Probing feedback in high-$z$ galaxies using extended UV luminosity functions

We fit the recently updated UV luminosity functions (LF) of high-$z$ ($1.5le zle8.0$) galaxies using our semi-analytical models of galaxy formation that take into account various feedback processes. In order to reproduce the overall redshift evolution we require the efficiency of converting gas into stars to decrease with decreasing redshift. Even for $zge 6$, our models require supernovae (SNe) feedback to reproduce the observed LF suggesting the prevalence of galactic winds that could have polluted the inter-galactic medium even at very high redshifts. The observed LF in the low luminosity end for $zu003c2.5$ shows an upward turn. In our models we reproduce this trend using passively evolving population of galaxies. Measuring stellar mass, age and metallicity of these galaxies using multi-band spectral energy distribution (SED) fitting will place strong constraints on the existence of such galaxies. While Active galactic Nuclei (AGN) feedback is essential to reproduce the LF at high luminosity end for $zu003c4$, it may not needed in the case of $zge 6$. We show that the expected turn around in the LF due to cooling criteria occurs at luminosity much lower than what has been probed with present day observations. With future deep observations, that can measure the LF more accurately, we will be able to distinguish between different modes of SNe feedback and get insights into the physical processes that drive the galaxy evolution.