Lights of the North! As in eons ago,
Not in vain from your home do ye over us glow!
William Ross Wallace (1819–1881)
Jan 25, 880 AD, was a remarkable night. The Arabian historian Ibn Abi Zar wrote about it more than 400 years later, from the ancient city of Fez, northeast of the Atlas mountains:
It was a great red splendour in the sky, from the beginning of the night to the end. An equal thing was never known before. It happened the night of the Saturday 19 of safar of the cited year [25 January 880].
African aurora was very rare. Ibn Hayyan described a later event seen in Iberia, where it is a little more common:
Toward the end of the night of the Thursday [27 April 942] a reddish colour as fire in the sky, from which a lot of rays that were reflected on the branches of the palm trees and on the top of the palaces emerged, appeared in the horizon. People thought that it was the daybreak light when Sun rises until the red colour became lost with the arrival of the morning and disappeared with the clarity of the day. It appeared from the Occident.
We have lost the magnificence of the night sky. From cities we can hardly see the brightest stars, and even in dark places 100 kilometers from any city, their lights reflected of the clouds above them can be seen, brightening the sky. Real dark skies have all but disappeared. Aurora would have been a powerful experience in the years before artificial light, the dancing, eery lights which soundlessly move across the sky and illuminating the landscape in ghostly red and green.
The oldest description of aurora seems to predate writing. It may be depicted in 30,000-year-old Cro Magnon cave paintings, some of these have strange swirling patterns called macaronis. The best example is in the ceiling of the cave of Rouffignac in France. It looks like aurora – but whether that is what it is is disputed. With artists, you can never really tell.
Much later came the written descriptions. The oldest generally accepted as an ancient description of aurora is from the book of Ezekiel, dated to 593 BC: …a stormy wind blew from the north, a great cloud with light around it, a fire from which flashes of lightning darted […] the fire flashed light, and lightning streaked from the fire […] a light all around like a bow in the clouds on rainy days…
The Greek philosopher Xenophanes mentioned “moving accumulations of burning clouds“. He lived at approximately the same time as Ezekiel, and in view of the rarity of bright aurorae this far south, it may well have been the same event. There are some 50 accepted descriptions of aurora from before the (non-existing) year zero, including one from 567 BC listed in Babylonian tablets. The event of 112 BC was remembered by Pliny the Elder: We sometimes see, them which there is no presage of woe more calamitous to the human race, a flame in the sky, which seems to descend to the earth on showers of blood; as happened in the third year of the 107th Olympiad, when Philip was endeavouring to subjugate Greece. Other particularly bright auroras happened in the Mediterranean in 349 BC and around 460 BC. Not everyone was lucky. Much later, Edmund Halley was “dying to see the aurora and expecting to die without seeing it.” He had to wait for many decades before he finally saw the great aurora of 17 March, 1716, which blazed the sky from America to Russia.
The spectacle of the aurora borealis can be mesmerizing, even hypnotic, draping the land in an unearthly colour. The light forms arcs, bands, rays, patches or veils; it is a dance of red and green, where curtains move, jump and wave across the sky, pulsing in brightness.
If you want to see aurora, pick a week around new moon, preferably autumn or winter when the nights are darkest, and travel to somewhere 15-20 degrees from the magnetic pole. The best auroras tend to be around midnight, with the sun is on the other side of the magnetic pole. Do give yourself time as not every night is clear! North Alaska is great, and Iceland has some good spectacles, with the added attraction of the occasional volcanic eruption to add fire to the glory. And if you would like to see the aurora australis, the southern tip of New Zealand may be your best bet. Avoid Mexico: it is almost never seen there. Similarly, Spain would not be an obvious choice as a visible aurora happens perhaps once a decade. Seeing a bright aurora in Morocco is unheard of. The African event described at the start of the post is most unusual indeed.
Aurora happens high in the atmosphere. Highly energetic particles from the Sun get captured in the Earth’s magnetic field, spiral around the field lines and enter the atmosphere where the field line bend towards the magnetic pole of the Earth. The high energy particles collide with the whispers of air 100 kilometer above the ground, and cause them to give light: oxygen glows green or red, nitrogen shines blue. Each colour occurs at its own height. Red oxygen (a wavelength of 630 nm) is seen high up. Green (557 nm) is lower, down to 100 km height. If red and green occur together, the eye picks up a pink glow. Blue light from nitrogen is the lowest layer. The layers terminate suddenly at a specific altitude: this sharp lower edge is clearly visible especially from space.
