3-D joint inversion of induced polarization and self-potential data for ore body localization

SUMMARY In mineral exploration, induced polarization and self-potential are two broadly used active and passive geophysical methods, respectively. In the case of ore bodies, both methods are associated with charge distributions associated with a secondary electrical field (induced polarization) and...

Full description

Saved in:
Bibliographic Details
Published in:Geophysical journal international Vol. 243; no. 3
Main Authors: Su, Z, Shen, J, Revil, A, Zhu, Z, Ghorbani, A
Format: Journal Article
Language:English
Published: Oxford University Press 01.12.2025
Subjects:
ISSN:0956-540X, 1365-246X
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:SUMMARY In mineral exploration, induced polarization and self-potential are two broadly used active and passive geophysical methods, respectively. In the case of ore bodies, both methods are associated with charge distributions associated with a secondary electrical field (induced polarization) and a source current density (self-potential). Both the chargeability and volumetric source current density distributions bring information regarding the shape of ore bodies. Therefore the joint inversion of these data sets is expected to better tomograms of ore bodies. A joint inversion approach is developed to combine both methods. The objective function to minimize includes two independent components plus a cross-gradient joint function. The use of the cross-gradient is justified from the underlying physics of the two geophysical problems at play. The structure of the cost function is tailored to overcome some problems like convergence and parameter determination in the inverse process. Two synthetic tests and a laboratory experiment are used to benchmark the proposed algorithm. We demonstrate that the joint inversion algorithm performs better than the localizations obtained from independent inversion approaches. To refine the interpretation of the shape of ores, we introduce an ore presence index using the chargeability and source current density resulting from the joint inversion algorithm. The K-Medoids clustering algorithm is used to automatically categorize the calculated ore presence index into different clusters. The cluster with larger values successfully identifies the ore bodies associated with strong chargeability and/or volumetric source current density.
ISSN:0956-540X
1365-246X
DOI:10.1093/gji/ggaf392