The quasi-steady structure of super-critical accretion flows around a black hole is studied based on the two-dimensional radiation-hydrodynamical (2D-RHD) simulations. The super-critical flow is composed of two parts: the disk region and the outflow regions above and below the disk. Within the disk region the circular motion as well as the patchy density structure are observed, which is caused by Kelvin-Helmholtz instability and probably by convection. The mass-accretion rate decreases inward, roughly in proportion to the radius, and the remaining part of the disk material leaves the disk to form outflow because of strong radiation pressure force. We confirm that photon trapping plays an important role within the disk. Thus, matter can fall onto the black hole at a rate exceeding the Eddington rate. The emission is highly anisotropic and moderately collimated so that the apparent luminosity can exceed the Eddington luminosity by a factor of a few in the face-on view. The massaccretion rate onto the black hole increases with increase of the absorption opacity (metalicity) of the accreting matter. This implies that the black hole tends to grow up faster in the metal rich regions as in starburst galaxies or star-forming regions.

We discuss distinctive features of luminous accretion disks shining at the Eddington luminosity in the context of galactic black-hole candidates (GBCs). We first note that the standard-disk picture is not applicable, although it is often postulated. Rather, the disk becomes advection-dominated while remaining optically thick (the so-called slim disk). The slim disk exhibits several noteworthy signatures: (1) The disk luminosity is insensitive to the mass-flow rates, Ṁ, and is always kept around the Eddington luminosity, LE, even if Ṁ greatly exceeds LE/c2. This reflects the fact that radiative cooling is no longer balanced by viscous heating and excess energy is carried by accreting matter to black holes. (2) The spectra of the slim disks are multi-color blackbody characterized by (i) a high maximum temperature, kTin ∼ a few keV, (ii) a small size of an emitting region, rin < 3 rg (with rg being Schwarzschild radius), due to substantial radiation coming out from inside 3 rg, and (iii) flatter spectra in the soft-X bands, vSv ∼ v0, because of a flatter effective temperature profile of the slim disk, Teff ∝ r−1/2 (in contrast with Teff ∝ r−3/4 in the standard disk). Thus, a small rin (≪ 3rg) does not necessarily mean the presence of a Kerr hole. Furthermore, (3) as Ṁ increases, Tin increases, while rin decreases as rin ∝ (Tin)−1 approximately. That is, the changes in rin derived from the fitting do not necessarily mean the changes in the physical boundary of the optically thick portions of the disk. Observational implications are discussed in relation to binary jet sources.

Aperiodic optical variability is a common property of Active Galactic Nuclei (AGNs), though its physical origin is still open to question. To study the origin of the optical -ultraviolet variability in AGN, we compare light curves of two models to observations of quasar 0957+561 in terms of a structure function analysis. In the starburst (SB) model, random superposition of supernovae in the nuclear starburst region produce aperiodic luminosity variations, while in the disk-instability (DI) model, variability is caused by instabilities in the accretion disk around a supermassive black hole. We calculate fluctuating light curves and structure functions, V (τ ), by simple Monte-Carlo simulations on the basis of the two models. Each resultant V (τ ) possesses a power-law portion, [V (τ )] 1/2 ∝ τ β , at short time lags (τ ). The two models can be distinguished by the logarithmic slope, β; β ∼ 0.74-0.90 in the SB model and β ∼ 0.41-0.49 in the DI model, while the observed light curves exhibit β ∼ 0.35. Therefore,we conclude that the DI model is favored over the SB model to explain the slopes of the observational structure function, in the case of 0957+561, though this object is a radio-loud object and thus not really a fair test for the SB model. In addition, we examine the time-asymmetry of the light curves by calculating V (τ ) separately for brightening and decaying phases. The two models exhibit opposite trends of time-asymmetry to some extent, although the present observation is not long enough to test this prediction.

We present the detailed global structure of black hole accretion flows and outflows through newly performed two-dimensional radiation-magnetohydrodynamic simulations. By starting from a torus threaded with weak toroidal magnetic fields and by controlling the central density of the initial torus, ρ 0 , we can reproduce three distinct modes of accretion flow. In model A with the highest central density, an optically and geometrically thick supercritical accretion disk is created. The radiation force greatly exceeds the gravity above the disk surface, thereby driving a strong outflow (or jet). Because of the mild beaming, the apparent (isotropic) photon luminosity is ∼ 22L E (where L E is the Eddington luminosity) in the face-on view. Even higher apparent luminosity is feasible if we increase the flow density. In model B with a moderate density, radiative cooling of the accretion flow is so efficient that a standard-type, cold, and geometrically thin disk is formed at radii greater than ∼ 7R S (where R S is the Schwarzschild radius), while the flow is radiatively inefficient otherwise. The magnetic-pressure-driven disk wind appears in this model. In model C the density is too low for the flow to be radiatively efficient. The flow thus becomes radiatively inefficient accretion flow, which is geometrically thick and optically thin. The magnetic-pressure force, in cooperation with the gas-pressure force, drives outflows from the disk surface, and the flow releases its energy via jets rather than via radiation. Observational implications are briefly discussed.

Narrow-line Seyfert 1 galaxies (NLS1s) exhibit extreme soft X-ray excess and large variability. We argue that both features can be basically accounted for by the slim disk model. We assume that a central black-hole mass in NLS1 is relatively small, M ∼ 10 5−7 M ⊙ , and that a disk shines nearly at the Eddington luminosity, L E . Then, the disk becomes a slim disk and exhibits the following distinctive signatures: (1) The disk luminosity (particularly of X-rays) is insensitive to mass-flow rates,Ṁ , since the generated energy is partly carried away to the black hole by trapped photons in accretion flow. (2) The spectra are multi-color blackbody. The maximum blackbody temperature is T bb ≃ 0.2(M/10 5 M ⊙ ) −1/4 keV, and the size of the blackbody emitting region is small, r bb < ∼ 3r S (with r S being Schwarzschild radius) even for a Schwarzschild black hole. (3) All the ASCA observation data of NLS1s fall onto the region ofṀ /(L E /c 2 ) > 10 (with L E being the Eddington luminosity) on the (r bb , T bb ) plane, supporting our view that a slim disk emits soft X-rays at ∼ L E in NLS1s. (4) Magnetic energy can be amplified, at most, up to the equipartition value with the trapped radiation energy which greatly exceeds radiation energy emitted from the disk. Hence, energy release by consecutive magnetic reconnection will give rise to substantial variability in soft X-ray emission.