Institution(s): 1. Institute of Astronomy, RAS
The individual stars observed in the dwarf galaxies orbiting the Milky Way are presumably red giants. Their chemical abundances are commonly determined under the classical LTE assumption, despite its validity is questionable for atmospheres of giant, in particular, metal-poor stars. Exactly metal-poor objects are important for understanding the early chemical enrichment processes of the host galaxy and the onset of star formation. We selected a sample of the -4 < [Fe/H] < -2 stars in the dwarf spheroidal (dSph) galaxies Sculptor, Sextans, and Fornax and the ultra-faint galaxies Bootes I and Segue I, with the high-resolution observational data available, and revised abundances of up to 12 chemical species based on the non-local thermodynamic equilibrium (NLTE) line formation. Stellar parameters taken from the literature were checked through the NLTE analysis of lines of iron observed in the two ionisation stages, Fe I and Fe II. For the Scl, Sex, and Fnx stars, with effective temperatures and surface gravities derived from the photometry and known distance (Jablonka et al. 2015; Tafelmeyer et al. 2010), the Fe I/Fe II ionisation equilibrium was found to be fulfilled, when applying a scaling factor of SH = 0.5 to the Drawinian rates of Fe+H collisions. Pronounced NLTE effects were calculated for lines of Na I and Al I resulting in up to 0.5 dex lower [Na/Fe] ratios and up to 0.65 dex higher [Al/Fe] ratios compared with the corresponding LTE values. For the six Scl stars, the scatter of data on Mg/Na is much smaller in NLTE, with the mean [Mg/Na] = 0.61 +- 0.11, than LTE, where [Mg/Na] = 0.42 +- 0.21. We computed a grid of the NLTE abundance corrections for an extensive list of the Ca I, Ti I-Ti II, and Fe I lines in the MARCS models of cool giants, 4000 K <= Teff <= 4750 K, 0.5 <= log g <= 2.5, -4 <= [M/H] <= 0.