Galaxy clusters: a probe to galaxy evolution and cosmology

This thesis presents some recent results concerning galaxy evolution and cosmology, based on the observation of galaxy clusters at optical wavelengths. We first introduce the main properties of galaxy clusters (Sect. 1.1) and how they can be used for cosmology within the standard cosmological model (Sect. 1.2). A large fraction of the presented results comes from the study of the DAFT/FADA galaxy cluster survey at redshifts 0.4 < z < 0.9 (Sect. 1.3). We divide our study in two parts according to the observable that is considered: galaxy luminosity or galaxy shape. The distribution of galaxy luminosities is called the galaxy luminosity function (GLF), which can be used to probe the evolution of cluster galaxies (Sect. 2.1). Computing the GLFs for a sub sample of 25 DAFT/FADA clusters, we find that faint blue star forming galaxies are quenched into red quiescent galaxies from high redshift until today. Comparing to the field shows that this transformation is more efficient in high density environments. We also study the fraction of baryons in galaxy groups and clusters (Sect. 2.2). We find that in groups, the stars contained in galaxies can reach masses of the same order as those of the intra-cluster gas, while in clusters they are usually negligible relatively to the gas. Taking both stars and gas into account we constrain the matter density parameter Ω_M. Galaxy shapes are distorted by foreground objects that bend light in their vicinity. This lensing signal can be exploited to measure the mass distribution of a foreground cluster. We review the basic theory of weak lensing and shear measurement (Sect. 3.1), and then apply it to a subsample of 16 DAFT/FADA clusters, with Subaru/SuprimeCam or CFHT/MegaCam imaging (Sect. 3.2). We estimate the masses of these clusters, and take advantage of the large fields of view of our images to detect filaments and structures in the cluster vicinity, observationally supporting the hierarchical scenario of cluster growth. Finally, we detect shear peaks in Euclid-like simulations, and use their statistics as a cosmological probe, similarly to cluster counts (Sect. 3.3). We forecast the cosmological constraints that this technique will achieve when applied to the Euclid space mission, and develop a tomographic analysis that adds information from redshifts. We conclude with a discussion of our perspectives on future studies in all the fields investigated in the present thesis. (author)