1: \begin{abstract}

2: Anisotropic mesh adaptation is studied for the linear finite element solution of eigenvalue problems

3: with anisotropic diffusion operators. The $\bM$-uniform mesh approach is employed with which any nonuniform mesh

4: is characterized mathematically as a uniform one in the metric specified by a metric tensor.

5: Bounds on the error in the computed eigenvalues

6: are established for quasi-$\bM$-uniform meshes. Numerical examples arising from

7: the Laplace-Beltrami operator on parameterized surfaces, nonlinear diffusion,

8: thermal diffusion in a magnetic field in plasma physics, and the Laplacian operator on an L-shape domain

9: are presented. Numerical results show that anisotropic adaptive meshes can lead to more accurate computed

10: eigenvalues than uniform or isotropic adaptive meshes. They also confirm the second order convergence of the error

11: that is predicted by the theoretical analysis.

12: 

13: The effects of approximation of curved boundaries on the computation of eigenvalue problems

14: is also studied in two dimensions. It is shown that the initial mesh used to define the geometry of the physical domain

15: should contain at least $\sqrt{N}$ boundary points to keep

16: the effects of boundary approximation at the level of the error of the finite element approximation,

17: where $N$ is the number of the elements in the final adaptive mesh.

18: Only about $N^{1/3}$ boundary points in the initial mesh 

19: are needed for boundary value problems. This implies that the computation of

20: eigenvalue problems is more sensitive to the boundary approximation than that of boundary value problems.

21: \end{abstract}

22: