A Multiwavelength Campaign on γ Cassiopeiae. III. The Case for Magnetically Controlled Circumstellar Kinematics

Abstract
In two previous papers, we have discussed simultaneous Rossi X-Ray Timing Explorer (RXTE) and Hubble Space Telescope GHRS observations made of γ Cas (B0.5e) over a full day on 1996 March 14-15. The light curves generated from these data show features that anticorrelate and led us to conclude that the star has multiple surface activity centers that rotate into view every rotational cycle of 1.123 days. In a second paper we found that dips in the UV continuum (UVC) light curve are probably caused by the passage of cool, co-orbiting clouds that are suspended above surface X-ray-active centers. In this paper we use difference spectra from our >21 hr GHRS time series to investigate ≤2% spectral variations within the photospheric Si IV 1394-1403 Å lines as well as smaller variations from features in neighboring wavelengths at 1382-1386 and 1404-1417 Å. Several difficulties arise in interpreting these variations as signatures of surface inhomogeneities, so we have studied both types of variations in the context of the kinematics of occulting circumstellar (CS) structures. By means of model-atmospheres codes and up-to-date line lists, we computed a grid of cloud opacity for various temperatures in our spectral range. Using these synthetic spectra, we are able to identify features as optically thick absorptions due to Fe II, Cr II, and C I lines from "cool" (TVsini, suggesting that they are formed in corotating cloudlets that are distinct from the cool clouds we studied in a previous paper. In contrast, warm and hot-plasma lines are "ultrasharp features" ("USFs") that maintain a constant velocity for several hours. The USFs are visible over a wide velocity range, in some cases having a velocity of at least +1500 km s-1. Both cooling and heating of circumstellar plasma is consistent with the existence of strong nonradiative processes operating in some regions above γ Cas. Additionally, the migrating subfeatures find a ready explanation in circumstellar cloudlets forced by magnetic forces into corotation, as was also inferred from the UVC light curve in Smith, Robinson, & Hatzes. The existence of the stationary ultrasharp absorption features, particularly at large positive velocities, is difficult to explain unless one invokes interactions between magnetic loops from the star and a putative field in the circumstellar disc. This picture holds the potential of explaining the hot, flaring character of this star's X-rays and predicts the existence of other X-ray-emitting γ Cas analogues as magnetic Be stars having dense CS discs.