An early warning system for solar flares

In 1859, Earth was attacked from space. A solar flare burst from the Sun, hitting the Earth with such power, the northern lights that normally dance in the sky over Alaska could be seen as far south as Mexico. The event took down telegraph systems across Europe and North America.

A similar-sized solar flare today would knock out electricity grids across entire continents. Last year, a report from the US National Academy of Sciences estimated recovery would take years, and cost more than $ US2 trillion.

Now, Queens University Belfast physicist David Jess and colleagues have devised a way to warn of monster solar flares early enough for the grids to be turned off. They published their algorithm in Nature Physics in November.

"We really do need to predict solar flares before they happen,” says Brett Carter, physicist and space weather scientist at Melbourne’s RMIT.  “They've found a very neat way to do this."

Accurately predicting the strength and timing of solar flares

has been impossible.

The Sun constantly burps out supercharged particles. When these burps hit the Earth’s atmosphere, interactions with our magnetic field create the southern and northern lights. But accurately predicting the strength and timing of the rare massive burps, or solar flares, such as the 1859 event, has been impossible.

A fast-growing sunspot, where the Sun’s chaotic magnetic field pushes its corona aside revealing its surface, is a warning sign. But only those triggered by super-strong magnetic fields will explode as full-fledged solar flares.

Currently, prediction relies on directly monitoring the strength of these underlying magnetic fields. But they twist and stretch constantly, and are faint when viewed from Earth-bound telescopes. This means that long-exposure times are needed and images are fuzzy, limiting their usefulness as warnings of an imminent threat.

Rather than directly monitoring the magnetic fields, Jess and his colleagues indirectly calculated magnetic field strength by watching how it wobbles.

The Sun’s magnetic field resembles a tangled mass of fishing lines, in which the jostling plasma creates waves. NASA’s Solar Dynamics Observatory, launched in 2010, sends back to Earth images of these oscillating fishing lines every 12 seconds. Jess and his colleagues developed an algorithm that combines this data with telescope data of sunspot temperatures to work out magnetic field strength across the surface of the Sun in real time. The result: clear pictures, more than 10 times faster than current prediction methods.

Presently the algorithm can't provide warning for an entire solar flare. The first part  – a burst of charged particles travelling at almost the speed of light – can’t be avoided. By the time the algorithm’s calculated the impending burst, it’s upon us. Thankfully, these particles deflect harmlessly off Earth’s atmosphere. The plan is to use artifical intelligence to understand and recognise the precursors responsible and warn us well before this part erupts.

But even in its current state, the algorithm can provide plenty of warning for the potentially highly damaging second stage: the coronal mass ejection, an eruption of 100 billion kilograms of material from the Sun. It travels much slower – only around five million kilometres per hour – and takes at least a day to reach Earth.

That’s plenty of time for automatic warnings to be sent to relevant organisations to disable satellites and shut down power grids likely to cop the brunt of it, says Jess.

Also in Cosmos: New eye in the sky