We have developed an analytical method based on forward-modeling techniques to characterize the high-mass end of the red sequence (RS) galaxy population at redshift $zsim0.55$, from the DR10 BOSS CMASS spectroscopic sample, which comprises $sim600,000$ galaxies. The method, which follows an unbinned maximum likelihood approach, allows the deconvolution of the intrinsic CMASS colour-colour-magnitude distributions from photometric errors and selection effects. This procedure requires modeling the covariance matrix for the i-band magnitude, g-r colour and r-i colour using Stripe 82 multi-epoch data. Our results indicate that the error-deconvolved intrinsic RS distribution is consistent, within the photometric uncertainties, with a single point ($<0.05~{rm{mag}}$) in the colour-colour plane at fixed magnitude, for a narrow redshift slice. We have computed the high-mass end ($^{0.55}M_i lesssim -22$) of the $^{0.55}i$-band RS Luminosity Function (RS LF) in several redshift slices within the redshift range $0.52<z<0.63$. In this narrow redshift range, the evolution of the RS LF is consistent, within the uncertainties in the modeling, with a passively-evolving model with $Phi_* = (7.248 pm 0.204) times10^{-4}$ Mpc$^{-3}$ mag$^{-1}$, fading at a rate of $1.5pm0.4$ mag per unit redshift. We report RS completeness as a function of magnitude and redshift in the CMASS sample, which will facilitate a variety of galaxy-evolution and clustering studies using BOSS. Our forward-modeling method lays the foundations for future studies using other dark-energy surveys like eBOSS or DESI, which are affected by the same type of photometric blurring/selection effects.
Download