A Multi-Index Quasi-Monte Carlo Algorithm for Lognormal Diffusion Problems
This work addresses the need for efficient uncertainty quantification in subsurface flow problems with many random parameters, offering a method that scales well with tolerance requirements.
The paper introduces a Multi-Index Quasi-Monte Carlo method for elliptic PDEs with random coefficients, achieving cost inversely proportional to the tolerance on root-mean-square error for a challenging 3D subsurface flow problem with lognormal diffusion.
We present a Multi-Index Quasi-Monte Carlo method for the solution of elliptic partial differential equations with random coefficients. By combining the multi-index sampling idea with randomly shifted rank-1 lattice rules, the algorithm constructs an estimator for the expected value of some functional of the solution. The efficiency of this new method is illustrated on a three-dimensional subsurface flow problem with lognormal diffusion coefficient with underlying Matérn covariance function. This example is particularly challenging because of the small correlation length considered, and thus the large number of uncertainties that must be included. We show numerical evidence that it is possible to achieve a cost inversely proportional to the requested tolerance on the root-mean-square error, for problems with a smoothly varying random field