Oxygen is of a major relevance in modern astrophysics, across difference fields including precision stellar physics, extragalactic astronomy, and evolution of the Milky Way. Most importantly, Oxygen determines much of the opacity in the solar interior, therefore its abundance is critical to the calculations of Standard Solar Models, which describe the evolution of our Sun from the pre-main-sequence to the present age of 4.5 Gyr. Oxygen is also the key element in gas-phase spectroscopic diagnostic on extragalactic scales, as well as a common tracer of nucleosynthesis in massive stars.
Over the past decades, there has been much debate about the true solar oxygen abundance. Some groups attempted to measure the O abundances from the photospheric spectrum. Other groups derived the solar O abundances from the solar wind data. The latter can be, however, measured less accurately compared to the photospheric methods. The most recent estimates, derived by means of the detailed Non-LTE analysis with 3D radiation-hydrodynamics simulations of solar convection, is log A(O) = 8.66±0.10 dex, which compares well, within the uncertainties to the earlier 3D NLTE estimate by (log A(O) = 8.78 ± 0.10 dex). These results are significantly lower than previous estimates based on LTE analysis and do not satisfy Standard Solar Models.
Here we present a spectroscopic re-analysis the solar photospheric O abundance. We take advantage of new atomic data (rates for collisional excitation and charge transfer with H atoms, as well as more complete and detailed photoionization cross sections than in previous studies) recent 3D magneto-hydrodynamic simulations of the solar surface, which include the chromosphere and photosphere (Carlsson et al. 2016), as well new high-resolution multi-wavelength spectroscopic observations of the Sun (PEPSI, HARPS, IAG, ESPRESSO). Furthermore, we investigate the O I and [O I] line formation using two 3D MHD models computed self-consistently, with and without chromosphere using the Bifrost code.