The Evolution of the Galaxy Luminosity Function: Theoretical Predictions vs. Observations

Abstract

We show that the canonical prescriptions adopted by current semi-analytic models (SAMs) of hierarchical galaxy formation lead to luminosity functions (LFs) too steep when compared with existing data (obtained from deep multicolor surveys in the HDF-N, HDF-S, NTT-DF in rest-frame B and UV (1700 Å) bands) at redshifts z > 1; the discrepancy grows to a factor ~ 8 at z ≈ 3. Such result is robust with respect to different recipes for the feedback if this is assumed to arise from the effect of Supernovae contained in the galaxies. When the depletion of satellite galaxies in common haloes due to binary aggregations is taken into account, the discrepancy of SAM predictions with the observed LFs at z > 1 and at z ~ 3 is appreciably reduced. At z ≈ 0 the effect of binary aggregations is to yield a flatter LF for −20 < M _B < −18, but with an upturn at the very faint end for M _B > −16.