Imagine yourself on a clear summer night, far from the city, admiring stars you rarely see in urban areas. Then suddenly, from the corner of your eye, you see a flash: brief, red, large, and high in the sky, like an aurora turned on and off for an instant. Was it just your imagination?

Puzzled observers had reported such sightings since at least 1886, but without producing a photograph or videotape. What should science do with such reports, given our proclivity to conjure up little green (or gray) “men,” report Venus to 9-1-1 as a UFO, and read more into observations of natural phenomena than facts warrant? Science needs to be patient, to wait for evidence. As the late Carl Sagan often said, extraordinary claims require extraordinary evidence. Science deals with facts, not forcing observations to conform to fanciful speculation. Mystics, poets, and the ancients were adept at speculation. And although they provided entertaining and amazing tales...that is all they were.

On the night of 6 July 1989 a “happy accident” occurred. The late auroral physics expert, Prof. John Winckler, and his grad students were testing a low-light camera (like the night scopes on TV news) soon to fly on a research rocket. Pointing out a U. of Minnesota observatory window (in central MN), the camera should have recorded northern stars and some fireflies. But on test tape playback came a total surprise: in two video fields appeared giant pillars of light towering high above nearby trees. Assuming these amazing events were associated with thunderstorms in northern MN 200 miles away, geometry showed they extended tens of miles high, far above the storm tops.

Meteorologists at Kennedy Space Center have one key task-keeping spacecraft and lightning apart. A tragedy was averted when lightning struck Apollo 12 just after liftoff, but the vehicle survived to make the second successful moon landing. Still, once was enough. The thought of lightning reaching up into space, threatening the space shuttle, was unnerving. All the reports of strange lights in the night sky above thunderstorms were suddenly reviewed-and the hunt was on. But what to call these “creatures”? Rocket lightning? Cloud-to-space lightning? Upward lightning? In science, terminology is critical. One can’t adopt a name that presumes more about the physics of a phenomenon than is known, and no one knew whether this was lightning, which way it actually traveled, or whether it was connected to the clouds or reached space. The deliberately fanciful name “sprite” was chosen. Fleeting spirits populate more than one of Shakespeare’s plays. But these sprites were not myth-they were real.

One night in July, 1993, playing a hunch, researchers at Colorado’s Yucca Ridge Field Station trained a low-light camera above a distant thunderstorm complex over Kansas. On the TV monitors appeared hundreds of sprites, dancing high above the cloud tops for hours. Soon sensitive color cameras in aircraft found red sprites, with tinges of blue in the downward extending tendrils! And sprites were huge, filling thousands of cubic miles of the thin atmosphere between 20 and 60 miles above the ground. They were the same brightness, and often similar in color to the aurora, but were fleeting, lasting only a hundredth of a second or less. So, little wonder people who witnessed sprites couldn’t be sure if they’d seen something. Video cameras proved they were there. But what caused them?

Atmospheric physicists, startled to find something so unexpected, began proposing, and disposing of, theories. Soon it became clear that sprites occurred only above rare positive cloud-to-ground lightning flashes (usually 10% of a storm’s total). A theory proposed around 1925 by Nobel Prize winning physicist C.T.R. Wilson suggested that when massive amounts of electrical charge were lowered to ground, the event would briefly upset the electrical balance of the middle atmosphere, causing a “spark.” Indeed observations showed this spark occurred at around 45 miles altitude, followed by a burst of upward and downward propagating electrical streamers. But for Wilson’s theory to be correct, the lightning involved had to be massive, larger than textbooks claimed. So another hunt started. And during summer 2000, an NSF sponsored program, the Severe Thunderstorm Electrification and Precipitation Experiment (STEPS), found such powerful lightning bolts dwell in the huge thunderstorms roaming the central U.S. on summer nights.

It’s possible to see sprites with the naked eye. If you’re in a rural area, and let your eyes become dark adapted, look above the area where large thunderstorms are occurring 100 to 300 miles away. You just might glimpse the brief life of a sprite-one of lightning’s children dancing though the thin air of the upper atmosphere near the edge of space.