2012ApJ...750...27N

During its 2005 outburst, GRO J1655-40 was observed twice with the Chandra High Energy Transmission Grating Spectrometer; the second observation revealed a spectrum rich with ionized absorption lines from elements ranging from O to Ni, indicative of an outflow too dense and too ionized to be driven by radiation or thermal pressure. To date, this spectrum is the only definitive evidence of an ionized wind driven off the accretion disk by magnetic processes in a black hole X-ray binary. Here we present our detailed spectral analysis of the first Chandra observation, nearly three weeks earlier, in which the only signature of the wind is the Fe XXVI absorption line. Comparing the broadband X-ray spectra via photoionization models, we argue that the differences in the Chandra spectra cannot possibly be explained by the changes in the ionizing spectrum, which implies that the properties of the wind cannot be constant throughout the outburst. We explore physical scenarios for the changes in the wind, which we suggest may begin as a hybrid MHD/thermal wind, but evolves over the course of weeks into two distinct outflows with different properties. We discuss the implications of our results for the links between the state of the accretion flow and the presence of transient disk winds.