Cosmological parameters determination using various observational data

This work focused on the constraints that can be made on cosmological models based on the standard cold dark matter and dark energy paradigm. With the increasing quality of different observational data, it is important to access how the astrophysical modeling, data treatment and assumptions taken can influence the final results on the cosmological parameters. Initially, I have studied the cosmological test that uses the principle of self-similarity in X-ray galaxy clusters, which translates into a supposed constant profile for the gas mass fraction at different redshifts that can be used to derive the 'correct' cosmological parameters through the angular diameter distance. The obtained results are highly dependent on the scaled radius considered to perform the test, indicating that such cosmological probe is not consistent and should not be used in precision cosmology. The second part of this thesis was dedicated to implement joint constraints using three of the main observational probes, which are the CMB anisotropies, Type Ia Supernovae, and large scale structure. The constraints were performed using the Bayesian method of Markov chain Monte Carlo (MCMC). The obtained results show a great level of precision in all cosmological parameters, typically 3 - 5%. By relaxing some of the assumptions, like considering a flat Universe or a cosmological constant, the constraints remain highly stable on most parameters. Again, possible systematics were studied when using these data. Using different descriptions (matter power spectrum or correlation function) and modelization on the LSS data can cause some the derived cosmological parameters to be shifted by a factor up to 1σ. Also, using compressed information from this data causes the error bars to be overestimated. At the same time, the possible evolution of the intrinsic luminosity of SNIa was constrained to be, at most, of the order of 0.1 mag at z = 0.5. Finally, we have constrained a model of dynamical dark energy equation of state based on the quintessence paradigm, which involves a rapid transition between values of w = p_DE/ρ_DE close to 0 at higher redshifts to a value close to -1 in the local universe. We show that the data impose a transition at redshifts z_t > 1.5 in all the studied cases, and z_t > 4 when the local value of w is equal to -0.95. (author)