A study on the multicolour evolution of red-sequence galaxy populations: insights from hydrodynamical simulations and semi-analytical models

Context. By means of our own cosmological-hydrodynamical simulation (SIM) and semi-analytical model (SAM), we studied galaxy population properties in clusters and groups, spanning over ten different bands from the ultraviolet to the near-infrared (NIR), and their evolution since redshift z = 2. Aims. We compare our results in terms of red/blue galaxy fractions and of the luminous-to-faint ratio (LFR) on the red sequence (RS) with recent observational data reaching beyond z = 1.5. Methods. Different selection criteria were tested to retrieve the galaxies that effectively belong to the RS: either by their quiescence degree measured from their specific star formation rate (sSFR; the so-called “dead sequence”), or by their position in a colour–colour plane, which is also a function of sSFR. In both cases, the colour cut and the lower limit magnitude thresholds were let to evolve with redshift so that they would follow the natural shift of the characteristic luminosity in the luminosity function (LF). Results. We find that the Butcher-Oemler effect is wavelength-dependent, with the fraction of blue galaxies increasing more steeply in optical-optical than in NIR-optical colours. Moreover, a steep trend in the blue fraction can only be reproduced when an optically fixed luminosity-selected sample is chosen, while the trend flattens when selecting samples by stellar mass or by an evolving magnitude limit. We also find that the RS-LFR behaviour, highly debated in the literature, is strongly dependent on the galaxy selection function: in particular, the very mild evolution that is recovered when using a mass-selected galaxy sample agrees with values reported for some of the highest redshift-confirmed (proto)clusters. For differences that are attributable to environments, we find that normal groups and (to a lesser extent) cluster outskirts present the highest values of both the star-forming fraction and LFR at low z, while fossil groups and cluster cores have the lowest values: this separation among groups begins after z ~ 0.5, while at earlier epochs all groups share similar star-forming properties. Conclusions. Our results support a picture where star formation is still active in SIM galaxies at redshift 2, in contrast with SAM galaxies, which have formed earlier and are already quiescent in cluster cores at that epoch. Over the whole interval considered, we also find that the more massive RS galaxies from the mass-selected sample grow their stellar mass at a higher rate than less massive ones. On the other hand, no dearth of red dwarfs is reported at z ≳ 1 from either model