Just as I began writing up some notes on solar storms, headlines across news platforms read: “Powerful solar flare to disrupt communications,” “Sun erupts with massive solar flare, triggers radiation storm on Earth,” and so on. It was an uncanny coincidence that perfectly illustrated the very risks I planned on writing about from this underestimated danger.
The sun, a colossal fusion reactor, loses about 4.3 million tons of mass every second. This process results in the release of energy through electromagnetic and particle radiation. This continuous emission of particles is termed as the “solar wind,” and the more energetic bursts are known as “solar eruptions” or “coronal mass ejections (CMEs).” When this solar material hits Earth’s magnetosphere, it can cause what is known as a geomagnetic storm.
Geomagnetic storms frequently go hand in hand with enhanced activity of the Aurora Borealis or Aurora Australis (northern and southern lights). However, their impact on Earth, especially in our society today, extends far beyond the visual.
The most severe solar storm ever documented is the Carrington Event, named after the British astronomer Richard Carrington who first observed the solar flare that instigated it. On the morning of September 1st 1859, Carrington detected a large solar flare, a sudden flash of brightness on the sun’s surface. Within hours, particles from the flare reached Earth, causing one of the most potent geomagnetic storms on record.
The effects of the storm were noted worldwide. The auroras, spurred by the storm, were visible even in locations close to the equator, illuminating night skies across the globe. Telegraph systems in Europe and North America malfunctioned, with some operators even reporting sparks flying from their equipment, causing small fires.
Thankfully, the Carrington Event happened in an era when our dependence on technology was minimal in comparison to today’s standards. A solar storm of similar magnitude today would have far more severe consequences.
Powerful solar storms like the Carrington Event occur roughly once every 100 years, with the last major one occurring in 1921. The Sun’s magnetic field undergoes a cycle, known as the solar cycle, every 11 years, during which its magnetic field entirely flips, leading to the swapping of the north and south poles.
NASA scientists predict that the next peak in the Sun’s current cycle will occur at some point in 2025, although it’s not confirmed whether it will be the once-in-a-century event that many fear. In fact, just earlier this month (July 2023), a sunspot called AR3354, which is ten times larger than Earth, unleashed an X-class flare aimed directly at our planet, resulting in radio blackouts in some parts of the US.
The Frontier Development Lab (FDL) is an applied artificial intelligence research accelerator, resulting from a partnership between NASA and the private sector. It’s an initiative that brings together researchers from various fields such as artificial intelligence, data science, and astronomy, all aiming to solve problems that could significantly impact scientific understanding or future space missions.
A notable project developed at FDL is a machine learning model known as DAGGER (formally, Deep Learning Geomagnetic Perturbation). DAGGER utilises a type of machine learning referred to as deep learning, which excels in learning from large, high-dimensional datasets.
DAGGER is likely trained on substantial volumes of historical geomagnetic data, solar activity data, and potentially other related data. Using this information, it learns to identify patterns and correlations in the data that precede geomagnetic disturbances.
After training, it can then analyse new data about current solar activity and provide swift and precise predictions. According to FDL, DAGGER can predict these events 30 minutes before they transpire, generate predictions in under a second, and update those predictions every minute.
The DAGGER team tested the model against two geomagnetic storms, one in August 2011 and another in March 2015. In both instances, DAGGER was able to forecast the storm’s impacts rapidly and accurately. And, as it accumulates more data, its accuracy will only improve.
The code used to create the DAGGER model is open source. This means that in the future, it could be utilised by relevant entities worldwide, such as power grid operators, communications companies, satellite controllers and more., to provide warnings and enable proactive measures in the face of an impending solar storm.
As a civilization heavily dependent on technology – a dependency only set to increase in the future – it’s critical that we prepare for potential damage caused by solar phenomena. Just as we reinforce our homes against earthquakes or construct drainage for storms, we must also prepare for solar disruptions. We are due a significant solar event, but with adequate preparation, it need not be a harbinger of doom. We can admire the awe-inspiring auroras it generates and rest assured, knowing our infrastructure is robust enough to weather the storms of our celestial neighbour.
