Modeling the Atmosphere of Solar and Other Stars: Radiative Transfer with PHOENIX/3D
Abstract
The chemical composition of stars is an important ingredient in our understanding of the formation, structure, and evolution of both the Galaxy and the Solar System. The composition of the sun itself is an essential reference standard against which the elemental contents of other astronomical objects are compared. Recently, redetermination of the elemental abundances using three-dimensional, time-dependent hydrodynamical models of the solar atmosphere has led to a reduction in the inferred metal abundances, particularly C, N, O, and Ne. However, this reduction in metals reduces the opacity such that models of the Sun no longer agree with the observed results obtained using helioseismology. Three dimensional (3-D) radiative transfer is an important problem in physics, astrophysics, and meteorology. Radiative transfer is extremely computationally complex and it is a natural problem that requires computation on the exascale. We intend to calculate the detailed compositional structure of the Sun and other stars at high resolution with full NLTE, treating the turbulent velocity flows in full detail in order to compare results from hydrodynamics and helioseismology, and understand the nature of the discrepancies found between the two approaches. We propose to perform 3-D high-resolution radiative transfer calculations with the PHOENIX/3D suite of solar and other stars using 3-D hydrodynamic models from different groups. While NLTE radiative transfer has been treated by the groups doing hydrodynamics, they are necessarily limited in their resolution to the consideration of only a few (4-20) frequency bins, whereas we can calculate full NLTE including thousands of wavelength points, resolving the line profiles, and solving the scattering problem with extremely high angular resolution. The code has been used for the analysis of supernova spectra, stellar and planetary spectra, and for time-dependent modeling of transient objects. PHOENIX/3D runs and scales very well on Cray XC-30 and XC-40 machines (tested up to 100,800 CPU cores) and should scale up to several million cores for large simulations. Non-local problems, particularly radiation hydrodynamics problems, are at the forefront of computational astrophysics and we will share our work with the community. Our research program brings a unified modeling strategy to the results of several disparate groups and thus will provide a unifying framework with which to assess the metal abundance of the stars and the chemical evolution of the galaxy. We will bring together 3-D hydrodynamical models, detailed radiative transfer, and astronomical abundance studies. We will also provide results of interest to the atomic physics and plasma physics communities. Our work will use data from NASA telescopes including the Hubble Space Telescope and the James Webb Space telescope. The ability to work with data from the UV to the far IR is crucial from validating our results. Our work will also extend the exascale computational capabilities, which is a national goal.
- Publication:
-
NASA ATP Proposal
- Pub Date:
- 2016
- Bibcode:
- 2016atp..prop...30B