Back to Volume
Paper: Chromospheres and Winds of Cool Supergiants
Volume: 412, The Biggest, Baddest, Coolest Stars
Page: 221
Authors: Bennett, P.D.
Abstract: Massive stars lose material via stellar winds on evolutionary timescales in the supergiant phases of their lives. In particular, stars between about 3 and 40 Mʘ spend a significant fraction of their post-main sequence lifetime as red supergiants of spectral classes G-M. The upper limit to red supergiant masses appears to be constrained by mass loss that increases rapidly with stellar mass. Although mass loss from red supergiants has been observed since the 1930s, we still don’t understand the mechanism that drives it. Unfortunately, most observations of mass loss from supergiant stars include the entire, unresolved, circumstellar envelope, and provide no spatial information, especially about the inner part of the wind where mass loss starts. One way to achieve spatially resolved observations is to observe the select group of red supergiants with main-sequence companions (typically B stars) in eclipsing orbits. As the hot, and much smaller (in physical size, not mass) companion moves behind the supergiant during eclipse ingress, the line of sight sweeps through successively deeper layers of the supergiant’s extended outer envelope. This circumstellar envelope superimposes an absorption upon the continuum of the hot companion, and this “chromospheric” eclipse spectrum can be used to infer the density, velocity and ionization state along the line of sight. By repeated observation of the spectrum of the binary through eclipse, the structure of the supergiant’s outer atmosphere can be derived. All of this was realized decades ago. But, despite this potential, the early promise of the binary method to reveal atmospheric structure in red supergiants has never been fully achieved. There are many reasons for this: the additional data reduction complication caused by the need to separate composite spectra, the uncertain perturbing effect of the companion on the structure of the circumstellar envelope being probed, the difficulty of removing light scattered by the circumstellar envelope into the line of sight (which appears as emission), and sometimes, because of the sheer complexity of the chromospheric absorption line spectrum observed (e.g. VV Cephei). The situation has improved considerably with the advent of ultraviolet observations from space, which mostly obviate the need to disentangle composite spectra. The International Ultraviolet Explorer has provided a huge amount of spectroscopic data of limited resolution and signal-to-noise, while the Hubble Space Telescope GHRS and STIS spectrographs have obtained a limited number of observations of superb spectral resolution and signal-to-noise. The Far Ultra-violet Spectroscopic Explorer has also contributed results shortward of Lymanalpha. In this talk, I will summarize the present state of the binary method, and present observations, results and prospects for several red supergiant binaries.
Back to Volume