Transverse Instability of Stokes Waves at Finite Depth

This paper rigorously proves that Stokes waves at finite depth exhibit transverse spectral instability, characterized by unstable eigenvalues lying approximately on an ellipse, for all but finitely many depth values, thereby extending previous infinite-depth results despite the increased mathematical complexity.

The Gist

Imagine the ocean as a giant, endless trampoline. When you jump on it, you create a wave that travels across the surface. In the world of physics, a specific type of perfect, repeating wave is called a Stokes wave. For over a century, scientists have studied these waves, mostly focusing on what happens if you push them from the front or the back (longitudinal instability).

But what happens if you push them from the side? This is the question of Transverse Instability.

The Big Discovery

In 1981, a scientist named McLean used a computer to simulate these waves and found something surprising: if you nudge a Stokes wave from the side, it doesn't just wobble; it can actually break apart or grow chaotic. However, for 40 years, no one could prove this mathematically. It was like seeing a magician pull a rabbit out of a hat but not being able to explain how the trick worked.

This paper, written by Creedon, Nguyen, and Strauss, finally provides the rigorous mathematical proof that these waves are indeed unstable from the side, but with a very specific catch: it depends on how deep the water is.

The Analogy: The Tightrope Walker

Think of a Stokes wave as a tightrope walker balancing on a high wire.

• The Wave: The walker is perfectly balanced, moving forward.
• The Depth (h): The height of the wire above the ground.
• The Instability: A gust of wind blowing from the side (transverse perturbation).

The authors prove that for almost any height (depth) of the wire, a strong enough side wind will make the walker lose balance and fall (the wave becomes unstable). However, there are a few "magic heights" where the walker is so perfectly balanced that the side wind has no effect—at least for a little while.

The "Isola" (The Floating Island)

One of the coolest concepts in the paper is the shape of the instability. When the wave does become unstable, the mathematical description of its growth forms a shape called an isola (Italian for "island").

Imagine a map of the ocean. Most of the map is calm water (stable). But in the middle of the ocean, there is a floating island of chaos (unstable).

• The Shape: This island is shaped like an ellipse (a stretched circle).
• The Size: The size of this island depends on how big the wave is. If the wave is tiny, the island of chaos is tiny. If the wave is bigger, the island grows.
• The Center: The center of this island sits on a line representing "no growth." The island floats slightly off-center, meaning the wave doesn't just grow; it also shifts its rhythm slightly as it breaks.

The "Magic Depth" Exception

The most fascinating part of this research is the exception. The authors found that while this instability happens for almost every depth of water, there is one specific depth (about 0.25 meters in their normalized units) where the instability vanishes.

Think of it like a specific frequency of sound that makes a glass shatter. If you change the water depth just a tiny bit, the glass shatters. But at that exact magical depth, the glass is immune to that specific sound. The paper proves that for all other depths, the "glass" (the wave) will eventually shatter.

Why is this hard?

The authors had to do some incredibly complex math to prove this.

1. The Transformation: They had to flatten the wavy surface of the ocean into a flat mathematical strip to make the equations solvable. It's like trying to study the wrinkles on a crumpled piece of paper by magically ironing it flat without tearing it.
2. The Calculations: They had to expand their equations into long series (like peeling an onion layer by layer) up to the third or fourth layer. This required using powerful computers (Mathematica) because the formulas were too messy for a human to write out by hand.
3. The Proof: They showed that for almost all depths, the "island of chaos" (the ellipse) exists and has a positive size, meaning the wave will grow unstable.

The Takeaway

This paper closes a 40-year-old gap in our understanding of water waves. It confirms that:

1. Side-winds matter: Stokes waves are inherently unstable if you push them from the side.
2. Depth is key: The depth of the water changes how they become unstable.
3. There is a "sweet spot": There is one very specific depth where this instability disappears, acting as a mathematical shield for the wave.

In short, the ocean is a bit more chaotic than we thought, but the chaos follows a very precise, beautiful, and predictable mathematical pattern—unless you hit that one magical depth.

Technical Summary

1. Problem Statement

The paper addresses the transverse instability of small-amplitude Stokes waves (periodic traveling water waves) in a fluid of finite depth ($h$).

• Context: Stokes waves are steady, two-dimensional solutions to the Euler equations for irrotational, incompressible, inviscid flow. While their longitudinal (modulational) instabilities are well-understood, their stability against three-dimensional perturbations (variations in the transverse direction $y$) has been a long-standing open problem.
• Background: Numerical studies by McLean (1981) and others suggested that Stokes waves are unstable to transverse perturbations, with unstable eigenvalues forming an "isola" (an isolated closed curve) in the complex spectral plane. A rigorous mathematical proof existed for the infinite depth case (Nguyen & Strauss, 2023), but the finite depth case remained unproven due to significantly increased algebraic complexity.
• Objective: To rigorously prove that the linearized water wave operator at a small Stokes wave in finite depth possesses unstable eigenvalues (with positive real parts) for almost all depth values $h$, thereby confirming the existence of transverse instability.

2. Methodology

The authors employ a combination of spectral theory, perturbation analysis, and conformal mapping techniques.

