Hyperbolic Anderson model with L\'evy white noise: spatial ergodicity and fluctuation

TL;DR

This paper investigates the spatial ergodicity and fluctuation behavior of the hyperbolic Anderson model driven by Le9vy white noise, establishing new limit theorems and ergodic properties in a non-Gaussian setting.

Contribution

It is the first to analyze spatial ergodicity and derive a quantitative CLT for HAM driven by Le9vy noise, extending limit theorem results beyond Gaussian cases.

Findings

Proved spatial ergodicity of the solution.

Established a quantitative CLT with explicit convergence rates.

Derived a functional CLT for the model.

Abstract

In this paper, we study one-dimensional hyperbolic Anderson models (HAM) driven by space-time pure-jump L\'evy white noise in a finite-variance setting. Motivated by recent active research on limit theorems for stochastic partial differential equations driven by Gaussian noises, we present the first study in this L\'evy setting. In particular, we first establish the spatial ergodicity of the solution and then a quantitative central limit theorem (CLT) for the spatial averages of the solution to HAM in both Wasserstein distance and Kolmogorov distance, with the same rate of convergence. To achieve the first goal (i.e. spatial ergodicity), we exploit some basic properties of the solution and apply a Poincar\'e inequality in the Poisson setting, which requires delicate moment estimates on the Malliavin derivatives of the solution. Such moment estimates are obtained in a soft manner by observing a natural connection between the Malliavin derivatives of HAM and a HAM with Dirac delta velocity. To achieve the second goal (i.e. CLT), we need two key ingredients: (i) a univariate second-order Poincar\'e inequality in the Poisson setting that goes back to Last, Peccati, and Schulte (Probab. Theory Related Fields, 2016) and has been recently improved by Trauthwein (arXiv:2212.03782); (ii) aforementioned moment estimates of Malliavin derivatives up to second order. We also establish a corresponding functional central limit theorem by (a) showing the convergence in finite-dimensional distributions and (b) verifying Kolmogorov's tightness criterion. Part (a) is made possible by a linearization trick and the univariate second-order Poincar\'e inequality, while part (b) follows from a standard moment estimate with an application of Rosenthal's inequality.