Under the assumption that low-mass dwarf stars with convective envelopes have a surface chemical composition that simply reflects that of their natal clouds, one can use such stars to trace the chemical evolution of the Galaxy. The Sun is then a convenient reference, but are there nearby stars with Solar composition? Or, in other words, is the Solar abundance pattern the norm in the local thin disk?
Inspection of some of the most recent synoptic studies of nearby stars suggests that the Sun's metallicity is slightly off from average (metallicity is here equated to the iron abundance [Fe/H]1). Nordstrom et al. (2004) obtained metallicities from Stromgren photometry for nearly 14,000 F- and G-type stars within 70 pc,
1 [Fe/H] = log10(N(Fe)/N(H)) + 12, where N is the number density.
The Metal-rich Universe, eds. G. Israelian and G. Meynet. Published by Cambridge University Press. © Cambridge University Press 2008.
finding that their distribution could be approximated by a Gaussian with a mean of [Fe/H] = -0.14 and a a of 0.19 dex. Allende Prieto et al. (2004) studied spec-troscopically the stars more luminous than MV = 6.5 (M > 0.76M0) within 14.5 pc of the Sun and concluded that their metallicity distribution is centered at [Fe/H] = -0.11 and has a a of 0.18 dex. Luck & Heiter (2005) derived spectroscopic metallicities for a sample of 114 FGK stars within 15 pc similar to that analyzed by Allende Prieto et al. (2004), finding a metallicity distribution with a consistent width (a = 0.16 dex), but centered at a value slightly closer to Solar (-0.07 for the complete sample, and -0.04 when thick-disk stars are excluded). Haywood (2002) has argued that sample selection based on spectral type discriminates against high-metallicity stars, proposing a metallicity distribution (based on photometric indices) for the Solar neighborhood that is centered at the Solar value.
Inevitably, one must ask whether there is any reason to expect the local metallicity distribution to be centered at the Solar value. Chemical differences among the Sun and its neighbors may be reasonable if the age or the Galactic orbit of the Sun is somewhat off from the average for nearby stars. The age distribution or, equiva-lently, the star-formation history of the Solar neighborhood is an unsolved problem, judging from the discrepant results obtained from analyses of the Hipparcos HR diagram (Bertelli & Nasi 2001; Vergely et al. 2002) and studies of stellar activity (e.g. Rocha-Pinto et al. 2000).
What about abundance ratios? Should we expect the ratios such as C/Fe to be fairly uniform at any given iron abundance? Chemical uniformity requires the interstellar medium to be extremely well mixed, but that is precisely what local spectroscopic studies find. Reddy et al. (2003) examined this issue by analyzing high-dispersion spectra of a few hundred stars and were unable to detect any cosmic scatter. The dispersion was as small as 0.03-0.04 dex for many elements, and could be entirely accounted for by considering the uncertainties in the atmospheric parameters. The immediate implication is that the local interstellar medium is well mixed and has been well mixed for many Ga. Such a conclusion is not contradicted by the results of studies of interstellar gas toward bright stars within and beyond the local bubble (e.g. Oliveira et al. 2005).
In this situation it seems only natural to expect the Sun to have abundance ratios similar to those of other low-mass dwarfs in the Solar vicinity with similar metallicity. That is indeed the case for most elements, but there are some striking offsets. The landmark study by Edvardsson et al. (1993) found nearby FGK-type stars with Solar iron abundance to be, on average, richer than the Sun in Na, Al, and Si. Part of this trend, but not all, could be linked to biases in other stellar parameters, such as mean Galactocentric distance and age. More recent studies of nearby low-mass stars kept finding offsets between the abundance ratios of stars with Solar iron abundances and the Sun. For example, Reddy et al. (2003) found small offsets, in the same sense as Edvardsson et al. for the ratios C/Fe, N/Fe, K/Fe, S/Fe, Al/Fe, and Si/Fe (and perhaps Na/Fe), but opposite trends for Mn/Fe and V/Fe. Allende Prieto et al. (2004) also found similar patterns in their sample for O/Fe, Si/Fe, Ca/Fe, Sc/Fe, Ti/Fe, Ni/Fe, and some neutron-capture elements (Na was not studied).
The lack of consistency among results regarding the existence and size of these chemical offsets is worrisome. Local samples of stars span variable ranges in spectral type, which may be associated with different systematic errors. In order to investigate further the nature of the observed offsets, we have observed a sample of Solar analogs selected from the Hipparcos color-magnitude diagram. We describe the results below.
Was this article helpful?