The lights form a fiery circle around the pole. There is surprisingly little aurora directly above the magnetic pole itself: they encircle it 10 to 20 degrees away. A place like Churchill, Canada has aurora every clear night. To the south it becomes rarer. It takes more energetic particles for the aurora to reach say New York. Further south, a bright aurora may become a once-in-a-lifetime event. Far from the pole the auroras that do occur are normally red, and they lack the crispness of the polar lights. Even so, they are impressive. The brightest I have seen it myself was one of those rare southern reaches, and it veiled half the sky in bright red as if the sky was reflecting a distant fire – which at first I thought it was.
Were you ever out in the Great Alone, when the moon was awful clear,
And the icy mountains hemmed you in with a silence you most could hear;
With only the howl of a timber wolf, and you camped there in the cold,
A half-dead thing in a stark, dead world, clean mad for the muck called gold;
While high overhead, green, yellow and red, the North Lights swept in bars?
— Then you’ve a hunch what the music meant… hunger and night and the stars.
(The Shooting of Dan McGrew)
Blame the Sun
The particles that cause the aurora come from the Sun. Local patches on the surface of the Sun get a strong magnetic field. Two different field lines can pass close to each other, and sometimes they connect. Such a magnetic reconnection is like breaking a taut guitar string. The tension is released and the two sides of the string whip away at high speed. In the Sun, the sudden release of magnetic stress accelerates the local gas, and it flies off in a flare. High energy, charged particles begin to stream away. A large flare can produce X-rays, gamma rays, and high energy protons.
Many miss us, but sometimes the flare heads directly for Earth. It takes the particles a day or so to reach us. Flares come in all kinds of sizes. The stronger the flare, the brighter and further south the aurora. Some warrant the name ‘solar storm’. And if storms are possible, how about hurricanes? A solar hurricane could be scary indeed!
Solar flares are assigned a size in a way that is a bit like the Richter scale for earthquakes. They are ordered alphabetically, as A, B, C, M and X, where each class is ten times more energetic than the previous one. A to C are too small to affect us. M-class can cause radio outages. X-class is more like it. Each class is subdivided by assigning a number from 1 to 9. The problem with this scale quickly became apparent: flares were seen that were more powerful than X9. So the numbering was extended for X-class flares only. The strongest one on record was the Halloween solar storm of 2003: it was measured at X28, 500 times stronger than an X1. But this was an underestimate: the detectors saturated at this level. Modeling suggests it may have as strong as X35. It mostly missed Earth, luckily. And there is no reason to assume that this is the strongest event that could ever happen. In fact, a bigger one had been seen almost 150 years earlier.
The Carrington event
The solar cycle number 10 had started in 1855. It was about average in strength, as measured in sun spot numbers. Of the cycles that have happened since, 7 were weaker than cycle 10, and 7 were stronger.
During August 1859, a huge sun spot came into view on the Sun. It was at mid northern latitudes on the Sun, and dominated the solar disk. There was a smaller spot in the southern solar hemisphere, but otherwise the surface had been clear. The new spot spread out over a 6th of the width of the Sun. At the start of September, it erupted.
Richard Hodgson in Highgate, London, a retired publisher and amateur scientist, saw the flare and later wrote
While observing a group of solar spots on the 1st September, I was suddenly surprised at the appearance of a very brilliant star of light, much brighter than the sun’s surface. [..] most dazzling to the protected eye, illuminating the upper edges of the adjacent spots and streaks, not unlike in effect the edging of the clouds at sunset.
