On the metallicity of the Milky Way thin disc and photometric abundance scales

The mean metallicity of the Milky Way thin disc in the solar neighbourhood is still a matter of debate, and we recently proposed an upward revision. Our star sample was drawn from a set of solar neighbourhood dwarfs with photometric metallicities. In a very recent study, it has been suggested that our metallicity calibration, based on Geneva photometry, is biased. We show here that the effect detected is not a consequence of our adopted metallicity scale, and we confirm that our findings are robust. On the contrary, the application to Strömgren photometry of the Schuster & Nissen metallicity scale is problematic. Systematic discrepancies of  ∼0.1–0.3 dex  affect the photometric metallicity determination of metal-rich stars, on the colour interval  0.22 < b − y < 0.59  , i.e. including F and G stars. For F stars, it is shown that this is a consequence of a mismatch between the standard sequence   m ₁( b − y )  of the Hyades used by Schuster & Nissen to calibrate their metallicity scale, and the system of Olsen. It means that although the calibration of Schuster & Nissen and Olsen's photometry are intrinsically correct, they are mutually incompatible for metal-rich F-type stars. For G stars, the discrepancy is most probably the continuation of the same problem, albeit worsened by the lack of spectroscopic calibrating stars. A corrected calibration is proposed that renders the calibration of Schuster & Nissen applicable to the catalogues of Olsen. We also give a simpler calibration referenced to the Hyades sequence, valid over the same colour and metallicity ranges.     

A cosmological study of the star formation history in the solar neighbourhood

We use a cosmological galactic evolutionary approach to model the Milky Way. A detailed treatment of the mass aggregation and dynamical history of the growing dark halo is included, together with a self-consistent physical treatment for the star formation processes within the growing galactic disc. This allows us to calculate the temporal evolution of star and gas surface densities at all galactic radii, in particular, the star formation history (SFH) at the solar radius. A large range of cosmological mass aggregation histories (MAHs) is capable of producing a galaxy with the present-day properties of the Milky Way. The resulting SFHs for the solar neighbourhood bracket the available observational data for this feature, the most probable MAH yielding the optimal comparison with these observations. We also find that the rotation curve for our Galaxy implies the presence of a constant density core in its dark-matter halo.     

Kinematics and history of the solar neighbourhood revisited

We use proper motions and parallaxes from the new reduction of Hipparcos data and Geneva–Copenhagen radial velocities for a complete sample of  ∼15 000  main-sequence and subgiant stars, and new Padova isochrones to constrain the kinematics and star formation history of the solar neighbourhood. We rederive the solar motion and the structure of the local velocity ellipsoids. When the principal velocity dispersions are assumed to increase with time as   t ^β  , the index β is larger for  σ _W (β _W ≈ 0.45  ) than for  σ _U (β _U ≈ 0.31)  . For the three-dimensional velocity dispersion, we obtain  β= 0.35  . We exclude saturation of disc heating after  ∼3 Gyr  as proposed by Quillen & Garnett. Saturation after  ≳4 Gyr  combined with an abrupt increase in velocity dispersion for the oldest stars cannot be excluded. For all our models, the star formation rate (SFR) is declining, being a factor of 2–7 lower now than it was at the beginning. Models in which the SFR declines exponentially favour very high disc ages between 11.5 and 13 Gyr and exclude ages below  ∼10.5 Gyr  as they yield worse fits to the number density and velocity dispersion of red stars. Models in which the SFR is the sum of two declining exponentials representing the thin and thick discs favour ages between 10.5 and 12 Gyr with a lower limit of  ∼10.0 Gyr  . Although in our models the SFR peaked surprisingly early, the mean formation time of solar-neighbourhood stars is later than in ab initio models of galaxy formation, probably on account of weaknesses in such models.     

Velocity distribution of stars in the solar neighbourhood

A two-dimensional velocity distribution in the UV plane has been obtained for stars in the solar neighbourhood, using Hipparcos astrometry for over 4000 'survey' stars with parallaxes greater than 10 mas and radial velocities found in the Hipparcos Input Catalogue. In addition to the already known grouping characteristics (field stars plus young moving groups), the velocity distribution seems to exhibit a more complex structure characterized by several longer branches running almost parallel to each other across the UV plane. By using the wavelet transform technique to analyse the distribution, the branches are visible at relatively high significance levels of 90 per cent or higher. They are roughly equidistant with a separation of about 15 km s⁻¹ for early-type stars and about 20 km s⁻¹ for late-type stars, creating an overall quasi-periodic structure which can also be detected by means of a two-dimensional Fourier transform. This branch-like velocity distribution might be caused by the Galactic spiral structure.     

