4.1. Source Counts and Cosmic IR Background

Source counts at 170 µm (e.g. Kawara et al., 1998; Puget et al., 1999; Dole et al., 2001) exhibit a steep slope of = 3.3 ± 0.6 between 180 and 500 mJy and, like in the mid IR range, show sources in excess by a factor of 10 compared with no evolution scenario. The brightness fluctuations in the Lockman Hole were used by Matsuhara et al (2000) to constrain galaxy number counts down to 35 mJy at 90 µm and 60 mJy at 170 µm , confirming the existence of a strong evolution down to these flux densities. Using a new data reduction method, Rodighiero & Franceschini (2004) extended the previous works of Kawara et al. (1998), Efstathiou et al. (2000) and Linden-Voernle et al. (2000) down to lower flux densities (30 mJy at 90 µm ) and found a clear excess of faint objects with respect to no evolution (see Fig. 4a). However, the resolved sources account for less than 10% of the Cosmic Infrared Background at 170 µm , which is expected to be resolved into sources in the 1 to 10 mJy range.

Figure 4. (a) Differential 90 µm counts dN/dS normalized to the Euclidean law (N S-2.5) (extracted from Rodighiero & Franceschini 2004). Results compared with those from the preliminary analysis of the ISOPHOT ELAIS survey (Efsthatiou et al. 2000, open circles) and with those from Kawara et al. (2004, filled triangles). The long-dashed line shows the expected contribution of non-evolving spirals as in the model of Franceschini et al. (2001). (a) Fluctuations of the CIB in a power spectrum analysis of the FIRBACK/ELAIS N2 field at 170 microns by Puget & Lagache (2000). Observed power spectrum: diamond; straight continuous line: the best fit cirrus power spectrum; dash line: cirrus power spectrum deduced from Miville-Deschênes et al. (2002); continuous curve: detector noise.

Sources below the detection limit of a survey create fluctuations. If the detection limit does not allow to resolve the sources dominating the CIB intensity, which is the case in the far IR with ISO, characterizing these fluctuations can constrain the spatial correlations of these unresolved sources of cosmological significance. An example of the modeled redshift distribution of the unresolved sources at 170 µm can be found in Fig. 12 of Lagache et al. (2003); the sources dominating the CIB fluctuations have a redshift distribution peaking at z ~ 0.9. After the pioneering work of Herbstmeierer et al. (1998) with ISOPHOT, Lagache & Puget (2000) discovered them at 170 µm in the FIRBACK data, followed by other works at 170 and 90 µm (Matsuhara et al. 2000; Puget & Lagache, 2000; Kiss et al. 2001). Fig. 4b shows the CIB fluctuations in the FN2 field by Puget & Lagache, (2000), at wavenumbers 0.07 < k < 0.4 arcmin^-1.