Metelka, Vasek1; Schofield, Anthony2; Wise, Tom3; Cole, David4; Otto, Alex1; Fabris, Adrian3; Hong, Wei5; Murr, James2
1CSIRO Mineral Resources, Perth, Australia; 2Geoscience Australia, Canberra, Australia; 3Geological Survey of South Australia, Adelaide, Australia; 4CSIRO Data 61, Melbourne, Australia; 5Univeristy of Adelaide, Adelaide, Australia
The National Drilling Initiative (NDI) and MinEx CRC research activities aim at providing new data that will drive our understanding of geological processes and mineral systems in key underexplored areas of the Australian continent. Ultimately, the newly acquired knowledge will underpin the future of the exploration industry as we push for discoveries under deeper cover.
Prospectivity analyses were conducted to aid with drill targeting for the NDI. These analyses were performed for two regions of interest: The East Tennant Creek – Mount Isa Area (TISA) of Northern Territory and Queensland and the Delamerian Orogen (Delamerian) Area of South Australia. Both regions, albeit with a different geological setting, were primarily omitted from substantive exploration efforts either due to thick regolith or younger sedimentary cover or lacking knowledge and missing significant discoveries. On paper; however, the geological settings and the proximity to known mineralisation suggest a high potential for discovery of new deposits under younger sedimentary/ regolith cover or deposits that formed mineral systems that were not considered previously.
We employed a multicommodity knowledge-based mineral systems analysis approach to identify critical components and model the overall propensity to mineralisation. Theoretical gold and copper-gold mineral systems models were established first. The models were then converted to geospatial representations, and quantitative prospectivity maps were created utilising fuzzy logic inference and compared to weighted overlay as well as data-driven methods (logistic regression, random forest), where training data could be applied from adjacent uncovered regions.
The results show several zones of interest that can serve the researchers and stakeholders as first-pass information for drill targeting. The approach also highlighted the importance of primary input data. It helped focus on areas where data quality or discrepancies existed as well as identified data sets that were crucial for mapping the components of a mineral system more accurately. When compared to the data-driven methods, which showed good classification accuracy metrics, the knowledge-driven results open up more space and enable conceptual, regional targeting.
Dr Vasek Metelka is a Senior Research Geologist with CSIRO Mineral Resources in Perth. He has over 15 years of R&D experience from research organisations in Australia and the Czech Republic. Throughout his career, Vasek focuses on interdisciplinary projects bridging geology, geochemistry, geophysics and remote sensing.