THE WOLF-RAYET SYSTEM WR-147 - A BINARY RADIO-SOURCE WITH THERMAL AND NONTHERMAL COMPONENTS

A revised distance of 630 pc is derived for the Wolf-Rayet star WR 147(WN8) from newly obtained near-infrared photometry. This distance is almost free of reddening corrections, and it indicates that WR 147 is not associated with the Cyg OB2 association (which lies at a distance of 2 kpc), as previously thought. A terminal wind speed of 900 km s-1 is found for WR 147 from the P Cygni profile of the He I (2(1)S-2(1)P) line at 2.058-mu-m. High-resolution VLA images were obtained at three different epochs at wavelengths 20, 6, 2, and 1.3 cm (although not all wavelength bands were imaged at each epoch) with angular resolutions of 1".60, 0".46, 0".14, and 0".11, respectively. These images show two radio components separated by 0".58 (approximately 364 AU) in an approximately N-S direction. The optical astrometry of Moran et al. shows that the Wolf-Rayet star WR 147 coincides with the southern radio component WR 147S whose spectrum and visibility function are consistent with a thermal wind. The northern radio component WR 147N is clearly nonthermal. The radio emission from WR 147S can be accurately modeled by a spherical wind whose density decreases with distance as r-2. We find an average mass-loss rate of 4.2 +/- 0.2 x 10(-5) M. yr-1 and a mean wind temperature of 9400 +/- 1000 K. Several mechanisms were investigated to determine if the observed X-ray luminosity could be produced by interaction of the wind from WR 147S with another star at the position of WR 147N. The only mechanism we found that could produce adequate X-ray luminosity is gravitational capture by a neutron star orbiting in a dense, slow, equatorial wind from WR 147S. This result lends support to the magnetic rotator model of Poe, Friend, & Cassinelli for W-R stars. We caution, however, that the X-ray emission could be produced by shocks in the wind of WR 147S or perhaps by colliding winds from WR 147S and WR 147N. The flux density of the WR 147 system varies by as much as 50% at 21 cm over a period of a week or two and as much as 20% at 6 cm over a period of several weeks. The variations are probably associated with WR 147N.