Assessing the impact of conceptual mineral systems uncertainty on prospectivity predictions

• DOI: 10.1016/j.gsf.2022.101435
• Publication Date: 2022-07-26T00:20:43Z
• Authors: Mark D Lindsay; Agnieszka M. Piec...; Mark W Jessell; Richard Scalzo; Jeremie Giraud; Guillaume Pirot; Edward Cripps
• Keywords: Prospectivity mapping; Mineral exploration; Conceptual model; Mineral deposit

The past two decades have seen a rapid adoption of artificial intelligence methods applied to mineral exploration. More recently, the easier acquisition some types data has inspired broad literature that examined many machine learning and modelling techniques combine exploration criteria, or 'features', generate predictions for prospectivity. Central design prospectivity models is 'mineral system', conceptual model describing key geological elements control timing location economic mineralisation. systems defines what constitutes training set, which features represent evidence mineralisation, how are engineered used. Mineral knowledge-driven models, thus all parameter choices subject human biases opinion so alternative possible. However, effect on rarely compared despite potential heavily influence final predictions. In this study, we focus uncertainty Fe ore in Hamersley region, Western Australia. Four important considerations tested. (1) Five different supergene hypogene guide inputs five forest-based classification model. (2) To uncertainty, then combined comparison. (3) Representation three-dimensional objects as two-dimensional tested address commonly ignored thickness units. (4) dataset composed known mineralisation sites (deposits) 'positive' examples, drilling providing 'negative' sampling locations. Each spatial assessed using independent performance metrics common AI-based subjected plausibility testing. We find produce significantly predictions, must be recognised. A benefit recognising robust geologically plausible can made may discovery.