Ball lightning: a real glowing sphere or just an illusion after a flash?

After a lightning flash, witnesses sometimes describe a glowing sphere about the size of an orange floating through the air, lasting a few seconds before fading or exploding. There are accounts of it passing through windows, appearing in aircraft cabins, and leaving a burnt smell. Ball lightning lies on an uncomfortable boundary: too many accounts to ignore, but too rare to measure on demand.

A strong flash of lightning can create an afterimage on the retina, while electric arcs, transformers, and burning objects create light sources easily mistaken for it. Therefore, a collection of accounts may contain multiple phenomena. Some outdoor spectral recordings show material signals from the soil, but the number of reliable samples remains very small.

Competing models

One hypothesis suggests that lightning striking the ground vaporizes silicon, creating a slow-burning aerosol that forms a glowing mass. Other models use microwaves trapped in plasma, chemical reactions, electric vortices, or sustained electric fields. Experiments can produce glowing spheres, but their lifespan, movement, and energy do not always match eyewitness accounts.

The "passing through glass" detail

This is the claim that requires the most caution. Electromagnetic waves can pass through gaps and create plasma on the inside, leading observers to believe an object has passed through; reflections on glass also deceive positioning. No calibrated multi-angle video has yet proven a physical mass passing through an intact pane of glass.

What needs to be recorded to solve it?

Time, distance, light spectrum, sound, electric fields, and images from multiple cameras must share the same clock. A single phone video is often overexposed and loses size information. Until the data is dense enough, ball lightning may be a family of rare phenomena rather than a single object.

What do witnesses usually describe?

Notable reports repeat several features: roughly spherical, bright but not always dazzling, lasting longer than a lightning bolt, and moving slowly regardless of the wind. Colors range from white and yellow to orange; the end can be silent or accompanied by an explosion. But estimated size is uncertain without knowing the distance.

The experience after a strong lightning flash is also subject to retinal afterimages. A bright area can "follow" the direction of gaze, creating the sensation of an object flying in the room. Therefore, cases without physical traces or independent witnesses should be classified differently from multi-angle videos and measured spectra.

Energy is a difficult problem

A sphere glowing for several seconds needs a sustaining source. If it contains high energy, why do many cases not heat nearby objects? If it is just diffuse plasma, why does its shape not dissipate immediately? Chemical aerosol models explain lifespan better but must account for movement; microwave models explain the energy region but require a source and a wave-trapping cavity.

Arc experiments can create burning spheres from silicon, while microwave chambers produce floating plasma. They are useful "relatives," but not necessarily replicas of the outdoor phenomenon. Real lightning has a more diverse energy spectrum, soil materials, and discharge geometry than a lab.

An ideal recording

High-speed cameras need to preserve details without overexposing to white; spectrometers measure emission lines; microphones record the timing of thunder; antennas measure electric fields; radar or dual cameras estimate distance. All must be synchronized to the millisecond and save the original files.

As lightning camera networks expand, the probability of recording ball lightning will increase. It is possible that the final result will be several different mechanisms creating similar images—a complex answer, but one that fits the diversity of accounts.

From mysterious stories to testable records

An engaging story often begins at a moment of awe, but an investigation must go back before and after that moment. Researchers reconstruct timelines, find records closest to the event, keep original image files, and flag details that only come from late recollection. The same detail copied across hundreds of pages may still originate from a single source.

With natural phenomena, "not found" also needs limits. Not finding meteorite fragments after decades in a swamp is different from not finding them in a clean survey area the very next day. Failing to record light with a north-facing camera says nothing about the southern sky. The scope of the search must go hand in hand with the conclusion.

Measuring shape, speed, and energy

Angular size on an image only becomes real size when the distance is known. Velocity also requires distance and time, while energy is usually inferred through its impact on the environment. Mistaking one of the three can turn car headlights into a hypersonic object or a small reaction into a giant energy source.

Calibration is a quiet but decisive step. Cameras need to know lens distortion; temperature sensors need reference points; station clocks need the same standard; chemical samples need blank samples to detect contamination. These details are less dramatic than a photo but are the boundary between clues and evidence.

A good model must accept being disproven

A scientific hypothesis does not just retell what has been seen but predicts what has not yet been measured. If ice pushes rocks, the movement must coincide with shallow water, cold nights, and wind; if convection creates rings, there must be a corresponding thermal structure and flow beneath the center. Incorrect predictions force a model to be modified or discarded.

Different models are sometimes correct in different cases. Nature is under no obligation to use a single mechanism for every glowing sphere, every soil ring, or every similar cloud. Accurate classification may be a greater advancement than finding a single overarching answer.

The role of probability and rarity

A chain of conditions, none of which is too unusual, can still be rare when they must occur at the same time. This explains why a phenomenon leaves traces but has few direct witnesses. It also creates bias: someone waiting for years is prone to recording cases close to what they expect, while equipment failing at the exact moment creates an appealing void.

Automated sensor networks and satellite data reduce reliance on luck. Even so, detection algorithms can produce false positives; every alert still requires original images and checks of familiar sources. More data does not automatically mean more good evidence.

When can we say a mystery has been solved?

