An ingenious way of measuring the ages of stellar populations in the halo of the Milky Way will allow astronomers to obtain direct information on the timing of the Galaxy's evolution.
In a paper1 published on Nature's website today, Kalirai describes a method for estimating the ages of the progenitors of newly formed white-dwarf stars in the halo of our Galaxy, the extended region outside the plane of the Galaxy's disk (Fig. 1). This new chronometer provides a means to determine the ages of stellar populations in the halo, and will increase our knowledge of how and where the Galaxy's stars have formed and evolved.
Advances in the understanding of how stars evolve, and detailed observations of large numbers of globular clusters (tightly bound, dense stellar systems) in our Galaxy, have yielded age estimates for individual clusters of between 10 billion and 13 billion years, with claimed precisions of between 0.5 billion and 1.0 billion years2.
Current methods for measuring the ages of the halo 'field' population — that is, of stars not in clusters — rely on one of two approaches. The first involves theoretical predictions of the relationship between the age, composition, temperature and luminosity of a given star, which place the star at a unique position in an observed two-dimensional diagram of colour versus apparent brightness3. The second approach uses empirical comparisons between the locations on the diagram of collections of stars that have just exhausted the supply of hydrogen in their core (main-sequence turn-off stars) with the locations of similar stars in globular clusters.
Both of these methods typically yield age estimates that are similar to those obtained for globular clusters, but with precisions no better than 1 billion to 2 billion years4 — twice as large as for the clusters. These estimates are also subject to other systematic uncertainties due to evolutionary effects on a star's atmosphere, such as atomic diffusion. Such uncertainties may alter the expected position of main-sequence turn-off stars on the diagram5.
Alternative techniques for calculating the ages of individual halo-field stars include use of the measured abundances of radioactive species, such as uranium and thorium, in ancient stars that have low metallicities (they are deficient in elements heavier than hydrogen and helium). However, such techniques are limited to precisions no better than about 2 billion to 3 billion years6. This limitation is due to the difficulty of measuring the weak signatures of these elements in stellar spectra, and to incomplete knowledge of the production ratios of such species in nuclear reactions that involve rapid neutron capture.
In his study, Kalirai used an ingenious approach. The technique is based on estimates of the masses of stars that have just turned into white dwarfs and are found in the halo. Because more-massive dwarfs originate from younger stars, whereas less-massive dwarfs come from older stars, the author could estimate the ages of their progenitors. This step required calibration of the relationship between mass and age, which was obtained by comparing these dwarfs with newly formed white dwarfs observed in a globular cluster called Messier 4, which is estimated to be 12.5±0.5 billion years old2. These recently minted white dwarfs are the immediate descendants of stars that have exhausted all sources for nuclear fusion in their cores. Over time, they cool to invisibility as they slowly release the internal heat of the former core of their parent main-sequence (hydrogen-fusing) star.
By focusing on white dwarfs that have just formed, Kalirai's mass and age estimates obviate the need to understand the complex cooling process of a white dwarf, which depends on the composition of the white dwarf's outer layers and on changes in its structure that occur over time. His age estimate for the progenitors of a few members of the stellar population in the inner-halo field that are found relatively close to Earth is 11.4±0.7 billion years; because of their low luminosities, only white dwarfs that are nearby can be studied in detail. Improved estimates will come from adding larger numbers of inner-halo white dwarfs, and, in particular, from comparisons with globular clusters that have different ages and metallicities from Messier 4.
The importance of this deft tool for estimating the ages of white dwarfs, and the stellar populations of the Galactic halo with which they are associated, lies in its ability to resolve details of how the halo-field stars formed and evolved during the period between 10 billion and 13 billion years ago, a spread in age comparable to the precisions of previous age-determination methods4,5,6. Recent observations7,8,9and numerical models of the formation of the Galactic halo10,11,12 strongly suggest that the halo consists of at least two stellar populations: an inner-halo and an outer-halo population. These would have differing spatial distributions, kinematics and composition, and would also contain gravitationally bound debris from recent mergers of the Galaxy with smaller individual galaxies, such as the Sagittarius dwarf galaxy.
