1: \begin{abstract}

2:   \textbf{Abstract:} When solving stochastic partial differential equations (SPDEs)

3:   driven by additive spatial white noise, the efficient sampling of

4:   white noise realizations can be challenging. Here, we present a new

5:   sampling technique that can be used to efficiently compute white

6:   noise samples in a finite element method and multilevel Monte Carlo

7:   (MLMC) setting. The key idea is to exploit the finite element matrix

8:   assembly procedure and factorize each local mass matrix

9:   independently, hence avoiding the factorization of a large matrix.

10:   Moreover, in a MLMC framework, the white noise samples must be

11:   coupled between subsequent levels. We show how our technique can be

12:   used to enforce this coupling even in the case of non-nested mesh

13:   hierarchies. We demonstrate the efficacy of our method with

14:   numerical experiments. We observe optimal convergence rates for the

15:   finite element solution of the elliptic SPDEs of interest in 2D and

16:   3D and we show convergence of the sampled field covariances. In a

17:   MLMC setting, a good coupling is enforced and the telescoping sum is

18:   respected.

19: \end{abstract}

20: