Next Contents Previous


Back in 2002 the Stellar Atmospheres and Populations Research Group (GrAPEs-for its designation in spanish) at the Instituto Nacional de Astrofísica, Óptica y Electrónica initiated a project aimed at providing updated stellar tools for the analysis of the UV spectra of a variety of stellar aggregates, mainly evolved ones. The overall project consists in four main steps, namely a)- the creation of a theoretical stellar database that we have called UVBLUE 3, b)- the comparison of such data base with observational stellar data, c)- the calculation of a set of synthetic SEDs of SSPs and their validation through a comparison with observations of a sample of Galactic globular clusters, d)- construction of models for dating local ellipticals and distant red galaxies. In Chavez (2009), we presented a summary of the results obtained in steps (a) and (b) and the reader is referred to that paper and the original references for a detailed description of the project (Rodriguez-Merino et al. 2005, Chavez et al. 2007). In what follows, we elaborate on the third step.

5.1. UV Spectroscopic Indices in Globular Clusters

In Chavez et al. (2009) we presented the first theoretical analysis of the UV integrated spectra of evolved SSPs (see also Maraston et al. 2009, for young populations). We focused on particular absorption lines and blends to establish, through the use of spectroscopic indices, their behavior in terms of age and chemical composition. We identified several interesting tendencies, such as the low general sensitivity of the indices to age and the remarkably distinct behavior of the indices Fe II 2332 and Fe II 2402, at super solar regimes (in fact, we propose these indices as a promising tool to establish the age in metal-rich systems). Synthetic indices were compared to IUE low resolution observations of prototypical simple populations, i.e. globular clusters, and the results were highly encouraging, indicating that theoretical SSPs might be confidently used in the analysis of more complex systems. There were two additional results that will be important in future analyses: we quantitatively showed that the presence of hot stars (e.g., blue stragglers and blue horizontal branch (B-HB) objects, which, by the way, are among the main contributors to the far-UV rising branch) can significantly dilute the mid-UV absorption indices, and that the enhancement of alpha-elements considerably modifies the overall SED of evolved populations.

Based on the results obtained so far, the project at its current stage is now focusing on the detailed analysis of local (mostly based on IUE observations) and distant evolved systems. This study will include, in a similar way as Cimatti et al. (2008), two steps: we are first conducting a UV analysis that will be later followed by a panchromatic study using, for instance, the modelling machinery developed by Panuzzo et al. (2005). We are also carrying out a detailed study of the far-UV indices and its validation process (as we did in the mid-UV) with the main goal of determining the metallicity of the objects responsible of the far-UV up turn.

5.2. The Sun, M32, and Distant Red Galaxies from a Purely UV Perspective

In Bertone & Chavez (2009), we presented a preliminary study of the mid-UV spectra of the Sun and M32 and determined, through a chinu2 analysis, their age and chemical composition. Briefly, this analysis consisted in comparing the observed SEDs of the Sun, extracted from the UARS/SUSIM archive 4, and that of M32, taken with the Faint Object Spectrograph onboard the Hubble Space Telescope (program ID=6636; PI: M. Gregg), to a set of theoretical integrated spectra calculated with the synthesis code developed by Buzzoni (1989). In the synthesis code, we have incorporated the UVBLUE stellar library and considered a red HB morphology with a Salpeter initial mass function (s = 2.35).

The results are listed in Table 1 (see Bertone & Chavez 2009, for more details). Interestingly, we obtained that for the Sun (or, equivalently, for a population whose mid-UV spectrum is dominated by stars like the Sun), the absolute chi-square minimum is found for the solar metallicity and an age of 10.1 Gyr. This result is in remarkably good agreement with the solar age at the turn-off (e.g., Jorgensen 1991, 10.5 Gyr). Similarly, for the central region of M32 we found a best fiducial age for the stellar population of the central region of M32 of 3.64 Gyr, at solar metallicity. This result is, again, in quite good agreement with the generally accepted age of 3-4 Gyr at solar (or slightly super-solar) metallicity (see, e.g., Worthey 2004, Schiavon et al. 2004).

Table 1. Age and metallicity for the SUN and M 32.

Sun M 32
Z Age (Gyr) chinu2 Age (Gyr) chinu2

0.0001 9.575 4347.0 10.000 153.9
0.001 15.000 2542.6 15.000 84.7
0.010 15.000 82.6 13.040 19.5
0.017 10.115-1.255+1.620 8.9 3.640-0.440+0.725 14.3
0.03 7.465-0.995+1.210 9.9 2.790-0.545+0.515 15.1
0.1 6.000 57.6 6.000 82.2

We have to note, however, that even though we obtained a "best value", in many instances, it is difficult to assess the significance of the difference of the minimum chinu2 values at the different metallicities. For example, the lowest chinu2 for Z = 0.01, solar, and 0.03 for M32 are quite similar. Moreover, these results indicate that the small difference in the metal content between Z = 0.01 and 0.017 produces a tremendous shift in the age of about 10 Gyr. This indicates that the age-metallicity degeneration is clearly present in UV spectra of stellar systems and, as mentioned before, operates in a different manner with respect to the optical.

A provocative exercise is to try to determine the age and chemical composition of distant ellipticals from a similar analysis, this is, solely based on their mid-UV spectrum. One of course can brandish that a panchromatic analysis (UV+optical+IR) should lead to an unambiguous determination of the parameters. Nevertheless, allow us, for now, to assume that we can not complement our UV data with optical and IR (or (sub)-mm data) as might be the case for the distant EROs for which we only have the IR fluxes (used for their selection from the surveys). Let us also assume that the UV light is indeed dominated by MS stars at the turn-off, as would be expected for systems such as globular clusters with red-clumped HBs or galaxies devoid of field counterparts of B-HB stars and their progeny. In other words, the global shape of the SEDs is not modified and the mid-UV spectroscopic indices are not diluted by the presence of hotter stars than the MS turn-off. This latter assumption might be tested with the measurement of the excess in the far-UV. Let us finally take for granted that the co-added spectrum depicted in Fig. 3 is representative of distant single objects.

Figure 4 shows the chinu2 distribution vs. age for the GMASS spectrum. For the analysis of this spectrum we have used the synthesis code of Bressan, Chiosi & Fagotto (1994) with the updates described in Chavez et al. (2009). The metallicities considered for this case range from Z = 0.0004 to 0.05 (as labelled in the figure). The reason for using this code is that it includes younger ages (< 2 Gyr) than that of Buzzoni (1989). The analysis indicates (see results in Table 2) that the smallest chinu2 is obtained for an age of 2.40 Gyr and a chemical composition of Z = 0.004. Nevertheless, analogously to the trends found for the Sun and M32, the minima are still more uneffective to segregate which of the results is more reliable.

Figure 4

Figure 4. Distribution of chinu2 vs. age at different metallicities for the composite GMASS spectrum of Cimatti et al. (2008). Different colors stand for five different metallicities, as described in the figure label.

At present, we are conducting the detailed analysis of the full sample of elliptical galaxies observed by IUE and instrisically distant red objects. The aim is not only to test other statistical methods (aside from the reduced chinu2), but to also test the validity of the different assumptions upon which the studies can de carried out.

Table 2. Best fit parameters for GMASS galaxies

Z Age (Gyr) chinu2

0.0004 11.75 2.81
0.0040 2.40 2.77
0.0080 1.45 2.96
0.0200 0.90 3.32
0.0500 0.55 3.69

3 Back.

4 Back.

Next Contents Previous