Table of Contents
Fetching ...

Gaia FGK Benchmark Stars: spectral library, metallicities and abundances of $α$ and Fe-peak elements of the third version

L. Casamiquela, C. Soubiran, P. Jofré, S. Vitali, S. Blanco-Cuaresma, N. Lagarde, D. Slumstrup, J. T. Palmerio, N. Brouillet, S. Elgueta, A. Rojas-Arriagada, C. Aguilera-Gómez, I. Hernández-Araya, O. L. Creevey, U. Heiter, L. Balaguer-Núñez, R. Carrera

TL;DR

GBSv3 delivers a comprehensive, homogeneous LTE abundance study for the Gaia FGK Benchmark Stars using a large, high-quality spectral library (R ≈ 42,000, S/N > 100) across 202 stars, derived with four radiative transfer codes and MARCS atmospheres. By fixing Teff and log g to fundamental values and performing line-by-line analysis of 13 Fe-peak and α elements, the work provides robust reference abundances and a detailed error budget, including line-fitting, line-to-line dispersion, and atmospheric-parameter uncertainties. The resulting abundance patterns align with Milky Way chemical evolution expectations, showing classic α-enhancement at low [Fe/H] and nuanced Fe-peak trends, while highlighting the importance of NLTE effects and careful line selection for cool stars. The study also delivers a publicly accessible spectral library and data products, enabling precise survey calibration and cross-survey homogenization, with future NLTE extensions planned.

Abstract

The accurate determination of chemical abundances in stars plays a pivotal role in understanding stellar structure and evolution, nucleosynthesis, and the chemical enrichment history of the Milky Way. Benchmark stars with precise and accurate atmospheric parameters and abundances are indispensable for calibrating spectroscopic surveys and testing stellar atmosphere models. This study focuses on the compilation of high-quality spectra and the determination of LTE chemical abundances of iron-peak and $α$ elements for the third version of the Gaia FGK Benchmark Stars (GBSv3). We compiled spectra of the GBSv3 from public archives and complemented these with our own observations. We use fundamental atmospheric parameters from Soubiran et al. 2024 to derive the chemical abundances and perform a spectroscopic analysis using the public code iSpec. We compile a homogeneous spectral library of high-resolution (42,000) and high signal-to-noise ($>100$) normalised spectra for 202 stars: including the 192 GBSv3, 9 stars with indirect measurement of the angular diameter from previous GBS versions, and the Sun. Using four radiative transfer codes, we derive chemical abundances of 13 chemical species (Fe I, Fe II, Mg I, Si I, Ca I, Ti I, Ti II, Sc II, V I, Cr I, Mn I, Co I, Ni I). We make an in-depth study of several sources of error. The GBSv3 contributes to the legacy samples of spectroscopic reference stars with improved statistics and homogeneity. This work offers the community a homogeneous spectral library and robust reference abundances for iron-peak and $α$ elements, supported by an extensive analysis of the associated uncertainties.

Gaia FGK Benchmark Stars: spectral library, metallicities and abundances of $α$ and Fe-peak elements of the third version

TL;DR

GBSv3 delivers a comprehensive, homogeneous LTE abundance study for the Gaia FGK Benchmark Stars using a large, high-quality spectral library (R ≈ 42,000, S/N > 100) across 202 stars, derived with four radiative transfer codes and MARCS atmospheres. By fixing Teff and log g to fundamental values and performing line-by-line analysis of 13 Fe-peak and α elements, the work provides robust reference abundances and a detailed error budget, including line-fitting, line-to-line dispersion, and atmospheric-parameter uncertainties. The resulting abundance patterns align with Milky Way chemical evolution expectations, showing classic α-enhancement at low [Fe/H] and nuanced Fe-peak trends, while highlighting the importance of NLTE effects and careful line selection for cool stars. The study also delivers a publicly accessible spectral library and data products, enabling precise survey calibration and cross-survey homogenization, with future NLTE extensions planned.

Abstract

The accurate determination of chemical abundances in stars plays a pivotal role in understanding stellar structure and evolution, nucleosynthesis, and the chemical enrichment history of the Milky Way. Benchmark stars with precise and accurate atmospheric parameters and abundances are indispensable for calibrating spectroscopic surveys and testing stellar atmosphere models. This study focuses on the compilation of high-quality spectra and the determination of LTE chemical abundances of iron-peak and elements for the third version of the Gaia FGK Benchmark Stars (GBSv3). We compiled spectra of the GBSv3 from public archives and complemented these with our own observations. We use fundamental atmospheric parameters from Soubiran et al. 2024 to derive the chemical abundances and perform a spectroscopic analysis using the public code iSpec. We compile a homogeneous spectral library of high-resolution (42,000) and high signal-to-noise () normalised spectra for 202 stars: including the 192 GBSv3, 9 stars with indirect measurement of the angular diameter from previous GBS versions, and the Sun. Using four radiative transfer codes, we derive chemical abundances of 13 chemical species (Fe I, Fe II, Mg I, Si I, Ca I, Ti I, Ti II, Sc II, V I, Cr I, Mn I, Co I, Ni I). We make an in-depth study of several sources of error. The GBSv3 contributes to the legacy samples of spectroscopic reference stars with improved statistics and homogeneity. This work offers the community a homogeneous spectral library and robust reference abundances for iron-peak and elements, supported by an extensive analysis of the associated uncertainties.
Paper Structure (21 sections, 4 equations, 17 figures, 2 tables)

This paper contains 21 sections, 4 equations, 17 figures, 2 tables.

Figures (17)

  • Figure 1: Distribution of the number of combined spectra per instrument and star (top), and S/N distribution of these combined spectra (bottom).
  • Figure 2: Literature comparison (with different colours and symbols) of the mean solar abundances obtained here for the four radiative transfer codes (in different panels). Each point represents the mean abundances obtained for all solar spectra, averaging all the measured lines.
  • Figure 3: Abundance of Fe per spectrum obtained in this study compared with the one from the literature adopted for the determination of fundamental $T_{\mathrm{eff}}$ and logg in Soubiran+2024 as a function of $T_{\mathrm{eff}}$ (fundamental value). The two panels and the different colours show the values from the different radiative transfer codes. Some outliers discussed in the text are indicated (the MOOG value of HIP8837 is outside of the Y-axis limits at $\sim1.3$).
  • Figure 4: Difference in Fe abundance per instrument (computed as $\mathrm{[Fe/H]_{mean}}-\mathrm{[Fe/H]_{instr}}$) as a function of effective temperature (left) and PASTEL $\mathrm{[Fe/H]}$ (right). The colour depicts the S/N of the spectra. The average and spreads of the differences for stars above 4000 K are written in each panel.
  • Figure 5: Comparison of the abundances in this study and the values from GBSv1. The mean difference and standard deviation are indicated.
  • ...and 12 more figures