Three Dimensional Sparse Inversion of Fluid Flow Electromagnetic Method

Abstract

Fluid Flow ElectroMagnetic (FFEM) method is a kind of electromagnetic method to monitor dynamic behavior of subsurface geothermal fluid flow by measuring transient changes of surface electromagnetic field associated with a movement of geothermal fluid flow. Subsurface fluid flow causes electric current flow called streaming current due to electrokinetic phenomena. It is very important to know the location of subsurface current sources to interpret observed FFEM data. Conventional smoothness-constrained 3-D inversions based on L2 norm minimization are not suitable for sparse structure such as a local anomaly like a dyke. In order to overcome such the problem that final solution results in smooth and wide structure, the L1 norm minimization considering parameter sparsity was applied to FFEM method. We developed a 3-D sparse inversion program based on First Iterative Shrinkage Thresholding Algorithm (FISTA). Some examples of 2-D and 3-D inversions of FFEM data are shown in this paper.