I

EXETER, United Kingdom — It was an overcast morning in southwest England, but Kirk Waite was staring at the sun. In the open-plan forecasting room of the Met Office, the U.K.’s national meteorological agency, the country’s senior weather-watchers were monitoring low-pressure systems sweeping in from the Atlantic and across the British Isles. Waite was studying weather of a very different sort.

He sat at a crescent-shaped desk arrayed with computer monitors displaying real-time images of our local star. Generated by the Solar Dynamics Observatory (SDO), a NASA satellite that has been taking second-by-second snapshots of the sun since 2010, the images had been processed by a spectral filter to display the solar surface in a range of lucent wavelengths measured in angstroms (Å).

At 6,173 Å, the sun appeared as a mostly featureless sphere with a couple of rogue blemishes: active sunspots that showed high levels of electromagnetic activity. The most spectacular projection was 171 Å, which depicted the star in the deep ultraviolet range. At this wavelength, it was an explosive yellow orb encircled by a chaotic nimbus of sunbursts and looping streamers. For around six years, the sun has been in what is called Solar Cycle 25, and the tumult shown at 171 Å was characteristic of the “solar maximum,” its most turbulent phase, which commenced last October.

This morning, Waite, the on-duty forecaster at the Met’s Space Weather Observation Centre (MOSWOC), was scrutinizing sunspot number 3998, an archipelago of blotches around four times the Earth’s diameter. On his central monitor, he brought up a greenish image of the sun at 94 Å, highlighting radiation in the X-ray range, and toggled the timestamp back around 24 hours to display 3998 just as it erupted in a torpedo-shaped expulsion of light. This was a solar flare, a massive discharge of radiation and solar energetic particles (SEPs).

Next, Waite turned to a pair of projections, each with a black circle overlaying the solar disc. These were coronagraphs, images taken from another observation platform stationed around a million miles from Earth. The spokes of light emanating from the disc showed the solar wind, a constant flow of ionized particles streaming off the sun’s visible surface, or photosphere, at a rate of one million tons per second.

But sun-watchers like Waite are more preoccupied by the more intermittent crescendos, outbursts with the explosive power to reach Earth with unusual velocity and volume. These can be divided into two distinct but intertwined phenomena: solar flares and coronal mass ejections (CMEs).

Orbital sensors had determined that the flare Waite showed me was an M3.3 — moderate but worth keeping an eye on. “It lasted around two hours, quite a long duration, which is often a sign that a CME might follow,” Waite said. Sure enough, a couple of hours later, the coronagraphs showed a blast of matter exploding from the right “limb” of the photosphere. Waite’s task during his 12-hour shift was to forecast “space weather” — to monitor these two types of solar eruptions and to predict their potential impact on Earth.

Using a mouse cursor, Waite lassoed the outer limits of the CME. Scrolling forward a few minutes — the ejection now greatly expanded — he drew another. The time-lapse between the two would provide an estimate of the CME’s ejection speed (this one was travelling at almost 800 miles per second), from which computer modelling could extrapolate an estimate of the eruption’s volume, density and its likelihood of hitting Earth.  

“The data says there’s an ever-so-slight chance of it just grazing us. So we can incorporate that into our forecast,” Waite said. In the subsequent hours, 3998 had become less complex. This particular event wasn’t large enough to cause significant disruption, Waite concluded. But the question of what these phenomena could do is a different story.

If you took a straw poll of the general public, chances are that few people would have any idea what space weather is, if they’ve ever heard the term at all. In contrast to terrestrial weather, space weather cannot be felt. It doesn’t warm your skin, drench your clothes or blow down your fence. Unlike the floods, droughts and hurricanes that have beset human civilizations since ancient times, it is not an age-old threat. For the first 10,000 years of human civilization, the sun’s flares and CMEs would have had no impact on life at all. 

