2023MNRAS.518.1789T

Black hole accretion discs can produce powerful outflowing plasma (disc winds), seen as blue-shifted absorption lines in stellar and supermassive systems. These winds in quasars have an essential role in controlling galaxy formation across cosmic time, but there is no consensus on how these are physically launched. A single unique observation of a stellar-mass black hole GRO J1655-40 was used to argue that magnetic driving was the only viable mechanism and motivated unified models of magnetic winds in both binaries and quasars. The alternative, X-ray heating (thermal-radiative wind), was ruled out for the low observed luminosity by the high wind density estimated from an absorption line of a metastable level of Fe XXII. Here, we reanalyse these data using a photoionization code that includes cascades from radiative excitation as well as collisions in populating the metastable level. The cascade reduces the inferred wind density by more than an order of magnitude. The derived column is also optically thick, so the source is intrinsically more luminous than observed. We show that a thermal-radiative wind model calculated from a radiation hydrodynamic simulation matches well with the data. We revisit the previous magnetic wind solution and show that this is also optically thick, leading to a larger source luminosity. However, unlike the thermal-radiative wind, it struggles to reproduce the overall ion population at the required density. These results remove the requirement for a magnetic wind in these data and remove the basis of the self-similar unified magnetic wind models extrapolated to quasar outflows.