Stellar Mass Loss
Early type stars exhibit stellar winds corresponding to mass losses as great as 10 -5 to 10 -4 solar masses per year. One of the mechanisms evoked is radiation pressure on lines, although it has proved difficult to fully explain the mass loss exclusively through this process. In general the radiation from hot massive stars exhibiting winds is polarized. This one simple statement says that the mass loss is not spherically symmetric and hence not simply driven by a spherically symmetric radiation field. The polarization is due to scattering of the stellar radiation from the ejected mass. In some cases mass escapes in quasi-periodic outbursts into an equatorial disk. In this case the degree of polarization varies but the position angle remains constant, perpendicular to the plane of the disk. In other cases mass is lost along polar plumes and in some cases there are both equatorial disks and polar plumes. We know this from observations extending across the visual and ultraviolet, because in the ultraviolet the position angle flips 90 degrees when the high UV dust extinction in the denser disk inhibits scattering from the disk. The presence of such non-symmetric mass loss poses problems for the acuurate determination of mass-loss rates, which in turn has a significant effect on the evolution of massive stars and of the galaxies in which they reside.
In other stars the matter is ejected in a stochastic manner, in random blobs, in which case the polarization is seen to vary in position angle as well as degree. If, however, the interstellar polarization is large compared to the variations of intrinsic polarization the position angle will not change much and the inference would be that the ejection is confined to a disk. Interstellar polarization falls off in the ultraviolet and thus provides the means to measure this intrinsic polarization and confirm the random ejection. Massive supergiants (SG), Be stars and Wolf-Rayet stars are among those that were observed during the Astro-1 mission. The results for Be stars were a surprise and have spawned a variety of theoretical studies and considerable insight into both the Be processes and the new insight we will be able to gain fromfuture observations. The large polarization variability observed in Be and SG stars places constraints on the mass-loss mechanisms, and combined optical and UV spectropolarimetry provides stern tests for proposed models.
Wolf-Rayet stars are an excellent example of the role of spectropolarimetry in modern astrophysics and parenthetically a Guest Investigator program for Astro-2. Wolf-Rayet stars are characterized by strong broadened emission lines showing supersonic flow not easily explained by radiation pressure. The line spectrum exhibits a wide range of ionization (HeI, HeII, CII, CIII, CIV, NIII, etc). Most models assume that the emission lines are produced in a radially expanding medium with a temperature stratification. The large line widths are then attributed to the Doppler effect of the expanding envelope. The lines of highest ionization have the smallest width. If the high ionization lines are formed near the star then the envelope accelerates outward, while if the ionization increases outward then the envelope is decelerating. Studies of the light curves of the eclipsing systems CQ Cepheus and V444 Cygni have given conflicting results on the nature of the stratification. This is expected, however, since polarization measurements establish that these stars are not spherically symmetric stratification. In fact what is expected is that lines formed closest to the star will exhibit the largest polarization (i.e. have a greater electron scattering path than lines formed in the outer regions). In order to provide a means of correcting for the interstellar polarization component and to discriminate between polar and equatorial ejection, polarization observations in the UV and optical are necessary. The current data from Astro-1 have clarified some features of this fascinating example of mass loss but additional data is required to adequately test the current models.
WUPPE observed many hot stars, Wolf-Rayet objects, binaries and novae during Astro-2 and the data are currently being studied.
- A. D. Code