It is only since humanity constructed a planet-scale network of electromagnetic technologies, and subsequently grew to depend on that network for just about everything, that the sun’s activity became a potential hazard. In basic terms, the primary danger of space weather is its capacity to produce an electromagnetic pulse (EMP). Upon making contact with the upper reaches of the atmosphere (the ionosphere), charged particles thrown out by the sun can instigate a “geomagnetic storm”, inducing currents in the Earth’s crust that overwhelm electrical equipment and its infrastructure, resulting in cascading malfunctions, power surges and blackouts. Anything that relies on electricity is vulnerable. Satellites, power grids, aviation, railways, communications, farming, heavy industry, military installations, global trade, financial transactions — the categories of vital systems that could be impacted by a sun-borne EMP are endless and interconnected, affecting every facet of our networked society.

The United Kingdom-based MOSWOC is one of only three institutions worldwide tasked with assessing and forecasting that risk. (The other two are in Boulder, Colorado, and Adelaide, Australia.) Each monitors solar activity 24 hours a day, 365 days a year. Low-severity space weather, like the expulsions Waite was scrutinizing during my visit, occurs all the time. During the solar maximum, MOSWOC usually records around 1,000 such events per year.

But playing at the back of every forecaster’s mind is the hypothetical centennial event, the moment when a sunspot might dispatch a solar storm at a scale that we know has happened historically, but never in our modern, technological age.

Across the desk from Waite, his colleague Krista Hammond recalled an event in May 2024 when a fusillade of CMEs careened into Earth’s atmosphere. “This was a really interesting one,” she told me, “because rather than just one enormous CME, we had multiple smaller ones, each travelling faster than the one before.” Geomagnetic storms are ranked from G1 to G5, and this one topped the scale, the first G5 in over two decades. As photos of unusually vivid and extensive auroras saturated social media, the forecasters at MOSWOC were working overtime. Waite, on scheduled leave at the time, came in anyway.

On this occasion, adverse impacts were mercifully slight. Starlink’s satellite constellation reported degraded service. In the U.S. and Canada, where agriculture relies on precise satellite positioning, some large farming operations were forced to pause work at the height of the planting season.

It could have been worse. But the event reignited concerns that space weather could precipitate a “black swan” event — a seemingly improbable, unexpected catastrophe, with severe ramifications. Space weather was first added to the U.K. National Risk Register in 2012. In 2023, its threat level was increased from level 3 (moderate) to level 4 (significant), which includes floods and emergent infectious diseases.

This is why the work at MOSWOC and its sister observatories is so crucial. It is also why governments and academics are deploying more resources to understand solar storms and to anticipate how a really big one might play out. The curious paradox at the heart of space forecasting is that the satellites and supercomputers that empower the observations are themselves vectors of vulnerability. The more umbilical our relationship to technology becomes — the more our lives and livelihoods become governed by algorithms and automation — the greater the risk of disaster.

II

For those tasked with observing and understanding the sun, apprehending it as a threat requires an inversion of some perennial assumptions. At a basic level, the sun exists as a largely beneficent force. Bestowing light and heat, it is the determinant of our “Goldilocks zone,” the sliver of cosmic conditions that render our planet habitable. Were it larger or smaller, closer to Earth or farther away, life as we know it would not exist.

In ancient cosmologies, the sun was a first-rank divinity — carried on the falcon-shaped head of the Egyptian god Ra or dragged across the sky by Apollo’s celestial chariot. For the Aztecs of Mesoamerica, the sun god Huitzilopochtli was a capricious overlord who demanded daily oblation. Most of the ritual human sacrifices for which the culture has become infamous were carried out in obeisance to the great orb, whose whims dictated the harvests and hence the fate of civilization.

It was Galileo Galilei who pioneered telescopic astronomy: By projecting the sun onto a flat pane, he found that he could transpose its blinding incandescence into an observable form. The German British astronomer William Herschel spent 40 years recording the evolution and passage of sunspots from 1779 to 1818, an admirable empirical endeavour only faintly undermined by his conjecture that each one comprised an “opening in the sun’s luminous atmosphere which is likely inhabited.” A couple of decades later, the German astronomer Heinrich Schwabe discovered the existence of the solar cycle, the approximately 11-year period in which the sun transitions between the solar minimum, when it exhibits its lowest level of activity, to the solar maximum and back again. This was later backdated, based on the earliest sequential records, to 1755: Solar Cycle 1.

