Stars are both drivers of the evolution of our Galaxy and witnesses of its past. At their centres, during most of their life they burn hydrogen and helium via nuclear fusion reactions into heavier elements, which astronomers broadly refer to as metals. The most massive stars burn their nuclear fuel very quickly on a cosmic timescale: they live only millions of years, while our Galactic disc already exists for about 10 billion years. During their violent death they expel most of their nuclear processed products. This mixes with the pristine gas clouds, from which new stars are continuously being formed, and enrich them with metals. Stars with masses similar to our Sun act as witnesses to this history: a star's outer layer hardly mixes with the material at their cores and so the stellar atmosphere largely reflects the chemical composition of the gas from which the star was formed. Their longer lifetimes, of the order of several billions of years, make them fossils of early cosmic epochs.

The enrichment of the star-forming interstellar gas is not a sudden process, metals will accumulate with time. Moreover, the relative composition of the metals changes: very old stars, i.e. stars that have been born during the youth of our Galactic disc, carry a surplus of so called "alpha elements" compared to iron. These alpha elements are multiples of helium nuclei such as oxygen, magnesium, silicon and calcium. The ratio of alpha elements to iron serves as a natural time indicator hinting at when a star was born.

Until recently, it was thought that nearby stars could be regarded as a book from which to read the local history of star formation and enrichment. This information alone is, however, not enough to fully characterize the history of the Milky Way disc: as shown by researchers at MPA and the University of Oxford, stars migrate heavily within the disc. They do not orbit around the centre of the Galaxy with a roughly constant radius but can shift both further-in and further-out with time, making the chemical signature very difficult to interpret. Stars found in the solar neighbourhood did not necessarily get formed there; they can be immigrants from elsewhere within the disc. There is an abundance gradient within the Galaxy, the star forming gas in the dense inner disc develops more quickly and is more metal rich than in the outskirts, and so the metal content of a star is linked to where it was born.

Taking migration of stars into account greatly modifies the interpretation of the history of the Milky Way. The xenophobic view without migration needs to invoke catastrophic happenings in the Galactic history, such as the impact of a smaller galaxy or at least a period when star formation ceased almost completely. In contrast, the new models with migration explain observations with a quite calm and simple Galactic history. The two models also imply very different relationships between the age of the stars and their metal content (”metallicity“): the classical perspective needs a pronounced change in the local metal content over time to build up the observed broad distribution of metallicities in the solar vicinity. The migration models do not require this as diversity is imported with the immigrants. So, while the classical perspective demands a strong evolution of the metal content over time, i.e. a significantly lower metallicity for older stars, combined with no significant development of the width of the metallicity distribution, our new models can allow for a nearly constant average metal content over large spans of time, but predict instead an increased spread of metallicity with age.

Whether stellar migration is indeed prominent can be decided by studying local samples. The MPA scientists therefore embarked upon a revision of the stellar parameters in the Geneva-Copenhagen survey, the most comprehensive catalogue of the solar neighbourhood, containing some 16,000 stars. The metal content of stars can be derived by studying the stellar spectra and comparing them with model spectra. As this approach is very time-consuming it is not feasible when dealing with thousands of stars. Some abundance information can, however, also be retrieved simply from colours: just as tiny amounts of dye can colour wine red, the presence of metals in a stellar atmosphere alters the star's colours. Using an improved scheme to derive stellar physical parameters from colours, the researchers at MPA have shown that the average metallicity of stars in the vicinity of the Sun is higher than previously thought, thus making the Sun a more common object.

The scientists were also able to re-derive the relationship between age and metallicity of the stars. As can be seen from Figure 1, the peak of the metallicity distribution hardly moves with age of the objects, but the width of the distribution increases. This is in conflict with the classical view, while it is naturally explained in the radial migration model.

The new view of migration simplifies our picture of Galactic history. To further discriminate this scenario from the classical view and to improve our understanding of the mixing process in the disc, more information is required. As noted above, the amount of alpha element is related to the age of a star. As stars get scattered around in the Galaxy, some information about their origin still remains in the concerned populations.

The model developed at MPA and the University of Oxford is the first analytical model that can relate chemical information with the motions of stars. The explanations for observed relationships are easy and straightforward in this model, while the classical model fails to explain them. Further correlations can be predicted, such as the mean rotational velocity in the solar neighbourhood as shown in Figure 2. Old stars with a content of alpha-elements are at the top, young stars at the bottom. Superimposed on this distribution is the velocity information: stars from the inner disc rotate more slowly (blue) and are less metal rich (right) than stars from the outer disc. This prediction was confirmed by observational data for young stars (the prediction for old stars is more complex). The international research group led by MPA could thus detect a systematic error in determination of the velocities and to recalculate them.

Moreover, the MPA astronomers have been able to estimate the content of alpha elements from colours. Thus they could explore the subtle links between the detailed chemistry and the movements of stars in the Galaxy with an unprecedented large sample.

Most excitingly, the new radial migration model also provides a natural explanation for the so called thick disc, a puffed up stellar population in the disc. Rather than postulating a cosmic collision between the Milky Way and another galaxy as in the classical view, the migration model explains its existence with the immigration of relatively old stars from the inner Galactic disc. In most definitions, stars with slow rotation are predominantly ascribed to this thick disc, which is linked to the stars' origin in the inner disc, while stars with higher rotation velocity are linked to an origin from the outer, more metal poor disc (see Figure 3). The slowly rotating stars clearly have a higher average content in alpha elements, which points to their higher age. At the same time, they are also more metal rich than the fast rotating population. This again points to a problem in the classical picture without migration.

The SAGA-Team

Publications

Schönrich, R. & Binney, J., "Origin and Structure of the Galactic disc(s)", 2009, MNRAS, 396, 203

Schönrich, R. & Binney, J., "Chemical evolution with radial mixing", 2009, MNRAS, 399, 1145

Schönrich, R., "Unruhe im Ruhestandard", Sterne und Weltraum, August 2010

Casagrande, L., Schönrich, R., Asplund, M., Ramírez, I., et al. 2010, A&A, submitted