Radiative feedback during inhomogeneous reionization

The reionization is a milestone in the history of the universe which is still poorly understood. We know that after recombination there was a time in which all the matter was neutral and the universe was opaque to light because of the atomic absorbtion lines; then the growth of the perturbations in the distribution of matter made matter itself begin to collapse and form structures. These structures emitted light, which is believed to be responsable of the reionization of the intergalactic medium. The aim of this thesis is to analyse some of the physical mechanisms which regulate reionization: in particular we focus on the feedback effect determined by the fact that the growth of structures is different in an ionized region than in a neutral one. There are two different mechanisms causing the baryons not to collapse efficiently onto halos in ionized regions, where the temperature is much higher than the temperature in neutral regions: the pressure opposes the gravitational attraction and the cooling processes that allow the baryons to collapse into stars are less efficient. The final effect is that halos growing in an ionized (and so hotter) region of the universe have a smaller baryonic mass than halos with the same total mass collapsing in a neutral region. This effect is expected to affect the history of reionization: it is a feedback mechanism in the sense that the more the universe is ionized, the more the accretion of baryonic matter on halos is suppressed; since the source of the ionizing radiation is the collapsing baryonic matter, further reionization is slowed down and we have a negative feedback mechanism. \newline Modelling the impact of radiative feedback during reionization presents great difficulties: it is necessary to follow the evolution of the density and of the velocity fields to the non-linear regime relevant in modelling the reionization process and the spectrum of the light emitted by the first structures (responsable of the heating of the intergalactic medium) is highly uncertain. In order to capture both the effect of radiative feedback on the single halos and the global history of the reionization process, we would have to resolve the structure of single halos in performing cosmological numerical simulations with a box side of 100 Mpc. This is clearly impossible for our computational possibilities, so we decide to adopt a hybrid approach: first we focus on single halos and then we use the information we get to calibrate a cosmological simulation. When we focus on single halos we look for an analytic expression of the critical mass scale at which a halo evolving in the presence of a UV background has lost half of its baryons with respect to the cosmic mean. After we have found the dependence of this mass scale on the collapse redshift, on the redshift when the ionizing UV background is turned on and on the intensity of the background, we perform a cosmological numerical simulation of the reionization process. To implement the effect of radiative feedback we regard only halos above the critical mass scale we have found as capable of emitting ionizing photons. In this way we can analyse properly the impact of radiative feedback on the global evolution of the neutral fraction and on the topology of the reionization process. In particular we compare our model of the radiative feedback effect with another model (often used in the literature) in which the radiative feedback is assumed to suppress instantaneously the emission of ionizing photons from halos below some critical mass: we find that the second model clearly overestimates the importance of the radiative feedback effect