Method and new tabulations for flux-weighted line opacity and radiation line force in supersonic media^★

¹ Institute of Astronomy, KU Leuven, Celestijnenlaan 200D, 3001 Leuven, Belgium
e-mail: luka.poniatowski@kuleuven.be
² National Solar Observatory, 22 Ohi’a Ku St, Makawao, HI 96768, USA
³ Bartol Research Institute, Department of Physics and Astronomy, University of Delaware, Newark, DE 19716, USA
⁴ University of Iowa, Iowa City, IA, USA
⁵ Astronomical Institute Anton Pannekoek, Amsterdam University, Science Park 904, 1098 XH Amsterdam, The Netherlands

Received: 11 December 2021
Accepted: 21 April 2022

Abstract

Context. In accelerating and supersonic media, understanding the interaction of photons with spectral lines can be of utmost importance, especially in an accelerating flow. However, fully accounting for such line forces is computationally expensive and challenging, as it involves complicated solutions of the radiative transfer problem for millions of contributing lines. This currently can only be done by specialised codes in 1D steady-state flows. More general cases and higher dimensions require alternative approaches.

Aims. We present a comprehensive and fast method for computing the radiation line force using tables of spectral-line-strength distribution parameters, which can be applied in arbitrary (multi-D, time-dependent) simulations, including those that account for the line-deshadowing instability, to compute the appropriate opacities.

Methods. We assume local thermodynamic equilibrium to compute a flux-weighted line opacity from ~4 million spectral lines. We fit the opacity computed from the line list with an analytic result derived for an assumed distribution of the spectral line strength and found the corresponding line-distribution parameters, which we tabulate here for a range of assumed input densities ρ ∈ [10⁻²⁰, 10⁻¹⁰] g cm⁻³ and temperatures T ∊ [10⁴, 10⁴⁷] K.

Results. We find that the variation in the line-distribution parameters plays an essential role in setting the wind dynamics in our models. In our benchmark study, we also find a good overall agreement between the O-star mass-loss rates of our models and those derived from steady-state studies that use a more detailed radiative transfer.

Conclusions. Our models reinforce the idea that self-consistent variation in the line-distribution parameters is important for the dynamics of line-driven flows. Within a well-calibrated O-star regime, our results support the proposed methodology. In practice, utilising the provided tables, yielded a factor >100 speed-up in computational time compared to specialised 1D model-atmosphere codes of line-driven winds, which constitutes an important step towards efficient multi-dimensional simulations. We conclude that our method and tables are ready to be exploited in various radiation-hydrodynamic simulations where the line force is important.