The very small accretion disks in ultra-compact X-ray binaries (UCXBs) are special laboratories in which to study disk accretion and outflows. We report on three sets of new (250 ks total) and archival (50 ks) Chandra/HETG observations of the dipping
neutron-star X-ray binary 4U 1916$-$053, which has an orbital period of $Psimeq 50$~minutes. We find that the bulk of the absorption in all three spectra originates in a disk atmosphere that is redshifted by $vsimeq 220-290$ $text{km}$ $text{s}^{-1}$, corresponding to the gravitational redshift at radius of $R sim 1200$ $GM/{c}^{2}$. This shift is present in the strongest, most highly ionized lines (Si XIV and Fe XXVI), with a significance of 5$sigma$. Absorption lines observed during dipping events (typically associated with the outermost disk) instead display no velocity shifts and serve as a local standard of rest, suggesting that the redshift is intrinsic to an inner disk atmosphere and not due to radial motion in the galaxy or a kick. In two spectra, there is also evidence of a more strongly redshifted component that would correspond to a disk atmosphere at $R sim 70$ $GM/{c}^{2}$; this component is significant at the 3$sigma$ level. Finally, in one spectrum, we find evidence of disk wind with a blue shift of $v = {-1700}^{+1700}_{-1200}$ $text{km}$ $text{s}^{-1}$. If real, this wind would require magnetic driving.
The mechanisms that drive disk winds are a window into the physical processes that underlie the disk. Stellar-mass black holes are an ideal setting in which to explore these mechanisms, in part because their outbursts span a broad range in mass accre
tion rate. We performed a spectral analysis of the disk wind found in six Chandra/HETG observations of the black hole candidate 4U~1630$-$472, covering a range of luminosities over two distinct spectral states. We modeled both wind absorption and extended wind re-emission components using PION, a self-consistent photoionized absorption model. In all but one case, two photoionization zones were required in order to obtain acceptable fits. Two independent constraints on launching radii, obtained via the ionization parameter formalism and the dynamical broadening of the re-emission, helped characterize the geometry of the wind. The innermost wind components ($r simeq {10}^{2-3}$ $GM/{c}^{2}$) tend towards small volume filling factors, high ionization, densities up to $n simeq {10}^{15-16} {text{cm}}^{-3}$, and outflow velocities of $sim 0.003c$. These small launching radii and large densities require magnetic driving, as they are inconsistent with numerical and analytical treatments of thermally driven winds. Outer wind components ($r simeq {10}^{5}$ $GM/{c}^{2}$) are significantly less ionized and have filling factors near unity. Their larger launching radii, lower densities ($n simeq {10}^{12} {text{cm}}^{-3}$), and outflow velocities ($sim 0.0007c$) are nominally consistent with thermally driven winds. The overall wind structure suggests that these components may also be part of a broader MHD outflow and perhaps better described as magneto-thermal hybrid winds.
