The optical colors of old populations are affected by the age-metallicity degeneracy (Worthey 1994): it implies that the spectrophotometric properties of an unresolved stellar population can not be distinguished from those of another population three times older and with half the metal content. As an example, we show, in Fig. 2, the comparison of the spectral energy distributions (SEDs) [calculated using the models of Worthey (1994) 1 and normalized to the K band] of two simple stellar populations with ages and abundances indicated in the labels of the figure. In the lower planel, we display the flux residual in magnitudes of the two energy distributions.
Figure 2. An example of the age-metallicity degeneracy. In the upper panel we plot the spectral energy distributions (normalized to the K band), of two simple stellar populations for the ages and chemical compositions as labeled in the figure. We have used the models of Worthey (1994) for a Salpeter initial mass function. In the lower panel we display the flux residual in magnitudes of the SEDs. Note that while the differences are small in the optical interval, in the UV the residuals reach up to one magnitude, indicating that if the age-metallicity degeneracy is present in the UV it is necessarily different from that in the optical.
Being arguably amongst the main parameters of a stellar population, there have been a number of studies aimed at finding the appropiate features(s) that unambiguously separate the effects of age from those of metallicity. In fact, Worthey (1994) conducted a detailed analysis of optical features in the form of spectrocopic indices (the so-called Lick indices) and found that akin the broad band colors, the indices, while partially diminishing the AMD, are also degenerate.
More recently, alternative spectral windows have been proposed as promising tools to lift the AMD. In particular, the rest-frame mid-UV flux and colors (Yi 2003, Dorman et al. 2003, Kaviraj et al. 2007) has been investigated on the basis that the UV properties are dominated by different stellar types at different evolutionary phases (MS) with respect to those dominating the optical (red giants). The overall results indicate that the UV indeed helps to better constrain the age of unresolved systems (as would be expected since the MS turn-off are much more sensitive to age than the red giant branch), but the determination of chemical composition was still better determined by the more sensitive optical features. The obvious path to further tackle the AMD problem was the use of mid-UV spectroscopic indices, however, there was the prevalent concept (somewhat justified, but quantitatively not investigated) that the use of synthetic UV indices at the appropiate resolution was still inadequate for the study of, for example, IUE spectra (see brief discussion in Chavez et al. 2007) and therefore investigations of absorption indices was conducted by using, for instance, Kurucz (1993) low resolution grid (e.g., Lotz et al. 2000), which appeared more reliable.
To date, the use of mid-UV synthetic indices is still in its infancy. Suffice here to mention that the citations to the relevant works where they were defined (Fanelli et al. 1990) are outnumbered by a factor of 15 (!) by the papers referencing the optical indices definition (Worthey et al. 1994), albeit they were defined roughly at the same time.
1 through the on-line application Dial a