MELBOURNE (Reuters) - Australian mining magnate Lang Hancock famously discovered the world’s biggest iron ore deposit when he noticed the red-stained cliffs of a canyon while flying over the Outback in the 1950s.
Now, the days when massive mineral deposits could be simply spotted by plane are gone, so miners are adopting new lab techniques and machine-led mapping to detect metal traces in everything from sand to gum tree leaves and groundwater.
Australian miners are taking a deeper look at the 80 percent of the country that lies “under cover” - obscured by meters of sediment and sand, particularly in the desert interior, across an area twice the size of India.
They are using a mix of new techniques and established methods in new ways, backed by greater computing heft, including a new generation of multi-layered maps produced by Australian government agencies.
“There has been a perception that Australia is a mature destination and all the big finds have been found,” said Will Robinson, managing director at Encounter Resources, which is exploring for gold using a new type of soil sampling developed by the CSIRO, Australia’s peak science body.
“You’re seeing more success as companies go out there and supply new technologies and new thinking into areas to find new big deposits,” he said. “It has reset the maturity clock.”
Having repaired their balance sheets on the back of a commodities price recovery, miners large and small are rushing to peg out prospective land in undercover areas.
Anglo American told Reuters it has recently been granted or is seeking permits over almost 11,000 square km of land in Queensland state, while iron ore miner Fortescue Metals Group raised its exploration tenements by a third last year.
Rio Tinto also sharply increased its exploration holdings under the remote Great Sandy Desert, where it announced a large copper find last week.
Encounter Resources last month reported a gold find under shallow sand using CSIRO’s Ultrafine Soil technique, sending the explorer’s shares up 30 percent.
“These ancient landscapes have buried deposits,” said Ryan Noble, a CSIRO scientist leading the development of the process.
The analysis relies on separating out large particles of sand, which in remote Australia can account for as much as 95 percent of a sample, from much smaller clays and iron oxides, which attract gold, copper and zinc.
Developments in analysis mean smaller particles of 2 microns can now be detected, about one-50th the size of a grain of salt, revealing signs of mineralization as deep as 10-20 meters (30-65 feet) under soil or sand.
“We think there is huge value for Australia and other places to reprocess the samples they have collected and generate new targets and open up new areas,” CSIRO’s Noble said.
Other new techniques are also starting to show results.
In South Australia, ASX-listed Marmota is using a biogeochemical leaf analysis which helped guide it to a high-grade gold strike at its Gawler Craton project in September.
Gum trees siphon up minute traces of gold from their roots to their leaves, which are then analyzed in a lab.
“The gums were able to pick up moderate mineralization down to about 60 meters (200 feet),” geologist Aaron Brown told Reuters.
The technology was developed during the boom years of 2012-2013 but has taken time to be used commercially as miners and investors look to reinvest in exploration after several lean years.
Australian government bodies are also getting on board.
The Australian government is part way through a four-year, A$100 million ($70.75 million) exploration program under the oversight of GeoScience Australia. It includes projects to collate soil geochemistry data with groundwater data sets into predictive maps for highly prospective mineral regions.
Geoscience Australia is also working with other international bodies such as the Geological Survey of India to help uncover hidden deposits on the subcontinent, underscoring the competition among global mining zones to attract capital in the new wave of exploration.
“The release of excellent pre-competitive geoscientific data by Geosciences Australia and the Geological Survey of Queensland has enabled us to evaluate previously under-explored parts of Australia, including undercover areas of interest,” Anglo American told Reuters.
Meanwhile, IBM, which launched a resources platform developed in Canada that uses machine learning to better predict where mineralization lies, is drawing interest from Australian miners.
IBM’s Watson for the natural resources industry uses artificial intelligence to combine data such as government drill results and aeromagnetic surveys with historical visual data including hand drawn maps in a way that has not been done before.
Geologists can ask technically specific questions of Watson about a resource and get an answer back.
“Traditionally geologists would spend 80 percent of their time sorting through data, trying to find it, correlate it, then organize it, and then running their models,” IBM partner David Dickson told Reuters.
“Now it’s flipped and 20 percent of the time is spent doing that, and geologists can spend more time on getting insights from the information and making better decisions on where the prospective regions are.”
Australia has a long tradition in mining and a significant potential of still unexplored reserves, he said.
“Gold for example, generally is very low-grade so you can spend a lot of money on exploration. The more we remove the uncertainty and variability from the exploration program, the more viable the program is.”
Reporting by Melanie Burton; Editing by Lincoln Feast