5.3. PNe as probes of the chemical evolution of galaxies
5.3.1. The universal Ne/H versus O/H relation
From a compilation of PNe abundances in the Galaxy and in the Magellanic Clouds, Henry (1989) found that the Ne/H versus O/H relation for PNe is very narrow and linear in logarithm. It is also identical to the one found for H II regions (Vigroux et al. 1987). This implies that Ne and O abundances in intermediate mass stars are not significantly altered by dredge up, and therefore that oxygen and neon abundances in PNe can indeed be used to probe the interstellar abundances of oxygen over a large portion of the history of galaxies.
5.3.2. Abundance gradients from PNe in the Milky Way
Table 8 presents a compilation of Galactic abundance gradients from PNe in units of d log(X/H) / dR in kpc-1. Column 9 shows the spanned range of galactocentric distances in kpc. Column 10 gives the total number of objects used to derive the gradients. Note that, as in the case of H II regions, the quoted uncertainties in the published abundance gradients include only the scatter in the nominal values of the derived abundances. In the case of PNe, distances are not known with good accuracy, they are usually derived from statistical methods, typically within a factor of 2 or more. However, if a gradient is found with erroneous distances, this means that a gradient is indeed most likely present, since one does not expect a conspiration of errors in distances to produce a spurious gradient. On the other hand, the values of the computed gradient strongly depend on the adopted PNe distance scale, as noted by Amnuel (1993). Only PNe arising from disk population stars are suitable to determine abundance gradients in the Galactic disk. Therefore, high velocity PNe (Type III according to the classification by Peimbert 1978) and a fortiori PNe belonging to the halo (Type IV PNe) are not suitable for this purpose.
|± .034||± .019||± .034||± .035|
|± .007||± .010||± .011||± .010|
|± .006||± .007||± .006||± .006|
|± .01||± .024||± .01||± .02|
|± .003||± .026||± .014||± .010||± .012||± .019||± .013|
|± .0033||± .045||± .0199||± .027||± .021|
|± 0.003||± .026||± .016||± .064||± .047||± .021|
|± .003||± .023||± .028||± .012||± .022|
|a Martins & Viegas (2000), Type II, homogeneous rederivation of abundances from compiled intensities|
|b Maciel & Quireza (1999), Type II, abundances compiled from the literature|
|c Maciel & Koppen (1994), Type II, abundances compiled from the literature|
|d Pasquali & Perinotto (1993), Type I + II , abundances compiled from the literature|
|e Amnuel (1993), Type In (according to his classification), abondances compiled from the literature|
|f Samland & al. (1992), Type II, homogeneous observational material an automated photoionization model fitting|
|g Köppen & al. (1991), Type II, homogeneous observational material and empirical abundance derivations|
|h Faundez-Abans & Maciel (1983), Type II, abundances compiled from the literature.|
It is to be noted that, while the existence of gradients seems established, there are significant differences in the magnitudes of these gradients as found by different authors. At present, it is not possible to say how accurate these gradients are. Note that accounting for possible "temperature fluctuations" would probably steepen the derived gradients (Martins & Viegas 2000).
From the most recent results, galactic gradients found for O, Ne and S using PNe appear to be similar to the ones found from H II regions and young stars (Maciel & Quireza 1999). This suggests that abundance gradients in the Galaxy have not changed during the last 3 Gyr. N and C gradients are different between PN and H II regions, which is expected of course. Their values have been reported in Table 8 only to be complete, but the existence of N or C gradients in PNe populations would rather tell something on the stellar populations from which the PN arise. As for the C gradients, they are highly unreliable anyway.
We can compare the average O/H in PNe and in H II regions of the solar vicinity, using the gradients given in Tables 4 and 8 and adopting for simplicity that the galactocentric distance of the Sun is 8.5 kpc. We find that 12 + log O/H = 8.81 ± 0.04 using the H II regions data from Shaver et al. (1983), 8.606 ± 0.06 using those from Afflerbach et al. (1997), and 8.63 ± 0.05 using Type II PNe from the compilation of Maciel & Quireza (1999). There is therefore no compelling evidence that O/H differs between Type II PNe and H II regions in the solar vicinity. This is a further indication that ISM abundances have remained constant during the last few Gyr and that there is no significant modification of O/H in PNe due to mixing in the progenitors.
Maciel & Köppen (1994) have examined whether abundance gradients in the Galaxy steepen with time, by comparing the gradients from Type I, Type II and Type III PNe. The evidence is marginal.
The question of possible vertical abundance gradients has been investigated by Faundez-Abans & Maciel (1988), Cuisinier et al. (1996) and Köppen & Cuisinier (1997), the latter study being the most detailed. Adopting careful selection criteria on the quality of the spectra and the location of the PNe in the Galaxy in a sample of 94 PNe, the latter authors find a systematic decrease of the abundances of He, N, O, S and Ar with height above the plane. The N/O ratio also exhibits a clear decrease with height. These findings are compatible with a simple empirical model that the authors work out for the kinematical and chemical evolution of the solar neighbourhood in which the progenitor stars are supposed to be born in the galactic plane and reach greater heights due to the velocity dispersion that increases with age.