Reflex instabilities II: The reflex motion of the central object drives at least seven instabilities in accretion discs

Abstract

This series of papers studies the stability of accretion discs to the reflex motion of the central object under the gravitational pull of the disc itself. In this second paper, we use linear analysis to reproduce or discover seven distinct instabilities. We outline their defining characteristics and propose a classification. Of the four instabilities that had already been discovered, three are artificial: one does not survive the transition to modern surface density profiles such as power-laws, one is due to the location of the domain's boundaries, and one is due to the use of damping layers near the domain's boundaries. Only the SLING instability of Adams et al. (1989) and the three new 'cavity' instabilities seem capable of affecting accretion discs. All four can be described to a high degree of precision by WKB methods. They may drive eccentricity growth, affect planet migration, create spiral and crescent substructures in the absence of any planet, and boost accretion onto the central object. The issue is that the three artificial instabilities are extremely hard to avoid in simulations, and would bias the same observables. This suggests that our community may not have the tools to model the reflex motion of the central object yet.