Observational tests of the evolution of spheroidal galaxies and predictions for SIRTF/Spitzer cosmological surveys

Granato et al. (2004) have elaborated a physically grounded model exploiting the mutual feedback between starforming spheroidal galaxies and the active nuclei growing in their cores to overcome, in the framework of the hierarchical clustering scenario for galaxy formation, one of the main challenges facing such scenario, i.e. the fact that massive spheroidal galaxies appear to have formed much earlier and faster than predicted by previous hierarchical models, while the formation process was slower for less massive objects. Adopting the choice by Granato et al. (2004) for the parameters governing the history of the star formation, of chemical abundances and of the gas and dust content of galaxies, we are left with only two, rather constrained, but still adjustable, parameters, affecting the timeand mass-dependent SEDs of spheroidal galaxies. After having complemented the model with a simple phenomenological description of evolutionary properties of starburst, normal late–type galaxies, and of AGNs, we have successfully compared the model with a broad variety of observational data, including deep Kband, ISOCAM, ISOPHOT, IRAS, SCUBA, radio counts, and the corresponding redshift distributions, as well as the 1–1000 μm background spectrum. Special predictions have been made for the especially challenging counts and redshift distributions of EROs. We also present detailed predictions for the GOODS and SWIRE surveys with the Spitzer Space Telescope. We find that the GOODS deep survey at 24 μm and the SWIRE surveys at 70 and 160 μm are likely to be severely confusion limited. The GOODS surveys in the IRAC channels (3.6 to 8 μm), reaching flux limits of a few mJy, are expected to resolve most of the background at these wavelengths, to explore the full passive evolution phase of spheroidal galaxies and most of their active star-forming phase, detecting galaxies up to z ≃ 4 and beyond. A substantial number of high z star-forming spheroidal galaxies should also be detected by the 24 μm SWIRE and GOODS surveys, while the 70 and 160 μm will be particularly useful to study the evolution of such galaxies in the range 1 ∼ z ∼ 2. However, starburst galaxies at z ∼ 1–1.5 are expected to be the dominant population in MIPS channels, except, perhaps, at 160 μm.