Global smooth solutions of 2-D quadratic quasilinear wave equations with null conditions in exterior domains

Technical Summary: Global Smooth Solutions of 2-D Quadratic Quasilinear Wave Equations with Null Conditions in Exterior Domains

Problem Statement
This paper addresses the global existence of small data smooth solutions for the initial boundary value problem (IBVP) of 2-D quadratic quasilinear wave equations in exterior domains with homogeneous Dirichlet boundary conditions. The equation is given by:
$$\Box u = Q(\partial u, \partial^2 u), \quad (t, x) \in [0, +\infty) \times K,$$
where $K = \mathbb{R}^2 \setminus \mathcal{O}$ is the exterior of a compact, smooth, star-shaped obstacle $\mathcal{O}$. The nonlinearity $Q$ is quadratic and satisfies the null condition.

While the global existence of such solutions is well-established for the corresponding Cauchy problem (whole space $\mathbb{R}^2$) and for 3-D exterior domain problems, the 2-D exterior domain case with null conditions remained an open problem. The difficulty arises from the slower decay rates of 2-D linear waves compared to 3-D, the lack of strong Huygens' principle, and the specific challenges posed by the boundary in exterior domains, which prevent the direct application of standard 3-D techniques like KSS (Klainerman-Sideris-Speck) estimates.

Methodology and Key Ingredients
The authors employ a continuous method (bootstrap argument) combined with several novel technical innovations to overcome the decay limitations inherent in 2-D exterior domains:

1. Precise Pointwise Estimates for Linear Waves:
   The authors derive new, improved pointwise estimates for the 2-D linear wave equation in exterior domains. Unlike previous estimates which suffered from logarithmic growth factors (e.g., $\ln(2+t+|x|)$) or insufficient time decay, the authors utilize Littlewood-Paley decompositions and properties of curve measures on the unit circle to establish sharper decay rates. Specifically, they prove that for solutions $w$ with compactly supported sources, the decay can be improved to:
   $$|w| \lesssim \frac{\ln^2(2+t+|x|)}{\langle t+|x| \rangle^{1/2} \langle t-|x| \rangle^\mu}$$
   and crucially, the gradient estimates remove the large logarithmic factor found in prior literature:
   $$|\partial w| \lesssim \frac{1}{\langle x \rangle^{1/2} \langle t-|x| \rangle} \sup \langle s \rangle |(1-\Delta)^3 g|.$$
   This removal of the $\ln(2+t+|x|)$ factor is essential for controlling the energy growth in the nonlinear setting.

2. Divergence Structures and Null Forms:
   The paper analyzes the algebraic structure of the null forms $Q_0$ and $Q_{12}$ under the null condition. The authors demonstrate that $Q_{12}$ can be written as a divergence term involving the rotation vector field $\Omega$. More significantly, they derive new algebraic identities to express the $Q_{0i}$ null forms as linear combinations of divergence terms, cubic null form terms, and quadratic error terms with improved spatial decay factors (e.g., $|x|^{-2}$).

3. Introduction of a "Good Unknown":
   To eliminate the problematic $Q_0$ type nonlinearity which is difficult to handle directly in the presence of boundaries, the authors introduce a new "good unknown" $V_a$. This variable is constructed by subtracting a specific term involving the product of derivatives from the standard vector field applied to the solution ($Z_a u$). This transformation effectively eliminates the $Q_0$ nonlinearity from the principal part of the equation and localizes the remaining terms spatially, allowing for better control near the boundary and at infinity.

4. Energy and Decay Estimates:
   The proof relies on a hierarchy of estimates:

   • Energy Estimates: Using a ghost weight technique (introduced in [2]) and elliptic estimates, the authors derive energy bounds. A critical step involves obtaining precise local energy decay estimates of the form $(1+t)^{-1}$ for the solution and its derivatives in a compact region near the obstacle.
   • Pointwise Decay: By combining the improved linear estimates with the divergence structure and the good unknown, the authors establish a sequence of pointwise decay estimates. These estimates are iteratively improved, gradually reducing the number of vector fields required to achieve specific decay rates (e.g., $\langle t-|x| \rangle^{-1}$ for derivatives).

Main Results
The paper proves Theorem 1.1, which states that if the obstacle $\mathcal{O}$ is star-shaped and the null condition holds, then for sufficiently small initial data $(u_0, u_1)$ satisfying compatibility conditions up to order $2N+1$ (with $N \ge 59$), the IBVP admits a global smooth solution $u$.

Furthermore, the solution satisfies the following optimal decay estimates:

• Derivative Decay:
  $$\sum_{|a| \le N+1} |\partial Z^a u| \le C \varepsilon \langle x \rangle^{-1/2} \langle t-|x| \rangle^{-1}$$
• Good Derivative Decay:
  $$\sum_{|a| \le N+1} |\bar{\partial} Z^a u| \le C \varepsilon \langle x \rangle^{-1/2} \langle t+|x| \rangle^{\epsilon_2-1}$$
• Solution Decay:
  $$\sum_{|a| \le N} |Z^a u| \le C \varepsilon \langle t+|x| \rangle^{\epsilon_2-1/2} \langle t-|x| \rangle^{\epsilon_2-1/2}$$
• Local Energy Decay:
  $$\sum_{|a| \le 2N-41} \|\partial^a u\|_{L^2(K_R)} \le C_R \varepsilon (1+t)^{-1}$$

These decay rates are shown to be essentially optimal, matching the behavior of the corresponding linear problem and the free wave equation in 2-D.

Significance and Claims
The paper claims to resolve a long-standing open problem regarding the global existence of small data solutions for 2-D quadratic quasilinear wave equations with null conditions in exterior domains. The authors emphasize that their work bridges the gap between the 3-D exterior domain results (where strong Huygens' principle aids decay) and the 2-D Cauchy problem results.

Key claims of significance include:

• Overcoming Logarithmic Loss: The paper explicitly addresses the failure of previous methods due to logarithmic growth factors in 2-D estimates. By deriving new pointwise estimates without these factors (via higher-order derivatives on the source term), they enable the closure of the bootstrap argument for global existence.
• Optimality: The authors assert that their local energy decay rate of $(1+t)^{-1}$ and the pointwise decay rates are optimal, noting that they match the known decay rates for the linear Dirichlet problem in 2-D.
• Generalization: The methodology is presented as robust enough to handle systems of equations and higher-order nonlinearities (cubic terms) under null conditions, as noted in Remarks 1.2 and 1.5.

The paper does not propose experimental applications or future implications beyond the mathematical resolution of the existence problem and the potential extension of these techniques to related hyperbolic systems.