[Paper Review] Intermittency and chaos for a stochastic non-linear wave equation in dimension 1

Daniel Conus, Mathew Joseph+2 more|arXiv (Cornell University)|Dec 8, 2011

Stochastic processes and financial applications16 references3 citations

TL;DR

This paper establishes intermittency and chaotic behavior for a one-dimensional stochastic nonlinear wave equation driven by space-time white noise. Using a comparison principle and moment estimates, it shows that the solution's supremum grows like $(\log R)^{1/2}$ as $R \to \infty$ when initial conditions are bounded away from zero, mirroring Gaussian-like behavior, while compactly supported initial data yield almost surely bounded maxima.

ABSTRACT

We consider a non-linear stochastic wave equation driven by space-time white noise in dimension 1. First of all, we state some results about the intermittency of the solution, which have only been carefully studied in some particular cases so far. Then, we establish a comparison principle for the solution, following the ideas of Mueller. We think it is of particular interest to obtain such a result for a hyperbolic equation. Finally, using the results mentioned above, we aim to show that the solution exhibits a chaotic behavior, in a similar way as was established by Conus, Joseph, and Khoshnevisan for the heat equation. We study the two cases where 1. the initial conditions have compact support, where the global maximum of the solution remains bounded and 2. the initial conditions are bounded away from 0, where the global maximum is almost surely infinite. Interesting estimates are also provided on the behavior of the global maximum of the solution.

Motivation & Objective

• To systematically analyze intermittency and chaotic behavior in the stochastic nonlinear wave equation with space-time white noise in one spatial dimension.
• To extend comparison principles—previously used in parabolic SPDEs—to hyperbolic SPDEs, a novel contribution for wave equations.
• To establish precise asymptotic estimates on the supremum of the solution under two distinct initial condition regimes: compact support and bounded away from zero.
• To compare the hyperbolic case with the well-known parabolic (heat equation) case, highlighting similarities and differences in the growth rates of the solution's maximum.

Proposed method

• Adopts the mild solution formulation via the stochastic wave kernel $\Gamma_t(x) = \frac{1}{2}\mathbf{1}_{[-\kappa t, \kappa t]}(x)$, integrating the noise through Walsh's stochastic integral.
• Applies a comparison principle for the wave equation, inspired by Mueller’s work, to control the solution’s growth and enable moment estimates.
• Uses moment bounds and tail probability estimates (via Chebyshev and exponential moment techniques) to analyze the solution’s supremum over expanding intervals.
• Employs a discretization strategy over a grid of points $x_j$ to estimate $\sup_{x \in [-R,R]} u(t,x)$, combining pointwise tail bounds with modulus of continuity estimates.
• Applies the Borel-Cantelli lemma to derive almost sure upper bounds on the growth rate of the solution’s maximum, using decay rates of tail probabilities.
• Derives a modulus of continuity estimate: $\mathbb{E}\left[\sup_{|x-x'| \leq \delta} \exp\left(\frac{|u(t,x)-u(t,x')|^2}{C\delta}\right)\right] \leq \frac{2}{\delta}$, which controls sample path regularity.

Experimental results

Research questions

• RQ1 Does the solution to the stochastic nonlinear wave equation exhibit intermittency and chaotic behavior similar to that observed in the parabolic (heat) equation?
• RQ2 How does the supremum of the solution behave as the spatial domain expands, i.e., as $R \to \infty$, under different initial condition regimes?
• RQ3 What is the precise growth rate of $\sup_{x \in [-R,R]} u(t,x)$ in the case where initial data are bounded away from zero?
• RQ4 How does the presence of a non-zero initial derivative $v_0$ affect the solution’s maximum and its dependence on the wave speed $\kappa$?
• RQ5 Can a comparison principle be established for hyperbolic SPDEs, and how does it facilitate the analysis of solution growth and tail behavior?

Key findings

• When initial conditions have compact support, $\sup_{x \in \mathbb{R}} u(t,x) < \infty$ almost surely, confirming boundedness of the global maximum.
• When initial conditions are bounded away from zero, the solution’s supremum over $[-R,R]$ grows almost surely as $\sim (\log R)^{1/2}$, with the rate depending on $\kappa$ and $\overline{v}_0$.
• The upper bound on the growth rate is $\limsup_{R \to \infty} \frac{\sup_{x \in [-R,R]} u(t,x)}{(\log R)^{1/2}} \leq \left(\frac{8}{q}\right)^{1/2}$ a.s., where $q = \min\left\{\frac{c}{\max\{\kappa, \overline{v}_0^2 \kappa^2\}}, \frac{1}{C}\right\}$.
• The solution exhibits chaotic behavior in the sense of rapid spatial fluctuations and high-peak formation, analogous to the parabolic case studied in [9].
• The growth rate is increasing in $\kappa$, in contrast to the parabolic case, though uniformity in $\kappa$ breaks down for small $\kappa$ due to limitations in the modulus of continuity estimate.
• When $v_0 \equiv 0$, the behavior aligns more closely with the parabolic case, and the dependence on $\kappa$ becomes symmetric in the upper and lower bounds.