Astron. Astrophys. 364, 26-42 (2000)

2. The sample

We selected a sample of candidate high-redshift ellipticals from a collection of EROs with published optical-infrared colors. We restricted our choice to the galaxies for which deep images in the red or near-infrared photometric bands were available from the HST archive - in particular observed by WFPC2 or NICMOS; roughly speaking, for objects at these two instruments map respectively the UV and optical rest-frame spectral regions of the emitted radiation. Four of the sample galaxies are in the Hubble Deep Field South (HDFS), and were studied by Benítez et al. (1999).

All the selected objects have and/or ; such thresholds are appropriate to select candidate elliptical galaxies at about , as shown in Fig. 1, where we plot the theoretical evolution of the observer-frame colors, as a function of z, for typical stellar populations that we may expect to find in elliptical galaxies.

Fig. 1. The curves show the and expected colors for passively evolving ellipticals. The models assume an exponential decaying star formation rate (SFR) with e-folding time , solar metallicity, Salpeter IMF, and they are based on the 1997 release of the Bruzual & Charlot (1993) spectral synthesis models. The assumed cosmology is and =50 km s-1 Mpc-1. The solid lines have Gyr and z3,4,5. The dashed line has Gyr and z5. The dotted lines at and show the color thresholds that can be defined to select elliptical candidates at . Redder color thresholds allow the selection of higher-z ellipticals (e.g. or correspond to ).

The resulting set comprises 63 galaxies, but 22 of them are either hardly visible or not detected at all in the final reduced images (see the next section), and have therefore been excluded from the following analysis; these galaxies are listed in Table 2. The final sample of 41 EROs is listed instead in Table 1, together with some relevant information on each object; colors and K magnitudes from the literature are reported in Table 3. The K-band magnitudes span rather homogeneously the range between 18 and 21, whereas the typical colors are and .

Table 1. The final sample (Lanzetta et al. 1998)

Table 2. Objects too faint for surface brightness analysis

Table 3. Photometric and structural parameters

Clearly, the sample was not selected according to any fixed limit in total flux or surface brightness, but it rather comprises objects observed in different passbands and with different sensitivities ¹: as a consequence, it cannot be considerd complete at any flux level. On the other hand, the selection is based only on the availability of deep HST images, so that we do not expect any particular bias to be present; we also note that, in spite of its incompleteness, the size of this sample is unprecedented for this class of objects. Finally, since we are mainly interested in the structural characteristic of these objects, rather than in their intrinsic photometric properties, the lack of information about the redshift of most of the selected galaxies (9 spectroscopic and 6 photometric redshifts are available from the literature) does not represent a major problem. The surface brightness distribution of local elliptical galaxies exhibits little shape variation from the near ultraviolet to the near infrared so that similar passively evolving galaxies at can be easily identified with our kind of analysis and with the photometric bands available. The possible effects of the different wavelength coverage of the WFPC2 and NICMOS images are discussed in Sect. 6.3.

© European Southern Observatory (ESO) 2000

Online publication: December 15, 2000
helpdesk.link@springer.de