A critical analysis of the UV luminosity function at redshift  ~7 from deep WFC3 data

¹ INAF - Osservatorio Astronomico di Roma, via Frascati 33, 00040 Monteporzio, Italy
e-mail: grazian@oa-roma.inaf.it
² Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218, USA
³ Department of Astronomy, University of Massachusetts, 710 North Pleasant Street, Amherst, MA 01003, USA

Received: 14 September 2010
Accepted: 18 April 2011

Abstract

Context. The study of the luminosity function (LF) of Lyman break galaxies (LBGs) at z = 7 is very important for ascertaining their role in the reionization of the Universe. These galaxies can be used also to investigate in detail the processes of formation and evolution of galactic structures in the infancy of our Universe.

Aims. In this work we plan to perform a detailed and critical analysis of the statistical and systematic errors in the z  ~  7 LF determination.

Methods. To this aim, we have assembled a large sample of candidate LBGs at z ~ 7 from different surveys, spanning a large variety of areas and depths. In particular, we have combined data from the deep (J < 27.4) and ultradeep (J < 29.2) surveys recently acquired with the new WFC3 NIR camera on HST, over the GOODS-ERS (~40 sq. arcmin) and the HUDF (~4 sq. arcmin) fields, with ground based surveys in wide and shallow areas from Hawk-I@VLT and HyperSuprimecam@Subaru. We have used public ACS images in the z band to select z-dropout galaxies, and other public data both in the blue (BVI) and in the red bands to reject possible low-redshift interlopers. We have compared our results with extensive Monte Carlo simulations to quantify the observational effects of our selection criteria as well as the effects of photometric scatter, color selections or the morphology of the candidates.

Results. We have found that the number density of faint LBGs at z ~ 7 is only marginally sensitive to the color selection adopted, but it is strongly dependent from the assumption made on the half light distributions of the simulated galaxies, used to correct the observed sample for incompleteness. The slope of the faint end of the LBGs LF has thus a rather large uncertainty, due to the unknown distribution of physical sizes of the z ~ 7 LBGs. The implications of these uncertainties have been neglected by previous works.

Conclusions. We conclude that galaxies at z ~ 7 are unable to reionize the Universe unless there is a significant evolution in the clumpiness of the IGM or in the escape fraction of ionising photons or, alternatively, there is a large population of z ~ 7 LBGs with large physical dimensions but still not detected by the present observations.