Ball Lightning: The Physics Mystery We Still Can't Explain

The Phenomenon Nobody Can Explain

Let me tell you about something that drives physicists a little bit crazy.

Ball lightning. Luminous spheres, typically the size of a grapefruit to a football, that appear during or after thunderstorms. They float through the air, drift through windows, hover in rooms, and then vanish — sometimes silently, sometimes with a bang. People have been reporting them for centuries. There are thousands of accounts. Airline pilots have seen them inside cockpits at altitude. Physicists have seen them. Farmers have seen them. They show up in historical records going back to the Middle Ages.

And we have absolutely no idea what they are.

That’s not an exaggeration for dramatic effect. As of right now, there is no widely accepted physical explanation for ball lightning. There are hypotheses — plenty of them, actually — but none that satisfactorily explains all the observed properties. Ball lightning remains one of the genuinely unsolved problems in physics, sitting uncomfortably in a field where we can predict the mass of the Higgs boson to ten decimal places but can’t explain a glowing ball that floats through somebody’s kitchen.

What People Report

The consistency across thousands of independent accounts is striking. Here’s what shows up again and again.

Ball lightning appears as a luminous sphere, usually 10–30 cm in diameter, though sizes from a few centimetres to several metres have been reported. The colour varies — white, yellow, orange, red, blue, and green have all been described, with yellow-white being most common. The luminosity is moderate: about as bright as a 100-watt light bulb. Not blinding, but clearly visible in daylight.

It lasts much longer than ordinary lightning — typically several seconds, occasionally up to half a minute. It moves slowly (walking pace or slower), often horizontally, and seems to follow structures: walls, fences, wire lines. Some reports describe it entering buildings through windows, chimneys, or even keyholes. It usually disappears silently by fading out, though a significant minority of reports describe it ending with an explosion or loud pop.

It almost always occurs during or immediately after thunderstorms, though a few credible accounts describe it in clear weather near high-voltage equipment.

The fact that these details recur across cultures, centuries, and continents — from 18th-century European naturalists to 21st-century Chinese researchers — makes it very hard to dismiss ball lightning as mass hallucination or misidentification. Something real is happening. The question is what.

The Problem with Plasma

The most intuitive hypothesis is that ball lightning is a self-contained ball of plasma — ionised gas, like a tiny captured lightning bolt. This sounds reasonable until you do the physics.

A plasma at atmospheric pressure cools incredibly fast. The hot ionised gas transfers energy to surrounding neutral air molecules through collisions, and at sea-level air density, the collision rate is enormous. A free-floating plasma sphere in air should cool to ambient temperature and stop glowing within milliseconds — maybe a few tens of milliseconds at most.

Ball lightning lasts seconds. Sometimes tens of seconds. That’s thousands of times longer than any uncontained atmospheric plasma has any business lasting. Something must be supplying energy to sustain it, or some confinement mechanism must be insulating it from the surrounding air. But what?

Several ideas have been proposed. Maybe it’s a magnetically confined plasma — a kind of natural tokamak, with internal currents generating magnetic fields that contain the hot gas. The problem is that the required field geometry (a magnetic bottle or similar) is extremely unlikely to form spontaneously in the chaotic environment of a thunderstorm. And even if it did, the configuration is inherently unstable. Magnetic confinement is hard enough in a billion-dollar fusion reactor. Nature doing it accidentally during a storm strains credibility.

The Silicon Vapour Hypothesis

In 2000, John Abrahamson and James Dinniss at the University of Canterbury proposed something different. Their idea: when lightning strikes soil, the intense heat vaporises silicon dioxide (the main component of most soils). The silica is reduced to pure silicon vapour in the superheated lightning channel. This silicon vapour is ejected into the air as a cloud of nanometre-scale silicon particles.

These silicon nanoparticles are unstable in air. They slowly oxidise — burn, essentially — at the particle surfaces, releasing heat and light. Because the particles are so tiny, they have an enormous surface-area-to-volume ratio, so the oxidation is sustained over seconds rather than being a single flash. The cloud is held together loosely by electrostatic charges acquired during the lightning strike.

The result: a glowing ball of slowly burning silicon nanoparticles. It fades as the particles are consumed. It might end with a pop if the oxidation accelerates and the remaining particles combust all at once.

