The GALAH Survey: Non-LTE departure coefficients for large spectroscopic surveys

TL;DR

This paper presents non-LTE calculations for 13 elements across a large grid of stellar atmospheres, improving the accuracy of chemical abundance determinations in large spectroscopic surveys like GALAH.

Contribution

It introduces new non-LTE departure coefficient grids for 13 elements and integrates them into the GALAH DR3 pipeline, enabling more precise stellar abundance analysis.

Findings

Non-LTE effects can alter abundances by up to 0.7 dex.

Accounting for non-LTE reduces abundance dispersion and removes dwarf-giant discrepancies.

Non-LTE analysis impacts interpretations of Galactic chemical evolution.

Abstract

Massive sets of stellar spectroscopic observations are rapidly becoming available and these can be used to determine the chemical composition and evolution of the Galaxy with unprecedented precision. One of the major challenges in this endeavour involves constructing realistic models of stellar spectra with which to reliably determine stellar abundances. At present, large stellar surveys commonly use simplified models that assume that the stellar atmospheres are approximately in local thermodynamic equilibrium (LTE). To test and ultimately relax this assumption, we have performed non-LTE calculations for $13$ different elements (H, Li, C, N, O, Na, Mg, Al, Si, K, Ca, Mn, and Ba), using recent model atoms that have physically-motivated descriptions for the inelastic collisions with neutral hydrogen, across a grid of $3756$ 1D MARCS model atmospheres that spans $3000\leq T_{\mathrm{eff}}/\mathrm{K}\leq8000$, $-0.5\leq\log{g/\mathrm{cm\,s^{-2}}}\leq5.5$, and $-5\leq\mathrm{[Fe/H]}\leq1$. We present the grids of departure coefficients that have been implemented into the GALAH DR3 analysis pipeline in order to complement the extant non-LTE grid for iron. We also present a detailed line-by-line re-analysis of $50126$ stars from GALAH DR3. We found that relaxing LTE can change the abundances by between $-0.7\,\mathrm{dex}$ and $+0.2\,\mathrm{dex}$ for different lines and stars. Taking departures from LTE into account can reduce the dispersion in the $\mathrm{[A/Fe]}$ versus $\mathrm{[Fe/H]}$ plane by up to $0.1\,\mathrm{dex}$, and it can remove spurious differences between the dwarfs and giants by up to $0.2\,\mathrm{dex}$. The resulting abundance slopes can thus be qualitatively different in non-LTE, possibly with important implications for the chemical evolution of our Galaxy.