2020MNRAS.496..223W

High-resolution spectra of typical wind diagnostics ([O I] 6300 Å and other forbidden emission lines) can often be decomposed into multiple components: high-velocity components with blueshifts up to several 100 km s^–1 are usually attributed to fast jets, while narrow (NLVC) and broad (BLVC) low-velocity components are believed to trace slower disc winds. Under the assumption that the line broadening is dominated by Keplerian rotation, several studies have found that the BLVCs should trace gas launched between 0.05 and 0.5 au and correlations between the properties of BLVCs and NLVCs have been interpreted as evidence for the emission tracing an extended magnetohydrodynamics (MHD) wind and not a photoevaporative wind. We calculated synthetic line profiles obtained from detailed photoionization calculations of an X-ray photoevaporation model and a simple MHD wind model and analysed the emission regions of different diagnostic lines and the resulting spectral profiles. The photoevaporation model reproduces most of the observed NLVCs but not the BLVCs or HVCs. The MHD model is able to reproduce all components but produces Keplerian double peaks at average inclinations that are rarely observed. The combination of MHD and photoevaporative winds could solve this problem. Our results suggest that the Gaussian decomposition does not allow for a clear distinction of flux from different wind regions and that the line broadening is often dominated by the velocity gradient in the outflow rather than by Keplerian rotation. We show that observed correlations between BLVC and NLVC do not necessarily imply a common origin in an extended MHD wind.