Much observing time has been devoted in the past few years to the problem of the detection of galaxies at high redshifts, as it was anticipated that any knowledge of their early luminosity and color evolution would set important constraints on the history of structure and star formation in the universe. As the best view to date of the optical sky at faint flux levels, the HDF imaging survey has rapidly become a key testing ground for models of galaxy evolution. The field, an undistinguished portion of the northen sky at high galactic latitudes (the data from a southern deep field are being analyzed as we speak), is essentially a deep core sample of the universe, acquired with the HST in a 10-day exposure. With its depth - reaching 5- limiting AB magnitudes of roughly 27.7, 28.6, 29.0, and 28.4 in U, B, V, and I (1) - and four-filter strategy to provide constraints on the redshift and age distribution of galaxies in the image, the HDF has offered the astronomical community the opportunity to study the galaxy population in unprecedented detail .
Figure 1. Differential galaxy number counts per square degree as a function of apparent magnitude in four bandpasses from near-UV to near-IR. The sources of the data points are indicated in each panel (see  for the complete reference list). Note the decrease of the logarithmic slope d log N / dm at faint magnitudes.
There are about 3000 galaxies in the HDF, corresponding to 2 x 106 deg-2 down to the faint limit of the images. The galaxy counts are shown in Figure 1 in four bandpasses centered at roughly 300, 450, 600, and 800 nm. A compilation of existing ground-based data is also shown, together with the predictions of no-evolution models, i.e. models in which the absolute brightness, volume density, and spectra of galaxies do not change with time. In all four bands, the logarithmic slope of the galaxy number-apparent magnitude counts, log N(m) = m, flattens at faint magnitudes, e.g., from = 0.45 in the interval 21 < B < 25 to = 0.17 for 25 < B < 29. The slope of the galaxy counts is a simple cosmological probe of the early history of star formation. The flattening at faint apparent magnitudes cannot be due to the reddening of distant sources as their Lyman break gets redshifted into the blue passband, (2) since the fraction of Lyman-break galaxies at B ~ 25 is only of order 10%. Moreover, an absorption-induced loss of sources could not explain the similar flattening of the galaxy counts observed in the V and I bands. Rather, the change of slope suggests that the surface density of luminous galaxies declines beyond z ~ 1.5.
2 For galaxies with
z > 2
(z > 3.5), the H I Lyman edge shifts into the 300 (450) nm HDF
bandpass. Neutral hydrogen, which is ubiquitous both within galaxies and in
intergalactic space, strongly absorbs ultraviolet light, creating a
spectral discontinuity that can be used to identify young, high-redshift
1 To get a feeling of the depth of this survey, note that AB = 29 mag corresponds to the flux at Earth from a 100 Watt light bulb at a distance of 10 million kilometers. Back.
2 For galaxies with z > 2 (z > 3.5), the H I Lyman edge shifts into the 300 (450) nm HDF bandpass. Neutral hydrogen, which is ubiquitous both within galaxies and in intergalactic space, strongly absorbs ultraviolet light, creating a spectral discontinuity that can be used to identify young, high-redshift galaxies . Back.