Beautiful photographs of auroras make for viral social media posts, but it turns out those lights in the sky and online networks have something else in common. The algorithms created to keep us hooked to Facebook and Twitter have proven useful for understanding auroral substorms, the electromagnetic events that accompany the polar lights. With the ever-present danger that disturbances of the magnetosphere will bring down electrical or communications systems, that understanding could end up keeping social media operating during what might otherwise be a blackout.

Storms on the Sun can release bursts of charged particles that collide with Earth's magnetic field, funneling them towards the planet's north and south magnetic poles. There they excite the electrons in atmospheric atoms, which release light when they return to their normal state, producing the auroras borealis and australis (the northern and southern lights). Especially violent interactions between solar particles and Earth's magnetic field produce magnetospheric substorms, sometimes associated with full-blown geomagnetic storms. These push the auroras towards the equator disrupting charged particles in Earth's atmosphere and generally wreaking havoc on radio communication and navigation systems.

In the most dramatic cases, a downward cascade of charge can cause power failures, most dramatically in 1989 when millions of Canadians lost electricity for hours.

Substorms, and how they relate to storms, remain a poorly understood subject, but we do have plenty of data on how they affect Earth's magnetic field. Dr Lauren Orr of the University of Warwick has expanded our knowledge of the topic in a paper published in Nature Communications. 

To investigate whether substorms are one large current system or lots of separate currents, Dr Orr and co-authors used data from over 100 magnetometers, devices used to track disturbances in Earth's magnetic field, in the Northern Hemisphere to find a correlation between magnetometer signals. Rather than re-inventing the wheel, Orr applied the algorithm that allows social networks to recommend new friends to data collected from 41 substorms between 1997 and 2001, linking "like-minded" magnetometers.

As a press release put it, “If you like this magnetometer, you might like this one too.” 

"This is a good way of letting the data tell us what's going on, instead of trying to fit observations to what we think is occurring,” Orr said. 

Using this technique, Orr observed substorms form from many small communities into what the paper calls “A large spatially-extended coherent system,” basically a single large system or "community" at its peak. This led the researchers to conclude that substorms are one large coherent system that can cover large quantities of Earth rather than a number of small disjointed currents.

The 11-year solar cycle means we haven't had to worry about substorms for a while, but as the Sun's activity climbs towards the next peak anything that can lead to better space weather predictions could be important. Orr's work may be part of that, which means everyone on social media will have played a part.