The event was described in more detail by Richard Carrington, who later had the event named after him. He worked at his observatory in Reigate, south of London, and wrote The image of the sun’s disk was, as usual with me, projected on to a plate of glass coated with distemper of a pale straw colour, and at a distance and under a power which presented a picture of about 11 inches diameter. I had secured diagrams of all the groups and detached spots, and was engaged at the time in counting from a chronometer and recording the contacts of the spots with the cross-wires used in the observation, when within the area of the great north group (the size of which had previously excited general remark), two patches of intensely bright and white light broke out, in the positions indicated in the appended diagram by the letters A and B, and of the forms of the spaces left white. My first impression was that by some chance a ray of light had penetrated a hole in the screen attached to the object-glass, for the brilliancy was fully equal to that of direct sun-light.
Seeing the outburst to be very rapidly on the increase, and being somewhat flurried by the surprise, I hastily ran to call some one to witness the exhibition with me, and on returning within 60 seconds, was mortified find that it was already much changed and enfeebled. Very shortly afterwards the last trace was gone, and although I maintained a strict watch for nearly an hour, no recurrence took place. The last, traces were at C and D, the patches having travelled considerably from their first position and vanishing as two rapidly fading dots of white light.
What his paper didn’t say is that he didn’t look for a repeat of the event. According to his note book, he was ‘too busy cutting trees’!
What Hodgson and Carrington had seen was a so-called white-light flare, a relatively rare event that can happen a few times per year around solar maximum. This was the very first ever recorded. But this was more than just such a flare, and what they saw was only the beginning. Almost at the same time, the magnetic instruments at Kew recorded a disturbance. The magnetometer quickly returned to its previous position. But 17.6 hours later, the disturbance came back and this time the instrument went off the scale. The geomagnetic hurricane had arrived.
The disturbance reached a level of 110nT in London. The observers did not fully appreciate how unusual this was. Nothing stronger was seen at this latitude until the super flare of 28 February 1942 when the force reached 112nT. The Halloween flare of 2003 reached 114nT. But in both these cases, this was the maximum seen over a range of stations across the globe: in a Sun storm, the measured strength of the magnetic disturbance can vary dramatically from location to location. The 1859 data is based on a single location only. It is very likely if more locations had been available, much higher strengths would probably have been measured.
Storms of light
The Carrington event refers to the solar storm of Sept 1 to 5, 1859. But there were in fact two storms in short succession. The first one started about 2300 UT on Sunday, August 28, intense but short-lived. The second one began on September 2 about 0500 UT, and lasted for three days. The white flare seen from London was the start of the second storm, but both had come from the same sun spot. And both storms were accompanied by worldwide auroras which reached as far as Honolulu, and reached to 18 degrees from the magnetic equator. This is not entirely unique: in later years the aurora on 4 February 1872 was even more widespread. But nothing like it had been seen in living memory.
The first aurora began on Sunday, Aug 28; it was seen from the Mediterranean to the Gulf of Mexico. The aurora re-appeared late on September 1 (September 2 in Europe), brighter and even more widespread. There are hundreds of reports and descriptions, some describing in detail how the spectacle changed from minute to minute. Picking just a few example, the Indianapolis Journal wrote: ‘It was far more brilliant than the one of Sunday night, and it certainly was much more luminous, though less marked by the darting rays and wonderful pulsations that made the first so splendid.’ John Cambell, also in Indianapolis wrote: At 3h 45m A.M. Magnificent corona in the zenith. Central portion spiral, red and white, changing instantly to a beautiful rose color, with spiral streams shooting forth into all parts of the heavens; the most brilliant streams flowing east and west.
Further south, the aurora was celebrated differently. “The influence of the Aurora Borealis has been felt in the Garden District. We see in the police reports, this morning, that several denizens of that delightful spot have been found drunk – many under a strange delusion, having taken the gutter for their own comfortable beds’’ (The New Orleans Daily Picayune, Wednesday, September 7, 1859).
The Physio-Meteorological Observatory at Havana approached it in a more scientifically detached manner: White rays with red and blue were then seen towards the west, which dilated longitudinally, oscillated laterally, were extinguished and resumed their brilliancy again by turns. […] At 2h the Aurora had attained its highest magnificence. The heavens then appeared stained with blood and in a state of complete conflagration. Havana also had aurora on Aug 28, and the Observatory commented how rare such a double event was: between 1784 to 1859, aurora had been seen from Havana a total of six times; two of these were the double storm of 1859.
