Article from resourceful: Issue 11
If the world of geophysics has a vision of El Dorado it would be something like this: the ability to see into the uppermost surface of the earth’s crust with the very same accuracy that we see the earth’s obvious features – its rivers, rocks, mountains and plains.
A fantasy? Nobody is pretending that 'seeing' 50 metres or even a kilometre beneath the surface of the earth is going to be easy.
The electromagnetic, gravitational and other sensing tools currently used to chart even a few metres below the ground are constantly being developed and improved. However, in order to see that kind of geological detail lower down, scientists have to aim higher.
For Australia’s mining and resources sector, the move towards a geophysical El Dorado has reached a critical phase. The country is running out of shallow, easy-to-access mineral deposits. They are either non-existent or close to fully prospected. We are still producing from mines that our forefathers discovered in the late nineteenth and early twentieth centuries. Prospecting needs to go underground.
How are Australian scientists hoping to change this? Already well under way is the UNCOVER initiative, a collaboration of government, research institutes, universities and industry. It is an overarching project that combines all actors in the search for valuable underground mineral resources.
Aligned with the UNCOVER vision is CSIRO’s new Deep Earth Imaging Future Science Platform. This new science initiative will be run by CSIRO and admits to being highly experimental with a unique set of characteristics. It will initially focus on invention and creativity over applied science or industrial innovation.
Currently as part of Deep Earth Imaging, CSIRO is looking to hire some of the best exploration geophysicists across the world to create a world-first 'exploration geophysics hub'. There are, of course, experts in this field everywhere, but they have never been cross-fertilised within the same incubator.
An exercise of this size and nature calls for other skills as well. Physicists, chemists and mathematicians will also have to be part of the mix. As important as the geophysicists will be the data scientists and software engineers, because Deep Earth Imaging is equally an exercise in developing predictive data analytics tools as it is about discovering subsurface phenomena.
The creation of this hub is not just restricted to minerals and mining. It will be just as involved in seeking out other important resources such as energy and water. The open-ended nature of CSIRO’s Future Science Platform investment is to mix and match a number of different disciplines, the combinational results of which are as yet unknown. The earth’s the limit.
“We’re looking for areas where we can develop generic science around the prediction of the subsurface – and it could be relevant to all of the above – water, energy and minerals,” CSIRO’s minerals exploration research director, Dr Rob Hough, says.
“We’re not looking at any one thing – this is more about putting together a suite of generic tools and software, which will allow us to simulate, model and predict the subsurface in a way we simply can’t do today.”
The Deep Earth Imaging platform is funded to the tune of $3 million per annum over the next three years.
The platform will be run by Dr Mike McWilliams, a former CSIRO executive, under whom there will be five team leaders. It will also include 18 post-doctoral, early-career researchers who have been carefully selected from around the world.
Dr Hough says the aim is to develop tools that can simulate and predict the geology of the subsurface and offer the best possible measure of quantification around the predictions.
“It’s about how we’re going to develop a beautiful multi-coloured image that accurately reflects the geophysics of any subsurface. And, determining if it’s possible to interpret the subsurface using these tools to not just see, but to predict, what the underlying geology is going to look like,” he says.
The team will be developing new analytical software tools that are founded in rock physics, but that also draw from predictive technology, machine learning, geological uncertainty analysis and geoscience modelling. The tools will manage real-time data streams and fuse multiple inputs from geology, hydro-geochemistry and geophysics.
Tools to actually see under the surface are currently a scientific work in progress. Airborne systems are becoming progressively sophisticated at collecting the electrical properties of the earth. There are also earth sensor tools and others measuring magnetics and gravity, all of which will be used to explain various expressions of the earth in relation to the other.
Recently, South Australia’s Department of Manufacturing, Innovation, Trade, Resources and Energy released information about surface cover over orebodies based on data already available from mining company reports. The industry is working on coordinating all the state geological maps, building knowledge of the larger mineral systems and their potential.
Researchers are looking deeper into the architecture of the earth – the igneous intrusions, the cracks and the potential mineral traps – in a way that would have been impossible just a few years ago.
There are new stratigraphic drilling methods that reveal the buried geology, which can be intelligently extrapolated across larger areas. The Deep Exploration Technologies Cooperative Research Centre is now pioneering coil tubed drilling – similar to petroleum industry methods – that allows analysis of rocks as drilling occurs. We have new drill sampling technologies that allow us to test samples live so that companies can make smarter decisions about the existence of nearby ores.
These are just a few examples of a whole new suite of emerging technologies available. These technologies provide more detailed information at greater speed in order to take much of the guess work and expense of mining exploration out of the equation.
Despite these increasingly sophisticated methods of seeing deeper and more accurately, Dr Hough says we still sit at a fairly primitive point in predicting subsurface geophysical properties and the geology. We often use one or two geological models produced by geophysics, test them by drilling some holes and then make a decision based on the results.
“What we now have to do is design real, live tools for forecasting what the subsurface geology is going to be. There may be several models, possibly even thousands of iterations from the geophysical data, but in terms of managing risk we need to be able to measure the uncertainty around our predictions.”
The ultimate development of the tools comes down to one thing: how will they be used to significantly reduce the risk of exploration beneath the surface. How will explorers – whether for oil, water or minerals – be able to use predictive models to eliminate the uncertainties? Dr Hough hopes to see the day when geophysical models can predict underground phenomena as accurately as weather models predict rainfall.
In the end the platform’s level of invention will be applied to industry.
A consultant to the Australian Mineral Industries Research Association (AMIRA), Tim Craske, says that while he applauds the ideals of the Deep Earth Imaging platform, it must not operate in a vacuum.
“Deep Earth Imaging is going to have to deliver something meaningful about the top couple of kilometres of the surface of the earth, otherwise it is in danger of being an academic exercise,” Mr Craske says.
“It’s not going to change how we do work unless it’s relevant.”
Mr Craske, who is working on the roadmap with CSIRO and industry in the UNCOVER project, would like to see a kind of open-source software philosophy applied to Deep Earth Imaging.
“As long as the products are planned for open source – which it sounds like they are – then this will be an exciting endeavour,” Mr Craske says.
“CSIRO needs to be putting stuff out for people to play with and use."
“What I hope to see is Deep Earth Imaging setting itself up a bit like Linux and Wikipedia – people can see it and pass comment on it. For this to become valuable, then value has to be created along the way as well as at the end.”
What Mr Craske doesn’t want to see is what he terms “the cathedral model” where the monks write the book but don’t show it to anybody until it’s finished. That’s the traditional software model, he says.
“What we need is the bazaar model where anybody can come along with a basket of stuff and get involved. Nobody is an owner and nobody is not an owner. They jump into that space without asking permission and generally the products get spectacularly better than anyone imagined in the first place.”
Dr Hough agrees that industry must be informed along the way as developments occur in the Deep Earth Imaging platform, and beyond this, he recognises that industry may possess valuable data and technology the platform could work with.
“Of course, we’ll be looking in time to partner with industry but we need to devise the methodologies before we can do that. We don’t want to define the business model from the start,” Dr Hough says.
“Don’t get me wrong, we may be at the pre-methodology phase but we have some very good ideas as to what our priorities and major research themes will be, as well as encouraging possibilities to attract people to work with us using the very latest data and technology. Just watch this space.”