Aims. We attempt to model the infrared galaxy evolution in as simple a way as possible and reproduce statistical properties such as the number counts between 15 μm and 1.1 mm, the luminosity functions, and the redshift distributions. We then use the fitted model to interpret observations from Spitzer, AKARI, BLAST, LABOCA, AzTEC, SPT, and Herschel, and make predictions for Planck and future experiments such as CCAT or SPICA. Methods. This model uses an evolution in density and luminosity of the luminosity function parametrized by broken power-laws with two breaks at redshift ∼0.9 and 2, and contains the two populations of the Lagache model: normal and starburst galaxies. We also take into account the effect of the strong lensing of high-redshift sub-millimeter galaxies. This effect is significant in the sub-mm and mm range near 50 mJy. It has 13 free parameters and eight additional calibration parameters. We fit the parameters to the IRAS, Spitzer, Herschel, and AzTEC measurements with a Monte Carlo Markov chain. Results. The model adjusted to deep counts at key wavelengths reproduces the counts from mid-infrared to millimeter wavelengths, as well as the mid-infrared luminosity functions. We discuss the contribution to both the cosmic infrared background (CIB) and the infrared luminosity density of the different populations. We also estimate the effect of the lensing on the number counts, and discuss the discovery by the South Pole Telescope (SPT) of a very bright population lying at high redshift. We predict the contribution of the lensed sources to the Planck number counts, the confusion level for future missions using a P(D) formalism, and the Universe opacity to TeV photons caused by the CIB. Material of the model (software, tables and predictions) is available online .

Context. Star-forming galaxies are highly biased tracers of the underlying dark matter density field. Their clustering can be studied through the cosmic infrared background (CIB) anisotropies. These anisotropies have been measured from 100 μm to 2 mm in the past few years. Aims. In this paper, we present a fully parametric model allowing joint analysis of these recent observations. Methods. To develop a coherent model at various wavelengths, we rely on two building blocks. The first one is a parametric model that describes the redshift evolution of the luminosity function of star-forming galaxies. It compares favorably to measured differential number counts and luminosity functions. The second is a halo model-based description of the clustering of galaxies. Starting from a fiducial model, we investigate parameter degeneracies using a Fisher analysis. We then discuss how halos of different masses and redshifts and how LIRGs and ULIRGs contribute to the CIB angular power spectra. Results. From the Fisher analysis, we conclude that we cannot constrain the parameters of the model of evolution of galaxies using clustering data only. The use of combined data of C , counts, and luminosity functions improves the constraints slightly but does not remove any degeneracies. In contrast, the measurement of the anisotropies allows us to set interesting constraints on the halo model parameters, even if some strong degeneracies remain. Using our fiducial model, we establish that the 1-halo and 2-halo terms are not sensitive to the same mass regime. We also illustrate how the 1-halo term can be misinterpreted with the Poisson noise term.Conclusions. We present a new model of the clustering of infrared galaxies. Our framework allows a coherent and joint analysis of various probes of infrared galaxies: number counts, luminosity functions, and clustering measurements; however, such a model has a few limitations, such as the parameters of the halo occupation that suffer from strong degeneracies.

An accurate measurement of the anisotropies and mean level of the cosmic infrared background at 100 µm and 160 µm