Modelling the SED Evolution of Dusty Galaxies and Applications

In Chap. 2 our spectrophotometric model, GRASIL, is described in details. Since the first input to provide to GRASIL is the star formation history of the model galaxy, we describe our chemical code, that will then be used in different applications throughout the thesis. Then we present our adopted dust model, i.e. our choice for chemical and physical properties of dust grains, their optical efficiencies and size distribution. This is fixed according to the observations of extinction and emission for the diffuse ISM in our galaxy. Then, once we set the geometrical distribution of stars and dust, the radiative transfer of stellar radiation through the galaxy is solved, assuming the ISM is subdivided into a dense component, i.e. molecular clouds wherein stars are born, and a diffuse (cirrus) one, mostly associated with older stars. The preliminary inclusion of nebular emission and the radio emission are also described. Further computational details are in App. A. Chap. 3 describes our library of single stellar populations (SSP). It provides the spectra for the generations of stars making up a galaxy model. In particular, we include the spectral effects of AGB dusty envelopes already in this spectral library. The method for this inclusion as well as the comparison with observations of late-type stars are presented. We discuss the possible breaking of the age-metallicity degenaracy intrinsic in integrated spectra of stellar populations, by considering optical and near/mid-IR flux ratios. As a first application of our model, we consider in Chap. 4 local starbursts, normal spirals and ellipticals. We show the capability of our model to provide quantitative estimates of the properties of galaxies with different star formation activity and extinction by reproducing their SED. Also one high-z galaxy is addressed here, HRlO at z=l.44. In Chap. 5 we apply our model to study a sample of K-selected early-type galaxies in the HDF, for which UBVIJHK data are available. These galaxies show no signatures of dust, so that their modeling require only stellar spectral synthesis. We find the SED indicate a fairly wide range of ages (1.5 to 3 Gyrs) for the stellar populations in these objects, corresponding to a rather protracted star formation activity. In particular, we argue the absence of sample objects at z > 1.3 could be due to dust enshrouded starbursts occurring during the first few Gyrs of the galaxy lifetime. The preliminary results of the same kind of study, for a sample of K-selected latetype galaxies in the HDF are in App. B. In this case we find a strong age-extinction degeneracy affects the analysis of the optical-near-IR SED available for the sample galaxies. We stress the importance of having a full coverage of the SED for galaxies undergoing star formation in order to be able to retrieve information. In Chap. 6 we apply GRASIL as the observational interface to semi-analytical models of galaxy formation. The latter.provide the main inputs for GRASIL, i.e. star formation and metallicity enrichment histories, and most geometrical parameters - scalelengths of star distributions in the bulge and disk components. We compute a synthetic catalogue of local galaxies, distinguished in normal and stabursts, that we compare with observations: SEDs, extinction properties, luminosity functions in different bands. We find a global good agreement. Some preliminarly results are also shown for high-z galaxy catalogues, compared with number counts and the cosmic background in different bands: the only real, and stimulating problem encountered is in the submm number counts, lower by an order of magnitue as compared to observations. A possible explanation is that star formation is distributed in too many low luminosity objects.