Then, on Sept. 1, 1859, Victorian brewery owner and amateur sun-watcher Richard Carrington found himself observing a sunspot from his garden in Surrey when he saw a sudden burst of light.

• • •

In the Met Office library, archivist Catherine Ross waited for me with a fabric-bound compilation of “Philosophical Transactions,” the academic journal of the Royal Society. On page 423 was a report first published in 1861 describing a strange celestial event two years earlier, accompanied by a sketch of some sunspots drafted by a precise hand.

This is what Carrington had been doing on that early fall morning in 1859. After cranking open the domed roof of his personal observatory, he projected an image of the sun from the viewfinder of his brass telescope onto a pane of glass painted with distemper and then set about engaging in what had become an obsessive undertaking: tracking and drawing the progress of sunspots across the solar surface.

At 11:18 a.m. came the flash. “Two patches of intensely bright light broke out,” he later recalled. “My first impression was that by some chance a ray of light had penetrated a hole in the screen attached to the object-glass. … [T]he brilliancy was fully equal to that of direct sun-light.”

Carrington later confessed to being “somewhat flurried by surprise.” But his unusual observation might have remained an astronomical footnote, dismissed as an equipment fault or hallucination, had it not been for the mass of solar material soon barrelling toward him, oblivious, at 5 million miles per hour. 

By the time it reached Earth, around 18 hours later, dawn was breaking over Carrington’s home. In America, it was still the dead of night.

The ensuing auroras were, to this day, among the most expansive ever recorded. Sightings of the Northern Lights were reported as far south as El Salvador, at just 13 degrees north of the Equator. Encroaching from the opposite pole, the Southern Lights flickered over Santiago.

Across the Western Hemisphere, local newspapers chronicled all manner of strange nocturnal happenings — of people awakening as they mistook the lights for sunrise and cocks crowing to greet a false dawn.

Eyewitnesses described seeing the heavens illuminated by “a livid red flame,” a radiance “so bright that one could easily read common print,” the sky cascading with “showers of nebulous matter like star-dust.” On America’s Atlantic coast, the Charleston Mercury reported: “No one could look at [the ocean] without thinking of the passage in the Bible which says, ‘the sea was turned to blood.’ The shells on the beach, reflecting light, resembled coals of fire.” So intense and haunting were the lights that several people admitted thinking it was the end of the world.

It was also one of the first times a solar event disrupted human technology. As the sky blazed overhead, telegraph operators in America and elsewhere found that Morse code messages sent across electrical wires failed to transmit. In many cases, their machines fizzed and spat fire or short-circuited completely. American Telegraph Company employees in Boston, having disconnected their malfunctioning apparatus, found that they could communicate with Portland, Maine, 100 miles north, using only the auroral current. Unseen, something alien had suffused the telegraph wires with an electrical charge.

What Carrington had witnessed would come to be known as a “white-light flare,” a prelude to a massive CME. As he watched, the points of light became “enfeebled” and then vanished. His famous sketch of the start and end points of the erupting sunspot suggested that, over the course of the eruption, it had migrated 35,000 miles across the photosphere.

In the final pages of the Royal Society report is a compilation of graphs showing the peaks and troughs of activity recorded by rudimentary magnetometers. Connecting these magnetic jolts to Carrington’s observation, the report surmised: “We have grounds for supposing this primary disturbing force to reside in our luminary.”

Retrospective analysis of the CME from the “Carrington Event,” as it would come to be known, estimated that it released energy equivalent to 10 billion megatons of TNT. It would go down in history as the benchmark by which future space weather would be measured, and a potential harbinger: If a Carrington-class solar storm were to strike Earth in the modern age, what then?

