The UV Luminosity Function of star-forming galaxies via dropout selection at redshifts z ~ 7 and 8 from the 2012 Ultra Deep Field campaign

TL;DR

This study measures the UV luminosity function of star-forming galaxies at redshifts 7 and 8 using ultra-deep Hubble data, revealing the faint end slope and improving understanding of early galaxy populations.

Contribution

It provides the deepest UV luminosity function measurements at z ~ 7-8 using enhanced HST near-IR data and refined selection techniques, extending the faint magnitude limits.

Findings

Faint end slope at z ~ 7: α = -1.87 ± 0.18

Faint end slope at z ~ 8: α = -1.94 ± 0.23

Probed galaxies down to M_UV = -17

Abstract

We present a catalog of high redshift star-forming galaxies selected to lie within the redshift range z ~ 7-8 using the Ultra Deep Field 2012 (UDF12), the deepest near-infrared (near-IR) exposures yet taken with the Hubble Space Telescope. As a result of the increased near-infrared exposure time compared to previous HST imaging in this field, we probe 0.65 (0.25) mag fainter in absolute UV magnitude, at z ~ 7 (8), which increases confidence in a measurement of the faint end slope of the galaxy luminosity function. Through a 0.7 mag deeper limit in the key F105W filter that encompasses or lies just longward of the Lyman break, we also achieve a much-refined color-color selection that balances high redshift completeness and a low expected contamination fraction. We improve the number of drop-out selected UDF sources to 47 at z ~ 7 and 27 at z ~ 8. Incorporating brighter archival and ground-based samples, we measure the z ~ 7 UV luminosity function to an absolute magnitude limit of M_UV = -17 and find a faint end Schechter slope of \alpha = -1.87+/- 0.18. Using a similar color-color selection at z ~ 8 that takes account of our newly-added imaging in the F140W filter, and incorporating archival data from the HIPPIES and BoRG campaigns, we provide a robust estimate of the faint end slope at z ~ 8, \alpha = -1.94 +/- 0.23. We briefly discuss our results in the context of earlier work and that derived using the same UDF12 data but with an independent photometric redshift technique (McLure et al 2012).