On dispersive estimates for one-dimensional Klein-Gordon equations

1. Problem Statement

The paper investigates the long-time behavior (dispersive decay) of solutions to the one-dimensional Klein–Gordon (KG) equation with a real potential $V(x)$:
$$\ddot{\psi}(x, t) = (\partial_x^2 - m^2 + V(x))\psi(x, t), \quad m > 0$$
The equation is reformulated in vector form as a first-order system $i\dot{\Psi}(t) = K\Psi(t)$, where $K$ is the associated operator. The potential $V(x)$ is assumed to be continuous and decaying as $|V(x)| \leq C\langle x \rangle^{-\beta}$.

The Core Challenge:
While dispersive decay for the Schrödinger equation is well-understood, the KG equation presents unique difficulties due to its relativistic nature. Specifically:

• Propagation Speed: Unlike the Schrödinger equation (infinite propagation speed), the KG equation has bounded propagation speed. Singularities do not disappear at infinity but persist, moving with finite velocity.
• Green's Function Behavior: The Green's function for the free KG equation does not decay uniformly in $L^1 \to L^\infty$ norms for all spatial separations (specifically when $|x-y| \sim |t|$), unlike the Schrödinger case.
• Resonances: The presence of resonances at the spectral edges $\lambda = \pm m$ (where the continuous spectrum begins) can degrade decay rates.
• Weighted Norms: Previous results often required restrictive conditions on the decay of the potential ($\beta$) and the weight of the initial data ($\sigma$) to establish decay in weighted Sobolev spaces.

Goal: To improve existing dispersive decay estimates by establishing decay in stronger norms (lower $\sigma$) and under weaker assumptions on the potential decay ($\beta$), particularly distinguishing between resonant and non-resonant cases.

2. Methodology

The author employs a novel approach combining Born expansion, spectral representation, and scattering theory techniques.

A. Decomposition of the Evolution Operator

The dynamical group $U(t) = e^{-itK}$ is decomposed into low-energy and high-energy parts using a smooth cutoff function $\zeta(\omega)$:

1. Low Energy: Handled via the scattering matrix and Jost solutions (building on previous work [3]).
2. High Energy: Analyzed using a Born expansion (perturbation series) of the resolvent. The evolution operator is expanded as:
   $$U(t) = U_0(t) - i \int_0^t U_0(t-s)V U_0(s) ds - \int_0^t U_0(t-s)V \left( \int_0^s U_0(s-u)V U_0(u) du \right) ds + W(t)$$
   where $U_0(t)$ is the free evolution and $W(t)$ is the remainder term involving the third power of the potential interaction.

B. Free Evolution Estimates (Section 2)

The author establishes decay estimates for the free Klein–Gordon operator $K_0$ in weighted spaces $F_\sigma = H^1_\sigma \oplus L^2_\sigma$.

• Lemma 2.2: Proves $L^1 \to L^\infty$ decay of order $t^{-1/2}$ for the low-energy part using the van der Corput lemma on oscillatory integrals.
• Lemma 2.3: Proves high-energy decay of order $t^{-\sigma}$ in weighted norms. This is achieved via integration by parts in the Fourier domain and interpolation (Stein-Weiss theorem) to extend results from integer $\sigma$ to real $\sigma$.

C. Scattering Theory and Resolvent Representation (Section 3)

The paper utilizes the properties of the Schrödinger operator $H = -\partial_x^2 + V$.

• Jost Solutions: The resolvent $R(\omega)$ of the KG operator is expressed in terms of the resolvent of the Schrödinger operator.
• Wiener Algebra: The scattering coefficients (transmission $T(k)$ and reflection $R_\pm(k)$) and the kernel function $\psi(x,y,k)$ are shown to belong to the Wiener algebra $A$ (Fourier transforms of $L^1$ functions). This integrability is crucial for bounding the low-energy part of the solution.

D. High-Energy Perturbation Analysis (Section 4)

The core innovation lies in estimating the terms of the Born expansion for the high-energy part:

1. First and Second Order Terms ($U_1, U_2$): These are treated as convolutions. By applying the free decay estimates (Lemma 2.3) and the decay of the potential $V$, the author derives decay rates dependent on $\beta$.
2. Remainder Term ($W(t)$): This term involves the third power of the potential. The author represents the kernel using Jost solutions and applies a specialized version of the van der Corput lemma (Lemma 4.5) to oscillatory integrals involving the scattering data.
   • Non-Resonant Case: By integrating by parts in the inner integral of the remainder term, the author gains an extra factor of $t^{-1}$, improving the decay rate.

3. Key Contributions and Results

The paper establishes two main theorems that significantly improve upon previous literature (specifically [4, 8]):

Theorem 1.1: General Decay (Resonant and Non-Resonant)

• Assumption: $|V(x)| \leq C\langle x \rangle^{-\beta}$ with $\beta > 2$.
• Result:
  $$\|e^{-itK}P_c\|_{F_\sigma \to F_{-\sigma}} = O(t^{-1/2}), \quad \sigma > 1$$
• Improvement: This holds even if the spectral edges $\pm m$ are resonances. Previous results often required stronger conditions or failed in the resonant case for these specific norms.

Theorem 1.2: Improved Decay (Non-Resonant Case)

• Assumption: $|V(x)| \leq C\langle x \rangle^{-\beta}$ with $\beta > 3$ and no resonances at $\pm m$.
• Result:
  $$\|e^{-itK}P_c\|_{F_\sigma \to F_{-\sigma}} = O(t^{-3/2}), \quad \sigma > 3/2$$
• Improvement:
  • Weaker Potential Decay: Reduces the requirement from $\beta > 5$ to $\beta > 3$.
  • Stronger Norms: Reduces the required weight $\sigma$ from $\sigma > 5/2$ to $\sigma > 3/2$.
  • No Derivative Assumptions: Removes the need for assumptions on the decay of $\nabla V(x)$, relying only on the decay of $V(x)$.

4. Significance

1. Optimization of Conditions: The results provide the sharpest known conditions for dispersive decay in 1D Klein–Gordon equations regarding the trade-off between potential decay ($\beta$) and the weight of the initial data ($\sigma$).
2. Robustness in Resonant Cases: The proof demonstrates that the $t^{-1/2}$ decay rate is robust even in the presence of resonances, provided the potential decays sufficiently fast ($\beta > 2$).
3. Methodological Advancement: The combination of the Born expansion with precise estimates on the remainder term using Jost solutions and Wiener algebra properties offers a powerful framework for analyzing relativistic wave equations.
4. Applications: These dispersive estimates are fundamental prerequisites for proving the asymptotic stability of solitons and establishing scattering theory for associated nonlinear Klein–Gordon equations. By relaxing the assumptions on $V$, the results extend the applicability of stability theorems to a broader class of physical potentials.

In summary, Kopylova's work refines the mathematical understanding of how relativistic waves disperse in 1D, overcoming technical barriers related to resonances and slow-decaying potentials that limited previous analyses.