Accelerated full-waveform time-domain electromagnetic forward modeling in underground full-space using an octree mesh and model order reduction

Read full text Summarise with VIRAT
Authors: Kailiang Lu, Jianhua Yue, Jianmei Zhou, Ya’nan Fan, Xiu Li, Qinrun Yang
Cited by: 0
References: 41

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Time-domain electromagnetic (TDEM) surveys commonly use diverse excitation waveforms, making accurate simulation of full-time three-dimensional (3D) responses necessary. This paper presents a fast 3D full-space TDEM forward solver that combines octree-based local refinement with a model order reduction (MOR) strategy. The governing equation is discretized using the mimetic finite-volume method in the magnetic flux density (b) formulation, and a source-decoupled shift-and-invert (SAI) Krylov-subspace MOR scheme is constructed to reduce the original high-dimensional system. Compared with tensor discretization, the octree mesh achieves a favorable trade-off between local resolution and memory consumption. Numerical results show that the proposed MOR solver accelerates simulations by factors of 10.5 and 4.2 relative to the conventional BDF and BDF-SAI schemes, respectively, while maintaining high accuracy. Comparisons between full-space and half-space responses reveal distinct electromagnetic field propagation characteristics in full-space environments. The results further indicate that anomalies outside the observation direction may affect measured responses, suggesting the need to mitigate interference from undesired directions in practical surveys.

Fields of Study & Subjects

Geophysical and Geoelectrical Methods Electromagnetic Simulation and Numerical Methods Geophysical Methods and Applications

Publication details

Journal / source
Journal of Geophysics and Engineering
Volume
23
Issue
3
Pages
1078–1093
Publication date
31 March 2026
Language
EN