Annu. Rev. Astron. Astrophys. 2000. 38: 667-715 Copyright © 2000 by Annual Reviews. All rights reserved

4.2. Galaxy counts

One of the benefits derived from the ready availability of the reduced HDF data is that a variety of techniques have been used to catalog the sources [Williams et al. 1996, Lanzetta et al. 1996, Colley et al. 1996, Sawicki et al. 1997, Dell'Antonio & Tyson 1996, Couch 1996, Metcalfe et al. 1996, Madau & Pozzetti 2000, Chen et al. 1999, Fontana et al. 1999]. Different algorithms have been used to construct the catalogs, but all at some level rely on smoothing the image and searching for objects above a surface brightness threshold set by the background noise. [Ferguson 1998b] compared a few of the available catalogs and found reasonable agreement among them for sources brighter than V₆₀₆ = 28, with systematic differences in magnitude scales of less than 0.3 (nevertheless, there are systematic differences at this level). The different catalogs apply different algorithms for splitting and merging objects with overlapping isophotes. These differences, together with the different schemes for assigning magnitudes to galaxies, result in overall differences in galaxy counts. At I₈₁₄ = 26 the galaxy counts in the catalogs considered by [Ferguson 1998b] all agree to within 25%, whereas at I₈₁₄ = 28 there is a factor of 1.7 difference between them. This highlights the fact that galaxy counting is not a precise science.

Figure 4 shows the counts in four photometric bands derived from the HDF and from ground-based observations. As a fiducial comparison, a no-evolution model is shown for a spatially flat model with _M = 0.3 and = 0.7. In all bands the number-magnitude relation at HDF depths is significantly flatter than N(m) m^0.4, but at blue wavelengths it exceeds the purely geometrical (no-evolution) predictions by roughly a factor of 3. Interpretation of the counts is addressed in more detail in Section 5.1.

Figure 4. Galaxy counts from the Hubble deep fields (HDF) and other surveys. The HDF galaxy counts (solid symbols) use isophotal magnitudes and have not been corrected for incompleteness. These corrections will tend to steepen the counts at the faint end, but in a model-dependent way. For the K-band, a color correction of -0.4 mag has been applied to the NICMOS F160W band magnitudes. The HDF-N counts from Thomson et al (1999) (filled squares) and the HDF-S counts from Fruchter et al. 2000 (filled circles) are filled circles are shown. For the U, B and I bands, the HDF counts are the average of HDF-N (Williams et al. 1996) and HDF-S (Casertano et al. 2000) with no color corrections. The groundbased counts (open symbols) are from Mcleod & Rieke (1995), Gardner et al. (1993; 1996), Postman et al. (1998), Lilly Cowie & Gardner (1991), Huang et al. (1997), Minezaki et al. (1998), Bershady (1998), Moustakas et al. (1997), and Djorgovski et al. (1995). (Smooth curves) A no-evolution model with M, , tot = 0.3, 0.7, 0.1 based on the luminosity functions, spectral-energy distributions, and morphological type mix of the "standard NE" model of Ferguson & McGaugh (1995).