A little further south, Kingston, Jamaica, reported: It appeared as if there was a colossal fire on earth which reflected its flames on the heavens. The whole island was illuminated. It looked as if Cuba was on fire, and many believe that a portion of this island had been destroyed by a conflagration. Other persons were of opinion that the light was that of an aurora, but the aurora has never before been seen in this latitude.
London and Brighton both reported a vividly white aurora. In Prague and Rome too the aurora was predominantly white. This is notable, given how colourful the light was in America. The aurora at Sept 1/2 also reached Hawaii. In Australia, the aurora australis was seen as far north as Brisbane.
The aurora was impressive, memorable, and perhaps frightening to those unfamiliar with it. But the sky light could not do anyone harm. There were other effects from the Sun storm which had a more direct impact, caused by the technological revolution that was going on: the rapid adoption of telegraphy.
The idea for the telegraph originated in the 1830’s, and the first telegraph was send in 1844. It was based on the discovery by Oersted that an electric current could change the direction of a magnetic needle, and on the development of the electric battery by Volta. The system was based on needles pointing at a display of letters and numbers. A push of the button closed the circuit with the battery, an electric current would move the needle, and words could be spelled out by varying the current. A person called Morse had been involved (obviously). By 1859, telegraph wires spread between al major cities. The first transcontinental wire came into service in 1861.
The telegraph used long wires, and these could pick up induction currents. During the 1859 storms, this is exactly what happened: the rapidly changing magnetic field put too much current on the lines, and the needles were moving too much to read any message. Many connections became impossible. Or worse: in Gothenburg, the telegraphist got an electric shock when disconnecting the earth. He came off lightly compared to some others. In Norway, the wires discharged sparks in the telegraph office, setting pieces of paper on fire. A telegraph operator at Baltimore even reported that The intensity of the spark at the instant of breaking the circuit was such as to set on fire the wood work of the switch board. Bright sparks were also noticed in Bordeaux and Toulouse.
Some telegraph lines did better than others. As one report states: On Friday, September 2nd, 1859, upon commencing business at 8 o’clock A.M. it was found that all the wires running out of the office were so strongly affected by the auroral current as to prevent any business being done, except with great difficulty. At this juncture it was suggested that the batteries should be cut off, and the wires simply connected with the earth. The Boston operator accordingly asked the Portland operator to cut off his battery and try to work with the auroral current alone. The Portland opera- tor replied, ‘‘I have done so. Will you do the same?’’ Boston operator answered, ‘‘I have cut my battery off and connected the line with the earth. We are working with the current from the Aurora Borealis alone. How do you receive my writing?’’ ‘‘Very well indeed,’’ rejoined the Port- land operator; ‘‘much better than with the batteries on. There is much less variation in the current, and the magnets work steadier. Suppose we continue to work, so until the Aurora subsides?’’ ‘‘Agreed,’’ said the Boston operator.
This is still the only reported case of Solar storms being used for renewable energy!
The cause of aurora
Aurora were still a mystery in 1859, and after the storm, a variety of explanations appeared in print. The Scientific American, not entirely justifying its name, proposed that it was caused by falling matter from erupting volcanoes. (The editors should have read VC.) The San Francisco Herald perhaps came closer, describing it as Nebulous matter . . . known to exist in planetary spaces’’ similar to meteors falling into the atmosphere. The New York Herald went for the least likely explanation and attributed it to to reflected light from icebergs or polar lights.
But the truth became clear once the full story appeared. The magnetic disturbance and the induced current in the telegraph lines showed that it had to be electric currents, high in the atmosphere. The alignment of the streamers with the earth magnetic field showed that the current followed these field lines., entering the atmosphere where those lines dipped down towards the poles. It clearly came from space. And the observations of Carrington showed that the cause was in the Sun. Nowadays, it is called space weather.
The uniqueness of the Carrington event
So how strong and how unique was the Carrington event? There have been many Sun storms since, a few of which caused auroras and electromagnetic disturbances. Several auroras have come even closer to the magnetic equator than in 1859. Magnetic disturbances have occurred, and although it seems that the 1859 event has not been paralleled, some came close. There have also been larger Sun spots, although not many.