Not long after the flare, coronographs show a colossal CME expanding at the edge of the sun. The on-duty forecaster’s first question is: What’s the trajectory? This one manifests as an advancing halo, like a giant smoke ring radiating outward, indicating that this coronal mass ejection is heading straight for Earth.

The forecasters feed the data into a computer model, which projects that the cloud contains a billion tons of plasma. On the geomagnetic storm scale, anything G3 and above warrants an alert. This one is a G5+, a once-in-a-century event. It’s the biggest Earth-directed CME in living memory.

The managing forecaster issues a “geomagnetic storm watch” alert that provides an estimate of the CME’s arrival and its projected repercussions. Observatory heads are notified; meetings are hastily convened with government emergency planners.

Amid all this activity, there is a waiting game. The event’s true magnitude won’t be known until it is right at the doorstep.

III

For decades afterward, no one was entirely sure what had caused all those geomagnetic jolts in September 1859. Looking at the sun on MOSWOC’s monitors, one thing that becomes clear is its astonishing dynamism. Often erroneously described as a hot ball of gas, the sun is in fact composed of plasma, the fourth state of matter, which is created when a gas is heated to the point where some of its electrons break free from their atoms.

Weighing around two octillion (2,000,000,000,000,000,000,000,000,000) tons, the sun measures 865,000 miles from pole to pole and has a volume equivalent to 1.3 million Earths. Temperatures within its core are believed to reach 28 million degrees Fahrenheit; its expansive energy is held in check by the countervailing force of gravity.

Decades of spectroscopy reveal that this gargantuan furnace contains all manner of trace elements — oxygen, magnesium, silicon, sulphur — but 98% of its mass is composed of the two simplest elements in the universe: hydrogen and helium. The fundamental physical reaction that generates the sun’s energy occurs when two atoms of hydrogen, each with one proton, fuse together to produce one atom of helium with two protons. The heat created by this fusion strips electrons from atoms in the main body of the sun, and this convective motion combines with the sun’s rotation to produce a powerful and complex electromagnetic field. This is the process that characterizes a “main sequence star.”

Born from a cloud of interstellar gas around 4.6 billion years ago, our sun is now believed to be around halfway through its lifespan. In 5.5 billion years, the hydrogen in its core will be all but exhausted and it will push against the confines of its gravitational field, becoming a red giant that engulfs Mercury and Venus before collapsing into a white dwarf, a stellar corpse so dense that a teaspoonful would weigh as much as a bull elephant. The Earth, if it survives this cataclysm intact, will be cast adrift into the void, no longer held in the sun’s orbit by its gravitational pull, no longer warmed and energized by its glow, a dead husk spinning in the darkness.

But back to more immediate concerns. At the start of a new solar cycle, the complex dynamics of the sun’s magnetic field cause its poles to switch places. During the solar minimum, the field can be thought of as a series of taut lines running north to south.

“You start off with a dipole not unlike the one you have on a solid celestial body like the Earth,” said Ryan French, speaking to me on a video call from his home in Boulder, Colorado, where he works as a solar physicist at the Laboratory for Atmospheric and Space Physics. This equilibrium is corrupted by the sun’s differential rotation. In contrast to the Earth’s consistent 24-hour rotation, the fluidity of all that roiling plasma causes the sun to spin faster at its equator, where it takes around 25 days to complete a full rotation. At the poles, the same rotation takes around 36 days. To illustrate what happens as a result, French held up an iPhone charging cable that he had lying on his desk. Gripping it taut between two fists, he then gradually moved his hands together, while simultaneously rotating each fist in opposite directions. The cable began to sag and concertina, then folded in on itself, forming a loop.

Similarly, over the course of a solar cycle, the lines of the sun’s magnetic field bunch, crisscross and tangle, particularly around the mid-latitudes. The intensity of the magnetism at these junctions — over 1,000 times more powerful than the sun’s general field — prevents the convection of plasma from the core to the photosphere, resulting in darker regions: sunspots. Burning around 3,600 degrees cooler than the average surface temperature of 9,900 degrees, each sunspot is a crucible of magnetic potential energy, a hotbed of space weather.

