Editing Rigel (section)

=== Spectroscopy ===

Rigel's [[Stellar classification#Modern classification|spectral type]] is a defining point of the classification sequence for supergiants.<ref name=morgan1978/><ref name=morgan1950/> The overall spectrum is typical for a [[Late-type star|late]] B class star, with strong [[absorption line]]s of the hydrogen [[Balmer series]] as well as neutral helium lines and some of heavier elements such as oxygen, calcium, and magnesium.<ref name=abetti/> The [[luminosity class]] for B8 stars is estimated from the strength and narrowness of the hydrogen spectral lines, and Rigel is assigned to the [[bright supergiant]] class Ia.<ref name=mkk/> Variations in the spectrum have resulted in the assignment of different classes to Rigel, such as B8 Ia, B8 Iab, and B8 Iae.<ref name=schultz/><ref name=bally/>

As early as 1888, the heliocentric [[radial velocity]] of Rigel, as estimated from the [[Doppler shift]]s of its spectral lines, was seen to vary. This was confirmed and interpreted at the time as being due to a spectroscopic companion with a period of about 22 days.<ref name=plaskett/> The radial velocity has since been measured to vary by about {{val|10|ul=km/s}} around a mean of {{val|21.5|u=km/s}}.<ref name=morrison/>

In 1933, the [[Hα]] line in Rigel's spectrum was seen to be unusually weak and shifted {{val|0.1|ul=nm}} towards shorter wavelengths, while there was a narrow [[emission line|emission spike]] about {{val|1.5|u=nm}} to the long wavelength side of the main absorption line.<ref name=struve/> This is now known as a [[P Cygni profile]] after a star that shows this feature strongly in its spectrum. It is associated with [[mass loss]] where there is simultaneously emission from a dense wind close to the star and absorption from circumstellar material expanding away from the star.<ref name=struve/>

The unusual Hα line profile is observed to vary unpredictably. It is a normal absorption line around a third of the time. About a quarter of the time, it is a double-peaked line, that is, an absorption line with an emission core or an emission line with an absorption core. About a quarter of the time it has a P Cygni profile; most of the rest of the time, the line has an inverse P Cygni profile, where the emission component is on the short wavelength side of the line. Rarely, there is a pure emission Hα line.<ref name=morrison/><!--cites five previous sentences--> The line profile changes are interpreted as variations in the quantity and velocity of material being expelled from the star. Occasional very high-velocity outflows have been inferred, and, more rarely, infalling material. The overall picture is one of large [[coronal loop|looping structures]] arising from the [[photosphere]] and driven by magnetic fields.<ref name=israelian/>