Astron. Astrophys. 359, 51–64 (2000) ASTRONOMY AND ASTROPHYSICS Deep far infrared ISOPHOT survey in “Selected Area 57” I. Observations and source counts M.J.D. Linden-Vørnle 1,2 , H.U. Nørgaard-Nielsen 2 , H.E. Jørgensen 1 , L. Hansen 1 , M. Haas 3 , U. Klaas 3 , P. ´ Abrah´ am 3 , D. Lemke 3 , I. Lundgaard Rasmussen 2 , and H.W. Schnopper 4 1 Niels Bohr Institute for Astronomy, Physics and Geophysics, Astronomical Observatory, Juliane Maries Vej 30, 2100København Ø, Denmark 2 Danish Space Research Institute, Juliane Maries Vej 30, 2100København Ø, Denmark 3 Max-Planck-Institut f ¨ ur Astronomie (MPIA), K ¨ onigstuhl 17, 69117 Heidelberg, Germany 4 Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cambridge, MA 02138, USA Received 29 October 1999 / Accepted 26 April 2000 Abstract. We present here the results of a deep survey in a 0.4 deg 2 blank field in Selected Area 57 conducted with the ISOPHOT instrument aboard ESAs Infrared Space Observa- tory (ISO 1 ) at both 60 μm and 90 μm. The resulting sky maps have a spatial resolution of 15 × 23 arcsrc 2 per pixel which is much higher than the 90 × 90 arcsec 2 pixels of the IRAS All Sky Survey. We describe the main instrumental effects en- countered in our data, outline our data reduction and analysis scheme and present astrometry and photometry of the detected point sources. With a formal signal to noise ratio of 6.75 we have source detection limits of 90 mJy at 60 μm and 50 mJy at 90 μm. To these limits we find cumulated number densities of 5±3.5 deg −2 at 60 μm and 14.8±5.0 deg −2 at 90 μm. These number densities of sources are found to be lower than previ- ously reported results from ISO but the data do not allow us to discriminate between no-evolution scenarios and various evo- lutionary models. Key words: infrared: galaxies – galaxies: evolution – galaxies: starburst 1. Introduction It is widely accepted that a significant part of the evolution of galaxies is hidden from UV/optical studies due to internal absorption by dust grains, and that the absorbed radiation is re-emitted in the infrared. The IRAS All Sky Survey has revealed more than 25.000 galaxies, of which only half were already known at optical wave- lengths (Soifer et al. 1987). The vast majority of these are local late-type spirals while ellipticals and S0 galaxies were rarely Send offprint requests to: M.J.D. Linden-Vørnle (michael@astro.ku.dk) 1 Based on observations with ISO, an ESA project with instruments funded by ESA member states (especially the PI countries: France, Germany, the Netherlands, and the United Kingdom) and with the participation of ISAS and NASA. detected. Only a few very luminous infrared galaxies were de- tected at significant redshifts. A first indication of a possible evolution of the population of infrared sources at the IRAS detection limit was found by Hacking & Houck (1987) (HH87 hereafter). They exploited all the data obtained during observations of the IRAS secondary calibration source, NGC 6543, a planetary nebula close to the North Ecliptic Pole. Although this area (6.25 deg 2 ) is far from being the most ‘cirrus clean’ area in the sky, the extensive cov- erage by IRAS resulted in detection of sources (S/N > 5) down to about 50 mJy in the 60 μm band, about 10 times fainter than the detection limit in the IRAS All Sky Survey. Following the pioneering work by HH87 several other deep IRAS surveys have been published. Gregorich et al. (1995) have analysed ∼ 20 deg 2 – socalled filler fields, not including the HH87 field – with eight or more observations and designated with ‘moderate cirrus’ or ‘low cirrus’ flags in the IRAS Faint Source Survey (FSS) (Moshir et al. 1992). Gregorich et al. (1995) find the source density at f ν (60 μm) = 50 mJy to be about twice as high as that found by HH87. Bertin et al. (1997) have extracted 60 μm data from the FSS from ∼ 400 deg 2 in four separate contiguous areas selected with good coverage, low cirrus indicators and a minimum of nearby galaxies. They have performed a detailed analysis of the statistical errors in the number counts at the faint limit – the Eddington bias – and give corrected number counts down to f ν (60 μm) = 100 mJy. At this limit they find a source density about 25% lower than HH87. They suggest that the large dis- crepancy between their results and Gregorich et al. (1995) is caused by cirrus contamination in their fields. Analysing the HH87 data, Hacking et al. (1987) found a significant excess of their 60 μm number density below 100 mJy, if no evolution out to z ≈ 0.2 is assumed. They fit the number densities with simple models assuming a power law in (1 + z) either as pure density or pure luminosity evolution, but are not able to distinguish between these two types of models. This excess of sources at faint infrared fluxes have invoked a considerable effort in constructing theoretical models for galaxy