EVIDENCE FOR A DISK IN THE WIND OF HD-93521 - UV LINE-PROFILES FROM AN AXISYMMETRICAL MODEL

Recently it has been suggested (Massa 1992; Howarth and Reid 1993), fr om the C IV ultraviolet resonance line profile of HD 93521, that there is a high-speed component in the polar outflow from the star as well as a low-speed component in the equatorial regions. In this paper we p resent theoretical calculations of the line profiles that would be pro duced by such a model. We find from Hubble Space Telescope Goddard Hig h Resolution Spectrograph observations of HD 93521 that the observed C IV profile can be produced if the star has an inclination angle very close to 90 degrees and if the star is surrounded by a thin disk, whos e half-width is approximately 3 degrees in latitude. The geometry of t his disk is similar to what one would expect from the wind-compressed disk model of Bjorkman and Cassinelli, so this star may provide an ide al observational test of that model. In addition to the C IV resonance line, we examine both the Si IV and N V resonance lines. The Si IV li ne exhibits low-velocity absorption that is similar to that seen in th e C IV line, but the emission is much weaker. The absorption component is well fitted by the geometry that is inferred from the C IV line. I t appears that the weaker emission from the polar component arises bec ause Si IV is one stage below the dominant state, Si V. On the other h and, the N V line has weaker absorption and much stronger emission tha n the C IV line. N V is a higher ionization state than C IV, so it is likely that N V is one stage above the dominant state, N IV. The corre lation of the emission strength of these three ions with their ionizat ion state suggests that there are latitudinal ionization gradients tha t occur because the density increases from the pole to the equator. Ap art from fitting individual line profiles, we also examine the differe nces between the two Hubble Space Telescope observations of HD 93521. We find evidence for a pair of narrow absorption components, seen at l ow velocity, as well as evidence for a discrete emission feature in th e blueshifted absorption cores of the lines. This blueshifted emission at low velocity can cause what instead appears to be an interstellar absorption line. Without multiple observations that can reveal the tem porary emission, one must be very careful when determining interstella r column densities to stars like HD 93521.