Exploring Inversion Solution Space: A case study over a Cu-Ag deposit in the Kalahari copper belt. Issue 1 (1st December 2018)

• Record Type: Journal Article
• Authors: Ellis, Rob; Lotter, Cas; Pithawala, Taronish
• Is Part Of: ASEG Extended Abstracts (Online). Volume 2018:Issue 1(2018)
• Page Start: 1
• Page End: 8
• Publication Date: 2018-12-01
• Subjects: inversion -- non-uniqueness -- DC -- IP -- 3D -- Kalahari -- exploring inversion solution space; Prospecting -- Geophysical methods -- Periodicals; Prospecting -- Geophysical methods; Periodicals
• DOI: 10.1071/ASEG2018abM1_4F
• Languages: English
• ISSNs: 2202-0586

Abstract

Abstract : Exploring inversion solution space is the process of mapping out the domain of inversion models which have a predicted response in satisfactory agreement with the observed data. This is particularly important in mineral exploration geophysics because the associated inverse problems are highly ill-posed. An inversion case study over a Cu-Ag deposit in the Kalahari copper belt provides a good demonstration of the process: in the case study 3D IP field data show a clear indication of chargeability so it is no surprise that a chargeability anomaly is recovered from a standard 3D IP inversion. Extensive drilling delineated a disseminated sulphide zone however the drilling delineates a shallow target while the inversion indicates an interesting deeper target. The immediate exploration question is whether to invest in a deep drillhole to test the deeper target? Or is the deeper target merely an artefact of non-uniqueness associated with the IP method? Exploring the inversion solution space produces a suite of models all of which fit the data satisfactorily. In particular, exploring the solution depth produces a suite of models with a variety of depths, including some models with a shallow chargeability anomaly. This significantly weakens expectations associated with the initial deep inversion model but improves exploration decision making. Exploring inversion solution space is achieved computationally by systematically modifying the regularization term in the traditional Tikhonov formulation. With an inversion workbench the process is easy, the results are remarkably helpful, and the outcomes surprisingly quickly become a natural part of the exploration team's interpretation process.