Astron. Astrophys. 363, 947-957 (2000)

7. Conclusion

A large sample of possible G dwarfs was examined. The sample was limited so as to be reasonably volume complete within 40 pc and south of . Masses were calculated for the stars in the sample, using models of stellar evolution by VandenBerg et al. (2000). The masses calculated from the interpolation are in general accurate within 10% (as estimated in Sect. 3.1). Binary stars, subgiants and other pollutants were removed from the sample. The sample was examined for any influence from chromospheric activity, but the sample did not appear to be affected. A new velocity correction method was developed, using the space velocities of the individual stars. This method did not produce a result that was markedly different from other methods (e.g. Sommer-Larsen 1991). After some investigations of the consequences, the sample was limited to stars with masses between 0.7 and 1.0 . Even after limiting the sample in these ways it is still significantly larger than most published until now.

The strict procedures undertaken to make sure that the sample used was free of any effects that could affect the resulting distribution (e.g. by using a mass interval and not a color interval, evolution effects were largely avoided; by removing multiple stars, their faulty photometric metallicities were not used etc.) makes it an even stronger result. It is the belief of the author, that even though these procedures have removed more than 75% of the stars from the raw sample, the end result is probably a more correct representation of the G dwarf metallicity distribution of the solar cylinder than any previously published.

When compared with prior observational work, this sample has a much smaller low metallicity tail. In Sect. 2 it is mentioned that the low metallicity tail might be over-represented in this study, which strenghens this point. It makes any reconciliation with the simple model impossible. The derived distribution can be fitted with infall models, such as the models by Pagel & Tautvaisiene (1995) and Lynden-Bell (1975). The careful selection procedure used in this work makes it even more unlikely that the G dwarf problem can be explained by selection effects.

That infall models are capable of explaining the G dwarf problem (perhaps with some delayed recycling and a relaxation of perfect mixing assumption) is a powerful result. Although it can not be ruled out that other effects are important, for now it appears to be sufficient to use an infall model to resolve the G dwarf problem. Thus the G dwarf metallicity distribution presented here is a solid constraint on numerical models taking into account the interaction of the different processes in the Galaxy throughout the lifetime of the Galaxy. This will hopefully one day lead to a better understanding of the formation and evolution of the Milky Way Galaxy.

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000
helpdesk.link@springer.de