2001ApJ...550..919B

We compare the predictions of hydrodynamic models of the LMC X-4 X-ray binary system with observations of UV P Cygni lines with the Goddard High Resolution Spectrograph and Space Telescope Imaging Spectrograph on the Hubble Space Telescope. The hydrodynamic model determines density and velocity fields of the stellar wind, wind-compressed disk, accretion stream, Keplerian accretion disk, and accretion disk wind. We use a Monte Carlo code to determine the UV P Cygni line profiles by simulating the radiative transfer of UV photons that originate on the star and are scattered in the wind. The qualitative orbital variation predicted is similar to that observed, although the model fails to reproduce the strong orbital asymmetry (the observed absorption is strongest for φ>0.5). The model predicts a mideclipse X-ray spectrum, caused almost entirely by Compton scattering, with a factor of 4 less flux than observed with ASCA. We discuss how the model may need to be altered to explain the spectral variability of the system.