paper

Strong observational constraints on Advection-Dominated Accretion in the cores of elliptical galaxies

arXiv:astro-ph/9807245 · doi:10.1046/j.1365-8711.1999.02334.x

Abstract

The growing evidence for supermassive black holes in the centres of nearby galaxies has brought into sharper focus the question of why elliptical galaxies, rich in hot gas, do not possess quasar-like luminosities. Recent studies suggest that advection-dominated accretion flows (ADAFs), with associated low radiative efficiency, might provide an explanation for the observed quiescence. Although ADAF models have been applied to a number of low-luminosity systems, compelling observational evidence for their existence is still required. Here, we examine new high-frequency radio observations of the giant, low-luminosity elliptical galaxies NGC 4649, NGC 4472 and NGC 4636 made with the Very Large Array (VLA) and the sub-millimetre common-user bolometer array (SCUBA) on the JCMT. At these frequencies the predictions are very precise and an ADAF is unequivocally characterised by a slowly rising spectrum with a sharp spectral cut-off produced by thermal synchrotron radiation. The new radio limits severely disagree with the canonical ADAF predictions which significantly overestimate the observed flux. While our observations do not rule out the presence of an ADAF in these objects, they place strong constraints on the model. If the accretion in these objects occurs in an advection-dominated mode then our radio limits imply that the emission from their central regions must be suppressed. We examine the possibility that the magnetic field in the flow is extremely low or that synchrotron emission is free-free absorbed by cold material in the accretion flow. We also discuss whether slow non-radiating accretion flows may drive winds/outflows to remove energy, angular momentum and mass so that the central densities, pressures and emissivities are much smaller than in a standard ADAF.

submitted to MNRAS, 13 pages, 6 figures in MNRAS LaTex style

References in corpus (10)

Cited by in corpus (53)