Optically thin circumstellar medium in $\beta$ Lyr A system

TL;DR

This paper presents a comprehensive model of the beta Lyr A system, combining diverse observational data and extending radiative transfer modeling to include optically thin components like jets and shells.

Contribution

It introduces an enhanced modeling approach that incorporates new observables and provides a detailed, multi-component model of the complex binary system.

Findings

Emission originates from an extended disk atmosphere, jets, and a symmetric shell.

Jets expand at approximately 700 km/s.

The system shows a low carbon abundance, about 1% of solar value.

Abstract

beta Lyr A is a complex binary system with an extensive observational dataset: light curves (from FUV to FIR), interferometric squared visibility, closure phase, triple product measurements, spectral-energy distribution (SED), high-resolution spectroscopy, differential visibility amplitude, and also differential phase. In particular, we use spectra from Ondrejov 2m telescope from 2013 to 2015, to measure the emission in Halpha, HeI, SiII, NeI, or CII lines, and differential interferometry by CHARA/VEGA from the 2013 campaign to measure wavelength-dependent sizes across Halpha and HeI 6678. This allows us to constrain not only optically thick objects (primary, secondary, accretion disk), but also optically thin objects (disk atmosphere, jets, shell). We extended our modelling tool Pyshellspec (based on Shellspec; a 1D LTE radiative transfer code) to include all new observables, to compute differential visibilities/phases, to perform a Doppler tomography, and to determine a joint chi^2 metric. After an optimisation of 38 free parameters, we derive a robust model of the beta Lyr A system. According to the model, the emission is formed in an extended atmosphere of the disk, two perpendicular jets expanding at ~700 km s^-1, and a symmetric shell with the radius ~70 R_S. The spectroscopy indicates a low abundance of carbon, 10^-2 of the solar value. We also quantify systematic differences between datasets and discuss alternative models, with higher resolution, additional asymmetries, or He-rich abundance.