Hole probability for nodal sets of the cut-off Gaussian Free Field

TL;DR

This paper investigates the probability that the cut-off Gaussian Free Field on a surface remains positive on a subset, analyzing its covariance structure and the supremum concentration as the cutoff parameter grows.

Contribution

It provides the first detailed asymptotic analysis of positivity probabilities for the cut-off Gaussian Free Field on surfaces, linking covariance behavior to geometric properties.

Findings

Asymptotic covariance function behavior as L→∞

Probability decay rate for positivity on subsets

Concentration of the supremum around 1√(2π)ln L

Abstract

Let ($\Sigma$, g) be a closed connected surface equipped with a riemannian metric. Let ($\lambda$ n) n$\in$N and ($\psi$ n) n$\in$N be the increasing sequence of eigenvalues and the sequence of corresponding L 2-normalized eigenfunctions of the laplacian on $\Sigma$. For each L \textgreater{} 0, we consider $\phi$ L = 0\textless{}$\lambda$n$\le$L $\xi$n $\sqrt$ $\lambda$n $\psi$ n where the $\xi$ n are i.i.d centered gaussians with variance 1. As L $\rightarrow$ $\infty$, $\phi$ L converges a.s. to the Gaussian Free Field on $\Sigma$ in the sense of distributions. We first compute the asymptotic behavior of the covariance function for this family of fields as L $\rightarrow$ $\infty$. We then use this result to obtain the asymptotics of the probability that $\phi$ L is positive on a given open proper subset with smooth boundary. In doing so, we also prove the concentration of the supremum of $\phi$ L around 1 $\sqrt$ 2$\pi$ ln L.