Probabilities of high excursions of Gaussian fields (Q393011)

Let \(\xi(t)\) be a differentiable (in the mean-square sense) Gaussian random field with \(\operatorname{E}[\xi(t)] = 0\), \(\operatorname{Var}[\xi(t)] = 1\), and continuous trajectories on the \(d\)-dimensional interval \({\mathcal T} \subset \mathbb{R}^d\). The paper studies the problem of large excursions of the random field \(\xi\). More precisely, the probability \(\mathbb{P} = \operatorname{P}\bigl(-v(t) \leq \xi(t) \leq u(t) : t \in {\mathcal T} \bigr)\) is studied, where it is assumed that \(u(t),v(t) \to \infty\) uniformly for all \(t \in \mathcal{T}\). It is shown that \(\mathbb{P} = e^{-Q} + Q\,o(1)\), where \(Q\) depends on \({\mathcal T},u,v\) and the covariance function of \(\xi\). As an application, some well-known results in the literature are recovered.