Semiregular variables in the solar neighbourhood

Period–luminosity sequences have been shown to exist among the semiregular variables (SRVs) in the Magellanic Clouds, the bulge of the Milky Way galaxy and elsewhere. Using modern-period and revised Hipparcos parallax data, this paper demonstrates that they also appear among the M giant SRVs of the solar neighbourhood. Their distribution in the   K , log  P   diagram resembles that of Bulge stars more closely than those in the Magellanic Clouds. The prevalence of mass-loss among local M-type SRVs and its dependence on period and spectral subtype are also discussed.   K −[12]  , a measure of circumstellar dust emission, increases clearly with V amplitude, M giant subtype and  log  P   .     

Determination of the IMF on the basis of a recently derived SFR history in the solar neighbourhood

The initial mass function (IMF) in the solar neighbourhood is determined on the basis of a recently derived history of the star formation rate (SFR) which shows the presence of a star formation burst about 8 Gyr ago. The observed present-day mass function (PDMF) is considered, and the average past distribution of stars of a given mass is estimated. Two cases are considered, namely (i) constant SFR, and (ii) variable SFR as derived from the new metallicity distribution of G dwarfs. The resulting IMF is compared with previous determinations by Scalo and Kroupa et al., and the variation with stellar mass of the slope of the IMF is compared with reference determinations in the literature.     

Hipparcos and the age of the Galactic disc

We use the Hipparcos colour–magnitude diagram of field stars with Tycho colours to make a new minimum age estimate for the Galactic disc. The method is based on fits to the red envelope of subgiants in the Hipparcos colour–magnitude diagram with synthetic isochrones covering the range of disc metal abundance. The colours and luminosities of the isochrones as a function of abundance are checked using new techniques involving 'red-clump' stars in the giant branch region and on the main sequence using G and K dwarfs. We derive a minimum disc age of 8 Gyr, in good agreement with other methods.     

Radial mixing and the transition between the thick and thin Galactic discs

The analysis of the kinematics of solar neighbourhood stars shows that the low- and high-metallicity tails of the thin disc are populated by objects which orbital properties suggest an origin in the outer and inner Galactic disc, respectively. Signatures of radial migration are identified in various recent samples, and are shown to be responsible for the high-metallicity dispersion in the age–metallicity distribution. Most importantly, it is shown that the population of low-metallicity wanderers of the thin disc (−0.7 < [Fe/H] < −0.3 dex) is also responsible for the apparent hiatus in metallicity with the thick disc (which terminal metallicity is about −0.2 dex). It implies that the thin disc at the solar circle has started to form stars at about this same metallicity. This is also consistent with the fact that 'transition' objects, which have α-element abundance intermediate between that of the thick and thin discs, are found in the range [−0.4, −0.2] dex. Once the metal-poor thin disc stars are recognized for what they are – wanderers from the outer thin disc – the parenthood between the two discs can be identified on stars genuinely formed at the solar circle through an evolutionary sequence in [α/Fe] and [Fe/H]. Another consequence is that stars that can be considered as truly resulting of the chemical evolution at the solar circle have a metallicity restricted to about [−0.2, +0.2] dex, confirming an old idea that most chemical evolution in the Milky Way have preceded the thin disc formation.     

Kinematics and metallicity analysis for nearby F,G and K stars

A sample containing 1 026 stars of spectral types F, G, and K, mainly dwarfs, from the solar neighbourhood with available space velocities and metallicities is treated. The treatment comprises a statistical analysis of the metallicity and velocity data and calculation of galactocentric orbits. Sample stars identified as members of the galactic halo are detached from the rest of the sample based on the values of their metallicities, velocity components and galactocentric orbits. In identifying halo stars a new, kinematical, criterion is proposed. Except one, these halo stars are the metal‐poorest ones in the sample. Besides, they have very high velocities with respect to LSR. On the other hand, the separation between the thin disc and thick one is done statistically based on LSR space velocities, membership probability (Schwarzschild distribution with assumed parameters) and galactocentric orbits. In the metallicity these two groups are not much different. For each of the three subsamples the mean motion and velocity ellipsoid are calculated. The elements of the velocity ellipsoids agree well with the values found in the literature, especially for the thin disc. The fractions of the subsystems found for the present sample are: thin disc 93%, thick disc 6%, halo 1%. The sample stars established to be members of the thin disc are examined for existence of star streams. Traces of both, known and unknown, star streams are not found (© 2009 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

