a1 Sandia National Laboratories, Org. 1442: Numerical Analysis and Applications, Livermore, CA 94550, USA.
a2 Georgia Institute of Technology, School of Computational Science and Engineering, Atlanta, GA 30332, USA.
a3 University of Texas at Austin, Department of Mathematics, Austin, TX 78712, USA.
a4 Stanford University, Department of Mathematics, Stanford, CA 94305, USA.. [email protected]
We present a parallel preconditioning method for the iterative solution of the time-harmonic elastic wave equation which makes use of higher-order spectral elements to reduce pollution error. In particular, the method leverages perfectly matched layer boundary conditions to efficiently approximate the Schur complement matrices of a block LDLT factorization. Both sequential and parallel versions of the algorithm are discussed and results for large-scale problems from exploration geophysics are presented.
(Online publication February 20 2014)
Mathematics Subject Classification: