Geophysical Electromagnetic (EM) simulations in heterogeneous media are computationally expensive. One reason being the multiple length scales that coexist in a realistic setting.
Discrete models require large meshes leading to solve huge systems of equations. Element Agglomeration Multigrid (AMGe) upscaling, developed at Lawrence Livermore National Laboratory, is a discretization framework for wide classes of partial differential equations and unstructured meshes. AMGe reduces significantly the computational cost while providing accurate solutions.
AMGe is implemented in the parallel distributed memory C++ library ParElag. Here, we show results of using ParElag on a synthetic example and how we plan to use it on an actual EM geophysical simulation.
Note: Project and results were presented in a poster.