Annu. Rev. Astron. Astrophys. 1997. 35: 389-443 Copyright © 1997 by Annual Reviews. All rights reserved

2.2. Evolutionary Predictions

The no-evolution models defined above have been very useful as a baseline for comparing the predictions and observations made by different workers, but, as the datasets available probe to higher redshift, evolutionary modeling has become increasingly important. These models take as ingredients the stellar evolutionary tracks (normally for a restricted metallicity range), the initial mass function, and a gas-consumption time scale adjusted to give the present range of colors across the Hubble sequence. Assuming each galaxy evolves as an isolated system, the rest-frame SED can be predicted at a given time, and thus an evolutionary correction can be determined with respect to the no-evolution equations above, or the predictions can be incorporated in ab initio models to generate simulated data sets. The cosmological model is a crucial, but often overlooked, variable in linking time and redshift. For H₀ = 70 and = 0, the redshift corresponding to a look-back time of, say, 7 Gyr, varies from z = 1-3 depending on .

A discussion of the reliability of these evolutionary predictions and the differences between the various approaches is beyond the scope of this review. Physical difficulties arise from the degenerate effects of age and metallicity and the uncertainties of post-main sequence stellar evolution. Results for various models have been intercompared by Mazzei et al (1992), Bruzual & Charlot (1993), and, most recently for populations of the same input age and metallicity, by Charlot et al (1996). Recent efforts have concentrated on improving the stellar tracks and spectral libraries (Bruzual & Charlot 1993) and including the effects of chemical evolution (Arimoto et al 1992). An all-inclusive database of progress in this area is presented by Leitherer et al (1996).

For single burst populations presumed appropriate for early-type galaxies, Charlot et al (1996) discuss surprisingly large discrepancies in the predicted behavior of such populations at times after 1 Gyr. The differences amount to at least 0.03 mag in rest-frame B-V, 0.13 mag in rest-frame V-K, and a 25% dispersion in the V-band mass/light ratio. The large uncertainties in the predicted optical-infrared colors and visual luminosity evolution imply a significant age range (4-13 Gyr) that is permissible even for the simplest case of a passively evolving red galaxy, emphasizing the continuing need to compare these models with representative high redshift data, as well as the need to improve our knowledge of post-main sequence stellar evolution.

For populations with constant star formation, both the evolutionary corrections and the discrepancies between the available models are less. This is because the same main sequence stellar types dominate the spectra at most times and their theoretical behavior is considerably better understood. Unfortunately, precise predictions are required for both types of model galaxy (as well as the large range in between) because faint blue galaxies could be either passively evolving systems seen at an early stage, systems of constant star formation, or bursts of star formation imposed on a quiescent system. A comparison of the predicted evolutionary behavior for two of these cases is shown in Figure 3. Such uncertainties represent a formidable obstacle to detailed modeling in the ab initio approach.

Figure 3. Evolution of the rest-frame B and K band luminosity with time, according to recent models by Bruzual & Charlot (1993; solid line), Bertelli et al (1994; dotted), and Worthey (1994; short-dash); long dash represents a Bruzual & Charlot model based on alternative evolutionary stellar tracks (see Charlot et al 1996 for details). (a) Comparison for a passively evolving galaxy following an instantaneous burst with solar metallicity (see Charlot et al 1996 for details); (b) comparison for a constant star formation model (see Charlot 1996 for details).