Rossby Wave Instability: How Disk Edges Spin Up Giant Vortices

What the Rossby Wave Instability Is

The Rossby Wave Instability is a global, non-axisymmetric hydrodynamic instability of a differentially rotating disk. It was introduced in a trio of papers by Richard Lovelace, Hui Li, Steve Colgate, and collaborators (Lovelace et al. 1999; Li et al. 2000, 2001), who showed that a thin gaseous disk becomes unstable if a certain radial profile develops a local extremum.

The key quantity is vortensity (also called potential vorticity), the ratio of the flow's vorticity to its surface density. In its barotropic form the RWI needs an extremum in Σ/ω_z; in the general adiabatic case the relevant function is

• L(r) = (Σ/ω_z) · (P/Σ^γ)^(2/γ), with Σ the surface density, ω_z the vertical vorticity, P the pressure, and γ the adiabatic index.

A local minimum in L — typically produced by a pressure bump — is the trigger. Physically, the bump acts as a resonant cavity in which Rossby waves (waves restored by the radial gradient of vortensity) become trapped, reflect, and reinforce one another, growing exponentially instead of dispersing.

The Mechanism: Trapped Rossby Waves Rolling Up

Rossby waves owe their existence to the conservation of vortensity: displace a fluid parcel radially across a vortensity gradient and it develops a restoring circulation, exactly as planetary Rossby waves arise from the latitude gradient of Coriolis force on Earth. In a disk, a sharp bump creates two opposing vortensity gradients — one on each flank of the bump — so Rossby waves can propagate on both sides.

At the bump these waves become trapped between corotation and the Lindblad resonances. Corotation sits at the extremum, and the trapped disturbance extracts angular momentum from one side and deposits it on the other, amplifying the perturbation. This over-reflection is the engine of the instability.

• Linear phase: a non-axisymmetric mode of azimuthal number m grows exponentially, ∝ exp(γ_g t), with growth rate γ_g a fraction of the orbital frequency Ω.
• Non-linear phase: the growing wave saturates and the flow rolls up into m discrete anticyclonic vortices that then merge, usually leaving one or a few large, long-lived vortices.

Because the vortices rotate opposite to the shear (anticyclonic), they are local pressure maxima.

Characteristic Numbers and a Worked Example

How steep must a bump be? Linear analyses show the RWI switches on when the density or pressure contrast at the edge exceeds roughly a factor of ~2 over a radial width comparable to the local scale height H. Equivalently, the bump width should be of order H = c_s/Ω, where c_s is the sound speed.

• Growth rate: γ_g ≈ 0.1–0.4 Ω for the fastest modes, so with an orbital period of ~500 yr at 60 AU around a 1 M_sun star, e-folding takes only a few orbits — a few thousand years, negligible next to the ~1–10 Myr disk lifetime.
• Preferred mode: broad, shallow bumps favor low m (m = 1–2, giving one big vortex); narrow, sharp edges excite higher m (up to ~5).
• Dust trapping: a vortex most efficiently concentrates grains with Stokes number St ≈ 1 — millimeter-to-centimeter pebbles at tens of AU — which is exactly the size ALMA sees piling up.

Example: a 5-Jupiter-mass planet at 20 AU carves a gap whose outer edge is a pressure bump; the RWI there spins up an m = 1 vortex ~10–20 AU across that survives hundreds of orbits and traps a lopsided crescent of pebbles.

How the RWI Is Observed

The RWI is not seen directly — we see its consequence: lopsided, crescent-shaped concentrations of millimeter dust in the (sub)millimeter continuum. Because an anticyclonic vortex is a pressure maximum, drifting pebbles pile up there while the gas stays comparatively smooth, producing a bright arc against a faint ring.

• Oph IRS 48 shows the most dramatic example: a highly asymmetric dust crescent imaged by ALMA (van der Marel et al. 2013), widely interpreted as a giant vortex-trapped dust trap near ~60 AU.
• HD 142527 and HD 34282 host strong azimuthal asymmetries consistent with vortex dust traps.
• Diagnostics beyond morphology include the gas kinematics (a vortex imprints a characteristic velocity signature in molecular line channel maps) and scattering-induced polarization, which probes the dust scale height inside the crescent.

Crucially, a dust crescent that is asymmetric in millimeter grains but nearly symmetric in gas and small grains is a fingerprint of RWI-driven trapping rather than, say, a simple eccentric cavity.

How It Differs From Its Cousins

The RWI is easy to confuse with other disk instabilities, but its trigger and outcome are distinctive.

• Rayleigh instability is axisymmetric (m = 0) and requires angular momentum to decrease outward; the RWI is non-axisymmetric and needs only a vortensity extremum, a far milder condition. A useful heuristic is that a bump can reach the RWI threshold while still only 'halfway to Rayleigh.'
• Magnetorotational instability (MRI) is magnetic and produces sustained small-scale turbulence; the RWI is purely hydrodynamic and produces a few coherent vortices. Ironically, the sharp viscosity transition at a dead-zone edge (where MRI shuts off) is a classic site for triggering the RWI.
• Gravitational instability needs a massive disk (Toomre Q ≲ 1) and makes spiral arms or bound clumps; the RWI works in low-mass disks and makes vortices, not spirals.

What all RWI sites share is a sharp radial feature — a planet-gap edge, a dead-zone boundary, or the inner rim near the star.

Significance, Open Questions, and Famous Cases

The RWI matters because it may solve a central problem in planet formation: the radial-drift (or 'meter-size') barrier. Left alone, pebbles spiral into the star in a few thousand years; a vortex pressure trap halts that drift and boosts the local solid-to-gas ratio by orders of magnitude, plausibly enough to trigger the streaming instability and build planetesimals.

• Famous case: Oph IRS 48, whose extreme crescent is the poster child for a giant dust-trapping vortex.
• Self-triggering loop: a planet opens a gap, the gap edge triggers the RWI, and the resulting vortex may itself concentrate solids to form more planets.

Open questions remain. In 3D, vortices are subject to the elliptic instability and can be partially destroyed, so their true lifetimes and vertical structure are debated. Dust feedback and self-gravity can weaken or disrupt a vortex once enough solids accumulate. And not every crescent must be a vortex — eccentric cavities, binary companions, and disk winds can mimic the signature — so distinguishing genuine RWI vortices from look-alikes with kinematics and polarization is an active observational frontier.

Rossby Wave Instability compared with related disk instabilities

Instability	Driving condition	Outcome	Characteristic mode/scale
Rossby Wave Instability (RWI)	Extremum in vortensity (density/pressure bump)	Few large anticyclonic vortices, dust trapping	Low m (1–5), grows in a few–tens of orbits
Rayleigh instability	Specific angular momentum decreases outward (dΩ²R⁴/dR < 0)	Axisymmetric overturning; violent	Axisymmetric (m = 0), local
Magnetorotational instability (MRI)	Weak B-field + outward-decreasing Ω, ionized gas	Sustained MHD turbulence, angular-momentum transport	Small-scale, needs coupling to field
Gravitational instability (GI)	Toomre Q ≲ 1 (massive, cool disk)	Spiral arms, fragmentation into clumps	Global spirals; can form bound objects
Vertical shear instability (VSI)	Vertical gradient in rotation + fast cooling	Weak vertical turbulence, small α	Small radial scale, corrugation modes