Abstract
Magnesium is one of the most important elements in stellar physics as an electron donor; in Galactic archaeology, magnesium serves to distinguish different stellar populations. However, previous studies of Mg I and Mg II lines in metal-poor benchmark stars indicate that magnesium abundances inferred from one-dimensional (1D), hydrostatic models of stellar atmospheres, both with and without the local thermodynamic equilibrium (LTE) approximation, can be problematic. Here, we present three-dimensional (3D) non-LTE calculations for magnesium in FG-type dwarfs and provide corrections for 1D-LTE abundances. 3D non-LTE corrections reduce the ionisation imbalances in the benchmark metal-poor stars HD84937 and HD140283 from −0.16 dex and −0.27 dex in 1D LTE to just −0.02 dex and −0.09 dex, respectively. We applied our abundance corrections to 1D LTE literature results for stars in the thin disc, thick disc, α-rich halo, and α-poor halo. We observed that 3D non-LTE results had a richer substructure in [Mg/Fe] − [Fe/H] in the α-poor halo, revealing two sub-populations at the metal-rich end. These two sub-populations also differ in kinematics, supporting the astro-physical origin of the separation. While the more magnesium-poor sub-population is likely to be debris from a massive accreted galaxy, Gaia-Enceladus, the other sub-population may be related to a previously identified group of stars, called Eos. The additional separation in [Mg/Fe] suggests that previous Mg abundance measurements may have been imprecise due to the 1D and LTE approximations, highlighting the importance of 3D non-LTE modelling.