Large time behavior for the classical wave equation with different regular data and its applications
Wenhui Chen, Ryo Ikehata
TL;DR
Extends the classical wave equation analysis by deriving optimal large-time $${L^2}$$-norm behavior under different regular data in $${\mathbb{R}}^n$$. The authors develop a kernel-based Fourier-splitting method, introducing a time-dependent scale $${\mathcal{D}}_{n,\sigma,s}(t)$$ and sharp estimates for the kernel $${\mathcal{I}}_{n,\sigma,s}(t)$$, to obtain precise growth/decay rates across dimension regimes. They identify critical thresholds $$n_0=2s$$ and $$n_1=2+2s$$ governing local stability and large-time stabilization, and show how $${L^{1,\kappa}}$$-type regularity with vanishing moments can yield global stability, while equality cases may cause blow-up. The results are then applied to the wave equation with scale-invariant terms, undamped $$\sigma$$-evolution, the critical Moore-Gibson-Thompson equation, and the linearized compressible Euler system. These findings illuminate how initial-data regularity controls long-time dynamics and offer tools for related undamped evolution equations.
Abstract
In this paper, we mainly consider large time behavior for the classical free wave equation $u_{tt}-Δu=0$ in $\mathbb{R}^n$. We derive some large time optimal estimates for the quantity of solution $\|u(t,\cdot)\|_{L^2}$ with initial data belonging to $L^2$ or with additional weighted $L^1$ integrabilities. Particularly, some thresholds are discovered for the (local or global in time) stabilization of this quantity. We also apply these results to the wave equation with scale-invariant terms, the undamped $σ$-evolution equation, the critical Moore-Gibson-Thompson equation, and the linearized compressible Euler system.
