The Evolution of the Optical and Near-Infrared Galaxy Luminosity Functions and Luminosity Densities to z ~ 2

Using Hubble Space Telescope and ground-based U through Ks photometry from the Great Observatories Origins Deep Survey, we measure the evolution of the luminosity function and luminosity density in the rest-frame optical (U, B, and R) to z ~ 2, bridging the poorly explored "redshift desert" between z ~ 1 and ~2. We also use deep near-infrared observations to measure the evolution in the rest-frame J band to z ~ 1. Compared to local measurements from the SDSS, we find a brightening of the characteristic magnitude, M*, by ~2.1, ~0.8, and ~0.7 mag between z ~ 0.1 and ~1.9, in U, B, and R, respectively. The evolution of M* in the J band is in the opposite sense, showing a dimming between redshifts z ~ 0.4 and 0.9. This is consistent with a scenario in which the mean star formation rate in galaxies was higher in the past, while the mean stellar mass was lower, in qualitative agreement with hierarchical galaxy formation models. We find that the shape of the luminosity function is strongly dependent on spectral type and that there is strong evolution with redshift in the relative contribution from the different spectral types to the luminosity density. We find good agreement with previous measurements, supporting an increase in the B-band luminosity density by a factor of ~2 between the local value and z ~ 1, and little evolution between z ~ 1 and ~2. We provide estimates of the uncertainty in our luminosity density measurements due to cosmic variance. We find good agreement in the luminosity function derived from an R-selected and a Ks-selected sample at z ~ 1, suggesting that optically selected surveys of similar depth (R ≲ 24) are not missing a significant fraction of objects at this redshift relative to a near-infrared-selected sample. We compare the rest-frame B-band luminosity functions from z = 0 to 2 with the predictions of a semianalytic hierarchical model of galaxy formation and find qualitatively good agreement. In particular, the model predicts at least as many optically luminous galaxies at z ~ 1-2 as are implied by our observations.