Natural disasters threaten citizens around the world with disruption to essential services, damage to property and infrastructure, and the loss of life. The EU-funded ChEESE project uses supercomputing to help forecast accurate disaster scenarios. As a result, authorities in La Palma were able to make informed decisions and save lives when the Cumbre Vieja volcano erupted. 

When it comes to natural disasters, every second counts. “Technology can buy us time, and during a natural disaster, time equals lives saved,” says Arnau Folch, coordinator of the EU-funded Centre of Excellence for Exascale in Solid Earth (ChEESE) project, based at the Barcelona Supercomputing Center (BSC).

Folch and his colleagues are leading an effort to advance technology’s usefulness during a natural disaster. “Our goal is to optimally leverage the tremendous power of supercomputers and apply it to hazard assessments and early warning forecasts for earthquakes, tsunamis and volcanoes,” adds Folch.  

The superpowers of supercomputers

As Folch explained, a supercomputer, also known as high performance computing (HPC), is a computer offering a much higher level of performance than what can be done using your typical desktop or laptop. “With supercomputers, one can solve hundreds, even thousands of scenarios very quickly, ensuring that the results are available under the strict time constraints required during emergency situations,” he said.

According to Folch, having this capability is particularly relevant during a natural disaster because of the many uncertainties that often exist. “With supercomputers, we can create probabilistic tsunami forecasts that combine thousands of physics-based scenarios to alert and predict impacts well before the wave hits the coast,” said Folch. “Similarly, high-resolution volcanic ash cloud forecasts can be issued to help civil aviation authorities plan accordingly.”

This is all well and good so long as authorities have access to collated and processed supercomputing simulations – which isn’t always the case. That’s why ChEESE has focused on what is called urgent computing. “Urgent computing is akin to a dedicated infrastructure that ensures authorities have immediate access to the HPC-based simulations and related technologies they need to make emergency computations,” said Folch.

More specifically, the urgent computing solution being developed by ChEESE utilises both capability and capacity computing. To understand the difference, Folch notes that capability computing tends to focus on the ‘big picture’ problems, while capacity computing is more about running many small-scale simulations.

“With capability computing, seismologists can simulate the generation and propagation of earthquakes at unprecedented high frequencies and predict such effects as tsunamis and landslides,” said Folch. “With capacity computing, many simulations, each representing a different scenario, can be combined to account for model uncertainties.”

When a volcano erupts, ChEESE delivers

Supercomputers, urgent computing, capability and capacity computing – that’s a lot of computing. But what can it actually do? On 19 September 2021, the world was able to find out.

That’s when the Cumbre Vieja volcano on La Palma island in the Canaries erupted. Since then, it has been producing devastating lava flows and emitting substantial amounts of ash into the air, destroying property, polluting the air, and disrupting air traffic. To date, more than 1 000 homes and buildings have been destroyed and more than 6 000 people have been evacuated. 

To help manage the ongoing emergency, local authorities have been using the ChEESE volcanic ash pilot demonstrator. “Since the beginning of the eruption, ChEESE has provided local authorities with daily operational forecasts,” said Folch.

The demonstrator is based on a supercomputer located at the BSC, which runs computations at several resolutions and covers a number of possible eruption scenarios. It includes a dedicated urgent computing system, which ensures that ChEESE can run its simulations every morning from 6-7am, delivering its forecast no later than 7.30am.

With these forecasts in hand, local authorities can accurately predict what could happen in the next two days. “Decision makers can understand how upcoming volcanic activity will impact civil aviation and, based on this, already begin rerouting traffic,” added Folch. “They can also predict issues with air quality and implement confinement orders for those living within the exclusion zone.”

The Cumbre Vieja eruption is proof that the use of supercomputers allows public authorities to act quicker and more efficiently, thus reducing potential damage.

“This eruption clearly shows the benefits of using urgent computing during emergency situations,” said Folch. “Not only does it inform authorities about expected scenarios and help them make better decisions, most importantly, it helps save lives.”