The largest sunspots, like the one Carrington was scrutinizing in 1859, grow to more than 100,000 miles from edge to edge, over 10 times the Earth’s diameter. As the solar cycle approaches its maximum, the tangling of the magnetic fields accelerates; plasma circulation is impeded until the trapped energy breaches the resistance tamping it down, and bursts free. 

French jerked his hands apart and the looped cord snapped taut. “All the energy that is stored in that magnetic field has got to go somewhere,” he said. “So when the loop snaps, all of the electrons and ions stored within it rush outward in every direction. Some of them collide with the sun’s lower atmosphere. The rest accelerate out into the solar system.”

Solar flares are the initial pop. Moving at or close to the speed of light, the radiation and particles they release can traverse the 93 million miles between the sun and Earth in a matter of minutes. By the time a flare registers on orbiting sensors, then, it has already arrived.

CMEs are a different beast. Often presaged by a flare, they travel more slowly, taking between 15 hours and five days to reach Earth. The material they carry is also heavier, more capable of penetrating the Earth’s atmosphere, and significantly more consequential. French differentiated between the two using a commonly cited metaphor. “You can think of it like a medieval cannon,” he said. A solar flare is the flash of ignited powder. The CME is the cannonball.

Nineteen and a half hours after the giant CME was observed exploding off the sun’s photosphere, its leading edge has just reached Lagrange-1, a point a million miles from Earth on a straight line to the sun.

Here, a flotilla of observation modules gauges the cloud’s polarity and hence its “geoeffectiveness,” or disruptive capacity. If its magnetic orientation matches Earth’s, even the most colossal CME will follow the Earth’s dipole around the planet, like water running over a boulder in a stream. But if its magnetic orientation is off-kilter or inverted, the particles will trigger “magnetic reconnection” with molecules in the Earth’s atmosphere, releasing a burst of energy. So it is with our hypothetical storm.

At the observatories, forecasters put out a “sudden impulse warning.” Mitigation strategies, long in the planning, now kick into gear. The CME will infiltrate the ionosphere in approximately 20 minutes.

IV

The decades following the Carrington Event were marked by other uncanny occurrences. In 1921, as another spectacular auroral display tripped across the night skies of the eastern United States, railway switching systems in Manhattan went haywire and sparks flew from overhead telegraph wires, igniting fires across New York state. In 1967, at the height of the Cold War, a series of powerful flares caused part of the U.S. Air Force Ballistic Missile Early Warning System to jam, almost provoking the military to scramble nuclear-equipped aircraft in response to a presumed Soviet attack. In 1989, a large CME took out the power grid across Quebec, leaving 6 million people without electricity for nine hours. In July 2012, sun monitors at the Met Office’s embryonic space-weather desk watched anxiously as a massive geomagnetic storm, comparable in size to the Carrington Event, tore through the orbital path that Earth had vacated just a few days earlier.

By the time of that near miss, the potentially cataclysmic implications of extreme space weather had become a subject of intensive research and conjecture. In 2008, America’s National Academy of Sciences held a workshop to study “the effects of ‘a space weather Katrina,’ a rare but, according to the historical record, not inconceivable eventuality.” The subsequent report noted that while the devastating 2005 hurricane cost an estimated $125 million, the damage wrought by a centennial solar storm could potentially run into the trillions.

Mark Gibbs, a 38-year veteran of the Met Office who has headed MOSWOC since its inception in 2012, admitted that his first reaction upon being approached to lead the U.K.’s new space-weather desk was to ask: “What is space weather?” But he quickly understood the importance of the new assignment. The London Olympics were just around the corner, Gibbs recalled, “and you could just imagine a big space event taking out satellite communications just as the 100-meter final was going on.” Today, he doesn’t do the forecasting legwork, but if a Carrington-class storm ever happened, he would be the one in the hot seat.

