Previous MHD simulations have shown that wind (i.e., uncollimated outflow) must exist in black hole hot accretion flows. In this paper, we continue our study by investigating the detailed properties of wind, such as mass flux and poloidal speed, and the mechanism of wind production. For this aim, we make use of a three dimensional GRMHD simulation of hot accretion flows around a Schwarzschild black hole. The simulation is designed so that the magnetic flux is not accumulated significantly around the black hole. To distinguish real wind from turbulent outflows, we track the trajectories of the virtual Largrangian particles from simulation data. We find two types of real outflows, i.e., a quasi-relativistic jet close to the axis and a sub-relativistic wind subtending a much larger solid angle. We confirm that the mass flux of wind is very significant and most of the wind originates from the surface layer of the accretion flow. The radial profile of the wind mass flux can be described byṀ wind ≈Ṁ BH (r/20r s ), withṀ BH being the mass accretion rate at the black hole horizon and r s being the Schwarzschild radius. The poloidal wind speed almost remains constant once they are produced, but the fluxweighted wind speed roughly follows v p,wind (r) ≈ 0.25v k (r), with v k (r) being the Keplerian speed at radius r. The mass flux of jet is much lower but the speed is much higher, v p,jet ∼ (0.3 − 0.4)c. Consequently, both the energy and momentum fluxes of the wind are much larger than those of the jet. We find that the wind is produced and accelerated primarily by the combination of centrifugal force and magnetic pressure gradient, while the jet is mainly accelerated by magnetic pressure gradient. Finally, we find that the wind production efficiency ǫ wind ≡Ė wind /Ṁ BH c 2 ∼ 1/1000, in good agreement with the value required from large-scale galaxy simulations with AGN feedback.

Previous hydrodynamical (HD) and magnetohydrodynamical (MHD) numerical simulations of hot accretion flows have indicated that the inflow (gas with inward radial velocity) accretion rate decreases with decreasing radius. Two models have been proposed to explain this result. In the adiabatic inflowoutflow solution (ADIOS), the inward decrease of accretion rate is because of the loss of gas in the outflow. In the alternative convection-dominated accretion flow (CDAF) model, the accretion flow is thought to be convectively unstable; the gas is then assumed to be locked in convective eddies, which results in the inward decrease of the accretion rate. In the present paper we investigate the nature of inward decrease of accretion rate using HD and MHD simulations. We calculate various properties of inflow and outflow, including the mass flux, radial and rotational velocities, temperature, and the Bernoulli parameter (Be). Systematic and significant differences between inflow and outflow are found. For example, for HD flows, the temperature of outflow is significantly higher than inflow; while for MHD flows, the specific angular momentum of outflow is nearly Keplerian, which is significantly higher than inflow. These results suggest that the inflow and outflow are not dominated by convective turbulence, but they are systematic inward and outward motion. We have also analyzed the convective stability of MHD accretion flow and found that they are convectively stable. These results indicate that the inward decrease of inflow rate is because of the mass loss in outflow. The different properties of inflow and outflow also suggest that the mechanisms of producing outflow in HD and MHD flows are buoyancy associated with the convection and centrifugal force associated with the angular momentum transport mediated by the magnetic field, respectively. The latter mechanism is similar to the Blandford & Payne mechanism but no large-scale open magnetic field is required; it is kind of "micro-Blandford & Payne" mechanism. We also study the effect of initial conditions in the simulations. We find that the value of Be, whose sign determines whether the outflow can escape to infinity or not, is mainly determined by the value of Be of the initial condition. We discuss some possible observational applications, including the Fermi bubble observed in the Galaxy center and winds widely observed in active galactic nuclei and black hole X-ray binaries.

Numerical simulations of hot accretion flow, both hydrodynamical and magnetohydrodynamical, have shown that the mass accretion rate decreases with decreasing radius; consequently the density profile of accretion flow becomes flatter compared to the case of a constant accretion rate. This result has important theoretical and observational implications. However, because of technical difficulties, the radial dynamic range in almost all previous simulations usually spans at most two orders of magnitude. This small dynamical range, combined with the effects of boundary conditions, makes the simulation results suspectable. Especially, the radial profiles of density and inflow rate may not be precise enough to be used to compare with observations. In this paper we present a "two-zone" approach to expand the radial dynamical range from two to four orders of magnitude. We confirm previous results and find that from r s to 10 4 r s the radial profiles of accretion rate and density can be well described byṀ (r) ∝ r s and ρ ∝ r −p . The values of (s, p) are (0.48, 0.65) and (0.4, 0.85), for viscous parameter α = 0.001 and 0.01, respectively. Or more precisely, the accretion rate is constant (i.e., s = 0) within ∼ 10r s ; but beyond 10r s , we have s = 0.65 and 0.54 for α = 0.001 and 0.01, respectively. We find that the values of both s and p are similar in all numerical simulation works, including previous and the present ones, no matter a magnetic field is included or not and what kind of initial conditions are adopted. Such an apparently surprising "common" result can be explained by the most updated version of the adiabatic inflow-outflow model (ADIOS). The density profile we obtain is in good quantitative agreement with that obtained from the detailed observations and modeling to Sgr A* and NGC 3115. The origin and implication of such a profile will be investigated in a subsequent paper.

A pair of giant gamma-ray Bubbles has been revealed by Fermi-LAT. In this paper we investigate their formation mechanism. Observations have indicated that the activity of the supermassive black hole located at the Galactic center, Sgr A*, was much stronger than at the present time. Specifically, one possibility is that while Sgr A* was also in the hot accretion regime, the accretion rate should be 10 3-10 4 times higher during the past ∼10 7 yr. On the other hand, recent magnetohydrodynamic numerical simulations of hot accretion flows have unambiguously shown the existence and obtained the properties of strong winds. Based on this knowledge, by performing threedimensional hydrodynamical simulations, we show in this paper that the Fermi Bubbles could be inflated by winds launched from the "past" hot accretion flow in Sgr A*. In our model, the active phase of Sgr A* is required to last for about 10 million years and it was quenched no more than 0.2 million years ago. The central molecular zone (CMZ) is included and it collimates the wind orientation toward the Galactic poles. Viscosity suppresses the Rayleigh-Taylor and Kelvin-Helmholtz instabilities and results in the smoothness of the Bubbles edge. The main observational features of the Bubbles can be well explained. Specifically, the ROSAT X-ray features are interpreted by the shocked interstellar medium and the interaction region between the wind and CMZ gas. The thermal pressure and temperature obtained in our model are consistent with recent Suzaku observations.

We study the dynamics of super-Eddington accretion flows by performing two-dimensional radiationhydrodynamic simulations. Compared with previous works, in this paper we include the T θφ component of the viscous stress and consider various values of the viscous parameter α. We find that when T θφ is included, the rotational speed of the high-latitude flow decreases, while the density increases and decreases at the high and low latitudes, respectively. We calculate the radial profiles of inflow and outflow rates. We find that the inflow rate decreases inward, following a power law form ofṀ in ∝ r s . The value of s depends on the magnitude of α and is within the range of ∼ 0.4 − 1.0. Correspondingly, the radial profile of density becomes flatter compared with the case of a constantṀ (r). We find that the density profile can be described by ρ(r) ∝ r −p and the value of p is almost same for a wide range of α ranging from α = 0.1 to 0.005. The inward decrease of inflow accretion rate is very similar to hot accretion flows, which is attributed to the mass loss in outflows. To study the origin of outflow, we analyze the convective stability of the slim disk. We find that depending on the value of α, the flow is marginally stable (when α is small) or unstable (when α is large). This is different from the case of hydrodynamical hot accretion flow where radiation is dynamically unimportant and the flow is always convectively unstable. We speculate that the reason for the difference is because radiation can stabilize convection. The origin of outflow is thus likely because of the joint function of convection and radiation, but further investigation is required.