Walter A. Lyons

Tornado chasing, quite the vogue these days, wears out one's car, burns up gasoline, and taxes the nerves of even the hardiest of twister trackers. But there are other mysterious, thunderstorm-related phenomena that can be viewed above distant thunderheads in the peace and tranquillity of a summer night while sitting in a rocking chair on your back porch. Less than ten years ago, nobody had heard of sprites, blue jets and elves, in fact, the terms hadn't even entered the meteorological vocabulary. Recent research, however, has confirmed over a century of frequent but generally ignored observations of an entire menagerie of strange, luminous, lightning-related flashes dancing high above thunderstorm tops. And you can, if you know how, observe some of these with the naked eye. Here's the story of how sprites were discovered and how you might be able to spot some on your own.

In decades past, the textbooks said weather stopped at the tropopause, the layer separating the turbulent troposphere from the quiescent stratosphere above. Not so. As long ago as 1886, people were publishing reports in which they struggled to describe momentary discharges of "lightning," for lack of a better term, that they had observed high above storm clouds.
A 1903 paper discussed "rocket lightning ... a luminous tail ... shooting straight up ... rather faster than a rocket..."
From Africa in 1937 came reports of "long and weak streamers of reddish hue...some 50 kilometers high..."
English scientific papers from the 1950s detailed what seemed to be flames appearing to shoot above thunderstorms near the horizon. Recently Earle Williams of the Massachusetts Institute of Technology took a closer look at a nighttime photograph of an Australian thunderstorm in his possession since the late-1980s. A lightning channel extending into the clear air above the storm top in turn appears to have "blue flame" fanning upward, perhaps even into the stratosphere. An airline passenger over Texas reported he saw over twenty faint plumes of light "extending from the top of the thunderhead above the pool of light from the lightning discharge..." Many of these observations were collected in books of meteorological esoterica alongside reports of turtles encased in hailstones, half meter wide snow flakes and showers of toads, fish, seeds and whatnot. Interesting, but what could be the relevance of such events variously termed "cloud-to-stratosphere lightning", "upward lightning" or "cloud-to-space" lightning?

One clue was offered in 1956 by Nobel Prize winning physicist C.T.R. Wilson who himself once saw "diffuse fan-shaped flashes of greenish color extending upward into the clear sky...". He speculated than such discharges between cloud tops and the ionosphere might be a normal accompaniment of lightning discharges to earth, but ones which are visible only under very special conditions. These might represent a heretofore unknown component of the global electrical circuit. Then, in 1989, by pure chance, University of Minnesota scientists John Winckler, Robert Franz and Robert Nemzek, while testing a low-light video camera for a high altitude scientific rocket shot, captured two fields of video showing giant twin pillars of light extending upward more than 30 kilometers above a distant thunderstorm. With the hard evidence now in hand, the race was on in mainstream scientific circles to find out what was going on way up there.

NASA scientists from the Marshall Space Flight Center soon found over a dozen examples of strange 'upward lightning' bolts in the routine video recordings of the horizon taken by the Space Shuttle's low-light payload bay camera. After reviewing the many dozens of anecdotal reports of these events, the author played a hunch as to how to "capture" these creatures on tape. The evidence suggested that these flashes occurred above the anvils of exceptionally large thunderstorm systems. At our rural laboratory, the Yucca Ridge Field Station, situated near Fort Collins on the High Plains east of the Colorado Front Range, we waited for the right conditions, big thunderstorm clusters that were far enough away so we could easily view the stratosphere and mesosphere region above their anvil canopies.

On the night of 7 July 1993, giant thunderstorms were boiling over Kansas and Nebraska, one of the series of mesoscale convective complexes that drowned the midwest in record setting flooding rains far to the east. We aimed a low-light video camera, cousin to the night scopes used by the military, to the east and began taping. For the first two hours, not much happened, except for the almost continuous flashing of lightning within the distant clouds. Then, suddenly, a bright flash occurred high above the storm tops (appearing white on the monochrome television system screen). Over the next several hours, over 240 high altitude flashes were captured. The very next night, University of Alaska scientists obtained similar images from a high-flying NASA aircraft over Iowa. Since then, thousands of flashes have been recorded on low-light video from the ground and from air. While easily visible on the television monitor on that very first night, I was unable to see anything with the naked eye while staring above the distant clouds. But several nights later, when the show started again, with some patience, and dark adapted eyes, there they were, bright reddish curtains dancing a gossamer ballet high above the storm clouds.

In 1994, while flying an extremely sensitive color camera normally used for auroral photography, University of Alaska scientists confirmed that the flashes indeed have a generally reddish color, but which often fades to purple or blue in the downward extending tendrils.

The flashes were named sprites after the creatures in Shakespeare's "The Tempest," in part because of their transient, ephemeral nature. But unlike the bard's characters, these sprites are very real indeed.

And the sprites were soon found to have company. At least two other distinct phenomena have been discovered to date. While flying near an especially active hailstorm in Arkansas, the University of Alaska team were startled to see blue beams of light shooting upward directly out of cloud tops at speeds over 100 kilometers a second. They reached heights of 40 or 50 kilometers (two or three times the cloud heights) before fading away. Around 50 of these "blue jets" were seen that night. But the blue jet seems to be very rare. In four years of ground monitoring, only one blue jet has been captured on tape. The discovery of the blue jet does explain many of the strange reports over the last century that did not seem to jibe with the characteristics of red sprites. It appears that the blue jet can be seen with the naked eye, if you are lucky enough to be around on a dark night when that rare storm produces them. Based only on a few sparse reports, intense hailstorms may be the best candidates for uncorking blue jets.