A strong solution reproduces the mechanism, fits multiple independent datasets, and does not ignore important details. It also explains why the phenomenon is rare, in the right place, and in the right season. With old traces, we sometimes only reach the level of "the most plausible solution" because the decisive data has been lost.

Solving the general mechanism does not mean knowing every individual event. We may know that Racetrack rocks move due to ice but not know the date of each old track; we may know the color of Blood Falls is due to iron but still have not mapped the subglacial water system. Accurate phrasing preserves the true sense of mystery.

Preserving sites is preserving data

Tire tracks on a playa, footprints on thin ice, soil samples taken from a circle, or lighting equipment brought into a valley can all destroy data. Visitors should stick to permitted routes, not move objects, and not create "evidence" for photos. A fake trace can live online longer than the real scene.

Respecting local communities is also important. They are not just background characters in a story but hold the history of observations, names, and environmental knowledge. Good research returns data and conservation benefits to where the phenomenon exists.

A checklist to self-evaluate new claims

Before a "shocking" video or news item, one can ask six things: is the original file available; can the time and coordinates be verified; are there any matching common light sources, flights, or weather; was the size measured or just estimated; has an independent group reviewed the data; and does the author state clearly what would make them change their mind. Lacking one item does not make a claim false, but the more that are missing, the lower the level of certainty.

It is also necessary to distinguish peer-reviewed papers, preprints, press releases, and online accounts. A single study is a starting point, especially when samples are scarce or the location is unique. Consensus forms when different methods converge, not when a single explanation is shared the most.

The part of the mystery worth keeping

Caution does not require a story to be dry. The moment a forest falls together, a glowing sphere hangs silently over a valley, or an ice finger grows down to the seabed still sends shivers down the spine even when there are equations behind it. A sense of wonder is the reason to ask questions; evidence is the way to ensure the answer does not betray the very thing we want to discover.

In each topic, the unknown is kept separate rather than covered by an absolute conclusion. It could be the composition of an object, the ratio between two mechanisms, an underwater heat source, or the exact trigger conditions. Named gaps with planned measurements are the mysteries most likely to be solved by the next generation.

How a phenomenon becomes a "mystery"

Rare phenomena often happen quickly, in hard-to-reach places, or in weather that degrades equipment performance. Witnesses only see one angle, and memory naturally fills in the gaps after multiple retellings. When the first photo spreads, different cases are easily lumped under the same name. Therefore, the first question is not "what is it" but "are we talking about one or multiple phenomena."

A strong observation needs the exact hour, location, direction of view, focal length, weather, and original data. Photos with background scenery help estimate angles; two cameras spaced apart allow distance calculation; light spectra reveal elements and temperature. Compressed videos online may look beautiful but have often lost the very data needed for verification.

Distinguishing feasible from what actually happened

An experiment producing an object similar to a phenomenon only proves that a mechanism can work. To say that mechanism occurred in nature, it must correctly predict the duration, energy, shape, weather conditions, and post-event traces. If it only explains color but not movement, the model is still incomplete.

Conversely, the lack of a complete model does not make all options equal. Hypotheses that fit physics, make predictions, and withstand measurements carry more weight than unfalsifiable ideas. The term "unidentified" describes the limits of data, not the name of a supernatural cause.

Common identification errors

Distance in the dark is very difficult to estimate. Bright stars, planets, aircraft lights, satellites, car headlights through the atmosphere, and insects close to the lens can take on unusual appearances. Long exposures stretch a point into a line; image stabilization makes the background shake differently from the object; reflections in glass create a "sphere" that moves with the camera.

In landscapes, circles also easily suggest a designing hand. But convection, waves, resource competition, and equal propagation speeds in all directions naturally create rings or repeating networks. Regular geometry is not in itself evidence of intelligence; what needs to be measured is the mechanism that maintains the pattern over time.

Which questions are still worth pursuing?

A good question must lead to a test: which sensors to place, which samples to take, which data can disprove the hypothesis. For rare events, continuous automated systems are often more valuable than short survey expeditions. For landscapes, multi-year satellite images reveal lifecycles and seasonal changes.

Science does not make a story less mysterious. A rock hitching a ride on a thin sheet of ice, a water column creating a kilometer-wide ring, or microbes living in dark brine beneath a glacier are all far stranger than many sensational solutions. The difference is we know which parts have been measured and which parts remain in the dark.

Reading new information with a clear mind

When there is a claim of "completely decoded," check the date, publishing institution, methods, and whether the data has been reviewed by another group. A press headline can turn a "fitting model" into a "solved mystery." Conversely, a storyteller may ignore dozens of identified cases just to keep a few ambiguous ones.

Reliable conclusions usually have boundaries: what they explain, what they do not yet explain, and what conditions would make them wrong. It is those very boundaries that keep curiosity alive without trading away the truth.

Staying safe when approaching phenomena

One should not step onto thin ice, walk into a thunderstorm, cross conservation fences, or take samples in extreme environments just for a photo. Observing from a safe position, recording the time and direction of view is often more valuable than a close-up shot that loses context. Fragile sites also need to be left intact so that natural traces are not distorted by footprints, vehicles, or items left behind by visitors.

References

• The Riddle of Ball Lightning – an overview
• Physical Review Research – ball lightning model