This hypothesis got a massive boost in 2012. A research team from Northwest Normal University in Lanzhou, China, was filming ordinary lightning with high-speed cameras and spectrographs during a thunderstorm. They happened to capture what appears to be ball lightning — a luminous ball about 5 metres across, lasting about 1.6 seconds, drifting horizontally across a field after a lightning strike hit the ground.

The spectrograph showed emission lines for silicon, iron, and calcium — exactly the elements you’d expect from vaporised soil. This is the only known spectral measurement of a natural ball lightning event, and it strongly supports the silicon vapour hypothesis.

But. (There’s always a but.)

The Lanzhou event was 5 metres across and lasted under 2 seconds. Many ball lightning reports describe objects of 20–30 cm that last 10 seconds or more. The silicon hypothesis struggles to explain the smallest, longest-lasting events. How does a tiny ball of nanoparticles stay coherent for 30 seconds without dispersing in air currents? Why does it sometimes pass through glass windows without breaking them? The hypothesis explains some observations beautifully but leaves others unexplained.

The Electrochemical Hypothesis

Another idea, proposed by various researchers, involves electrochemical reactions. During a lightning strike, strong electric fields might initiate chemical reactions in atmospheric aerosols — tiny droplets of water containing dissolved salts and organic matter. These reactions could produce metastable chemical species that release energy slowly through chemiluminescence (light from chemical reactions rather than heat).

This would explain the modest luminosity, the slow decay, and the relatively cool temperature that some reports suggest (ball lightning has been described passing near flammable materials without igniting them — hard to reconcile with a hot plasma). But it doesn’t explain the sharply defined spherical shape, the coherent motion, or the occasional violent ending.

Microwave Cavity Hypothesis

A more exotic proposal suggests that ball lightning might be a standing electromagnetic wave — essentially a microwave cavity formed in the atmosphere. Peter Kapitsa proposed this in the 1950s, and variations have been explored since.

The idea is that certain atmospheric conditions during a thunderstorm might create a resonant cavity that traps electromagnetic waves at microwave frequencies. The trapped microwaves ionise the air inside the cavity, creating a luminous plasma sphere sustained by the electromagnetic energy.

The physics of microwave cavities is well understood — your microwave oven uses one. But forming a natural cavity in open air, with no metallic walls, that persists for seconds? That requires atmospheric conditions we’ve never observed or modelled convincingly. It’s a clever idea that lacks a plausible mechanism for cavity formation.

Why It’s So Hard to Solve

Ball lightning is the worst kind of scientific problem: rare, unpredictable, short-lived, and impossible to produce on demand.

You can’t point a telescope at it (it’s in the atmosphere, not the sky). You can’t schedule observations (it appears randomly during storms). You can’t reproduce it in a laboratory (not convincingly, anyway). You’re stuck with eyewitness accounts — notoriously unreliable in terms of quantitative details like size, duration, and brightness — and the single lucky spectrographic recording from Lanzhou.

Compare this to other atmospheric phenomena. Ordinary lightning can be photographed thousands of times per storm with high-speed cameras. Sprites and jets in the upper atmosphere were confirmed by systematic camera surveys. Tornadoes can be chased and instrumented. Ball lightning doesn’t cooperate. It shows up when and where it pleases, stays for a few seconds, and leaves no physical trace.

This is why the problem remains open after more than 200 years of serious investigation. It’s not that physicists haven’t tried. It’s that the phenomenon doesn’t give them enough data to work with.

What Would Solve It

The path forward is probably more accidental recordings like the Lanzhou event. As cameras become cheaper and more ubiquitous — dashcams, security cameras, smartphone cameras running during storms — the probability of capturing ball lightning on video with useful spectral or temporal data increases.

What we really need is a close-range recording with a calibrated spectrograph that captures the full spectrum, a high-speed camera that resolves the internal structure, and environmental sensors that measure the local electric field, temperature, and composition. That’s a lot of instrumentation to have pointing at the right place at the right time. But with millions of cameras rolling worldwide during every thunderstorm, it might just be a matter of time.

Until then, ball lightning remains what it has been for centuries: a real phenomenon that physics can’t quite get its hands around. And honestly? In a field where we sometimes pretend we’ve got everything figured out, it’s a useful reminder that nature still has a few tricks we haven’t cracked.