After the red sprites and blue jets came the elves. In 1995, scientists from the University of Tohoku (Japan) and Stanford University, working with other science teams at the Yucca Ridge Field Station, confirmed the presence of elves (emissions of light and VLF perturbations from EMP sources). These were actually predicted by theorists before they were ever caught on tape. The elves appear as giant expanding disks of light between 70 and 100 kilometers altitude. They are caused by the passage through the ionosphere of the electromagnetic pulse (EMP), the intense radio waves emitted from powerful lightning flashes. Though huge, sometimes expanding to more than 400 kilometers in diameter, the elves are so transient (less than one-thousandth of a second), it is unlikely the human eye could detect them.

But the red sprites can be seen by the naked eye. They are by far the most common of these mesospheric creatures, and we know where they "live". So a plan for some serious "sprite hunting" is relatively easy to develop.

Sprites come in a bewildering variety of sizes and shapes. They can look like giant red blobs, picket fences, upward branching carrots, or tentacled octopi. The sprite luminosity can extend upward as high as 95 km, with the brightest part usually located between 50 and 75 km altitude. The often bluish tendrils can sometimes extend downward below 30 km, close to, but probably not touching, the cloud tops. Sprites can occur singly or in clusters which sometimes fan out for over 150 kilometers. Sprites appear to be uniquely associated with cloud-to-ground (CG) flashes of positive polarity, usually those having peak currents larger than most of the other positive CG events in the storm. By comparison to the pencil-thin channel of their parent positive CG flash, the volume illuminated by a large sprite can reach hundreds or even thousands of cubic kilometers.

So now that we know about all the 'electrical action' above the clouds, a natural question is - can we see and photograph them? The answer is yes... and no. Taking standard photographs will not work unless you have a film with an ISO of 2 million (don't bother to ask for it in the photo store). To take images, you need a low-light video system. This is well within the reach of well-funded scientific investigators, but not your average storm watcher. Yet under ideal viewing conditions you can indeed see sprites with the naked eye. Here's how.


Sprites occur high above very large thunderstorm systems. Since they are so high up, it is much easier to see them if they are at least 50 to 100 kilometers away. Sprites have been visually detected as far as 400 km out, but those occurrences are rare. Not every thunderstorm produces sprites, even if it has vigorous lightning. To improve your chances, check out the radar echoes on your local TV or the Weather Channel. Look for thunderstorm clusters that combined are at least 150 kilometers on a side. If you have access to data from the National Lighting Detection Network, look above the part of big storms where the positive CGs are occurring. This is most often in the large stratiform or anvil region of storm systems.

To actually view the sprites, find a location with a good view of the horizon. The further away from the city lights, the better. It is best to choose a dark night with no moonlight. In the eastern and southern United States, unfortunately, haze and air pollution can sometimes blot out the sprites. Let your eyes adapt to the dark for at least ten minutes. Look in the direction of the big storms. If you can see the illuminated tops of the distant storms, shield your eyes (a piece of cardboard can help) from the lightning flashing within the clouds. Concentrate your gaze at an altitude about four to five times the height of the cloud top, not the storm itself. Then be patient. In the more active storms, sprites can occur every one or two minutes, but every five to ten minutes is more common. They only last from one one-hundredth to one-tenth of a second. Blink and you can miss one. Due to a quirk in human night vision, you are often more likely to perceive them out of the corner of your eye. What will you see? To many it looks like the aurora borealis turning on and off in an instant. The true sprite color is salmon red, but at such low light levels the eye can play tricks on you and you might perceive them as green, orange or white. If you are looking in the right place and think you saw something, you probably did.

The best places in North America for sprite watching? Probably above the northern High Plains and upper midwest in a broad belt from Colorado to North Dakotas over to Minnesota and down into Texas. But they do occur above big storms worldwide, and have been spotted from aircraft and the Space Shuttle above Panama, Peru, Africa, Australia and Indonesia, to name a few places.

The more scientists look above thunderstorms, the more they find. Researchers have been making measurements with satellites, spectrometers, and photometers, and probing with radars and radio waves. It is clear that these progeny of thunderstorm lightning flashes can influence upper atmospheric electrical structure, radio transmissions, and perhaps the chemistry of the stratosphere and mesosphere. The mystery is just beginning to be unraveled. Theoreticians are furiously proposing and testing many mathematical models for sprites, jets and elves. Learned discussions abound in technical journals about electromagnetic pulses, breakdown from quasi-electrostatic fields generated by massive shifts of hundreds of coulombs of charge within giant storm clouds, and runaway electrons accelerating to energies above a million electron volts in the intense electric fields above the storms. But for most of us, the sprite is a chance to spend a calm evening on the porch, or perhaps on the tailgate of a chase vehicle that never did quite catch the tornado, and just contemplate what's up there. And to wonder. What else might nature be willing to tell us if we keep looking very carefully, and we don't blink.

Walter A. Lyons is President of FMA Research and Forensic Meteorology Associates, Ft. Collins, CO. He has 30 years experience in basic and applied research. His latest book, The Handy Weather Guide was a Paperback Book of the Month Club Selection. To learn more about sprites, check out www.fma-research.com.

This page last updated on 5-20-00.