Exact parabolic asymptotics for singular \(n\)-D Burgers' random fields: Gaussian approximation (Q1805781): Difference between revisions

An \(n\)-dimensional Burgers' equation with a random initial condition of a gradient type \[ \frac {\partial u}{\partial t} + (u,\nabla)u = \nu \Delta u, \quad u(0,\cdot)=\nabla \xi , \quad t\geq 0, \;x\in \mathbb R^{n}, \] is considered, where \(\nu >0\) and \(u=(u_{1},\ldots ,u_{n})\). The initial velocity potential \(\xi \) is assumed to be a Gaussian zero-mean, measurable, almost surely differentiable, homogeneous and isotropic random field on \(\mathbb R^{n}\) with \(E\xi ^{2}(x) =1\). Let its covariance function be of the form \(L(| x|)/| x| ^{\alpha}\), \(0<\alpha <n\), \(L\) being a locally bounded function slowly varying for large values of \(| x| \). The parabolic scaling limit for \(u\) is known, that is, it is known that the finite-dimensional distributions of the random field \(L^{-1/2}(\sqrt {t})t^{1/2+\alpha /4} u(t,a\sqrt {t})\), \(a\in \mathbb R^{n}\), converge weakly as \(t\to \infty \) to a certain homogeneous Gaussian random field \(X(a)\) whose covariance function may be described explicitly. In the paper under review, the rate of this convergence (in the uniform Kolmogorov distance) is studied.