is a crucial time for understanding reionisation
Hot ionised plasam after BB
Expanded cooled recombination
Current universe is ionised
How, what, when the universe was reionised?
The complete absorbtion indicates there is a relatively large fraction of neutral hydrogen between us and the quasar
Marks the end of the epoch of reionisation
Pentricci et al. 2002

3. Introduction
The number density of AGN at z>6 appears too low to fully reionise the IGM (Dijkstra et al. 2004)
Main candidate for reionisation is star formation (Eyles et al. 2007)
What set the stage for galaxy formation at the end of the ‘dark ages’
Dijkstra estimated the level of soft x-rays produced by AGN that would be required to reionise the univese and compared that to the
measured the redshifted hard x-ray background (now soft x-rays) and subtracted out the lower redshift x-ray sources and found the number density of the AGN to be too low to be solely responsibloe for the reionisation of the universe
We want to search for UV bright star forming objects at high redshift
Pentricci et al. 2002

4. The method
Lyman Break caused by absorption due to intervening clouds of neutral hydrogen short-ward of ~1216Å
Search for a significant spectral break in photometry
For higher redshift objects look to longer wavelength filters
At z>6.5 LB is shifted to NIR wavelengths
This technique has been spectroscopically confirmed to find galaxies at z=6
Extending the technique to higher redshifts

5. The data
New VLT/HAWK-I Y-band science verification data (Fontanna et al. and Venemans et al.)
Covers an area of ~119 sq. arcmin
5 sigma limiting magnitude 25.9 for deepest region
VLT ISAAC J & Ks
HST/ACS B,V,i’,z’
Spitzer/IRAC
All data is over GOODS-South field
Fills in a wavelength gap at 1 micron between the z (0.85 microns) and j (2.14 microns) band filters (UKIRT)
Onsists of 2 pointings, 1 deeper

6. Selection & Contaminants
SExtractor used in dual-image mode to create catalogs
Matched to GOODS-MUSIC catalog within 0.36’’
Criteria:
Colour cut of (z’-Y)AB > 1mag
Colour cut of (Y-J)AB > 0.75 mag
Signal/Noise > 5
Exposure map value > 2.5 hours observation
Main Contaminants
Detector artifacts
Ghost image haloes
Low redshift interlopers (identified via optical detections)
Transients
Transients - 6 possible z’drops, split the Yband data in to two halves to see if all the cands were visible with simillar photometry in both, 2 of these cands were only present in the second half of the data and so were eliminated from our selection, as transient objects, possibly SN, both were found to be undetected on the first 3 nights of observations and then of simmillar magnitudes for the last 2 nights.
V drop
Brown Dwarfs (Manucci)

7. Completeness
More than 5000 artificial compact sources were created convolved with the PSF
Span AB mags
Our 50% completeness limit of YAB=25.9 was comprable to the 5 sigma limiting magnitude

8. Z’-drops 2200 Stanway et al. identified as a dusty z=2.73 galaxy
9697 has colors inconsistent with being a brown dwarf
2200 Stanway et al. identified as a dusty z=2.73 galaxy
9136 z=7.2 candidate
9697 z=6.9 candidate

9. Brown Dwarfs
Compared photometry of our candidates to brown dwarf colours
9136 is inconsistent with being a brown dwarf
9697 at the extremes of its errors has colours consistent with an L dwarf
Would need spectroscopy to prove or disprove candidates
(Patten et al. 2004)

10. Y-drops Only well detected in J-band
Can be explained by spectral slope arguments
Possible result of line emission, requires Ly EWrest > 100Å
No robust Y-drops

11. Expected numbers
The number of robust candidates can constrain the z~7 luminosity function
The number of robust candidates can constrain the z~8 luminosity function
Y-drops
Steidel z=3 predicts 10.5±3.2
Bouwens z=6 predicts 1.1±1
Once again we find evolution in the LF since z=3
Need to go deeper and /or wider to detect Y-drops and constrain LF’s at z=8
Ultra-VISTA and VIDEO will probe this region
z’-drops
Steidel z=3 and Bouwens z=6 predict 29.5±5.4 & 5.2±2.3 z’drops
We find evolution in the LF since z=3 and we are consistent with the Bouwens et al. z=6 LF

12. Summary
We found 2 z’-drop candidates both with strong IRAC detections, 2 robust candidates
Comparing observations with expected numbers we rule out the case of no evolution in the luminosity function since z=3, but we are consistent with the Bouwens z=6 LF and the McLure z=6 LF
We found a maximum of 4 possible Y-drop candidates (only detected in J) cannot rule out line emission or spectral slopes
Need deeper and /or wider observations to probe the z>7 luminosity function. This is a possibility with Ultra-VISTA and VIDEO