4.3. Abundance gradients in the Galaxy from H II regions

Abundance gradients in disk galaxies constitute one of the more important observational constraints for models of galaxy chemical evolution. As a matter of fact, abundance gradients were first recognized to exist in external galaxies, where radial trends of emission line ratios were noted as far back as in the fourties (Aller 1942) and were attributed to abundance gradients in the early seventies (Searle 1971, Shields 1974).

In our own galaxy, gradients are more difficult to determine, due to distance uncertainties and because many H II regions are highly obscured by dust lying close to the galactic plane. The first determination of an abundance gradient in our galaxy from H II regions was made by Peimbert, Torres-Peimbert & Rayo (1978). It is worth the effort to derive abundance gradients in the Milky Way because it is a benchline for chemical evolution of galaxies. Only in the Milky Way can one have direct access to abundance measurements from so many sources as H II regions, planetary nebulae, individual B, F, G stars etc..., which all probe different epochs in the Milky Way history. Esteban & Peimbert (1995) and Hou et al. (2000) provide excellent reviews on this topic. Table 4 presents a compilation of Galactic abundance gradients from H II regions in units of d log(X/H) / dR in kpc^-1. Column 9 indicates the spanned range of galactocentric distances in kpc. Column 10 lists the total number of objects used to derive the gradients. Note that the errors quoted for the gradients include only the scatter in the nominal values of the derived abundances about the best fit line. They do not take into account the uncertainties in the abundances and the possible errors on the galactocentric distances. Most abundances were obtained using empirical methods.

It must be noted that, even in the case of similar methods, some details in the procedures employed may lead to significantly different results. For example, the much larger oxygen gradient found by Peimbert et al. (1978) probably results from their using the temperature fluctuation scheme (with t² = .035).

A possible flattening of abundance gradients in the outer disk has been mentioned by Fich & Silkey (1991) and Vilchez & Esteban (1996) but Rudolph et al. (1997) and Deharveng et al. (2000) find no clear evidence for that.

The situation with the N/O ratio is not clear. N/O ratios determined from N⁺⁺/O⁺⁺ using far infrared (FIR) lines (Simpson et al. 1995, Afflerbach et al. 1997, see also Lester et al. 1987 and Rubin et al. 1988) are up to twice the values derived from N⁺/O⁺ using optical data. Actually, what is found is that N⁺⁺/O⁺⁺ is larger than N⁺/O⁺, so it cannot be an ionization correction factor problem. Rubin et al. (1988) suggest that the discrepancy may be due to the neglect of the recombination component of the [O II] 3727 emission. Such an explanation can indeed hold at low T_e (say below 6000 K) but is not expected to work at high T_e. Another possibility suggested by Rubin et al. (1988) is that the [OIII] 52µm and [OIII] 88µm lines are optically thick, thus increasing the derived N⁺⁺/O⁺⁺. FIR lines from N⁺⁺ and O⁺⁺ have now been observed by ISO (Peeters et al. 2002), but in their analysis Martin-Hernández et al. (2002) do not use them to derive abundance gradients. It is not clear why, since they have constructed photoionization model grids to correct for unseen ions.

The only data on a possible carbon abundance gradient comes from optical recombination lines measures in 3 objects! Obviously more work is needed in this respect.