A. Mathematical Formulation

• Governing Equations: The system is modeled using the Zakharov-Craig-Sulem formulation, reducing the 3D Euler equations to evolution equations on the free surface involving the Dirichlet-Neumann operator (DNO), $G(\eta)$.
• Linearization: The system is linearized around a small-amplitude Stokes wave (parameterized by amplitude $\varepsilon$). The perturbation is assumed to have the form $e^{\lambda t + i\alpha y}$, where $\alpha$ is the transverse wavenumber and $\lambda$ is the growth rate.
• Conformal Mapping: To handle the moving boundary, the authors use a Riemann mapping (conformal map) to flatten the fluid domain from a strip with a wavy top boundary to a fixed rectangular strip. This transforms the DNO into a pseudo-differential operator $\mathcal{S}_{\varepsilon, \beta}$ (where $\beta = \alpha^2$) acting on a fixed domain.

B. Spectral Reduction (Kato's Method)

• Resonance Condition: The analysis focuses on the case where the unperturbed operator ($\varepsilon=0$) has a double imaginary eigenvalue $i\sigma$. This occurs at a specific resonant transverse wavenumber $\beta^*$ (derived from the condition that two distinct Fourier modes collide on the imaginary axis).
• Reduction to $2 \times 2$ Matrix: Using Kato's perturbation theory, the authors construct a perturbed basis for the eigenspace associated with the double eigenvalue. This reduces the infinite-dimensional spectral problem to the analysis of a $2 \times 2$ matrix $\mathbf{L}_{\varepsilon, \delta}$ (where $\delta$ is a perturbation of $\beta$ around $\beta^*$).
• Hamiltonian Structure: The reduced matrix retains the Hamiltonian structure of the original system, allowing the eigenvalues to be expressed in terms of the matrix entries.

C. Asymptotic Expansions

• High-Order Expansions: The core technical challenge is expanding the entries of the matrix $\mathbf{L}_{\varepsilon, \delta}$ in powers of the amplitude $\varepsilon$ and the detuning parameter $\delta$.
• Third-Order Requirement: Unlike the infinite depth case where second-order terms might suffice for certain insights, the finite depth case requires expansions up to third order ($O(\varepsilon^3)$) to capture the leading-order instability mechanism.
• Computational Assistance: Due to the extreme algebraic complexity of the coefficients (involving hyperbolic functions of depth and wavenumbers), the authors utilized Mathematica to derive and verify the explicit expressions for the expansion coefficients.

3. Key Contributions and Results

A. Main Theorem (Theorem 1.1)

The paper proves that for any given depth $h \in (0, \infty)$, except for a finite number of exceptional values, small-amplitude Stokes waves are transversely unstable.

• Instability Mechanism: There exist parameters $\varepsilon$ (amplitude) and $\delta$ (transverse wavenumber detuning) such that the linearized operator has a pair of eigenvalues:
  $$\lambda_{\pm} = i\left(\sigma + \frac{1}{2}T(\varepsilon, \delta)\right) \pm \frac{1}{2}\sqrt{\Delta(\varepsilon, \delta)}$$
  where $\Delta(\varepsilon, \delta) > 0$ for small $\varepsilon$, implying $\text{Re}(\lambda_+) > 0$.
• Growth Rate: The growth rate of the instability scales as $O(\varepsilon^3)$.
• Spectral Shape: The unstable eigenvalues lie approximately on an ellipse (an isola) centered on the imaginary axis in the complex plane. The semi-axes of this ellipse scale as $O(\varepsilon^3)$.

B. The Exceptional Depth

• The proof relies on a key coefficient, $b_{3,0}$, which appears in the expansion of the off-diagonal term of the reduced matrix.
• Lemma 5.3: The authors prove that $b_{3,0}(h)$ is an analytic function of depth $h$.
  • As $h \to 0^+$, $b_{3,0} \to +\infty$.
  • As $h \to \infty$, $b_{3,0} \to b_{3,0, \infty} < 0$ (consistent with the infinite depth result).
• Conclusion: By the Intermediate Value Theorem, $b_{3,0}$ must have at least one zero. Since it is analytic, it has only a finite number of zeros.
• Numerical Finding: Numerical computation identifies exactly one exceptional depth $h_{crit} \approx 0.25065$. At this specific depth, the $O(\varepsilon^3)$ instability vanishes (though higher-order instabilities may still exist).

C. Validation

• The analytical results (the shape and location of the instability isola) are compared with numerical computations using the Floquet-Fourier-Hill method.
• Figure 1 shows excellent agreement between the analytical ellipse and numerical data for depths $h=1, 1.5, 2$, with an error of $O(\varepsilon^4)$.

4. Significance

1. Rigorous Confirmation: This work provides the first rigorous mathematical proof of transverse instability for Stokes waves in finite depth, resolving a problem that has been open since the 1980s.
2. Finite vs. Infinite Depth: It highlights the subtle differences between finite and infinite depth. While the qualitative behavior (instability) is the same, the finite depth case requires higher-order expansions and reveals the existence of specific "exceptional" depths where the leading-order instability mechanism fails.
3. Methodological Advancement: The paper demonstrates the power of combining rigorous functional analysis (analytic implicit function theorem, Kato perturbation) with symbolic computation to handle the intricate algebra of water wave problems.
4. Physical Insight: It confirms that even in finite depth, the dominant instability mechanism for small Stokes waves is the transverse "Type II" instability (as classified by McLean), leading to the formation of 3D wave patterns (like "fish-scale" patterns) from initially 2D waves.

In summary, Creedon, Nguyen, and Strauss have successfully extended the spectral instability theory of Stokes waves to the physically relevant case of finite depth, proving that transverse instability is a generic phenomenon, with only a finite set of depth values acting as exceptions.