Next Contents Previous

2. RESOLVING THE CFIRB at 170 µm

2.1. The FIRBACK Deep Survey

FIRBACK is a survey of 4 square degrees in 3 high galactic latitude fields, chosen to have as low a HI column-density as possible, typically NHI appeq 1020 cm-2, and as much multiwavelength coverage as possible. Observations were carried out with the ESA Infrared Space Observatory, ISO, (Kessler et al. 1996) using the ISOPHOT photometer (Lemke et al. 1996) with the C200 camera and C160 broadband filter centered at lambda = 170 µm. A detailed description of the reduction, data processing, and calibration will be discussed in Lagache & Dole (2001); the analysis of the complete survey is discussed in Dole et al. (2001).

Eighty-six sources are detected above the sensitivity limit of the survey (200 mJy, 5sigma). The number of sources detected above 120 mJy (3 sigma) is 235. The first result of this survey is the high number of sources observed when compared to no, or moderate, evolution models for infrared galaxies. Extensive simulations were done to establish noise properties, incompleteness (80% at 200 mJy) and Eddington bias; the cumulative number counts were then established from this catalog. These number counts are extremely steep, N(> S) propto S-2.2, indicating very strong cosmological evolution of infrared galaxies. This is illustrated in Figure 3 (taken from Dole et al. 2000). Two models from Guiderdoni et al. (1998) are shown. It can be noticed that the semi-empirical model ``E'', which was built to explain the infrared background, fits the counts well but falls slightly short of explaining the steepness of the counts.

Number counts in a small field (0.25 square degrees) have already been published (Puget et al. 1999). This field was observed as a ``feasibility demonstration'' of the FIRBACK survey at a time when the limitations to weak source detection with ISOPHOT were not yet well understood. The counts given in Puget et al. (1999) are compatible with the counts shown in Fig. 3.

Integrating these counts for all sources detected above 120 mJy (3sigma) gives a brightness of 1.6 nW m-2 sr-1 or 8% of the CIRB at 170 µm. If we extrapolate the counts to weaker fluxes with a slope -2, we find that the background is fully accounted for by sources brighter than about 10 mJy. As we expect the counts to flatten in a progressive way we predict that the background at 170 µm is likely to be dominated by sources of a few mJy.

Figure 3

Figure 3. FIRBACK galaxy counts superposed on the Guiderdoni et al. (1998) models ``A'' (dotted line) and ``E'' (continuous line)(from Dole et al. 2000).

Considering the rather low angular resolution allowed by a 60 cm telescope at this wavelength, the identification of these sources with optically detected sources is difficult. Starburst galaxies are expected to be radio sources with a well defined ratio of infrared to radio flux. Radio surveys have been carried out in the FIRBACK fields with the VLA for the two northern fields (Ciliegi et al. 1998) and the Australian Telescope for the southern field. Furthermore, 15 µm observations have been carried out with the ISOCAM instrument aboard ISO (Oliver et al. 2000). These two sets of observations are the best tools for initiating the identification process when counterparts are available - which is the case for about 50% of our sources. Confirming the identification is best done with ground based millimeter or submillimeter observations with SCUBA, CSO or the IRAM 30 meter telescope. The last step is optical identification using the radio positions if a millimeter or submillimeter detection has confirmed the FIRBACK-radio tentative identification. This process is frustratingly slow. On the one hand, a substantial fraction of the sources are relatively nearby moderate starburst galaxies with bright optical counterparts (redshift less than 0.3, far infrared luminosities of a few 1011 Lsun). On the other hand, a small fraction of the sources have very weak or no optical counterpart and are likely to be associated with distant (redshift larger than 1) ultraluminous starburst or dust enshrouded AGNs.

2.2. The Lockman Hole Survey

A survey of comparable depth has been carried out in the Lockman hole (Kawara et al. 1998; Matsuhara et al. 2000). The number counts at 170 µm are in excellent agreement with the counts from the FIRBACK survey. It should be noticed that the Lockman hole is one of the fields with the lowest cirrus content. The agreement in the number counts confirms that, for point sources, the contamination of the extragalactic source counts by small scale cirrus structure is negligible. This is also in agreement with the conclusions of the analysis reported in the next section, which discusses the fluctuations of the background. The difference in the power spectra of the cirrus component and of the extragalactic component leads to a negligible cirrus contribution to the power spectrum at the highest spatial frequencies explored by this survey.

Next Contents Previous