Ball lightning has been one of the most elusive phenomena in atmospheric science since the first systematic attempts to explain it in the mid-19th century. The phenomenon — a luminous, roughly spherical object that appears during thunderstorms and moves through the air before disappearing, sometimes with an explosion — has been reported by thousands of witnesses over centuries, yet for most of scientific history, many researchers questioned whether it existed at all. The first scientific discussion of ball lightning is generally attributed to French physicist François Arago, who compiled and published witness accounts in 1838. Since then, over a dozen competing hypotheses have been proposed.
The vaporized silicon hypothesis (Abrahamson and Dinniss, 2000) suggests lightning strikes vaporize soil silica into glowing nanoparticles. The plasma toroid model proposes self-sustaining rings of ionized gas. The microwave cavity hypothesis suggests ball lightning is a standing wave of electromagnetic radiation.
The buoyant plasma theory envisions it as a hot, ionized gas bubble. Despite this profusion of theories, no single model explains all observed characteristics of ball lightning, and reproducible laboratory creation of the phenomenon has proved extraordinarily difficult.