3.6 Chemical Evolution Models
Assuming stellar yields, IMF and star formation history, models of the chemical evolution of galaxies can be constructed. Various assumptions such as instantaneous recycling and closed box (i.e. the total mass of the system is conserved and perfectly mixed at all times) allows one to construct simple analytic models (cf. Pagel 1997). In the simplest case, with a constant net yield, the ISM abundance will be a simple function of gas mass fraction (µgas = Mgas / Mtot). However, many of the effects mentioned above are likely to complicate the real picture.
Many dwarfs are believed to undergo short bursts of star formation separated by long quiescent epochs. Several additional ingredients were added to the closed-box models when it was realised that they could not account for the observed Z - µgas distribution by simply changing the number of bursts from galaxy to galaxy. Models with normal or differential winds (selectively enriched in heavy elements) have been applied to starburst galaxies (see Matteucci 1996 for a review). They seem to be successful in reproducing the observed He/H vs. O/H distribution with a number of bursts between 7 and 10 in general, differential winds and various amounts of primary and secondary nitrogen from intermediate stars. This is in agreement with Pilyugin's conclusions (Pilyugin 1992, 1993) although other possibilities have been explored. For instance, one can vary the IMF from galaxy to galaxy by changing the slope of the lower mass cut-off (Marconi et al. 1994) while Olofsson (1995) proposed that different behaviour of N/O versus O/H can be attributes to an effect of mass loss as a function of metallicity. On the other hand individual galaxies do not fit into these schemes whenever one enters into details. A few galaxies for instance tend to have large N/O for their their O/H. The galaxy IZw18 falls into this category and one is forced to assume that N is produced as a primary element in massive stars. Even this assumption does not relax the need for a strong star formation efficiency (expressed as the inverse of the time scale of star formation) and strong O-enriched winds (Kunth et al. 1995). Note that recent work, e.g. by Izotov and Thuan (1999) on low metallicity BCGs, shows a reduced scatter for N/O and C/O at low abundances, to some extent removing the need for all the mechanisms originally invoked to produce a scatter. Clearly, the way that chemical evolution in metal-poor galaxies proceeds is far from beeing a settled issue.
More sophisticated ``chemodynamical'' models attempt to describe both the star formation history, its impact on the abundances and the interaction with the ISM in a self consistent way. Lately, these models have been applied to dwarf galaxies with some success (Hensler and Rieschick 1998). Recent models of the chemical evolution versus redshift in the Universe by Cen and Ostriker (1999) predict that metallicity shows a stronger dependence on the local density (i.e. galaxy mass) than on redshift, hence objects with high and low abundances are found at all z. In the local Universe, these models predict that any region denser than the cosmic average (a minimum requirement for a galaxy) should have a metallicity of ~ 0.01 Z or higher.