The X-ray and UV energy density absorbed by the torus of type II AGN is reemitted at longer wavelengths, most likely in the far-IR. It is thus plausible that obscured AGN may significantly contribute to the extragalactic background light in this band. In principle, such a calculation could be performed following the guidelines discussed in Section 3 for the synthesis of the XRB spectrum. In practice, such an approach is unfeasible, mainly because the average AGN spectral shape at long wavelengths is still rather uncertain, both from an observational and from a theoretical point of view.
In order to overcome these difficulties, a simple integral argument can been employed to compute the AGN contribution to the far-IR background. The reprocessed luminosity is estimated from the hard X-ray luminosity using the XRB intensity as an upper limit of the total energy output (Fabian & Iwasawa 1999). If one further assumes reasonable values for the effective temperature of the reprocessing dust and the cosmological evolution of the luminosity function, as determined from X-ray surveys (Almaini, Lawrence, & Boyle 1999; Brusa, Comastri, & Vignali 2001; Risaliti, Elvis, & Gilli 2002), it is also possible to estimate the spectral shape of the long wavelength background due to X-ray emitting AGN.
The different approaches come to a similar result: obscured AGN contribute from a few percent up to a maximum of 10 - 15% at wavelengths longer than about 100 µm, being possibly higher in the mid-IR (15 - 60 µm; Risaliti et al. 2002). These findings are in relatively good agreement with observations in the infrared and submillimeter bands. The cross-correlation of ISO and XMM-Newton sources in the Lockman Hole (Fadda et al. 2002) indicates an AGN contribution of about 15-20% at 15 µm (see also Alexander et al. 2002), which is similar to that estimated by Matute et al. (2002) from the optical identifications of the ELAIS field. In the submillimeter, Barger et al. (2001c) found that the ensemble of X-ray sources in the 1 Ms exposure of the CDF-N with SCUBA observations contribute about 15% of the extragalactic light at 850 µm. According to the deepest investigation so far carried out using the 2 Ms CDF-N data (Alexander et al. 2003), many SCUBA sources are rather faint at high energies, suggesting that most of the submillimeter background is due to stellar processes.
An attempt to estimate the contribution of Compton thick sources to the long wavelength background is discussed by Brusa et al. (2001). Assuming the XRB model of Comastri et al. (1995) and the observed correlations between the hard X-ray and far-IR luminosities of bright nearby AGN as a function of the absorbing column density, these authors concluded that, although not energetically dominant, the most important contribution to the far-IR background comes from Compton thick sources (Fig. 10).
Figure 10. Model predicted AGN contribution at far-IR and X-ray wavelengths (solid thick line). Dashed line represents contribution of Compton thin sources, while dotted line that of Compton thick sources. Data points are from a compilation of measurements of the extragalactic background intensity from Fixsen et al. (1998; submillimeter data and best fit curve - thin black curve), Lagache et al. (2000; far-IR data), Dwek & Arendt (1998; DIRBE mid-IR data), Pozzetti et al. (1998; optical and near-IR data) and Marshall et al. (1980; X-ray data).
If anything, heavily obscured Compton thick sources appear to be the most favored class of AGN to be detected in the far-IR/submillimeter bands. The issue of the nature of the far-IR background will greatly benefit from upcoming Spitzer Space Telescope observations.