Eclipse observations of an accretion disc wind

Abstract

The Goddard High Resolution Spectrometer on the Hubble Space Telescope has been used to monitor, through eclipse, the C iv 1550-A and He ii 1640-A lines in the spectrum of the nova-like cataclysmic variable UX UMa. Previous work has suggested that the Civ line arises in an accretion disc wind. The HST data confirm this, in that the line shows a broad, asymmetric emission profile. However, superimposed on the emission are previously unsuspected narrow absorption components at the rest wavelengths of the doublet. The absorption disappears during the continuum eclipse, causing the net continuum-subtracted flux near line centre to increase by about 20 per cent during this time. The mid-eclipse line profile shows the classic asymmetric shape expected for a bipolar wind viewed edge-on. The various wavelength segments within the C iv line exhibit different eclipse light curves. The flux at line centre declines gradually, beginning at least 0.05 in orbital phase before the continuum eclipse, then rising sharply coincident with continuum ingress due to loss of the absorption components. Flux is also lost from the red wing of the C iv line beginning just before (Deltaphi ~ 0.02) the onset of continuum eclipse. The red wing flux declines more slowly than the continuum itself, reaching a minimum after the time of continuum mid-eclipse. This suggests that the emitting gas may have a significant rotational component about the white dwarf. However, the blue wing shows no comparable signature of rotation. These data show that the shallowness of the line flux eclipse is primarily due to occultation of regions of net absorption, with the implication that the line-forming region can be smaller than the companion star. To explain the appearance and eclipse evolution of the C iv line profile solely in terms of outflow will require a highly collimated disc wind. Alternatively the narrow, superposed absorption components may originate elsewhere in the system. These data also provide the first clear demonstration that He ii 1640 A is mainly a wind-formed feature.