0 A local infrared perspective to deeper ISO surveys

We present new techniques to produce IRAS 12 μm samples of galaxies and stars. We show that previous IRAS 12 μm samples are incompatible for detailed comparison with ISO surveys and review their problems. We provide a stellar infrared diagnostic diagram to distinguish galaxies from stars without using longer wavelength IRAS colour criteria and produce complete 12 μm samples of galaxies and stars. This new technique allows us to estimate the contribution of non-dusty galaxies to the IRAS 12 μm counts and produce a true local mid-infrared extragalactic sample compatible with ISO surveys. We present our initial analysis and results. 1 The importance of the local infrared picture The recent ISO mission has produced a number of deep mid-infrared extragalactic surveys [1,2,3,4] many of which are presented elsewhere in these proceedings. In order to accurately evaluate the apparent source evolution found in these surveys it is essential to have a stable and exact local infrared picture that is compatible with ISO surveys. 2 Previous work and problems There have been a number of previous IRAS 12 μm extragalactic samples produced from either the Point Source Catalog, hereafter PSC, or the Faint Source Catalog, hereafter FSC. The FSC was constructed by co-adding the individual PSC scans and is consequently deeper at 12 μm by approximately one mag; the FSC is considered complete to f12>0.2 Jy. Due to the greater depth of the FSC only those samples constructed from it will be considered here [5,6 hereafter RMS and FSXH]. In essence none of these samples are truly compatible with the deeper ISO surveys because they apply longer wavelength IRAS colour selection criteria and do not objectively classify galaxies. Some of these samples additionally suffer from inaccurate source flux estimation, no correction for the overdensity due to large scale structure and inaccurate K-correction. Due to the lack of space here these latter two points are not considered although we refer the interested reader to [6,14] for excellent coverage of these problems. 2 Alexander and Aussel 2.1 The colour selection problem Selecting objects at 12 μm without colour selection will produce an abundance of stars over galaxies due to the Jeans tail of stellar emission. Without exception every extragalactic 12 μm sample to date has had (the majority of) stars removed by applying longer wavelength IRAS colour criteria. This technique is clearly incompatible with ISO surveys where no colour criterium is applied and will cause a bias towards dusty galaxies. This also enforces that every galaxy must have a longer wavelength flux, producing incompleteness even within the selection boundaries. For example, in RMS the primary selection is f12>0.22 Jy but every galaxy must also have f60>0.5f12 or f100>f12. However due to the completeness of FSC (f60>0.2 Jy and f100>0.6 Jy) this sample cannot be complete for f12<0.4 Jy or f12<0.6Jy respectively. 2.2 The classification problem To produce accurate extragalactic luminosity functions and understand the galaxy contributions to fainter source counts it is necessary to classify galaxies in an objective way, the most common technique is with optical line ratios [7,8]. To date the only classified 12 μm sample is RMS although their classification was taken from various catalogues which differ in the definition of extragalactic type and completeness. As a comparison to this classification we have obtained line ratios from the literature for 349 of the 483 RMS galaxies with δ>0 degrees. This gives completenesses of 72%, 78% and 93% for objects f12>0.22, 0.3 and 0.5 Jy respectively. Due to the spectroscopic incompleteness at lower fluxes and the colour selection incompleteness we only consider those of f12>0.5 Jy here, see table 1; our classification follows that of [7,9]. Table 1. Extragalactic classification