Rossby Wave Instability with Self-Gravity

TL;DR

This paper investigates how self-gravity influences the Rossby wave instability in thin discs, identifying key parameters that determine stability and providing criteria for when self-gravity must be considered in models.

Contribution

It extends the analysis of Rossby wave instability to include self-gravity effects across a range of disc models, deriving conditions for instability based on disc parameters.

Findings

Self-gravity significantly affects instability for low Toomre Q values and certain azimuthal wavenumbers.

Instability can occur at surface density depressions or bumps depending on self-gravity and Q.

Self-gravity is negligible for high Q discs with certain mode numbers, simplifying modeling efforts.

Abstract

The Rossby wave instability (RWI) in non-self-gravitating discs can be triggered by a bump at a radius $r_0$ in the disc surface mass-density (which is proportional to the inverse potential vorticity). It gives rise to a growing non-axisymmetric perturbation [$\propto \exp(im\phi)$, $m=1,2..$] in the vicinity of $r_0$ consisting of anticyclonic vortices which may facilitate planetesimal growth in protoplanetary discs. Here, we analyze a continuum of thin disc models ranging from self-gravitating to non-selfgravitating. The key quantities determining the stability/instability are: (1) the parameters of the bump (or depression) in the disc surface density, (2) the Toomre $Q$ parameter of the disc (a non-self-gravitating disc has $Q\gg1$), and (3) the dimensionless azimuthal wavenumber of the perturbation $\bar{k}_\phi =mQh/r_0$, where $h$ is the half-thickness of the disc. For discs stable to axisymmetric perturbations ($Q>1$), the self-gravity has a significant role for $\bar{k}_\phi < \pi/2$ or $m<(\pi/2) (r_0/h)Q^{-1}$; instability may occur for a depression or groove in the surface density if $Q\lesssim 2$. For $\bar{k}_\phi > \pi/2$ the self-gravity is not important, and instability may occur at a bump in the surface density. Thus, for all mode numbers $m \ge 1$, the self-gravity is unimportant for $Q > (\pi/2)(r_0/h)$. We suggest that the self-gravity be included in simulations for cases where $Q< (r_0/h)$.