For most of the 20th century, ball lightning was a subject scientists avoided. Witnesses across centuries described the same thing: a glowing sphere, often the size of a football, drifting slowly through the air for a few seconds before vanishing, sometimes passing through a window or wall. Physicists had no theory that worked, no way to reproduce it in a lab, and no measurements. Then, in July 2012, a team on the Tibetan Plateau recorded one by accident — and in January 2014 they published the first spectrum of natural ball lightning.
What witnesses have reported for centuries
The description is remarkably consistent. A luminous sphere, usually between a few centimetres and several metres across, red, orange, yellow or blue-white, moving horizontally at walking pace for one to a few seconds. Reports describe it passing through closed windows and walls, then either fading or ending with a loud bang.
The earliest known reference dates to 7 June 1195, in the chronicle of the English monk Gervase of Canterbury: “A marvellous sign descended near London,” a dense dark cloud emitting a white substance that grew into a spherical shape under the cloud, from which a fiery globe fell towards the river. In 2022, physicist Brian Tanner and historian Giles Gasper of Durham University identified this entry as probably describing ball lightning, noting how closely the 12th-century description matches modern reports.
Why science ignored it for so long
Ball lightning sat awkwardly between testimony and physics. The accounts were numerous and came from trained observers, including physicists. Roger Clifton Jennison, a British physicist, reported watching one from close range while sheltering in a stone hut during a storm, describing a glowing sphere that entered and hovered inside.
But a sighting cannot be put in an instrument, and no laboratory had produced a stable glowing sphere lasting seconds under atmospheric conditions. Without a measurement, there was nothing to test a theory against. Ball lightning stayed in the category of things that were reported but not explained, and many scientists quietly suspected the reports were mistaken or exaggerated.
The accidental recording on the Tibetan Plateau
In July 2012, Jianyong Cen, Ping Yuan and Simin Xue of Northwest Normal University in Lanzhou, China, were studying ordinary cloud-to-ground lightning on the Tibetan Plateau. They had set up two slitless spectrographs — instruments that spread light into its component wavelengths — at a distance of 0.9 km. A normal lightning strike hit the ground. Ball lightning formed afterwards, and the instruments happened to be pointed at it.
The result, published in Physical Review Letters on 17 January 2014, is the first optical and spectral measurement of natural ball lightning. The team recorded 1.64 seconds of video from formation after the parent lightning strike to optical decay. A high-speed camera at 3,000 frames per second captured the last 0.78 seconds of the event.
According to the paper’s summary, the ball moved horizontally at an average speed of 8.6 m/s, had a diameter of about 5 m, and covered roughly 15 m during the 1.64 seconds it was visible. The temperature was assessed as lower than that of the parent lightning strike, which the paper puts in the range below 15,000 to 30,000 kelvin.
What the spectrum showed
This is the part that matters. A spectrum is chemistry made visible: every element emits light at characteristic wavelengths, so reading a spectrum identifies what is glowing.
The parent lightning strike showed the emission lines of ionised nitrogen, which is what a hot bolt of air produces. The ball lightning was different. Its spectrum showed neutral atomic silicon, calcium, iron, nitrogen and oxygen — and the paper states that radiation from soil elements is present for the entire lifetime of the ball lightning.
Soil elements, glowing, for the whole life of the phenomenon. That points at the ground, not the air.
The 45-year-old hypothesis it supports
One explanation had been sitting in the literature since the 1970s: the vaporised silicon hypothesis. The idea is that lightning striking soil vaporises silica, that the silicon is separated from the oxygen, and that the resulting silicon vapour condenses into a floating aerosol bound by its charge, glowing as the silicon recombines with oxygen.
It was elegant but untested. Then, in 2007, a Brazilian team led by Gerson Silva Paiva published an experiment in Physical Review Letters: by evaporating pure silicon with an electric arc, they produced “luminous balls with lifetime in the order of seconds.”
The 2012 field recording and the 2007 lab experiment fit together. Neither proves the hypothesis on its own — one natural observation is a small sample, and lab conditions are not a thunderstorm — but the soil-element signature is exactly what the hypothesis predicts.
What is still not settled
One recorded event is not a general explanation. The paper reports that light intensity and the oxygen and nitrogen emission oscillated at 100 Hz, possibly caused by the electromagnetic field of a nearby 50 Hz high-voltage power line — a detail that makes this particular occurrence specific to its location.
No accepted theory explains every reported property, including the window-passing accounts, and ball lightning has not been reliably reproduced in the lab under atmospheric conditions. The silicon mechanism remains the explanation best supported by measurement, not a settled fact.
What changed in 2014 is the category. Ball lightning moved from testimony to data. It is now a phenomenon with a measured spectrum, a measured temperature and a measured speed — still unexplained in full, but no longer something science has to take on trust.
Frequently asked questions
Is ball lightning real?
It has now been measured. In July 2012 researchers on the Tibetan Plateau recorded 1.64 seconds of video and a spectrum of what the published paper identifies as natural ball lightning, and the results appeared in Physical Review Letters in January 2014.
Does anyone know what causes it?
No theory explains all reported properties. The best-supported explanation by measurement is vaporised silicon from soil struck by lightning, which matches the soil-element signature in the recorded spectrum and a 2007 laboratory experiment. It is not proven.
Can it pass through walls?
Many witness accounts describe that, but the 2014 measurement did not capture such an event, and no measurement reviewed for this article supports it. The recorded ball moved horizontally in open air.
Can it be created in a lab?
A 2007 experiment produced luminous balls lasting seconds by evaporating silicon with an electric arc, which supports the silicon hypothesis. Reliable reproduction under normal atmospheric conditions has not been reported.
Sources
- Jianyong Cen, Ping Yuan and Simin Xue, “Observation of the Optical and Spectral Characteristics of Ball Lightning,” Physical Review Letters 112, 035001, published 17 January 2014: journals.aps.org
- Philip Ball, “Focus: First Spectrum of Ball Lightning,” Physics 7, 5, 17 January 2014: physics.aps.org
- Gerson Silva Paiva, Antonio Carlos Pavão, Elder de Vasconcelos, Odim Mendes Jr and Eronides Felisberto da Silva Jr, “Production of Ball-Lightning-Like Luminous Balls by Electrical Discharges in Silicon,” Physical Review Letters 98, 048501, 2007: doi.org
- Wikipedia, “Ball lightning,” for the historical accounts and the summary of the 2014 measurements: en.wikipedia.org
- BBC Weather, “Is this England’s earliest report of ball lightning?”, 27 January 2022, on the 1195 Gervase of Canterbury account: bbc.co.uk
The 2014 paper’s full text is behind a paywall; the abstract and the paper’s summary as cited were consulted, along with the Wikipedia summary of the measurements. Ball lightning remains unexplained in full. Check the primary literature before treating any single mechanism as settled.
