On the morning of 1 September 1859, a gentleman astronomer named Richard Carrington was doing what he did every clear day: projecting an image of the Sun onto a screen in his private observatory at Redhill, south of London, and tracing the sunspots by hand. At around eleven o’clock two patches of blinding white light appeared inside a large spot group, brightened for a few minutes, and faded.
He ran to fetch a witness. By the time he came back the light was almost gone. Seventeen hours later the sky over Havana turned red, telegraph operators in Boston were receiving messages on lines with the batteries disconnected, and somewhere in the Rocky Mountains a camp of gold miners got up and started cooking breakfast because they were convinced it was dawn.
What Carrington Actually Saw
Carrington was thirty three, the son of a brewer, and rich enough to build his own observatory and spend his time on a single question: whether the Sun rotates as a solid body. It does not, and he was the one who proved it, by tracking sunspots at different latitudes and showing that the equator turns faster than the higher latitudes.
What he recorded that morning was a white light solar flare, the first ever observed. These are rare. Most flares are visible only in particular wavelengths, and it takes an exceptionally violent one to outshine the surface of the Sun in ordinary light. A second observer, Richard Hodgson, saw the same event independently from Highgate, which is the reason the observation was taken seriously rather than treated as an artefact of one man’s telescope.
Carrington was cautious in print. He noted that the magnetometers at Kew had jumped at the same moment, and then, in a line that historians of science still enjoy quoting, warned against concluding too much from one coincidence. He was right to be careful. He was also, as it turned out, watching cause and effect.
The Seventeen Hour Gap
The Kew instruments twitched within a couple of minutes of the flare, which is light and radiation arriving at light speed. The storm itself took considerably longer, and that delay is the useful part of the story.
A flare on that scale throws a mass of magnetised plasma off the Sun, a coronal mass ejection in modern language, and the plasma travels at a fraction of the speed of light. A typical ejection needs two to four days to cross the ninety three million miles to Earth. The 1859 cloud did it in about seventeen and a half hours, which is close to the fastest transit ever recorded.
The best explanation is that it did not travel through undisturbed space. An earlier ejection had swept the path a few days before, producing the auroras that were already being reported in late August, and the second cloud ran through the corridor the first had cleared. This is why solar physicists worry less about single events than about sequences of them.
A Night Nobody Could Sleep Through
The auroras of 1 and 2 September were seen from Cuba, Hawaii, Jamaica, Colombia, Panama, Queensland and Santiago. The Baltimore American reported that the light was bright enough to read ordinary newsprint by. Ships’ logs in the Pacific record crews woken by a sky the colour of blood.
Aurora at the tropics is not a small curiosity. The lights sit above the magnetic poles because charged particles from the Sun spiral down the field lines and hit the upper atmosphere there. To push that ring down to the latitude of Havana, the storm has to compress and deform the Earth’s magnetic field on a scale that essentially nobody alive has experienced since.
Local astronomical societies collected accounts for months afterwards. Several observers described the display as silent, which mattered because there was a long standing folk claim that the aurora crackles, and several others insisted they had heard it, which is an argument that has never entirely been settled.
The Telegraph Network Behaves Very Strangely
In 1859 the only electrical infrastructure worth the name was the telegraph. Tens of thousands of miles of iron wire strung across North America and Europe, grounded at both ends, and therefore an accidental instrument for measuring geomagnetically induced currents on a continental scale.
It behaved badly for two days.
- Operators reported shocks strong enough to knock them off their chairs, and several offices had sparks jumping from the equipment to the ground
- In Pittsburgh the currents were strong enough to melt and fuse contacts. In several offices the sparking set fire to the paper tape and to the wooden desks it sat on
- Some lines simply refused to carry a signal, drowned in a background current that swamped anything an operator could send
- Most memorably, two operators on the Boston to Portland line disconnected their batteries entirely and kept working for around two hours on the current the storm was putting into the wire, an exchange that survives because one of them wrote it down
The telegraph was barely fifteen years old as a public service. Gauss and Weber had strung their experimental line across Göttingen in 1833, and Gauss had also organised the international network of magnetic observatories whose descendants recorded the 1859 storm. The instrument and the science that explained it grew from the same root.
Nobody Could Explain It for Seventy Years
The idea that the Sun could reach across empty space and disturb a compass needle in London was not accepted. William Thomson, later Lord Kelvin, calculated the energy that would be required and concluded it was absurd, on the reasonable assumption that whatever crossed the gap had to be radiation spreading in all directions.
The objection collapses once you allow that the Sun throws matter, not just light, and that the matter is channelled rather than diffuse. That took decades. Kristian Birkeland fired electron beams at a magnetised globe in a vacuum chamber in the 1900s and produced glowing rings around the poles, which was the right picture and was dismissed as a laboratory trick. Sydney Chapman and Vincenzo Ferraro worked out in the 1930s how a cloud of solar plasma would deform the Earth’s field. The solar wind was not confirmed until spacecraft went out and measured it in the early 1960s.
The Sun kept giving up its secrets to people looking at light rather than at particles. Helium was identified in the solar spectrum in 1868, nine years after Carrington’s flare, by observers who had no idea what the Sun was doing to the Earth or why.
What Would Happen Now
The honest answer is that we have a fair idea of the physics and very little experience of the consequences, because the grid that a storm like 1859 would hit did not exist the last time one arrived.
Smaller storms give the shape of it. In March 1989 a geomagnetic storm collapsed the Hydro Québec network in ninety seconds and left six million people without power for nine hours. The Halloween storms of 2003 forced flights off polar routes, knocked out a Japanese research satellite and caused a blackout in Malmö. On 23 July 2012 an ejection of roughly Carrington scale crossed Earth’s orbit and missed, by about nine days. It was measured in passing by the STEREO A spacecraft, which is how we know what it was.
The vulnerable component is the high voltage transformer. Induced direct current saturates the core, the transformer overheats, and the units are custom built with lead times measured in months. Losing a handful is an inconvenience. Losing several hundred at once is a different kind of problem, and it is the scenario that keeps appearing in insurance modelling and in national risk registers.
Satellites are the other exposure. In February 2022 SpaceX lost thirty eight Starlink satellites to a modest storm that puffed up the upper atmosphere and increased drag on them before they could raise their orbits. That was a minor event by any historical measure.
How Often
Ice cores hold nitrate spikes that may record large storms, and tree rings hold carbon 14 spikes that certainly record something violent. The event of 774 AD, found in Japanese cedar by Fusa Miyake, appears to have been several times larger than 1859. Nobody in 774 noticed anything except an unusually red sky, because there was nothing on Earth for it to break.
Estimates for a Carrington scale event range from once a century to once every five hundred years, and the error bars on that are wide enough to be nearly useless for planning. What has changed is not the Sun. It is that we have wrapped the planet in exactly the kind of long conducting wire that first revealed the problem in 1859, and then made everything else depend on it.
Carrington’s Own Ending
He never became the astronomer he intended to be. His father died, he took over the brewery, sold it, bought a house on Devil’s Dyke in Sussex and built another observatory, and then his life came apart in a sequence of scandals and illness. He died in 1875 at forty nine, and his wife had died a few months earlier in circumstances that were investigated and never satisfactorily explained.
What survives is a sheet of paper with two small white shapes drawn on a sunspot group, sketched by a man who had the presence of mind to note the time, fetch a witness, and then refuse to claim more than he could prove. The event that carries his name is the benchmark against which every space weather forecast is still measured.
No products in the cart. 