Gamma-ray binaries emit most of their radiated power beyond ∼10 MeV. The non-thermal emission is thought to arise from the interaction of the relativistic wind of a rotation-powered pulsar with the stellar wind of its massive (O or Be) companion star. A powerful pulsar creates an extended cavity, filled with relativistic electrons, in the radiatively driven wind of the massive star. As a result, the observed P Cyg profiles of ultraviolet (UV) resonant lines from the stellar wind should be different from those of single massive stars.
We propose to use UV emission lines to detect and constrain the colliding wind region in gamma-ray binaries. We compute the expected orbital variability of P Cyg profiles depending upon the interaction geometry (set by the ratio of momentum fluxes from the winds) and the line of sight to the system. We predict little or no variability for the case of LS 5039 and PSR B1259−63, in agreement with currently available Hubble Space Telescope (HST) observations of LS 5039. However, variability between superior and inferior conjunction is expected in the case of LS I+61 303.