Astron. Astrophys. 326, 950-962 (1997)

5. Discussion and conclusion

The understanding of galaxy evolution requires models of spectral evolution. The model proposed here benefits from many improvements and new input data, in particular in the NIR. It becomes sufficiently reliable to be entirely published as a useful tool for many evolution studies. The wide wavelength range (220 Å- ) of the library of stellar spectra and the capacity of the algorithm to follow the evolution on very rapid (1 Myr) and very long timescales (20 Gyr) provide constraints on the weights of various stellar populations in starbursts and evolved galaxies. The inclusion of extinction, notably in E/S0, and nebular emission has been attempted in consistency with the evolution of stars. A library of standard synthetic evolving spectra has been built and shows a fair agreement with galaxies from observed optical spectra and colors in the near-UV, optical and NIR. Several samples of templates are used to define a set of reference synthetic spectra at , adopted as typical of 8 spectral types of the Hubble sequence. An age of about 13 Gyr is derived from optical-NIR colors for spheroidals and early spirals and decreases for later types.

Predicted galaxy counts agree with most recent optical and NIR bright counts as well as with new determinations of the normalization of the luminosity function, confirming that the evolution of nearby galaxies is well described by pure luminosity evolution models.

Uncertainties in the stellar evolutionary tracks are discussed in this paper, and we tried to reduce at maximum those related to stellar atmospheres. Other main uncertainties concern the effects of extinction and metallicity. The optical thickness of spirals, as well as the state of the ISM and the resulting extinction in ellipticals, are still a matter of debate. The influence of the metallicity on extinction curves, stellar spectra and evolutionary tracks could be important in metal-rich ellipticals as well as metal-poor H II galaxies. Finally, the star formation (SFR, IMF) in galaxies should be derived from a chemo-dynamical model.

In spite of these uncertainties, this model should prove useful to the astronomical community. The code sources, input data, especially our selected stellar library ², and the atlas of evolving synthetic spectra of standard galaxies (spectra, colors, equivalent widths of emission lines, numbers of SNII) with corresponding k- and e-corrections are available on an AAS CD-ROM (see Fioc & Rocca-Volmerange (1996) and Rocca-Volmerange & Fioc (1996) in Leitherer et al. (1996b)) and the latest version corresponding to this paper may be obtained at addresses given in the abstract. It may be easily used by anyone, even unfamiliar with spectral evolution questions, but also, thanks to its structure, adapted by those interested in this subject. We hope to allow fruitful comparisons with other existing models. Further improvements will also be available by ftp. Comments, questions and suggestions may be sent to pegase@iap.fr.

© European Southern Observatory (ESO) 1997

Online publication: April 8, 1998
helpdesk.link@springer.de