In a series of our previous studies, a model of radiatively inefficient accretion flows (RIAFs) in a global magnetic field (so-called resistive RIAF model) has proved its ability to account for many physical processes taking place in such accretion flows, as realized in the nuclei of galaxies believed to be accreting at a very small fraction of each Eddington accretion rate. Within the present status of this model, however, the model cannot describe the launch of a self-confined bipolar jet from the vicinity of the disk's inner edge, although it allows the existence of a thermal wind widely distributed over the disk surfaces. This is because the electric field (and hence the Poynting flux) vanishes everywhere in the disk, whereas such a jet in a globally ordered magnetic field is most likely to be accelerated electrodynamically. We show in the present paper that this defect can be overcome naturally if we reformulate the problem so as to admit a quasi-stationary change of the magnetic field (and hence the appearance of a non-irrotational electric field), and also restore all of the terms of order $ \epsilon$$ \equiv$ ($ v_r/v_{\varphi}$ )$ ^2$$ \lesssim$ 1 (where $ v_r$ and $ v_{\varphi}$ denote the radial and azimuthal components, respectively, of the fluid velocity), which have been neglected altogether in our previous scheme. The restored effects are the inertial and magnetic draggings on the infalling matter. As an illustrative example, a model solution that is correct up to $ {\cal O}(\epsilon)$ is derived under a set of plausible restrictions. The new solution predicts the appearance of a localized Poynting flux in a region near the disk inner edge, strongly suggesting that a jet is launched from this region. Another interesting prediction is the appearance of a rapid change of the magnetic field, also localized to this region.
Read full abstract