Vortex atmospheres of traveling vortices: rigorous definition, existence, and topological classification

In incompressible and inviscid fluids, the vortex atmosphere refers to the collection of fluid particles outside the support of a traveling vortex that are nevertheless carried along with it. This phenomenon has been recognized since the nineteenth century, e.g., in the classical works of O. Reynolds [Nature, 1876] and O. Lodge [Lond. Edinb. Dubl. Phil. Mag., 1885], yet rigorous mathematical definitions and proofs have remained largely undeveloped, with most subsequent studies relying on thin-core approximations or asymptotic analyses. In this paper, we give a rigorous definition of a vortex atmosphere and establish its existence and uniqueness. We further compare the planar atmosphere surrounding a 2D vortex dipole with the axisymmetric atmosphere surrounding a 3D vortex ring. In particular, we emphasize and prove the topological distinctions observed by W. Hicks [Lond. Edinb. Dubl. Phil. Mag., 1919]: under natural assumptions, every 2D dipole with its atmosphere forms an oval-shaped region, whereas for 3D rings, both spheroidal and toroidal configurations may occur. Our proof is based on showing that each atmosphere can be characterized precisely as a specific superlevel set of its corresponding stream function.

💡 Research Summary

The paper provides a rigorous mathematical treatment of the “vortex atmosphere” – the region of irrotational fluid that travels together with a moving vortex while lying outside the vortex core. Working within the incompressible, inviscid Euler framework, the authors first define traveling vortex solutions in two settings: a two‑dimensional counter‑rotating vortex dipole and a three‑dimensional axisymmetric vortex ring without swirl. The core of a vortex is the bounded region where the vorticity is non‑zero; the atmosphere is the part of the surrounding fluid that is carried at the same constant speed as the core.

A central contribution is the definition of a “vortex domain” (Definitions 2.3 and 2.4). This is the maximal bounded open set Ω that satisfies a translation invariance property under the flow map Φ(t,·): for all t ≥ 0, Φ(t,Ω) = Ω shifted by the traveling speed W in the direction of motion. Any proper superset of Ω fails this property, which guarantees uniqueness of the vortex domain (Remark 2.5). The vortex atmosphere is then simply Ω minus the closure of the core. Existence of such a maximal set is proved in Proposition 2.6 using a variational approach that exploits energy maximization under Steiner symmetry and simple‑connectedness of the core.

The authors show that the boundary of the vortex domain can be described precisely as a super‑level set of the stream function ψ associated with the flow. In the irrotational region ψ satisfies Laplace’s equation, so its level sets are smooth curves (or surfaces) determined by the harmonic structure. This characterization allows the authors to translate geometric questions about the atmosphere into analytical properties of ψ.

For the two‑dimensional dipole, the odd symmetry of the vorticity forces the velocity at the dipole midpoint to exceed twice the translation speed. Consequently the midpoint lies inside the atmosphere, and the super‑level set {ψ ≥ c} is a single simply‑connected oval surrounding the pair of cores. This confirms Hicks’s 1919 observation that a dipole’s vortex domain is always oval‑shaped (Theorems 3.4 and 4.1). A special case, the Sadovskii dipole (touching counter‑rotating patches), yields an empty atmosphere, illustrating that the atmosphere need not exist for every dipole (Theorem 4.2).

In contrast, for an axisymmetric vortex ring the situation is richer. The authors identify three distinct configurations of the vortex domain, determined by the comparison between the axial velocity at the geometric centre of the ring (v_c) and the translation speed W:

1. v_c < W – the centre lags behind the ring; the super‑level set forms a toroidal (donut‑shaped) domain.
2. v_c = W – the centre moves exactly with the ring; the domain becomes a “revolved lemniscate” (∞‑shaped cross‑section).
3. v_c > W – the centre outruns the ring; the domain is spheroidal (ball‑like).

These three possibilities correspond to Hicks’s toroidal, lemniscate, and spheroidal configurations (Theorems 3.6, 3.7, 4.5). The proof hinges on monotonicity and sign properties of the velocity components derived from the harmonic stream function; the sign of the first and second derivatives of ψ in the radial and axial directions dictates whether particles near the centre are captured or left behind.

Section 3 presents concrete examples (Hill’s spherical vortex, Chaplygin–Lamb dipole, Norbury’s rings, etc.) and verifies that they satisfy the theoretical classification. Section 4 extends the analysis under the additional assumption that the core cross‑section is simply connected, a condition met by all known traveling vortex solutions.

Finally, the paper lists open problems: existence of traveling vortices with multiply‑connected cores, the effect of viscosity, non‑Steiner‑symmetric vorticity distributions, and possible extensions to three‑dimensional non‑axisymmetric rings.

In summary, the work establishes a solid mathematical foundation for the vortex atmosphere concept, proves existence and uniqueness of the associated vortex domain, and provides a complete topological classification of its possible shapes. By linking the geometry of the atmosphere to super‑level sets of a harmonic stream function, the authors bridge historical experimental observations with modern rigorous analysis, opening new avenues for both theoretical investigation and practical applications involving vortex transport.