328 Highlights in modern observational cosmology
Figure 11.5.Number counts for different morphological types as derived from the HDF-N
survey (Driveret al1998). The full and broken curves are predictions from no-evolution
and passive evolution models respectively (form=1,=0).
how the stellar mass is assembled over time in a hierarchical structure formation
scenario.
11.3.4 Colour selection techniques
The measurement of the redshift of distant (sayz>1), faint galaxies is a time-
consuming task and becomes impossible at magnitudes fainter than∼25, even
with 8–10 m class telescopes equipped with modern spectrographs. As outlined
in section 11.3.3, statistical studies of the nature and evolution of galaxies require
an estimate of their SED and their redshift at magnitude selections well beyond
the spectroscopic limit. This has stimulated intensive activity over the last few
years, aimed at exploitingcolour selection techniquestoisolate and studygalaxy
populations at different redshifts. The basic idea has been to use multi-colour
imaging, in as many passbands as possible, to constrain the SED of galaxies by
detecting spectral features and measuring the continuum slope, thus estimating
the redshift.
The most successful colour selection method in recent years, which has
become known asLyman break technique, was devised to detect the ubiquitous
Lyman limit discontinuity at 912A, which is redshifted into the HST bandpasses ̊
atz&2(oratz& 2 .5 for redder ground-based filters) (e.g. Steidelet al1996).
This technique is illustrated in figure 11.7 (see the review by Dickinson 1998). A
galaxy with an unreddened UV continuum (i.e. a star-forming galaxy or an AGN)
has a nearly flat spectrum infν, and a sharp break due to photolectric absorption
of intervening neutral hygrogen (in the galaxy itself and in the intergalctic space
along the line of sight) shortward of 912A (lyman limit). The integrated effect ̊