Disk outflows and the accretion rate gap

We argue that the observed `accretion rate gap' - between black holes in radio-loud active galactic nuclei (AGN) accreting at close to the Eddington limit and those accreting at considerably lower rates - can be explained in terms of the adiabatic inflow-outflow (ADIOS) scenario for radiatively inefficient accretion. Whenever the accretion rate falls below a threshold value (corresponding to a luminosity L_crit) that depends on the viscosity parameter, alpha, the inner region of the accretion disk - extending from the marginally stable orbit to ~ 1000 Schwarzschild radii - is susceptible to becoming hot and radiatively inefficient. If this happens, the disk luminosity decreases by a factor of ~100, as most of the matter originally destined to be swallowed is instead expelled in a wind. According to our conjecture, accretion flows onto black holes never radiate steadily in the range ~ 0.01 L_crit < L < L_crit, hence the inferred accretion rate gap. We expect the gap to exist also for black holes in X-ray binaries, where it may be responsible for state transitions and the luminosity fluctuations associated with X-ray nova outbursts.