Galaxies at very high-z

Galaxy Luminosity Function and Cosmic Hydrogen Reionization

Searching for and understanding galaxies at very high redshifts has always been of great interest to me. Based on the fact that cosmic hydrogen should have been fully reionized by z ~ 6, we argued (Yan & Windhorst 2004, ApJ, 600, L1) that the luminosity function of galaxies at z ~ 6 must have a very steep faint-end slope (with α < -1.6 and possibly much steeper, approaching α ~ -2.0). In other words, the faint ("dwarf") galaxy population must have played the dominant role in keeping the cosmic hydrogen ionized at z ~ 6. In order not to have the cosmic hydrogen fully reionized at a too early epoch (say z ~ 7), the luminosity function must truncate at a certain level so that there would not be too many ionizing photons.

The very steep faint-end slope was later confirmed by the z ~ 6 dropout searches in the Hubble Ultra Deep Field ("HUDF"); in our own work on the z ~ 5.5-7 candidates in the HUDF (using data from both ACS and NICMOS), we specifically pointed out (Yan & Windhorst 2004, ApJ, 612, L93) that α is likely between -1.8 and -1.9. The installation of the WFC3 to HST enabled the dropout search to higher redshifts (up to z ~ 10), and the steep α persists in the results from the HUDF WFC3-IR data (e.g., Yan et al. 2010, RAA, 10, 867).

The WFC3 has also enabled high-z dropout search over much larger areas than in the HUDF, and therefore we now have a better statistics at the bright-end of the luminosity function. We saw a somewhat unexpected result -- a "bright-end excess" at z ~ 7--8: there seems to be more z ~ 7--8 dropouts observed than a standard Schechter function fit would predict at the bright end. This is seen in the data from both our Hubble Infrared Pure Parallel Extragalactic Survey ("HIPPIES", of which I was the PI; Yan et al. 2011, ApJ, 728, L22) and the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey ("CANDELS", of which I was one of the co-I's; Yan et al. 2012, ApJ, 761, 177 )