3. Jet modelling
The detailed shape of the shifted H lines, in particular their extended wings towards small absolute velocities potentially contains valuable information on jet geometry and kinematics. We therefore designed a simple kinematic jet model derived from that used by Becker et al. (1998). The jet is described as a cone of half opening angle in which all atoms move with velocity . Within the cone, material is assumed to be flowing uniformly per unit solid angle. In order to compute line profiles we further assumed that the outflow emission region is optically thin and convolved the model profile with a Gaussian of FWHM = 333 km s-1 ( = 142 km s-1) representing the instrumental profile.
Since we only have a snapshot observation at an unknown orbital phase, we cannot constrain the orientation of the jet with respect to orbital plane or with respect to the axis joining the two stars. In our case, the only relevant angles are and the angle i between the line of sight and the jet axis. In this simplified geometry, the jet axis is aligned with the z axis and the line of sight is contained in the x-z plane. A flow making an angle with respect to jet axis and an angle with respect to the x axis will have a projected component on the line of sight.
The jet components extend in velocity from about 3,800 to 5,800 km s-1 with a peak at 5,200 km s-1. Any effect related to orbital motion is likely to be negligible since the K amplitude of the He ii line is only 80 km s-1 (Motch 1996) and since the duration of the observation ( 1 h) is small with respect to the suspected orbital period ( 3.8 d).
Considering the uncertainties resulting from the He i/He ii line contamination, we did not try to fit a model jet profile to the redshifted H component. The width of the blue component profile and its asymmetry, namely its larger extension towards low absolute velocities, can be accurately represented by a well opened jet seen at rather low inclination. As shown on Fig. 4, the fit is surprisingly good considering the simplicity of the model ( = 38.1). At the 99% confidence level, the formal accepted ranges of inclinations i are to and to and those of half-opening jet angles are to and to . Large inclination angles correspond to small jet opening angles. The range of values compares well with that derived for RX J0019.8+2156 by Becker et al. (1998). In the framework of this simple model the difference in profile shape between the blue and red components could reflect different opening angles (assuming inclinations are identical).
Alternatively, the width and asymmetry of the blueshifted profile could be due to an intrinsic spread of material velocity in the outflow. This would allow much narrower opening angles, more consistent with the conception of a jet. In this picture, the difference in profile shape and extent between the blue and red components may be for instance interpreted in terms of occultation of the low velocity part of an accelerating jet by the accretion disc. Finally, line profiles may also be intrinsically broadened by Keplerian velocity in the inner parts of the accretion disc (Becker et al. 1998).
© European Southern Observatory (ESO) 1998
Online publication: September 8, 1998