The data mastermind helping detect secret nuclear tests and clean energy minerals

Associate Professor Janice Scealy always knew she wanted to be a researcher.  

As a teenager, while most of her classmates would go home at the sound of the bell, she would often head to the university library in her hometown of Wollongong, where her mother worked.  

“I would go straight into the journal and book sections to find information to help complete my school assignments,” Scealy says.   

She didn’t know it then, but her inquisitiveness would eventually lead her to become one of the country’s most prominent statisticians.  

“I was an all-rounder at school and did well academically in different subjects such as physics, history and English, but never really knew what I wanted to do. I certainly did not know about statistics back then,” she says.  

Driven by the need to find a profession with decent job prospects, she started a degree in telecommunications engineering, quickly realising wiring circuits inside labs was not her idea of fun.  

That’s when she decided to pivot to applied maths and statistics, discovering she could solve real-world problems without ever having to pick up a soldering iron again.  

Today, as a researcher at The Australian National University (ANU) College of Business and Economics (CBE), Scealy is helping scientists make sense of their data: from seismologists who study earthquakes to geochemists who explore rocks and minerals.  

“Statistics allows me to use all those skills I had in different disciplines at school. You have to understand datasets from different domains and communicate with diverse experts to get it right,” she says.

Was it a nuke or a quake?

In 2024, Scealy’s research rippled far beyond Australia, drawing the attention of thousands of experts worldwide. 

Together with Earth scientist Dr Mark Hoggard from ANU, she developed a method capable of distinguishing an earthquake from a nuclear test with 99 per cent accuracy.  

This game-changing breakthrough is especially relevant in the current geopolitical context.  

With the expiration of the nuclear arms reduction treaty between Russia and the United States (US), monitoring countries who might be going nuclear on the sly is a matter of national security. 

But before the ANU-led study, this was not possible – at least with certainty.  

Since nuclear tests were moved underground in the 60s, the only evidence experts get come from seismic waves. The problem? Earthquakes generate these too – and there are thousands every day. 

This ambiguity came to a head in 2017, when a North Korean nuclear test baffled standard monitoring tools. The blast generated larger waves than expected due to the site’s complex topography and wasn’t effectively classified.  

“Statistics allows me to use all those skills I had in different disciplines at school. You have to understand datasets from different domains and communicate with diverse experts to get it right.”

The need to upgrade these systems brought Scealy to the scene. 

A specialist in analysing data on spheres and curved surfaces, she spotted shortcomings in existing monitoring methods. 

“When I read one of my Earth scientist colleague’s papers, I realised the model they were using did not fit their data,” she says.  

Drawing on her earlier work modelling the Earth’s magnetic field, she adapted her methods to analyse data from both earthquakes and nuclear explosions. 

The result was a classification algorithm that calculates which of the two scenarios is more probable with astounding accuracy. 

“It certainly wasn’t a five-minute job, even if it sounds like it. It was a purely collaborative effort. No one could have written this paper on their own,” she says.  

Currently, Scealy is collaborating with researchers in the Lawrence Livermore National Laboratory and in Los Alamos in New Mexico – the lab where Oppenheimer developed the atomic bomb – to expand the algorithm’s ability to distinguish between chemical and nuclear explosions – a development that could help improve global non-proliferation efforts. 

Her work could also sharpen scientific understanding of earthquakes. 

“If we’re able to more precisely determine the physical mechanisms of earthquakes, we can better understand the potential for tsunamis, landslides and volcanic eruptions and minimise their impact by improving the predictive accuracy of early warning systems,” she says.  

Powering the green transition

Critical minerals are, arguably, the building blocks of modern life. 

They’re used in electric vehicles, solar panels, computers and many other technologies. 

World leaders, including US President Donald Trump, haven’t been shy about talking up their strategic value or signalling how far they’re willing to go to secure them.   

Australia is home to some of the world’s largest reserves of critical minerals, such as lithium, cobalt and rare earths. In 2024, the government committed funding over a decade to map where existing deposits may be.  

This is yet another area where Scealy’s statistical expertise could become crucial.  

To be able to discover new minerals, she explains, the data must be appropriately interpreted to avoid measurement errors and uncertainties propagating into extraction decision-making.  

Her research could help Australia identify, with confidence, previously undetected critical mineral deposits – essential to the country’s transition to net zero. 

“I want to develop new statistical discrimination methods for analysing publicly available data to better identify deposits with reduced misclassification rates,” she says.  

Last year, Scealy was invited to join the Rio Tinto Centre for Future Materials with fellow ANU researchers, where she will support the discoverability of these precious resources.  

“We are hoping to predict new copper-favourable regions using statistical and machine learning techniques trained on spatio-temporal geodynamic data,” she says.  

Copper is central to the renewable energy transition. Experts estimate that meeting the world’s electrification goals will require 115 per cent more copper to be mined over the next 30 years than has been mined throughout history. 

Shooting for the stars 

Statisticians rarely make headlines. They work behind the scenes, helping shape decisions that affect us all. 

“Some people think it’s trivial, but they’re completely wrong,” Scealy says.  

“Today, everyone talks about innovations in artificial intelligence (AI) without realising it is mostly statistics. Even Donald Trump knows the value in them as he’s used electoral data to market himself and target voters. 

“A statistician is an essential person to have on board in any project if you want to avoid making biased decisions.” 

With the advent of AI, Scealy thinks her field matters more than ever.   

“AI is like a black box that not a lot of people understand. Statisticians are the ones who know how the algorithm works,” she says.  

Looking ahead, Scealy wants to continue supporting earth scientists, potentially helping to create a new sub-discipline that fuses geophysics with the precision of statistics. 

“A lot of statisticians have historically specialised in biology, creating a huge field called bioinformatics. There aren’t many who specialise in the physical sciences, and I see huge potential there. I would love to create a physical science version of bioinformatics,” she says.  

When asked where she sees her research going in the future, she wants to shoot for the stars – literally. 

“It would be interesting to analyse seismic data from Mars and work with astrophysicists,” she says.