Point Vortex Dynamics on Closed Surfaces

• Point vortex dynamics on closed surfaces are the study of singular vorticity on compact, boundaryless 2D Riemannian manifolds with an intrinsic Hamiltonian structure.
• The dynamics incorporate geometric effects such as curvature, topology, and the Robin function, which collectively influence vortex drift, stability, and energy exchange.
• Efficient computational methods, leveraging conformal mappings and uniformization, enable simulations that bridge classical fluid behavior with quantum hydrodynamic applications.

Point vortex dynamics on closed surfaces is the study of the motion and hydrodynamics of localized singularities of vorticity—idealized as Dirac measures—confined to compact, boundaryless two-dimensional Riemannian manifolds. This subject generalizes the classical planar N-vortex problem to arbitrary closed geometries, such as spheres, tori, surfaces of revolution, and non-orientable surfaces, and explores both the microscopic Hamiltonian dynamics and the emergent collective (coarse-grained) behaviors in the large-vortex limit. Curvature, topology, and global geometric properties fundamentally affect both the equations of motion and the stationary or equilibrium properties of vortex ensembles. Recent advances have clarified the precise geometric and topological origins of curvature-vortex coupling, odd viscosity, energy exchange mechanisms, and stability or existence conditions for equilibria.

1. Geometric and Analytical Foundation

A closed surface $M$ is a compact, smooth, two-dimensional manifold without boundary, equipped with a Riemannian metric $g$. Let $dV = \sqrt{\det g}\,d^2x$ denote the area form, and $\nabla_a$ the Levi-Civita connection. The Laplace–Beltrami operator is $\Delta = g^{ab}\nabla_a\nabla_b$. For the point vortex system, the vorticity is modeled as a sum of weighted Dirac measures: $\omega(x) = \sum_{i=1}^N \Gamma_i \delta(x, x_i),$ where $x_i \in M$ and $\Gamma_i \in \mathbb{R}\setminus\{0\}$ are the positions and strengths (circulations) of the vortices. On compact surfaces, incompressibility typically requires $\sum_i \Gamma_i = 0$ to ensure global solubility of the stream function equation $-\Delta \psi = \omega$ (Bogatskiy, 2019, Padilla, 3 Feb 2026).

The hydrodynamic Green's function $G(x, y)$ of $-\Delta$ is defined by

$-\Delta_x G(x, y) = \delta(x, y) - \frac{1}{V},$

with $\int_M G(x, y) dV(x) = 0$. Locally,

$G(x, y) = -\frac{1}{2\pi} \ln d(x, y) + h(x, y),$

where $h$ is smooth (Bogatskiy, 2019, Ahmedou et al., 2022).

2. Hamiltonian Structure and Equations of Motion

The equations governing point vortex dynamics on closed surfaces are Hamiltonian,

$$\dot{x}_i^a = \frac{1}{\Gamma_i}\epsilon^{ab}(x_i)\nabla_{b, i}H_N,$$

where $\epsilon^{ab}$ is the contravariant Levi–Civita tensor and the Kirchhoff–Routh energy is

$$H_N(\{x_i\}) = -\frac{1}{2}\sum_{i\neq j}\Gamma_i\Gamma_j G(x_i, x_j) + \frac{1}{2} \sum_{i=1}^{N} \Gamma_i^2 G^R(x_i).$$

Here, $$G^R(x) = \lim_{y\to x}(G(x, y) + \frac{1}{2\pi} \ln d(x,y))$$ is the Robin function, regularizing the vortex self-energy (Bogatskiy, 2019, Ahmedou et al., 2022, D'Aprile et al., 2015, Gustafsson, 2018, Padilla, 3 Feb 2026). In multiple connected surfaces, global circulation variables appear, coupled through harmonic 1-forms and period matrices (Gustafsson, 2022, Gustafsson, 2018).

Hamilton’s equations can equivalently be written as

$$\Gamma_i \sqrt{g(x_i)}\,\epsilon_{ab}(x_i)\dot{x}_i^b = \nabla_{a,i} H_N,$$

which defines a symplectic flow on the phase space of vortex positions, possibly augmented by global circulation variables in genus $g > 0$ (Gustafsson, 2022, Gustafsson, 2018).

3. Curvature, Affine Connections, and Odd Viscosity

Curvature fundamentally modifies vortex dynamics relative to the plane. The Gaussian curvature $K$ appears explicitly in both microscopic dynamics (Robin function and velocity drift) and macroscopic continuum descriptions:

• The microscopic drift of a single vortex at $z_k$ in local chart $z$ is given by the difference of two affine connections (Levi–Civita for the metric and that derived from the Robin function):

$$\lambda(z_k)^2 \frac{dz_k}{dt} = \frac{\Gamma_k}{4\pi i}\left( \overline{r_{\text{metric}}(z_k)} - \overline{r_{\rm robin}(z_k)} \right),$$

where $$r_{\mathrm{metric}}(z) = 2\,\partial_z \log \lambda(z)$$, and $r_{\mathrm{robin}}(z) = -2 c_1(z)$ with $c_1(z)$ from the local Robin expansion (Gustafsson, 2018, Gustafsson, 2022).

• For large ensembles, hydrodynamic equations include curvature anomalies. On a genus-zero closed surface, the coarse-grained vortex velocity field $v^\alpha$ satisfies the covariant equation:

$$\partial_t(\rho v^\alpha) + \nabla_\beta T^{\alpha\beta} + \rho\nabla^\alpha p = \eta K\left( \eta\frac{\sigma}{\rho}\nabla^\alpha \sigma - 2\sigma\epsilon^{\alpha}_{\;\;\beta}v^\beta \right),$$

where $\eta = \kappa/8\pi$, with $\kappa = 2\pi\hbar/m$; $K$ is the Gaussian curvature and the right side encodes the macroscopic curvature anomaly (Xiong et al., 2023).

