What's the deal with ball lightning?

Cecil mentioned ball lightning in this recent column.

I searched the archives for a column on ball lightning, but I couldn’t find one. Can anyone tell me what we know about this phenomenon? Do we understand what causes it? I’ve also read some claims that ball lightning may not even be real. Do some people still think it doesn’t exist?

Okay, from what I can tell, the consensus seems to be No one really knows what ball lightning is.

With that in mind, this may shed some light:

Great Balls of Steam

By Carl Zimmer
David Turner does mostly bread-and-butter chemistry. The University of Bristol researcher is an expert on steam turbines, and he can, among other things, describe the conditions inside nuclear reactor turbines and the possible hazards of an explosion. But recently Turner realized that his work could help solve a more exotic puzzle. The peculiar chemistry of steam could help explain a strange weather phenomenon known as ball lightning.
Over the past 200 years there have been thousands of reports of people seeing these globes of light. The glowing, grapefruit-size spheres seem almost alive, floating down the aisles of passenger planes, gliding down chimneys, dodging objects in their path. When ball lightning passes close to people, they claim not to feel any heat, and yet apparently it can melt a hole in a glass window. It lives for a few seconds or minutes and then either fades away or explodes.
Many explanations have been advanced for ball lightning, including some from the esoteric frontiers of science. Perhaps a nugget of antimatter lies at the heart of ball lightning, some researchers have suggested, or a magnetic monopole—a particle predicted by theoretical physics but never seen. Or perhaps a lightning ball is a natural nuclear fusion reactor whose energy we could somehow harness. But the most popular theory of late has been the tamest one: it holds that ball lightning arises from unusual conditions in the same thunderstorms that create ordinary lightning bolts.
In a thunderstorm, an intense electric field between the positively charged ground and the negatively charged cloud excites air molecules, causing them to lose electrons and become charged ions. A bolt of lightning further energizes the molecules until they become a plasma—a soup of hot, charged molecules and electrons. Perhaps, researchers have suggested, the electric or magnetic field created by a small lump of plasma could trap it in the shape of a ball. Short-lived plasma fireballs have even been created in laboratory experiments, giving the idea some support.
Yet the plasma model has its drawbacks. A hot ball of gas shouldn’t keep close to the ground the way ball lightning does; it should rise like a helium balloon, quickly dissipating its heat until it vanishes. What’s more, the reports that ball lightning has a cool surface make no sense at all if it is a fireball.
But those reports, Turner says—indeed, all the commonly reported traits of ball lightning—fit nicely into the new model he has proposed. In Turner’s model, ball lightning is a reactor, but not a fusion reactor. It is a floating, self-sustaining chemical reactor, in which certain chemical reactions between the plasma and the surrounding air release heat and others absorb it. As a result, instead of simply dissipating into the air, the initial heat of the plasma gets recycled back into the blazing interior of the ball, while the outside of the ball becomes a cool, watery skin.
The ions making up the plasma, Turner says, fly around crazily, moving away from the core of the ball. Certain reactive ions, such as oxygen or hydroxide (OH), combine almost immediately, forming stable compounds like water or ozone and shedding their energy as heat and light. But three kinds of ions are much more stable and don’t combine so quickly. They are positively charged hydrogen and negatively charged nitrites (NO2) and nitrates (NO3). Their chemistry, in Turner’s view, explains most of ball lightning’s properties.
Traveling farther from the hot core into cooler air, these three types of ions start attracting water molecules. (A water molecule has electric poles: the side of the molecule that has the two hydrogens attached is slightly positive, while the other side is negative.) As the water molecules huddle around the ions, they condense to form liquid droplets. They thereby surrender heat. Some of the nitrites—the least stable of the three ions—react with some of the hydrogen to form nitrous acid and release even more heat. These two reactions, condensation and combination, keep the interior of the ball lightning hot.
But the formation of nitrous acid is also what gives the ball its cold skin. As nitrites travel farther from the core, the ones that still haven’t turned into nitrous acid keep gathering more water. From his previous research into steam, Turner knew that swarms of water molecules can have strange effects. If a nitrite is surrounded by six or more water molecules, he calculates, it actually has to absorb energy from its surroundings in order to combine with a hydrogen ion and create nitrous acid; basically it needs the energy to push the water out of its way. Sucking in heat, the nitrites now chill their surroundings instead of heating them. Hence the cool skin.
The skin is watery primarily because of nitrates, the second of the three ions: they are so stable that they rarely react with anything; instead they just keep attracting more and more water molecules. Soaking up water like a sponge, they weigh the ball down, counteracting the lighter-than-air plasma inside and keeping the ball close to the ground. They also keep it round: as more nitrogen and oxygen get incorporated into nitrate-laced drops of water on the outside of the ball, the interior becomes starved for nitrogen and oxygen, which begin to rush in from the outside. The imploding wind forces the ball into a spherical shape, even as it is providing the reactor at the center with fresh raw material.
The third ion, hydrogen, is what causes the ball to wander. Hydrogen ions that don’t combine with nitrites give the skin of the ball a strong positive charge. The intense electric fields in a thunderstorm can thus push the ball around. It keeps on wandering until its heat finally leaks away—although on occasion a ball has been known to rupture and explode more dramatically.
Turner himself has never seen ball lightning, but the tidiness of his model has helped convince him that it exists—something that for a long time was questioned by some researchers, who tended to lump ball lightning with UFOs, ESP, and other popular but dubious phenomena. When Turner first read the accounts of eyewitnesses, he too found a lot of the details hard to believe. But his work has converted him. “As a rule we tend not to believe what we can’t explain,” he says. “I believe a lot of the accounts now because this model explains them.”
Source: Discover Magazine, July 1993.