Current understanding suggests that the Galaxy was assembled from hierarchical mergers of small proto-Galactic fragments. Stars of the inner-halo population, to which the motions of Kalirai's white dwarfs suggest they belong, are thought to have originated from the collective mergers of proto-Galactic fragments of relatively high mass and metallicity11,13. These progenitor fragments were able to attain higher metallicities because they could retain an interstellar medium throughout many bursts of star formation, with each burst polluting this material with the nucleosynthetic products from which newer stars formed. Such fragments may have been similar to the more-massive surviving satellites of the Milky Way (such as the Fornax, Sculptor, Sextens and Carina dwarf spheroidal galaxies).
By contrast, stars of the outer-halo population are likely to have formed from fragments of relatively low mass and metallicity11. Such low metallicities suggest that outer-halo stars formed in these fragments early in the history of the Galaxy, and that the fragments underwent a single burst of star formation that either consumed or drove out any interstellar medium capable of forming newer stars14. These outer-halo progenitors may resemble the ultra-faint dwarf spheroidal galaxies — such as Ursa Major II and Leo IV — that have been discovered by the Sloan Digital Sky Survey15. Such spheroidal galaxies have subsequently been shown to contain relatively large numbers of stars that have extremely low metallicities. Some of the stars in these ultra-faint dwarf galaxies have metallicities close to the lowest yet found in the Galactic halo16.
The identification and analysis of white dwarfs among halo-field stars in addition to those investigated by Kalirai will, in principle, allow a distinction to be made — on the basis of their differing kinematics — between dwarfs that could be associated with inner-halo progenitors and those with outer-halo progenitors. Assuming that the interpretation of a dual halo applies, one would expect to see differences in the inferred ages of the different populations (the inner halo being somewhat younger than the outer halo), and in the derived spread of the ages of the white dwarfs associated with the two populations. Strong constraints could then be placed on the duration of star formation in the inner-halo population, and the expectation that outer-halo progenitors experienced only a single burst of star formation could be tested empirically. Ongoing and future large-scale surveys, both from the ground and in space, will supply the required samples of halo white dwarfs for such investigations. Kalirai's white-dwarf chronometer provides a valuable tool for exploring this anticipated wealth of information.
Kalirai, J. S. Nature http://dx.doi.org/10.1038/nature11062 (2012).
Dotter, A. et al. Astrophys. J. 708, 698–716 (2010).
Holmberg, J., Nordström, B. & Andersen, J. Astron. Astrophys. 501, 941–947 (2009).
Jofré, P. & Weiss, A. Astron. Astrophys. 533, A59 (2011).
Salaris, M., Groenewegen, M. A. T. & Weiss, A. Astron. Astrophys. 355, 299–307 (2000).
Frebel, A. et al. Astrophys. J. 660, L117–L120 (2007).
Carollo, D. et al. Nature 450, 1020–1025 (2007).
Carollo, D. et al. Astrophys. J. 712, 692–727 (2010).
Beers, T. C. et al. Astrophys. J. 746, 34 (2012).
Font, A. S. et al. Mon. Not. R. Astron. Soc. 416, 2802–2820 (2011).
Tissera, P. B., White, S. D. M. & Scannapieco, C. Mon. Not. R. Astron. Soc. 420, 255–270 (2012).
McCarthy, I. G. et al. Mon. Not. R. Astron. Soc. 420, 2245–2262 (2012).
Bullock, J. S. & Johnston, K. V. Astrophys. J. 635, 931–949 (2005).
Qian, Y.-Z. & Wasserburg, G. J. Proc. Natl Acad. Sci. USA (in the press); preprint available at http://arxiv.org/abs/1202.3202 (2012).
York, D. G. et al. Astron. J. 120, 1579–1587 (2000).
Frebel, A. & Norris, J. E. in Planets, Stars and Stellar Systems Vol. 5 (ed. Gilmore, G. ) (Springer, in the press); preprint available at http://arxiv.org/abs/1102.1748 (2012).