Covariance regularity and \(\mathcal {H}\)-matrix approximation for rough random fields (Q522237)

The authors consider the linear operator equation \(Au=f\) where \(A\) is a bounded invertible, strongly elliptic pseudodifferential operator of order \(r\in {\mathbb{R}}\) with analytic coefficients, covering all linear, second-order elliptic partial differential equations as well as their boundary reductions. The domain can be an open bounded domain \(D\subset {\mathbb{R}}^n\) with smooth boundary \(\partial D\) or a smooth, closed and compact, Riemannian \(n\)-manifold \({\mathcal{M}} \subset {\mathbb{R}}^{n+1}\). The data \(f\) is assumed to be in \(L^2 (\Omega; H^t )\), i.e., an \(H^t\)-valued random field with finite second moments, with \(H^t\) denoting the (isotropic) Sobolev space of order \(t\) modelled on the domain \(D\) or manifold \({\mathcal{M}}\). It is shown that the random solution's covariance kernel \(K_u =(A^{-1} \otimes A^{-1} )K_f\) is an asymptotically smooth function provided that the covariance function \(K_f\) of the random data is a Schwartz distributional kernel of an elliptic pseudodifferential operator. As a consequence, numerical \(\mathcal {H}\)-matrix calculus allows the deterministic approximation of singular covariances \(K_u\) of the random solution \(u=A^{-1} f \in L^2 (\Omega;H^{t-r} )\) with work versus accuracy essentially equal to that for the mean field approximation with splines of fixed order.