1: \begin{abstract}

2: In a previous paper (J.\ Comp.\ Phys.\ \textbf{230} (2011), 3668--3694,

3: see ref.~\cite{GrKuNuScSl:11}), the authors proposed a new practical

4: method for computing expected values of functionals of solutions for

5: certain classes of elliptic partial differential equations with random

6: coefficients. This method was based on combining quasi-Monte Carlo

7: (QMC) methods for computing the expected values with circulant embedding

8: methods for sampling the random field on a regular grid.

9: It was found capable of

10: handling fluid flow problems in random heterogeneous media with high

11: stochastic dimension, but \cite{GrKuNuScSl:11} did not provide a

12: convergence theory. This paper provides a convergence analysis for the method

13: in the case when the QMC method is a specially designed randomly shifted

14: lattice rule.

15: The convergence result depends on the eigenvalues of the underlying

16: nested block circulant matrix and can be independent of the number of

17: stochastic variables under certain assumptions.

18: In fact the QMC analysis applies to general factorisations of the covariance

19: matrix to sample the random field.

20: The error analysis for the underlying fully discrete finite element method

21: allows for locally refined meshes (via interpolation from a regular sampling

22: grid of the random field). Numerical results on a non-regular domain with

23: corner singularities in two spatial dimensions and on a regular domain in three

24: spatial dimensions are included.

25: \end{abstract}

26: