Comparison of Eigenvectors for Coupled Seismo-Electromagnetic Layered-Earth Modeling

Research output: Contribution to journalArticleScientificpeer-review

9 Citations (Scopus)
65 Downloads (Pure)

Abstract

We study the accuracy and numerical stability of three eigenvector sets for modelling the coupled poroelastic and electromagnetic layered-Earth response. We use a known eigenvector set, its flux-normalized version and a newly derived flux-normalized set. The new set is chosen such that the system is properly uncoupled when the coupling between the poroelastic and electromagnetic fields vanishes. We carry out two different numerical stability tests: the first test focuses on the internal system, eigenvector and eigenvalue consistency; the second test investigates the stability and preciseness of the flux-normalized systems by looking at identity relations. We find that the known set shows the largest deviation for both tests, whereas the new set performs best. In two additional numerical modelling experiments, these numerical inaccuracies are shown to generate numerical noise levels comparable to small signals, such as signals coming from the important interface conversion responses, especially when the coupling coefficient is small. When coupling vanishes completely, the known set does not produce proper results. The new set produces numerically stable and accurate results in all situations. We therefore strongly recommend to use this newly derived set for future layered-Earth seismo-electromagnetic modelling experiments.
Original languageEnglish
Pages (from-to)152-190
Number of pages39
JournalGeophysical Journal International
Volume206
Issue number1
DOIs
Publication statusPublished - 2016

Keywords

  • Numerical solutions
  • Electromagnetic theory
  • Magnetic and electrical properties
  • Theoretical seismology
  • Wave propagation
  • Acoustic properties

Fingerprint

Dive into the research topics of 'Comparison of Eigenvectors for Coupled Seismo-Electromagnetic Layered-Earth Modeling'. Together they form a unique fingerprint.

Cite this