Among nature's most bizarre spectacles are the eerie, floating orbs of electrical fire called ball lightning, which sometimes pop up during thunderstorms. Thousands of eyewitnesses have reported ball lightning over the centuries, but its origins have left scientists stumped. Now, some eye-catching experiments suggest that ordinary lightning bolts trigger ball lightning when they strike soil and cast up lacy webs of silicon embers.
Most ball lightning lasts from a few seconds to a minute before fading away or exploding. The spheres are as small as marbles or as large as beach balls, and they drift in the air or skid along the ground. Despite their blazing appearance, they don't emit much heat and they're usually less bright than a 100-watt light bulb. The best previous guess at what causes the pyrotechnics was hot ionized gases confined by electric fields, but such plasma blobs should flame out in fractions of a second.
A better answer may lie underfoot, say chemical engineers John Abrahamson and James Dinniss of the University of Canterbury in Christchurch, New Zealand, in tomorrow's Nature. Other research had shown that when a lightning bolt smashes into the ground, the intense heat can forge globules of pure silicon metal from the carbon and silicon dioxide in soil. Donning their safety goggles, the duo zapped a thin layer of dirt with a laboratory stand-in for lightning: electric arcs packing a punch of 10,000 to 20,000 volts. The hot silicon vaporized and blasted into the air as chains of particles just a few ten-thousandths of a millimeter wide. Each hot particle then burned slowly as it reacted with oxygen in the air.
The temperatures of the glowing webs of silicon should vary with the soil conditions and the duration of the lightning strikes, Abrahamson says, leading to hot orbs that explode dramatically or cooler ones that just fizzle. The researchers haven't yet seen their gauzy networks of silicon coalesce into miniballs of lightning in the lab, but they're still trying.
The new model "came out of left field, but it's very original," says materials scientist Graham Hubler of the Naval Research Laboratory in Washington, D.C., who saw ball lightning as a teenager during a thunderstorm in upstate New York 40 years ago. The model, he says, explains "90% to 95% of the sightings of ball lightning" that scientists have gathered from reliable witnesses.