Storms can be ranked on different aspects: the geomagnetic disturbance, the aurora strength, sun spot size, sudden ionospheric change. On each aspect, there have been solar storms as strong as the 1859 event since. But 1859 is the only one that was among the top few on every aspect. Putting the numbers together, it seems that the energy in the flare was twice as large as any flare since. It is estimated as between X40 and X50. An event like this is estimated to have happened perhaps only twice per millennium. However, the real strength of the event is still being discussed. It is also important to note that solar flares cannot really be described by a single number: it depends on energy, hardness of the radiation, proton flux, and direction of the embedded magnetic field. Flares with the same classification can be quite different in their impact.
Why was this event so extreme? The solar cycle of the time was not particularly strong, but it now seems that strong storms can happen during any cycle. But in 1859 there were two solar storms in short succession. That is unusual, and it may be part of the reason for the strength of the second storm. If a fast mass ejection meets a slower one and sweeps it up, the interaction between the two can amplify the resulting shock. The first storm was strong but not too exceptional. The second one went off the scale.
Our modern world
Our technology has developed quite a bit since 1859. In some ways we are better placed. Telegraph cables have become optical fibres, and these are safe from solar interference. But new risk factors have appeared. We depend on satellites and on power lines which both are open to the effects of solar storms. This became apparent during the solar storm of 1989, when an X10 (or so) flare caused a storm that affected the power lines of Hydro-Quebec. Quebec is at high magnetic latitude, and the very long power lines were perfectly placed to pick up the induction current. Large currents began to flow through the grounding points, causing the circuit breakers to flip. Within 90 seconds, the lines linking the hydro-electric dams in the north to the cities of the south were cut. It left much of Quebec without power for nine hours: restarting a complex power grid is not easy. This was a large but not an extreme storm. What could a repeat of the Carrington really do? How resilient are we?
The first effects would be on satellites. These are right in the firing line, with little protection. Communications would be disrupted by surface charge building up on the satellites. In a bad storm, such charge could leak in and damage the electronics. A study in 2005 estimated that we could loose 10% of the satellites, ranging from low-earth orbit to geostationary ones. Many of the GPS satellites could fail – and repair crews are not readily available. The study estimated the total cost to replace the satellites at 50 billion dollar, while the companies providing commercial satellites could face losses of the order of 30 billion dollar. There would also be heating of the outer atmosphere, and over several years satellites in low-earth orbits would be slowly pulled down by the expanding air. Even the International Space Station could be at risk, and would require an altitude boost.
Radio communications are badly affected, especially those using high frequencies. Mobile phone connections becomes difficult. GPS signals from the satellites give high timing errors: GPS positioning becomes unreliable and intermittent. Compass-based systems would fare better but could still mis-point by several degrees. Inertial systems would be unaffected. Airplanes at high latitudes would also suffer from the communication problems (they would also be subject to significant radiation doses). During the 2003 halloween storm, planes had to be re-routed further south. During that event, the FAA could not provide GPS navigational guidance for approximately 30 hours.
How about our power lines? Here is the biggest risk. Quebec was actually a fairy mild case, where the circuit breakers prevented real damage. Since that time, capacitors have been installed on the lines to prevent a re-occurence. But a much larger Carrington event could still overload the grid, causing severe voltage regulation problems and, potentially, widespread power outages. Worse, the currents can cause internal heating in extra-high-voltage transformers, causing failure or permanent damage. The 1989 storm damaged two such transformers in the UK. The manufacture of transformers can take many months, and if many are damaged, replacement could take the best part of a year. Modeling of an extreme storm suggested that 130 million people in the US could lose power, and that 350 transformers could be badly damaged. The worst affected regions would be Scandinavia, the UK, the Pacific Northwest, and the Northeast region including Quebec, New York, and New England, with power outages of weeks to months and an economic cost of 1 trillion dollars. This is very unlikely scenario, and it assumes that there has been no real preparation. But that is what risk analysis is for.
The dark side of the Sun
The storm of 1989 became a wake-up call, after decades of a quiet sun. It was minor compared to the Carrington event? But we don’t know enough about it to really know what happened. And we still don’t know what is the worse the Sun can throw at us. The beautiful aurora has a dark side.
Albert Zijlstra, October 2018