4.2. Nature of the ISOPHOT Galaxies
Determining the nature of the far IR galaxies has been a longer process than in the mid IR, mainly because of the difficulty to find the shorter wavelength counterparts in a large beam. Various techniques have been used to overcome this problem, one of the most successful being the identification using 20 cm radio data (e.g. Ciliegi et al. 1999). Another technique is the far IR multiwavelength approach (Juvela et al. 2000) that helps constraining the position and the SED; it also helps to separate the cirrus structures from the extragalactic sources. A variation is to use ISOCAM and ISOPHOT data, like the ELAIS Survey (Rowan-Robinson et al. 2004, see also Oliver in this book). Finally, the Serendipity Survey (Stickel et al. 1998, 2000), by covering large and shallow areas, allows to detect many bright objects easier to follow-up or already known.
Far IR ISO galaxies can be sorted schematically into two populations. First, the low redshift sources, typically z < 0.3 (e.g Serjeant et al. 2001; Patris et al. 2003, Kakazu et al. 2002), have moderate IR luminosities, below 1011 L, and are cold (Stickel et al. 2000). Second, sources at higher redshift, z ~ 0.3 (Patris et al., 2003) and beyond, z ~ 0.9 (Chapman et al. 2002) are more luminous, typically L > 1011 L, and appear to be cold. Serjeant et al. (2001) derived the Luminosity Function at 90 µm, and started to detect an evolution compared to the local IRAS 100 µm sample.