Chandra obtained two High Energy Transmission Grating spectra of the stellar-mass black hole GRO J1655-40 during its 2005 outburst, revealing a rich and complex disk wind. Soon after its launch, the Neil Gehrels Swift Observatory began monitoring the same outburst. Some X-ray Telescope (XRT) observations were obtained in a mode that makes it impossible to remove strong Mn calibration lines, so the Fe Kα line region in the spectra was previously neglected. However, these lines enable a precise calibration of the energy scale, facilitating studies of the absorption-dominated disk wind and its velocity shifts. Here we present fits to 15 Swift/XRT spectra, revealing variability and evolution in the outflow. The data strongly point to a magnetically driven disk wind: both the higher-velocity (e.g., $V\simeq {10}^{4}\,\mathrm{km}\,{{\rm{s}}}^{-1}$ ) and lower-velocity (e.g., $V\simeq {10}^{3}\,\mathrm{km}\,{{\rm{s}}}^{-1}$ ) wind components are typically much faster than is possible for thermally driven outflows ( $V\leqslant 200\,\mathrm{km}\,{{\rm{s}}}^{-1}$ ), and photoionization modeling yields absorption radii that are two orders of magnitude below the Compton radius that defines the typical inner extent of thermal winds. Moreover, correlations between key wind parameters yield an average absorption measure distribution that is consistent with magnetohydrodynamic wind models. We discuss our results in terms of recent observational and theoretical studies of black hole accretion disks and outflows, as well as future prospects.
