Formation And Evolution Of Massive Early-Type Galaxies At High Redshift

In this thesis I present the results obtained during my PhD studies at USTC and SISSA in the field of galaxy formation, with focus on the problem of formation of Early Type Galaxies (ETGs) at high redshift. The interest to this issue has been increasing with time, since in the last decade observational evidence brought the idea that the formation and evolution of quasars is strictly related to that of ETGs. In order to tackle the galaxy formation most of the efforts has been put in developing semi-analytical and numerical codes based on the assumption that the hierarchy of galaxy Dark Matter halos traces the history also of baryons (stars and gas) through continuous merging. On the other hand, the group in SISSA and in Padua Observatory put forward the conccpt that the galaxy and quasar phenomenology at high redshift is related chiefly to physical processes (such as cooling, collapse, star formation, accretion onto the central Black Hole seed and feedback) occurring in the gas associated to the phase of fast collapse/merging of the central30-40%of the mass of DM halos, while the later phase of slow accretion is little affecting the evolution of stars and quasars.My main contributions to the line of research in SISSA have been i) the analysis of the relationship between star formation and dust in the primeval galaxies at z-2; ii) the exploration of the effects of large gas outflows triggered by stellar and quasar feedback on the structure of newly formed ETGs. I also have been collaborating in the analysis of the very exciting results obtained with the HERSCHEL-ATLAS survey.Main results are as follows:i) We construct a sample of spectroscopically con-firmed Z850-band-selected galaxies at redshift z-2from the GOODS-MUSIC sample, with a well sampled SED from the rest frame UV to near IR.which can allow us to have the analysis of their physical parameters by means of the usual spectral popula-tion synthesis technique. We have combined a local (U)LIRGs sample with a sample of normal galaxies and we have re-calibrated a relation between L8μm and LIR. We find that the usual spectral population synthesis technique with a fixed extinction law cannot reproduce the observed IR luminosity for the whole sample. We have outlined a new method that adds IR luminosity to the global SEDs and uses a generalized Calzetti law with variable R.V. With this new method we could determine in a unprecedent robust way the physical parameters,namely SFR, attenuation and age. We confirm that low luminosity galaxies harbor, on average significantly older stellar populations and are also less massive than the brighter ones. Thus we confirm that the observed downsiz-ing effect is consistent with a picture where less massive galaxies actually form first, but their star formation lasts longer, consistent with the anti-hierarchical galaxy for-mation scenario.ⅱ)We show that the observed compact size (kpc or even sub-kpc) of high redshift ETGs is expected by theoretical predictions based on the largely accepted assumption that most of the stars have been formed during dissipative collapse of cold gas.The available observational data strongly suggests that most of the size evolution occurs at z≥1, while at z<1sizes increase by no more than40%. Moreover, a large fraction of high-z passively evolving galaxies have projected stellar mass within their effective radii a factor of2larger than those of local ETGs with the same stellar mass, within the same physical radius. It is also shown that the size distribution of high redshift ETG progenitors is broader than the corresponding distribution for local ETGs. While a significant fraction of massive high redshift ETGs already exhibit sizes as large as those of their local counterparts with the same mass, for a bunch of ETGs the size has to increase by a factor of4-6to match the local half mass radius. All the above results are easily accounted for if most of the size evolution is due to a puffing up driven by the rapid expulsion of large amounts of mass. The contribution of dry minor merger to the size evolution can be up to a factor of2which is constrained by the mass function evolution.