1: \begin{abstract}

2: A grid-overlay finite difference method is proposed for the numerical approximation of the fractional Laplacian on arbitrary bounded domains.

3: The method uses an unstructured simplicial mesh and an overlay uniform grid for the underlying domain

4: and constructs the approximation based on a uniform-grid finite difference approximation and a data transfer

5: from the unstructured mesh to the uniform grid.

6:  The method takes full advantages of both uniform-grid finite difference approximation in efficient matrix-vector

7: multiplication via the fast Fourier transform and unstructured meshes for complex geometries and mesh adaptation.

8: It is shown that its stiffness matrix is similar to a symmetric and positive definite matrix

9: and thus invertible if the data transfer has full column rank and positive column sums. Piecewise linear interpolation

10: is studied as a special example for the data transfer. It is proved that the full column rank and positive column sums of linear interpolation

11: is guaranteed if the spacing of the uniform grid is smaller than or equal to a positive bound

12: proportional to the minimum element height of the unstructured mesh.

13: Moreover, a sparse preconditioner is proposed for the iterative solution of the resulting linear system for the homogeneous Dirichlet problem

14: of the fractional Laplacian. Numerical examples demonstrate that the new method has similar convergence

15: behavior as existing finite difference and finite element methods and that the sparse preconditioning is effective.

16: Furthermore, the new method can readily be incorporated with existing mesh adaptation strategies. Numerical results obtained

17: by combining with the so-called MMPDE moving mesh method are also presented.

18: \end{abstract}

19: