How do you detect the distant sound of an exploding submarine?

When the ARA San Juan went missing off the coast of Argentina, a global network of highly-sensitive listening posts designed to track nuclear explosions was called into service
Foto24/Gallo Images/Getty Images

It only takes 11 underwater sensors to listen to all the oceans on the planet. When something as small as a submarine vanishes, this ability to listen out for relatively tiny sounds becomes crucial.

This was the case of the ARA San Juan submarine, which last made contact with the Argentinian naval authorities on November 15. The submarine was making a routine sailing from Ushuaia in the south of Argentina to the northerly port of Mar del Plata when it vanished.

A week of searching followed, but with no success. Then, on November 23, the Comprehensive Nuclear-Test-Ban Treaty Organisation (CTBTO) detected an explosion off the coast of Argentina, in the same spot where the submarine was last detected.

Although the San Juan was 6,000 kilometres from the nearest sensor on Ascension Island, the CTBTO was able to detect a tiny spike in sound waves that indicated a small explosion. "In the case of the Argentinian submarine, looking for an explosion signal this small was like looking for a needle in a haystack," says Lassina Zerbo, Executive Secretary of CTBTO.

Zerbo and his team also worked to detect the missing MH370 Malaysian airplane that went down in 2014. "For months we were working, at points we would think we had a signal, we would send people to check and it would just be an old ship." This time, however, the International Monitoring System was successful in detecting a small explosion, providing the beginning of an answer to the families of the 44 crew on-board.

But how did they do it?

What is the Comprehensive Nuclear-Test-Ban Treaty Organisation?

The CTBTO set up the International Monitoring System Network in 1996 as a means of holding the world's nuclear powers to account. The network listens out for vibrations 24 hours a day and it's impossible for a major explosion to happen without the CTBTO noticing.

Today, the CTBTO has 337 monitoring systems on and off land, spanning every continent and every ocean. While the network monitors constantly for any sound that could indicate a nuclear launch or explosion, in recent years its role has shifted. It now listens to marine life, earthquakes and disappearances at sea. For that, it uses a network of just 11 hydrophone stations.

How do hydrophones work?

Hydrophones are sensors that look like microphones. They are placed at various points in the ocean and deployed in threes to make it easier to determine which direction sound waves are coming from. The hydrophones detect changes in water pressure and convert these changes into electrical signals.

The hydrophone stations are some of the most expensive monitoring mechanisms in existence. The most recent hydrophone station to be built was completed last year off the coast of the Crozet Islands in the Southern Indian Ocean and cost $34 million.

The CTBTO also has T-Phase monitors, which track seismic movements rather than hydro-acoustic sound waves. The seismic monitors track the pressure and impact of the movement underwater rather than just the sound frequency, giving the CTBTO a better understanding of an explosion's impact.

Computers at the onshore monitoring stations then record the signals from the hydrophones and transmit them to various satellites. These satellites then transmit the data to the International Data Centre in Vienna, Austria.

Analysing the data

The Centre receives ten gigabytes of data a day from its monitoring stations. Data is analysed, filtered and then searched for sound spikes. "We knew from the media attention that a submarine had disappeared, so we were able to look in a smaller area," says Zerbo. "This small signal would not usually catch our attention."

"Knowing that our sensors can catch any noise under the water, be it an explosion or just the submarine hitting the sea bed, we set about sweeping all the data we collected," he explains. This process took six days. Experts were told to focus on the smallest signal or impulse that could be an indication of any underwater explosion associated with the time and location of the submarine.

The Centre was forced to write and rewrite the programming script used by the monitors in order filter the data appropriately. "We weren't even sure if they were looking for a small explosion or just the burst of the submarine engine," Zerbo says. Eventually, a short, single anomaly was detected, indicating an unusual, non-nuclear event.

The team concluded this was made by the missing submarine, although quantifying the magnitude of the explosion is tricky. "Estimating a possible magnitude is even more difficult because this sound has propagated over a large distance from the source location to the receiver, but this is something that the engineers and scientists are currently still working on," the CTBTO said in a statement on Friday.

This article was originally published by WIRED UK