5.4. Age-metallicity Relations

In order to derive a reliable, detailed, age-metallicity relation suitable to constrain the quantitative nature of the chemical evolution history of a galaxy, including the importance of infall, gas and metal loss, burstiness, etc., one would ideally want very well-resolved temporal sampling. This is not yet possible with the currently available, sparse data.

Present-day abundances are traced well by HII regions and massive stars. Progress is being made at intermediate and old ages using planetary nebulae and field red giants. Planetary nebulae have recently been used for an independent derivation of the age-metallicity evolution of the LMC at intermediate ages (Dopita et al. 1997). While extragalactic planetary nebulae cannot normally easily be age-dated, Dopita et al. extended their spectroscopic data set for the LMC to the ultraviolet to try to directly measure the flux from the central star and also used the size information for the nebulae. They were then able to not only derive abundances but also ages using full photoionization modeling, and found their results in good agreement with stellar absorption-line spectroscopy. In less massive dIrrs, the number of planetary nebulae tends to be small, making it more difficult to derive a well-sampled age-metallicity relation. In order to derive ages for individual field stars, one has to complement the spectroscopic abundances by photometric luminosities and colors and rely on isochrone models. Considerably more accurate information can presently be obtained when using star clusters as single-age, single-metallicity populations. Disadvantages of relying on star clusters are that one often only has very few such objects in a galaxy, and that their properties are not necessarily representative of the field populations.

In all Irrs and dIrrs studied in some detail to date, there is evidence for the expected increase in chemical enrichment with younger ages. The LMC's cluster age-metallicity relation clearly demonstrates this, although there is the famous cluster age gap in the age range from ~ 4 to ~ 9 Gyr (Da Costa 2002, and references therein). The SMC has the unique advantage among all Local Group dIrrs to have formed and preserved clusters throughout its lifetime. While spectroscopic abundance determinations and improved age determinations are still missing for many clusters, the SMC shows a very well-defined age-metallicity relation with what appears to be considerable metallicity scatter at a certain given age (see Fig. 3, adopted from Da Costa 2002, and discussion in the previous subsection). While the age-metallicity relation appears to be flatter than predicted by closed-box models in LMC and SMC, Da Costa (2002) notes that the presently available data do not yet permit one to distinguish between simple closed-box or leaky-box evolution models, bursty star formation histories, or infall.

Figure 3 also shows data points for clusters and field stars in NGC6822, for which rough age estimates and spectroscopic abundance determinations are available from a variety of different sources. As noted by Chandar et al. (2000), the clusters generally seem to be more metal-poor than the SMC clusters and NGC6822's field population. It is unclear whether these differences would be reduced if all metallicities were determined using the same method, but overall the graph seems to indicate a trend of increasing metallicity with decreasing age. Undoubtedly, these kinds of studies will be refined in the coming years.