Despite his years of experience, he conceded that there is still much that forecasters don’t know about how such a centennial event might transpire. Smaller pulses like the geomagnetic storm last May that Waite and Hammond at MOSWOC characterized as a “once-in-20-year event” have served as salutary trial runs. “It was big enough to test all our processes,” Gibbs said. “We were in constant contact with government, we were briefing stakeholders. It meant that we could test everything without the place being on fire.”

One man thinking about what might happen the day the fire arrives is Jim Wild. A professor of space physics at England’s University of Lancaster and a member of the Space Environment Impacts Expert Group, he’s spent the last 15 years drawing up “reasonable worst-case scenarios” that form the basis of government mitigation and resilience strategies. His research is mainly concerned with what is arguably a solar storm’s most far-reaching hazard: the potential of geomagnetically-induced currents (GICs) to cripple terrestrial electrical networks.

When an Earth-directed CME careens into the magnetosphere, additional amperage is injected into the complex network of electrical currents encircling the planet. The extent of a particular region’s vulnerability to GICs depends on its underlying geological profile. “The intensity of the induced electric field is inversely proportional to the conductivity of the ground,” Wild told me. In areas sitting on top of resistive basement rock, “the excess current will inevitably find its way into long pieces of electrically conducting infrastructure — lengths of metal that are bonded to the Earth.” Once inside such networks, the surplus electricity can overload the system, causing malfunctions and, in extreme cases, rendering them inoperable. This is how global telegraph networks were overwhelmed in 1859.

Recently, Wild and other researchers found that GICs could trigger “wrong-side” failures in railway signalling, erroneously flipping trackside signals from red to green. But he’s more worried about the power grid — specifically the vulnerability of extra-high voltage (EHV) transformers. 

EHVs are huge structures, sometimes four stories high, that are responsible for ratcheting voltage up for long-term transmission and down for domestic use, depending on need. When a GIC infiltrates an EHV transformer, the collision between the auroral charge’s direct current (DC) and the alternating current (AC) of our national power grid generates voltage variations and excessive levels of internal heating that can trip out transformers or even destroy them completely. This process, operating in a chain reaction as the rogue current saturates one part of the network and then another, can culminate in “voltage collapse” — a metastasizing power cut or complete system shutdown. Blackout.

Recovering from serious damage to EHV transformers is an agonizing process in the best of times. Each one is custom-made, constructed by only a handful of specialist manufacturers, and the supply chain is grindingly slow. A new transformer can take anywhere from several months to two years to build and install. They are so large that moving them into position requires whole highways to be shut down.

In 2016, the University of Cambridge Centre for Risk Studies published the “Helios Solar Storm” study, a stress test based on a scenario in which a Carrington-class storm sparked a U.S.-wide power system collapse. According to the study, such an eventuality would cause a third of America’s transformers to go offline, with 3% incurring major damage, plunging almost half the national population into darkness. In the most extreme scenario, up to 10 million people would be stranded without power for 300 days.

Wild and many of his counterparts at what he characterizes as “the ground end of things” remain hopeful that it will never come to that. Experts have spent much of the last two decades coordinating with government and private sector contingency planners on emergency mitigations.

But making predictions and planning suitable responses is becoming more complicated with newer, runaway technologies, especially those “embedded ubiquitously across multiple systems,” Wild explained. Among the hardest to quantify is the risk to satellite services, which operate outside the protection of the Earth’s atmosphere. Our global satellite constellation is diffuse and growing fast, spread across multifarious government and private operators, with minimal central oversight or control. Radiation from solar flares can invade and corrupt onboard circuitry, while the injection of space-borne matter from CMEs densifies the atmosphere, causing satellites to veer off course or even bump entirely out of orbit. In February 2022, when a flotilla of new Starlink satellites was hit by a G2 storm, at least 40 of the 49 suffered catastrophic failure. Last May’s solar storm temporarily doubled the rate at which the NASA Hubble Telescope descended Earthward, or its orbital decay, from 130 to 260 feet per day.  

