Using non-solar-scaled opacities to derive stellar parame ters Toward high-precision parameters and abundances

C. SAFFE; M. FLORES; P. MIQUELARENA; F. M. LÓPEZ; M. JAQUE ARANCIBIA; A. COLLADO; E. JOFRÉ; R. PETRUCCI

ASTRONOMY AND ASTROPHYSICS

EDP SCIENCES S A

Lugar: Paris; Año: 2018

0004-6361

P { margin-bottom: 0.08in; }Aims.In an effort trying to improve spectroscopic methods of stellarparameters determination, we implemented non-solar-scaled opacitiesin a simultaneous derivation of fundamental parameters andabundances. We want to compare the results with the usual solar-scaledmethod using a sample of solar-type and evolved stars.Methods.We carried out a high-precision stellar parameters and abundancedetermination by applying non-solar-scaled opacities and modelatmospheres. Our sample is composed by 20 stars (includingmain-sequence and evolved objects), with six stars belonging tobinary systems. The stellar parameters were determined by imposingionization and excitation equilibrium of Fe lines, with an updatedversion of the FUNDPAR program, together with plane-parallel ATLAS12model atmospheres and the MOOG code. Opacities for an arbitrarycomposition and vmicro werecalculated through the OS (Opacity Sampling) method. Detailedabundances were derived using equivalent widths and spectralsynthesis with the MOOG program.We applied the full line-by-line differential technique using the Sunas reference star, both in the derivation of stellar parameters andin the abundance determination. We start using solar-scaled models ina first step, and then continue the process but scaling to theabundance values found in the previous step (i.e. non-solar-scaled).The process finish when the stellar parameters of one step are thesame of the previous step, i.e. we use a doubly-iterated method.Results.We obtained a small difference in stellar parameters derived withnon-solar-scale dopacities compared to classical solar-scaled models. The differencesin Teff, log g and [Fe/H]amount up to 26 K, 0.05 dex and 0.020 dex for the stars in oursample. These differences could be considered as the first estimationof the error due to the use of classical solar-scaled opacities toderive stellar parameters with solar-type and evolved stars. We notethat some chemical species could also show an individual variationhigher than those of the [Fe/H] (up to ∼ 0.03dex) and varying from one specie to another, obtaining a chemicalpattern difference between both methods. This means that condensationtemperature Tctrendscould also present a variation. We include an example showing thatusing non-solar-scaled opacities, the solution found with theclassical solar-scaled method indeed cannot always verify theexcitation and ionization balance conditions required for a modelatmosphere. We discuss in the text the significance of thedifferences obtained when using solar-scaled vs non-solar-scaledmethods.Conclusions.We consider that the use of the non-solar-scaled opacities is notmandatory e.g. in every statistical study with large samples ofstars. However, for those high-precision works whose results dependon the mutual comparison of different chemical species (such as theanalysis of condensation temperature Tctrends), we consider that it is whortwhile its application. Todate, this is probably one of the more precise spectroscopic methodsof stellar parameters derivation.