Stability for multiple Lamb dipoles

Technical Summary: Stability for Multiple Lamb Dipoles

Problem Statement
This paper investigates the long-time dynamics of nonnegative vorticities evolving under the incompressible Euler equations in the upper half-plane $\mathbb{R}^2_+$. While global well-posedness is established by Yudovich theory, the asymptotic behavior of solutions remains a challenging problem. Specifically, the authors address the stability of finite linear superpositions of Lamb dipoles—traveling wave solutions that move to the right. The central question is whether a configuration of $N$ well-separated Lamb dipoles, ordered by velocity (faster dipoles to the right of slower ones), remains stable over time, or if interactions lead to collisions, filamentation, or structural breakdown.

Methodology
The authors employ a robust strategy combining sharp energy estimates near the Lamb dipoles with a Lagrangian bootstrapping scheme. The proof proceeds through the following key components:

1. Variational Characterization: The stability of a single Lamb dipole is grounded in its property as an energy maximizer within a class of vorticities with fixed $L^2$ norm (enstrophy) and weighted $L^1$ norm (impulse). This variational principle, established in prior work [1, 2], provides a baseline for Lyapunov stability.
2. Decomposition and Bootstrap: The solution $\omega(t, \cdot)$ is decomposed into $N$ disjoint pieces $\omega_i(t, \cdot)$ based on time-dependent "borders" $L_i(t)$, which track the expected positions of the dipoles. The authors introduce bootstrap assumptions on the enstrophy ($K_i$), impulse ($\mu_i$), and energy ($E_i$) of each piece.
3. Lagrangian Analysis of Interactions: A critical difficulty arises from "filamentation," where vorticity is advected from the front to the rear of a dipole, creating long tails. The authors analyze the exchange of circulation, enstrophy, impulse, and energy between adjacent dipoles.
   • Gain–Error Interaction: The most technically demanding part involves controlling the interaction between the "gain" (vorticity in a slower dipole that originated from a faster one) and the "error" (vorticity near the interface). Because the kernel defining impulse exchange is singular when particles are close, the authors switch to Lagrangian coordinates. They track the trajectories of "gain particles" to show that their contribution to impulse exchange is integrable in time and small, provided the initial separation is large.
   • Velocity Decay: The proof leverages the linear separation of dipoles. As the distance between dipoles grows linearly in time, the interaction velocity decays as $O(t^{-2})$, ensuring that the cumulative effect of mutual interactions remains bounded.
4. Closing the Bootstrap: By establishing that the time derivatives of enstrophy, impulse, and energy are sufficiently small (specifically, that the flux of these quantities across the moving borders is controlled), the authors close the bootstrap assumptions, proving that the solution remains close to the superposition of Lamb dipoles for all time.

Key Contributions and Results

• Theorem A (Stability of Linear Superposition): The primary result establishes the Lyapunov stability of a superposition of $N$ Lamb dipoles. Given initial data sufficiently close to a superposition of $N$ Lamb dipoles with strictly ordered velocities ($\bar{V}_1 > \bar{V}_2 > \dots > \bar{V}_N$) and large initial separation, the solution remains close to a time-dependent superposition of Lamb dipoles for all $t > 0$. The centers of the dipoles $p_i(t)$ satisfy an estimate $|p_i(t) - \bar{p}_i - \bar{V}_i t| \leq C\varepsilon^{1/2}(1+t)$, indicating that the dipoles maintain their relative order and do not collide.
• Theorem B (Separation of a Single Dipole): A secondary result demonstrates that a single Lamb dipole can be "separated out" from a more complex solution if the remaining part of the vorticity is initially slower and confined to a specific geometric region. This result holds without requiring detailed information about the slower part, relying only on a bound on its maximum possible velocity under rearrangement.
• Quantitative Control of Fluxes: The paper provides explicit bounds on the fluxes of enstrophy, impulse, and energy across the moving interfaces between dipoles. It quantifies how filamentation affects these quantities, showing that while filamentation occurs, the net transfer of circulation and impulse is insufficient to destabilize the configuration due to the rapid linear separation.

Significance and Claims
The authors claim that their strategy is robust and capable of generalization. They note that the result is nontrivial even for exact superpositions, as the velocity field generated by each dipole is not compactly supported and induces "lift-up" and "push-down" effects on neighbors. The condition that faster dipoles must be to the right of slower ones is essential; reversing this order would lead to collisions and instability.

The paper distinguishes its approach from previous works on point vortices and gKdV solitons. Unlike point vortices, Lamb dipoles have $O(1)$ energy and enstrophy, and unlike gKdV solitons, the family of Lamb dipoles is a multi-parameter family (amplitude, impulse, radius), complicating the variational analysis. The authors explicitly state that asymptotic stability (convergence to a specific state) is not known even for the single dipole case ($N=1$) and is not claimed here; the result is strictly one of Lyapunov stability (persistence of proximity).

The work connects to the broader theory of scattering for nonnegative vorticities in the half-plane, suggesting that solutions of Theorem A provide nontrivial limiting measures in the rescaled sense, potentially consisting of a finite sum of Dirac deltas corresponding to the individual dipoles. The authors also acknowledge that extending these results to other traveling waves (e.g., Sadovskii vortices) or countable sequences of dipoles presents significant challenges and remains an open problem.