One area of particular vulnerability is the Global Positioning System (GPS). Mostly confined to high Earth orbit, these satellites are less susceptible to atmospheric drag. However, the accuracy of their measurements and transmissions is still predicated on the atmosphere retaining a consistent chemical and electrical character. As CME material adulterates that barrier, it can distort the wavelength of signals traveling between satellites and the ground, an effect known as “scintillation.” This inevitably results in location errors and “loss of lock,” or signal, corrupting GPS accuracy by tens or perhaps hundreds of feet.

Today, GPS reaches into every corner of communications, including navigation, logistics, utilities, travel and commerce. High-tech farms use it to plant and harvest; aviation authorities depend on it to coordinate air traffic. At ports around the world, autonomous cranes use GPS positioning to load containers onto ships. Oil and gas exploration utilizes the same triangulations to calibrate drilling operations. Financial trades, once recorded manually on paper certificates, now rely on timing signals transmitted from atomic clocks within GPS satellites, which are accurate to within a billionth of a second. Media broadcasts, cellphone networks and a multitude of other sectors similarly depend on the transmission and recombination of timestamped digital data. A solar storm on a scale sufficient to frazzle satellite electronics and scintillate their signals could cause all of these vital functions to falter.

A 2021 report from the U.S. National Security Telecommunications Advisory Committee on the hazards of our growing societal reliance on GPS satellites concluded that “their vulnerabilities pose a near-existential threat.” In the coming years, with the anticipated growth of AI and autonomous vehicles, that dependence is only expected to grow.

At the bow-wave of disaster, there is beauty. As the electrified nebula slams into atmospheric atoms, its south-oriented polarity triggers magnetic reconnection. On the dayside of the Earth, few people will have a clue that anything unusual is happening. But on the nightside, the sky erupts into shimmering cirri of light: auroras. 

Auroral greens, blues and purples typically appear when solar particles interact with oxygen and nitrogen molecules at altitudes of between 60 and 150 miles. However, the intensity of the barrage on this occasion excites oxygen at much higher altitudes, generating light at lower frequencies, which manifests to the human eye as an apocalyptic red.

Meanwhile, GICs are rampant everywhere above the 40th parallel, and contingency planners face screens that are flashing with alerts. Power-grid operators hasten to reroute electricity away from saturated parts of the system. But there’s not enough time for all the necessary mitigations to be implemented.

Within seconds, extra-high voltage transformers in power plants and substations in multiple geographical regions are rattling and groaning under the strain. Power cuts soon blanket entire conurbations.

V

So just how bad could a Carrington-class storm be? While the protocols have been in gestation for some time, the answer remains a subject of dispute, inviting prognostications ranging from the blasé to the apocalyptic.

Speaking to a congressional hearing in 2014, Peter Vincent Pry, then the director of the Electromagnetic Pulse Taskforce, contended that a prolonged power grid outage could kill 90% of Americans through starvation, disease and other consequences of rapid societal collapse. Pry, who died in 2022, wasn’t particularly known for his sangfroid; his corpus of work includes books titled “Electric Armageddon,” “Blackout Wars” and “Apocalypse Unknown.” 

But even more sober experts have conceded that the threat posed by space weather is primarily due to its totalizing potential — its capacity to induce cascading failures in multiple social and economic functions. A 2019 FEMA report presented two possible catastrophes that could affect the entire U.S. — a pandemic and a solar storm. In the event of a truly massive storm affecting multiple nations all at once, Gibbs told me, “The concept of mutual aid goes out the door.”

The Helios stress-test predicted worst-case economic losses of $2.7 trillion in the U.S. alone, around 8% of its national GDP, and insurance industry losses of $333.7 billion. But even the most rigorous worst-case scenarios remain poorly constrained. “We can model it, estimate the impact, and try to prepare,” Wild said. “But we can’t predict the implications of a solar storm on this scale with total surety because we’ve never experienced one.”

