The data revolution will enable greater certainty and reduce risks in deep exploration. TIM THWAITES reports
Article from resourceful: Issue 11
Future Science Platform geologists and geophysicists will work alongside researchers from one of CSIRO's newest business units, Data61.
Data61 was formed in October 2015 from a merger of CSIRO's existing Digital Productivity Flagship with the National Information and Communications Technology research centre, NICTA. Developing and commercialising new data-driven approaches to complex problems is exactly what Data61 is all about.
"We bring together many skills, including mathematics, machine learning, spatial analysis, computational modelling, statistics and software engineering," Data61's director of analytics research, Dr Simon Barry, says.
"That means we can conceptualise the problem, and work out how we can make inferences from the available data. We can develop algorithms and computational approaches to learn more about these complicated systems."
"With this work comes all the challenges related to accessing data, formatting it, developing processing techniques, and finally reporting and visualising the results – in short, to using data to learn about the world."
Although the data team is only just starting to work out how it is going to approach the problem of imaging the deep earth, Dr Barry says some significant barriers are already apparent.
Good data is limited, and any set of data can be consistent with a whole range of concepts about the underlying geology it represents. There is inherent uncertainty.
"Nevertheless, we are confident we will be able to make more accurate predictions as to where economic deposits are likely to be, and we will also be able to better represent the uncertainty in those predictions. This will feed into better risk-based decision-making."
Improved estimates of uncertainty are important because they suggest whether the expense of gathering further data and analysing it is most likely to pay off, which should lead to a better allocation of limited resources.
The potential products expected to emerge from Deep Earth Imaging include better exploration tools, algorithms and software platforms for analysing data, all of which can eventually be incorporated into industry software. Equally important is the indirect benefit of coming up with new ways of thinking about how to approach these sorts of problems.
"Data61 has strong links with the universities, and we are hoping to use those links to bolster our work and ensure it is at the cutting edge," Dr Barry says.