The metallicity distribution of G dwarfs in the solar neighbourhood and the chemical evolution of the Galaxy

Based on a new metallicity distribution for G dwarfs within 25 pc of the sun this paper discusses in detail four galactic chemical evolution models in an inhomogeneous interstellar medium. It is shown that both the so-called simple model and the collapse model deviate greatly from the observations, while the two models labelled PIE (Prompt Initial Enrichment) and PPY (Proportional Yield) not only give much better fit, but also alleviate the so-called G-dwarf problem. The indications are that the outer halo has little influence on the chemical evolution near the Sun.     

The recent star formation history of the Hipparcos solar neighbourhood

We use data from the Hipparcos catalogue to construct colour–magnitude diagrams for the solar neighbourhood, which are then treated using advanced Bayesian analysis techniques to derive the star formation rate history, SFR ( t ), of this region over the last 3 Gyr. The method we use allows the recovery of the underlying SFR ( t ) without the need of assuming any a priori structure or condition on SFR ( t ), and hence yields a highly objective result. The remarkable accuracy of the data permits the reconstruction of the local SFR ( t ) with an unprecedented time resolution of ≈50 Myr. An SFR ( t ) that has an oscillatory component of period ≈0.5 Gyr is found, superimposed on a small level of constant star formation activity. Problems arising from the non-uniform selection function of the Hipparcos satellite are discussed and treated. Detailed statistical tests are then performed on the results, which confirm the inferred SFR ( t ) to be compatible with the observed distribution of stars.     

Spectroscopic analysis of nearby lower-main-sequence stars

Abundances of O, Na, Mg, Al, Si, K, Ca, Sc, Ti, V, Cr, Fe, Ni and Ba are determined for 30 nearby lower-main-sequence stars in the Northern sky using high-resolution, high signal-to-noise ratio spectra. Our results show an equilibrium of  [Fe/H]_I  and  [Fe/H]_II  and a much smaller star-to-star scatter of the abundance ratios as a function of metallicity compared with the results of Kotoneva et al. The non-local thermodynamic equilibrium (non-LTE) corrections for oxygen are considered and found to be small  (∼−0.04 dex)  . A flat trend of [O/Fe] exists over the whole metallicity range. The non-LTE effects for some important elements are discussed, and it is found that the abundance pattern for our programme stars is very similar to that of F and G dwarfs.     

Non－LTE Analysis of the Sodium Abundance of Metal－Poor Stars in the Galactic Disk and Halo

We performed an extensive non-LTE analysis of the neutral sodiumlines of Na I 5683/5688, 5890/5896, 6154/6161, and 8183/8195 in disk/halo starsof types F-K covering a wide metallicity range (-4 [Fe/H] +0.4), using ourown data as well as data collected from the literature. For comparatively metal-rich disk stars (-1 [Fe/H] +0.4) where the weaker 6154/6161 lines are thebest abundance indicators, we confirmed [Na/Fe] ～ 0 with an "upturn" (i.e., ashallow/broad dip around -0.5 [Fe/H] 0) as already reported in previousstudies. For the metal-deficient halo stars, where the much stronger 5890/5896 or8183/8195 lines subject to considerable (negative) non-LTE corrections amountingto 0.5 dex have to be used, our analysis suggests mildly "subsolar" [Na/Fe] valuesdown to ～ -0.4 (with a somewhat large scatter of ～±0.2 dex) on the average at thetypical halo metallicity of [Fe/H] ～ -2, followed by a rise again to a near-solar ratioof [Na/Fe] ～ 0 at the very metal-poor regime [Fe/H] ～ -3 to -4. These resultsare discussed in comparison with the previous observational studies along with thetheoretical predictions from the available chemical evolution models.