This thesis was devoted to the study of clusters and superclusters of galaxies. In particular, those that show a detectable Sunyaev-Zeldovich (SZ) effect. We aimed to fully characterize these structures and the relative contributions that their physical components (diffuse and collapsed phases) play in building up the total Sunyaev-Zeldovich signal. This characterization allows us, in turn, to achieve two goals. On the one hand, we can optimize the strategies to detect WHIM (Warm-Hot Intergalactic Medium) via its Sunyaev-Zeldovich effect. The detection of this phase is fundamental within the framework of the missing baryons problem, as numerical simulations indicate that it could allocate a large fraction of the undetected baryons in the present-day Universe. On the other hand, it also allows us to better understand the origin of an anomalous feature detected with the Very Small Array interferometer in the Cosmic Microwave Background in the direction of the Corona Borealis Supercluster: the so-called Cold Spot (CrBH).

From the theoretical point of view, we used one of the largest cosmological numerical simulations currently available: the Mare Nostrum Universe. This simulation provided the ideal environment to study the Sunyaev-Zeldovich signal individually, without the contribution of other phenomena that are observed in the Cosmic Microwave Background. We developed codes that generate synthetic Sunyaev-Zeldovich maps from the simulation. In addition, the complete simulation was analysed to search for simulated superclusters that resemble the Corona Borealis Supercluster, thus obtaining an statistical sample to test the origin of the Cold Spot; i.e., the probability that a feature like the Cold Spot could be caused by each of the components in a supercluster (WHIM, clusters of galaxies or galaxy groups).

We have quantified the two components of the thermal and kinetic Sunyaev-Zeldovich effect in the local Universe and estimated the associated anisotropy that is produced by either galaxy clusters, galaxy groups and diffuse gas. Even though the collapsed phase accounts for roughly 66% of the SZ signal, the WHIM signal is of the order of 10 ¿K and could be detected with Planck, ACT or SPT. In the thermal Sunyaev-Zeldovich maps we
obtain that the diffuse gas has a very small contribution (l < 25%) while for the kinetic component, the diffuse gas dominates at large scales (l smaller than 40), providing  ~90% of the power. The contribution of the diffuse phase is significant also in the shape of the kinetic SZ two-point correlation function, causing a non-gaussian tail that extends up to ¿=50º. These propierties of the distribution could be used as tools to trace the WHIM. Simulated superclusters similar do the Corona Borealis supercluster yield Sunyaev-Zeldovich anisotropies smaller than the measurements in the Cold Spot. The probability that the WHIM can cause the detected signal is smaller than 1%, rising up to a 3.2% when the contribution of galaxy groups is included. Thus, simulations indicate that the SZ component of the Cold Spot most-likely arises from an unknown cluster of galaxies at high redshift and aligned with the supercluster.

From the observational point of view, a multiwavelength observational strategy for detecting clusters of galaxies at high redshift (z < 1) was developed. To achieve such detection, observations in the optical and near infrared bands are required. We applied this strategy in the region of the Cold Spot, for which we obtained deep images in the Sloan filters g', r', i', z' and in J and Ks. These images were taken in two campaigns at the WHT with LIRIS and ACAM and one campaign in service mode with OSIRIS@GTC. These images were then reduced and analysed in order to search for an unknown cluster of galaxies located behind the Corona Borealis Supercluster and on the same line of sight of the spot, thus completing a previous study in the local Universe (z < 0.2), carried out with SDSS data. These analyses included a search for galaxy overdensities at different colour cuts, a study of the abundance of extremely red objects (EROs) and, after the determination of the photometric redshifts of the galaxies in the region based on their colours, a study of the galaxy number density distribution in redshift bins. Each of the three methods used indicate that there is a population of red galaxies at high redshift (z=0.72), within a three arcminute radius of the Cold Spot, that could be compatible with a cluster of galaxies with a mass of 2.1 × 10^14 solar masses, and that could cause an estimated SZ signal of -18 ¿K. In addition, OSIRIS data also seem to indicate that a second galaxy cluster candidate might exist, at a similar redshift (z=0.73) and located at a distance of the order of 3.5 arcminutes from the first one. We find that its mass is of the order of 3.7 × 10^14 solar masses and it could yield a therma SZ signal of -34 ¿K.

If both cluster candidates were spectroscopically confirmed, we could be looking at a high redshift supercluster of galaxies aligned (but not conected) with the Corona Borealis supercluster, and that could account for the observed SZ signal of the Cold Spot (of the order of -42 ¿K).