DEEP SPITZER 24 μm COSMOS IMAGING. I. THE EVOLUTION OF LUMINOUS DUSTY GALAXIES-CONFRONTING THE MODELS

We present the first results obtained from the identification of similar to 30,000 sources in the Spitzer/24 mu m observations of the COSMOS field at S-24 (mu m) greater than or similar to 80 mu Jy. Using accurate photometric redshifts (sigma(z) similar to 0.12 at z similar to 2 for 24 mu m sources with i(+) less than or similar to 25 mag AB) and simple extrapolations of the number counts at faint fluxes, we resolve with unprecedented detail the buildup of the mid-infrared background across cosmic ages. We find that similar to 50% and similar to 80% of the 24 mu m background intensity originate from galaxies at z less than or similar to 1 and z less than or similar to 2, respectively, supporting the scenario where highly obscured sources at very high redshifts (z greater than or similar to 2) contribute only marginally to the cosmic infrared background. Assuming flux-limited selections at optical wavelengths, we also find that the fraction of i(+)-band sources with 24 mu m detection strongly increases up to z similar to 2 as a consequence of the rapid evolution that star-forming galaxies have undergone with look-back time. Nonetheless, this rising trend shows a clear break at z similar to 1.3, probably due to k-correction effects implied by the complexity of spectral energy distributions in the mid-infrared. Finally, we compare our results with the predictions from different models of galaxy formation. We note that semianalytical formalisms currently fail to reproduce the redshift distributions observed at 24 mu m. Furthermore, the simulated galaxies at S-24 (mu m) > 80 mu Jy exhibit R-K colors much bluer than observed and the predicted K-band fluxes are systematically underestimated at z greater than or similar to 0.5. Unless these discrepancies mainly result from an incorrect treatment of extinction in the models they may reflect an underestimate of the predicted density of high-redshift massive sources with strong ongoing star formation, which would point to more fundamental processes and/or parameters (e.g., initial mass function, critical density to form stars, feedback, ...) that are still not fully controlled in the simulations. The most recent backward evolution scenarios reproduce reasonably well the flux/redshift distribution of 24 mu m sources up to z similar to 3, although none of them is able to exactly match our results at all redshifts.