Victoria is believed to have the richest orogenic gold endowment of any state or region globally. However, the majority of its gold remains hidden beneath younger sediment cover, with few near-surface signs of mineralisation at depth.
In recent years, explorers have been progressively moving into these undercover areas, as surface discoveries dry up. Many geologists have relied on advanced datasets to interpret the underlying geology.
But with so many datasets at hand and with complicated relationships between each, narrowing the search for economic mineralisation remains a complex task in undercover regions.
Machine learning to the rescue
Geologist Ben Jupp and his colleaugues at SRK Australia and north America, are helping to address this by using machine learning to conduct prospectivity analysis.
Pulling together large datasets that are often hard to compute and visualise, the tool can help geologists undertake prospectivity analysis in a fraction of the time it typically takes.
“Machine learning enables large datasets to be analyzed in relatively short timeframes. For geologists this process can be very time consuming,” said Ben.
“It also helps us better understand what the data is trying to tell us. It paints a clearer picture, helping us rule out inconsistencies and draw comparisons we may not have otherwise drawn. In turn, it can point to high quality targets for further analysis.”
Over time, its accuracy improves, as the tool learns patterns in the data.
“In very simple terms, machine learning applies algorithms to establish or ‘learn’ patterns in the data. From these patterns the machine can make predictions. The method is objective and not subject to human biases. This makes it an appealing tool for prospectivity analysis,” said Ben.
A combined approach
For prospectivity studies, SRK has been using both machine leaning and mineral systems expertise. It combines advanced data processing and geological interpretations in both the 2D and 3D space.
“The integrated workflow gives us clearer insight into the underlying mineral system. The 2D and 3D modelling can help us define high quality target areas for follow up investigations and further drill targeting,” Ben said.
A growing need
In Victoria, the need for new exploration methods – like those powered by machine learning – is growing.
Primary mineralisation and surface geochemistry has driven much of the state’s gold discovery over the last fifty years. With these low hanging fruits now largely mined to depletion, explorers are sourcing more innovative methods to fuel the next wave of discovery.
While many have already turned to methods such as magnetics and gravity to help guide their undercover exploration, machine learning has yet to become mainstream.
“Machine learning is only just beginning to serve as an additional tool to evaluate and understand their data and narrow the search window for mineralization,” Ben said.
“However, I expect it will begin to take off in the coming years, when explorers become more aware of the competitive advantage it brings.”
A barely tapped treasure trove
The rewards for a competitive advantage in gold exploration are high.
To date, Victoria has contributed 1.5 percent of the world’s gold from only 0.15 percent of the world’s land mass – or just 0.03 percent if you factor in the small land mass occupied by its existing productive goldfields. This makes Victoria’s goldfield rocks 2 orders of magnitude (100 x richer) than the global average. Geoscience Victoria believes there may be as much undiscovered gold throughout the state as there is already extracted.
“All the data suggests that we are sitting on a gold mine here in Victoria, if not several – now we just need all the help we can get to find where it sits,” Ben concluded.
Hear more from Ben Jupp about the Victoria Gold Mining & Exploration Conference. The event will be held 22-23 February at the Rendezvous Hotel Melbourne.
Joining Ben on the stage are key industry representatives from the Minerals Council of Victoria, Agnico Eagle, Southern Cross Gold and many more.
Learn more and register your place.