Advection-Dominated Accretion and the Black Hole Event Horizon

As the luminosity of an accreting black hole drops to a few percent of Eddington, the spectrum switches from the familiar soft state to a hard state that is well-described by a distended and tenuous advection-dominated accretion flow (ADAF). An ADAF is a poor radiator, and the ion temperature can approach 10^{12} K near the center, although the electrons are cooler, with their temperature typically capped at ~10^{9-11} K. The foundational papers predicted that the large thermal energy in an ADAF would drive strong winds and jets, as later observed and also confirmed in computer simulations. Of chief interest, however, is the accreting gas that races inward. It carries the bulk of the accretion energy as stored thermal energy, which vanishes without a trace as the gas passes through the holes event horizon. One thus expects black holes in the ADAF regime to be unusually faint. Indeed, this is confirmed by a comparison of accreting stellar-mass black holes and neutron stars, which reside in very similar transient X-ray binary systems. The black holes are on average observed to be fainter by a factor of ~100-1000. The natural explanation is that a neutron star must radiate the advected thermal energy from its surface, whereas a black hole can hide the energy behind its event horizon. The case for an event horizon in Sagittarius A*, which is immune to caveats on jet outflows and is furthermore independent of the ADAF model, is especially compelling. These two lines of evidence for event horizons are impervious to counterarguments that invoke strong gravity or exotic stars.

Estimating the Spins of Stellar-Mass Black Holes by Fitting Their Continuum Spectra

We have used the Novikov-Thorne thin disk model to fit the continuum X-ray spectra of three transient black hole X-ray binaries in the thermal state. From the fits we estimate the dimensionless spin parameters of the black holes to be: 4U 1543-47, a* = a/M = 0.7-0.85; GRO J1655-40, a* = 0.65-0.8; GRS 1915+105, a* = 0.98-1. We plan to expand the sample of spin estimates to about a dozen over the next several years. Some unresolved theoretical issues are briefly discussed.