Research

PhD

NLTE radiative transfer in stellar photospheres and chromospheres: application to atoms of magnesium, calcium and iron in late-type stars

Abstract

Stellar abundance analyses generally rely on the assumption of Local Thermodynamic Equilibrium (LTE) in stellar atmospheres. This assumption, which is valid in stellar interiors, is however not appropriate to describe radiative transfer in the atmospheric layers (photosphere and chromosphere). It is therefore necessary to adopt a Non-LTE (NLTE) description of radiative transfer (replacing the equations of Boltzmann and Saha by the equations of statistical equilibrium) to improve our understanding of the physical processes in the stellar atmospheres; a description which needs a large quantity of atomic data.

My thesis work consisted in constructing model atoms from the most recent atomic databases. The main atoms that were considered are those of α-elements: neutral (I) and ionized (II) magnesium (Mg) and calcium (Ca). These elements are astrophysically interesting because they permit us to probe the history and the chemical enrichment of galactic and extra-galactic stellar populations. I therefore developed a model atom contruction code, FORMATO, which combines data from online atomic databases (NIST, VALD, TOPbase, R. Kurucz database), semi-empirical and semi-classical formulae, and quantum computations from the literature. These atoms are used to study the NLTE formation of spectral lines with the code MULTI.

I first concentrated on the NLTE effects for the MARCS model atmospheres of late-type stars, in particular, giants and super-giants. I used the magnesium and calcium model atoms for computing a grid of NLTE corrections to apply to the equivalent widths of the principal spectral lines of these elements, while paying special attention to spectral lines corresponding to the wavelength range of RVS/Gaia. These corrections can be used by the automated abundance analysis methods based on equivalent widths for current and forthcoming large surveys (e.g. RAVE and Gaia).

I also applied the NLTE radiative transfer of Ca II to study the chromospheres of red giant K type stars observed with interferometry(VEGA@CHARA). The CaII triplet line cores are formed in the chromospheres of such stars. Using a model atmosphere including a model chromosphere of the red giant ß Cet, I computed the NLTE limb-darkening laws for this triplet. The mean radius of the chromosphere is 15 % to 30 % larger than the photospheric radius. This is the first time that a chromospheric radius has been obtained in this way.

Finally, as part of the Carina Project, we highlighted NLTE effects on ionization equilibrium of iron (Fe) in a sample of 44 red giants of the Carina dSph galaxy, based on a comparative study between FeI and FeII lines in LTE and NLTE. The discrepancy obtained by LTE abundance analysis is 0.10 dex between the FeI - FeII metallicity distribution of Carina giants. This difference is explained by the over-ionization of FeI in giants that are more and more metal-poor and is supported by NLTE computations which also highlight large discrepancies in excitation equilibrium of FeI and FeII.

As a consequence of this thesis project, we may test the influence of inelastic collisions with neutral hydrogen on the NLTE discrepancies of magnesium since quantum computations are now beginning to appear. Further work also includes the construction of other light element models in order to study abundance trends in stellar populations and to contribute to the development of NLTE stellar atmosphere models.

Related refereed publications

  • Merle T., Thévenin F., Pichon B., Bigot L., 2011, MNRAS, 418, 863 « A grid of NLTE corrections for magnesium and calcium in late-type giant and supergiant stars: application to Gaia»
  • Berio P., Merle T., Thévenin F. et al., 2011, A&A, 535, 59 « Chromosphere of K giant stars. Geometrical extent and spatial structure detection »
  • Fabrizio M., Merle T., Thévenin F., Bono G. et al., 2012, PASP, en cours de révision « The Carina project. V. The impact of NLTE effects on the iron content »