Odd (Hall) viscosity is recovered as $\eta_o = \Gamma/(8\pi)$ in the continuum limit, with emergent non-dissipative stress terms arising from both interactions and the curvature of the Robin function (Bogatskiy, 2019, Xiong et al., 2023).

4. Stationary Solutions and Geometric Potential

On the round sphere ($S^2$), the simplest nontrivial stationary solution has uniform background vortex density. The coarse-grained stationary flow features a dipole-like charge density ($\sigma$) and azimuthal flow: $$\sigma(z) = \rho_0 \frac{R^2 - |z|^2}{R^2 + |z|^2} \leftrightarrow \rho_0 \cos\theta,$$

$$v^\phi = R\left(4\pi\eta\rho_0 - \frac{\eta}{R^2}\right),$$

analogous to the Rossby–Haurwitz waves of geophysical fluid dynamics (Xiong et al., 2023). The difference between the vortex velocity field and the underlying fluid velocity is purely a curvature effect, vanishing in the infinite-radius (planar) limit. Curvature acts as a geometric potential: positive vortices are attracted to positive curvature regions, and antivortices repelled, generalizing the classical geometric potential familiar from few-vortex mechanics (Xiong et al., 2023).

On arbitrary closed surfaces, relative equilibria and rigidly-rotating solutions have been constructed, including two-ring vortex configurations on surfaces of revolution, which underlie rotating periodic solutions of the Gross–Pitaevskii equation (Chen, 2014).

5. Existence and Classification of Equilibria

The existence and classification of equilibrium configurations depend essentially on the topology of the surface:

• On higher-genus orientable or nonorientable surfaces (other than $S^2$ and $\mathbb{RP}^2$), variational and min–max schemes guarantee, for generic vortex strengths avoiding collapse, the existence of critical points (equilibria) of the Kirchhoff–Routh Hamiltonian (Ahmedou et al., 2022, D'Aprile et al., 2015). The precise topological distinction is whether the fundamental group is nontrivial.
• On the sphere ($S^2$), special algebraic or combinatorial constraints on the vortex strengths and positions are required. For example, equilibrium of three vortices is possible only when the configuration lies on a great circle, satisfying a specific linear condition.
• On non-orientable surfaces (e.g., the Klein bottle), Hamiltonian vortex dynamics can be formulated locally, but global properties (such as conserved center-of-vorticity and relative equilibria) reflect the twisted geometry and altered Green’s function structure (Balabanova, 2022).

The equilibria correspond to steady solutions of the incompressible Euler equations, and their limiting sets match those in singular mean-field Liouville equations describing vortex concentration (D'Aprile et al., 2015, Ahmedou et al., 2022).

6. Connections to Hydrodynamic Limits and Quantum Vortices

In the hydrodynamic (mean-field or large-N) limit, point vortex systems approach coarse-grained vortex fluid descriptions. On genus-zero closed surfaces, the macroscopic equations inherit a curvature anomaly, and the integrated effect can be traced, via Gauss–Bonnet, to the topological Euler characteristic. This topological coupling underpins the anomalous odd viscosity and bulk torque in vortex matter (Xiong et al., 2023, Bogatskiy, 2019).

For quantum fluids (e.g., Bose–Einstein condensates), the Gross–Pitaevskii equation on a closed surface supports vortex solutions whose centers evolve according to point-vortex dynamics in the $\varepsilon \to 0$ limit; this has been rigorously demonstrated for symmetric surfaces of revolution with rotating vortex rings (Chen, 2014).

Bernoulli-type laws and hydrodynamic conservation laws extend naturally to include contributions from both regular and singular vorticity, with the pressure augmented by the kinetic energy of the singular part (Shimizu, 2020).

7. Computational Methods and Numerical Algorithms

Efficient simulation of point vortex dynamics on closed genus-zero surfaces employs the conformal uniformization theorem: each such surface admits a conformal map to the sphere, allowing pullback of the spherical Green’s function and implementation of the metric Hamiltonian

$$H_M(\{q_i\}) = H_{S^2}(\{p_i=f(q_i)\}) - \frac{1}{4\pi}\sum_i \Gamma_i^2 \ln h(p_i),$$

with $h$ the conformal factor (Padilla, 3 Feb 2026). The velocity ODEs on $S^2$ are advected along geodesics and mapped back via the inverse uniformization, enabling highly efficient O(N²) algorithms, suitable for moderate vortex numbers. Larger ensembles employ tree codes or fast multipole methods.

Best practices include enforcing vortex neutrality, area-weighted sampling, and employing RK4 integrators with small time steps in stiff regimes. End-use applications include visualization of vortex flow on triangulated genus-zero surfaces (“bunny” or “bear” meshes), as well as accurate reproduction of classical leapfrogging and merging phenomena (Padilla, 3 Feb 2026).

---

The interplay of geometry, curvature, and topology with vortex dynamics on closed surfaces yields a hierarchy of phenomena not present in the plane, manifesting in modified energy landscapes, emergent odd viscosity, curvature-driven drift, and new families of stationary solutions, with rigorous connections to quantum hydrodynamics and geometric analysis (Xiong et al., 2023, Bogatskiy, 2019, Ahmedou et al., 2022, Gustafsson, 2022, Padilla, 3 Feb 2026, Gustafsson, 2018, Chen, 2014).