As we enter what some have dubbed “the golden age of heliophysics,” physicists and forecasters are seeking new data and building new projections. Last Christmas Eve, NASA’s Parker Solar Probe, a 10-foot-long module bristling with sensors and encased in 4.5 inches of carbon-composite shielding, passed within 3.8 million miles of the photosphere, the deepest incursion of any human spacecraft into the sun’s outer atmosphere. Having picked up speed by sling-shotting around Venus, it was zipping along at 430,000 miles per hour, fast enough to circle the Earth’s equator in a little over three minutes. Parker is just one component of NASA’s Heliophysics System Observatory, a growing fleet of spacecraft tasked with furthering our understanding of solar dynamics.

Earlier this month, the European Space Agency (ESA) announced that its “Solar Orbiter” had sent back the first pictures ever taken of the sun’s southern pole. In 2031, ESA will launch its “Vigil” observation platform to Lagrange-5, a cosmic coordinate equidistant from the sun and Earth. Once in position, it will transmit a “side-on” image of the sun, enabling observatories to see solar activity up to five days ahead of visuals generated from Lagrange-1.

Yet for all the brainpower and innovation deployed in its direction, the true extent of the energy expelled by the sun is rife with unknowns. Heliophysicists remain stumped by the “coronal problem,” where regions of the corona blaze at temperatures of up to 1.8 million degrees — orders of magnitude hotter than the surface itself. Nor can they be sure that the centennial event for which they are preparing will be Carrington-sized, or something bigger.

Over the following days, human society enters unknown territory. A power-grid failure on this scale, known as a “black sky” event, disrupts or incapacitates every utility that depends on electricity and computer technology. In the modern age, that means virtually everything.

Failures cascade through the infrastructure. Lights go out, telecommunications crash, supply chains grind to a halt. Shops close as card transactions cease to function; food spoils from a lack of refrigeration. People in affected areas are advised to shelter in place. Not that they have much choice. Gas pumps have stopped working, all air traffic is grounded and trains are at a standstill. If the blackout hits populous areas during a period of extreme high or low temperatures, the collapse of climate control systems poses a grave threat to life at a time when critical hospital equipment is compromised.

For the general public, this has come out of nowhere. The loss of communications leaves governments with limited means to broadcast information to anxious populations. During a societal disruption on this scale, it is all but inevitable that crime will surge, with looting widespread. Conspiracy theories proliferate. Rumors swirl of a malicious foreign attack; fringe millenarian groups declare that we have entered the final rapture.

The sun, the source of all life, has temporarily plunged society back into a pre-technological age.

VI

A couple of years ago, on a quiet stretch of the Drouzet River in the southern French Alps, a team of dendrochronologists extracted samples from a group of partially fossilized stumps of ancient Scotch pines that had spent millennia encased in alluvial sediment.

Within a single tree-ring carbon-dated to 14,300 years ago, they discovered an abrupt spike in the presence of the carbon radionuclide C-14. C-14 is cosmogenic, produced when cosmic rays traveling at close to the speed of light collide with atmospheric nitrogen. Absorbed into the trees’ tissue, researchers explained in a subsequent paper in Philosophical Transactions, its presence provided “a mirror image of solar activity.”

The discovery marked the latest addition to a growing catalogue of “Miyake Events,” named after the cosmic ray physicist Fusa Miyake, who in 2012 first made the connection between tree-ring archives and cosmogenic C-14 levels in a group of Japanese cedars. Miyake, who dated the spike in the cedars to 774, posited that it was evidence of a superflare, a solar expulsion many times larger than the Carrington Event in 1859. The Drouzet River spike was twice as large again.

Raimund Muscheler, a professor of geology at Sweden’s University of Lund who has been working to corroborate tree-ring isotope records with ice core archives, told me that there is no way of knowing whether the flares responsible for the anomalous readings were accompanied by CMEs of commensurate magnitude. Nevertheless, Muscheler fears that the Carrington benchmark may be a wholly inadequate metric of the threat. “Considering the enormous events we can observe in the paleodata, it seems probable that present-day contingency plans don’t go nearly far enough,” he said.

At the time of writing, the latest American Space Weather Implementation Plan, updated last year by the Biden administration, has been removed from the U.S. government’s websites.