<i>Small Scale CMB Anisotropies with Planck: Constraints on Primordial Magnetic Fields and the Impact of Foreground/Secondary Anisotropy Residuals</i>

Abstract

The main subject of the thesis is the study of small sale Cosmic Microwave Background (CMB) radiation anisotropies. In particular two topics are investigated. The first concerns the study of the impact of a stochastic background of primordial magnetic felds (PMF) on CMB anisotropies intemperature and polarization and the derivation of the constraints on PMF with CMB data; the second topic concerns the study of the impact of small scale foreground and secondary anisotropy residuals on cosmological parameters extracted with CMB data in the contest of the Planck mission. The first part of the thesis is devoted to the investigation of all the types of perturbations induced by PMF: scalar, vector and tensor. We present the cosmological perturbation evolution equations with PMF contributions and the original results we obtained for magnetized initial conditions and for the analytical Fourier spectra of the PMF energy momentum tensor relevant components. In order to investigate the impact of magnetized perturbations on CMB anisotropies, it has been developed an extension of the public Einstein-Boltzmann code CAMB, which computes the angular power spectrum for CMB anisotropies, where all the PMF contributions have been included. The results show that the dominant PMF contribution is given by vector perturbations on small angular scales where primary CMB is suppressed by the Silk damping. In order to derive the constraints on PMF with CMB data we have developed an extension of the Markov Chain Monte Carlo public code CosmoMC, this extension is connected with the modified version of CAMB and includes PMF parameters together with the standard ones. We show the constraints we obtained with current CMB data and the forecasts we made for Planck and the satellite COrE proposal. The results show that with current data of WMAP7, ACBAR, QUaD and BICEP the PMF amplitude is constrained to less than few nGauss, with Planck the constraints are improved by a factor two and with a mission like COrE it would be possible to constrain PMF amplitude to less than nGauss. In the thesis it is investigated also the non-Gaussian contribution of PMF to CMB anisotropies. In particular, a stochastic background of PMF has a fully non-Gaussian contribution to CMB anisotropies and induce a non-zero bispectrum. We have derived the CMB magnetized bispectrum generated by scalar magnetized perturbations on large angular scales and we have derived the constraints on PMF with current non-Gaussianity data, we show that these constraints are comparable to the ones derived with CMB angular power spectrum data. The second part of the thesis is devoted to the investigation of the im¬pact of foreground and secondary anisotropy residuals on small scales on the cosmological parameters with Planck data. In particular the contributions which are relevant for Planck frequencies and capabilities are: the residual point source contributions and the residual Sunyaev Zeldovcih effect. For what concerns the point source contribution we have considered both the Poissonian term for all unresolved sources and the clustering term for the infrared galaxies. In order to marginalize over the residual contributions in cosmological parameters extraction, we have developed hybrid theoretical/empirical parametrizations for each residual considered. Together with the spectral shape of the sig¬nals it is parametrized also their frequency dependence, this allowed the reductio n of the number of parameters necessary to describe the residuals to the minimum of three, one for each contribution. We have developed an extension of the CAMB+CosmoMC code which includes the marginalization over the considered residual contributions with the approach we have presented. The code is used to investigate the impact of the foreground and secondary anisotropy residuals on the cosmological parameters and in particular, the results show the importance of the marginalization in order to not introduce biases. In the thesis it is addressed also the topic of the frequency channel combination. In particular, we have developed an original empirical method to combine the frequency channels. We show the comparison between the results obtained with this technique and the ones obtained with the standard inverse noise variance weighting method.