The mass growth of galaxies over the last 8 Gyr

We measured the stellar mass growth of galaxies over the last 8 Gyr (z < 1.2) by deriving B and K-band luminosity functions and inferring mass evolution from these. Accuracy was achieved using improved methods for determining both total apparent magnitudes and K-corrections. The very large sample size of 335,000 galaxies in Bootes is ~10 times larger than samples in previous studies covering a similar redshift range, and this reduced cosmic variance errors to a minimum. Our accurate method of accounting for all the light coming from an individual galaxy was based on analysis of growth curves of magnitude with aperture diameter for isolated galaxies. Our measurements of galaxy luminosity used an improved method of determining K-corrections which is both simpler and more accurate than many previous methods, being based on a quadratic function of just one suitably chosen observed colour. The method relies on the extensive and accurate new set of 125 empirical galaxy template SEDs from Brown et al. (2013). Our sample was divided into red and blue subsamples using an evolving cut in restframe colour-magnitude space and we studied these separately as well as together. To quantify luminosity evolution we fitted Schechter functions to our B and K-band luminosity functions and compared the evolution of these with previous studies in the literature. Assuming published K-band stellar mass to light ratios based on optical colour, we derived stellar mass functions. We found that the stellar mass density of red galaxies increased by a factor of ~1.9 from z = 1.1 to z = 0.3, while that of blue galaxies increased only by a factor of ~1.4. These measurements are consistent with migration of blue galaxies to the red sequence as they cease to form stars. We also found that the most massive red galaxies hardly changed in mass from z = 1.1 to z = 0.3, whilst the most massive blue galaxies decreased in mass by a factor of ~0.6. This further confirmed the phenomenon of "downsizing" reported by many authors whereby the most massive star-forming galaxies ceased star formation first and moved to the red sequence. For red galaxies we compared these results with inferences from our measurements of B-band luminosity function evolution. In addition to using the new K-correction method for our B and K-band luminosity functions, we also derived K-corrections for the five Sloan Digital Sky Survey filters and make tables of the parameters for these publicly available to the astronomical community. Using a sample of SDSS galaxies we demonstrated that our method provides superior accuracy in several respects to the commonly used kcorrect method, with clearer restframe colour separation of red and blue galaxies, a much tighter red sequence, and no spurious cloud of extremely red galaxies at z~0.2 which is not present in the low redshift Universe.