Astron. Astrophys. 341, 641-652 (1999)

1. Introduction

With the implementation of efficient CCD cameras at 4m class-telescopes deep imaging in high galactic latitude fields has become a powerful tool to study galaxy evolution. Since the pioneering work of Tyson (1988), deep ground -based observations have been extended to different color bands (Metcalfe et al., 1995; Smail et al. 1995, Hogg et al. 1997) with image quality of 1 arcsec FWHM or better.

A new benchmark for deep survey work has been set by the Hubble Deep Field observations (Williams et al., 1996). Although the size of the HST primary mirror is 2.5m only and the CCDs in the WFPC2 instrument have relatively poor blue and UV sensitivities, the combination of very long integration time, low sky background and sub-arcsec angular size for most of the faint galaxies in the field, has led to limiting magnitudes which are more than a factor of ten fainter than the deepest ground-based surveys.

The HST observations, beside their intrinsic scientific value, acted as a very efficient catalyst for complementary photometric work in other bands, notably the IR, and for spectroscopic observations of galaxies down to =25 with the Keck telescope. The project also demonstrated the scientific advantage of dedicating a sizable chunk of observing time to the deep exploration of a single, size-limited field.

The other crucial development in this field has been the identification by the "Lyman break" technique (Steidel and Hamilton, 1993; Steidel, Pettini and Hamilton, 1995) and subsequent spectroscopic follow-up of a large number (a few hundreds known at the beginning of 1998) of galaxies with the Keck telescope (Steidel et al. 1996, Steidel et al. 1998). These results have permitted to address for the first time the issues of star formation rate and clustering at these redshifts, extending the previous spectroscopic survey work (Lilly et al. 1995; Cowie, Hu and Songaila 1995) beyond z= 1.5.

The Lyman break technique is an example (so far the one with the highest success rate) of the photometric redshift techniques. Since accurate photometry can be obtained for objects at least two magnitudes fainter than the spectroscopic limit, photometric redshifts, that is redshifts which are obtained by comparing broad band observations of galaxies with a library of observed templates or with stellar population synthesis models, are the only practicable way to extend the studies of the population of galaxies at high redshifts (4 and beyond) to luminosities below . Photometric redshifts have been derived from ground-based observations to relatively bright magnitude limits and redshifts (Koo 1985, Connolly et al. 1995). The data set of the HDF, which goes much deeper, has revived the interest in this type of work with significant results (Sawicki et al 1997, Connolly et al. 1997 and references therein to earlier work). Giallongo et al. (1998) have used ground -based observations with the SUSI CCD camera at the ESO New Technology Telescope to measure photometric redshifts for galaxies down to a limiting magnitude of .

The observations presented in this paper were obtained for the program "Faint Galaxies in an ultra-deep multicolour SUSI field", P.I. S.D'Odorico, approved for ESO Period 58 and executed in service mode also at the ESO NTT in February through April 1997 in photometric nights with seeing better than 1 arcsec. The scientific goals were the study of the photometric redshift distribution of the faint galaxies and of gravitational shearing in the field. The field of view of the SUSI CCD camera is comparable to the HDF, and the goal was to reach limiting magnitudes in the four bands which would enable photometric redshift estimates to or about 1.5 magnitude fainter than in the previous work by Giallongo et al.(1998). The final coadded calibrated frames have been made available since January 98 at the following address:
http://www.eso.org/research/sci-prog/ndf/ .

The chosen field, hereafter referred to as NTT Deep Field or NTTDF, is at 80 arcsec south of the QSO BR1202-072 (McMahon et al 1994). It is partially overlapping with the field centered on the QSO and studied in the same 4 optical bands and in K band by Giallongo et al. (1998). The high redshift QSO at the center of the field has several known metal systems in its line of sight spanning from to (Wampler et al. 1996, Lu et al. 1996) making this field quite interesting for a future comparison between the absorbers and the properties and distribution in redshifts of the field galaxies.

In this paper, we describe the observations, the reduction procedures and the objects catalogue in Sect. 2, and the galaxy counts and colors in Sect. 3. The data in the I band are compared with a deep HST observation of the same field and in the same band and more in general with the Hubble Deep Field results in Sect. 4. The selection of high redshift galaxy candidates is discussed in Sect. 5 and the conclusions are presented in Sect. 6.

In a forthcoming paper (Fontana et al. 1998, to be submitted) the data are combined with infrared observations and used to derive the photometric redshifts of the galaxies in the field.

© European Southern Observatory (ESO) 1999

Online publication: December 16, 1998
helpdesk.link@springer.de