Elves: Ionospheric Heating By the Electromagnetic Pulses from Lightning:

A primer

C. P. Barrington-Leigh


Here I describe the history and science of the optical manifestation (known as "elves") of heating of the lower ionosphere by the electromagnetic pulses launched by cloud-to-ground lightning in thunderstorms.

Contents / Frequently Asked Questions:

What are elves?

When were elves discovered?

Is it possible to see an elve with your bare eyes?

What colour are elves?

What causes elves?

How do scientists observe or measure elves?

How do elves relate to sprites?

Why were elves and sprite halos confused?

Do elves move faster than the speed of light?

Do elves occur during the day, too?

Are any scientific publications (primary literature) about elves available online?

Have elves been reported by online popular press?

Who else is studying elves?


What are elves?

"Elves" is the frivolous name given to a very brief (less than 1 millisecond) but very broad (expanding out to over 500 km diameter) flash which occurs in the lower ionosphere at about 90 km altitude in response to a very strong (high current) conventional lightning stroke between a thundercloud and the Earth. Elves are predicted to occur only during the night. This flash occurs primarily at optical (visible) wavelengths and in the near infra-red. The singular of "elves" is by convention "elve," but when pronounced, sounds like "elf".

When were elves discovered?

In 1992, Boeck et al. reported observing a transient enhancement in the airglow layer above and coincident with some lightning activity, as viewed with a low-light video camera aboard the Space Shuttle. Figure 1 shows an image of this event, which was probably the first unambiguous recording of light from an elve. The existence of elves was actually predicted by Inan et al. [1991], shortly before it was observed.

Figure 1: Intensified video image from the Space Shuttle. Visible are the Earth's limb, at least one star, the airglow region at ~100 km altitude, and a wide (~500 km) enhancement to the airglow, overlying a lightning-brightened cloud.

Since that one event, all subsequent elves until 1999 were recorded from the ground. In 1999, elves were reported from balloon and aeroplane observations.

The name "elves" first appeared in the peer-review literature in Fukunishi et al.'s [1996] photometer measurement of an elve, which was shortly followed by Inan et al.'s [1997] space-time resolved measurements.

Is it possible to see an elve with your bare eyes?

No. The brightness of a surface at a given wavelength may be measured in Rayleighs. While an exceptional elve may be as bright as 10 million Rayleighs (much brighter than aurora) over a ~1 degree field of view, this luminosity only occurs for about 50 millionths of a second. Thus a human eye might receive a few thousand photons, only some of which would be in the visible spectrum.

Unfortunately, without being able to resolve the time dynamics or spatial structure of the flash, even a perceptible elve would be quite indistinguishable to a human eye from the very common phenomenon of "skyflash" due to distant lightning. This skyflash results from light emitted by a thundercloud discharge; the light scatters in all directions throughout the atmosphere and can therefore be seen (for about one thousandth of a second) at great distances from which the clouds themselves are well out of view below the horizon. This skyflash can sometimes be perceived and has unfortunately been referred to as "heat lightning," a name which leads to common misunderstanding of its origin.

What colour are elves?

Elves are NOT green! If viewed from the ground, and at a great distance, elves would appear reddish even if they were coloured white, simply as a result of atmospheric scattering (the same reason the sun appears red when it is low on the horizon). However, the optical emissions constituting elves are largely red and near-infrared, with a lesser component of near-ultraviolet. Their true colour in a visible sense, then, is certainly red.

What causes elves?

The light emitted by elves comes primarily from nitrogen molecules (the primary constituent of our atmosphere) which have been excited by a collision with an energized electron in the nighttime lower ionosphere. The electrons, in turn, are energized by the electromagnetic pulse which is launched by the huge currents which flow (for less than a ten-thousandth of a second) in the brightest phase of cloud-to-ground lightning. The loud static burst that you hear in your A.M. radio (simultaneous with a lightning flash) is evidence of this same electromagnetic pulse, whose total energy may exceed 10 gigaWatts. Most of the electromagnetic pulse travels no higher than the lower ionosphere, because it is absorbed in energizing the electrons there. Only an unusually powerful lightning stroke, with a peak current greater than 60,000 Amps, launches an electromagnetic pulse strong enough to cause detectable emissions in elves.

The interaction of the lightning electromagnetic pulse with the lower ionosphere can be simulated in detail by computer. By clicking on Figure 2, you can launch an MPEG video sequence showing some features of the lower ionosphere while it is being hit by an electromagnetic pulse originating near the ground by a vertical lightning current. There are two panels. Both show only the altitude range 70 km to 100 km, and they show a horizontal range of 350 km, where the horizontal measure is the horizontal distance from the lightning. In this model, the lightning is located (well out of view) between 0 and 10 km altitude and at 0 km horizontal distance. In the top panel, the colour scale shows the strength of the electric field in the electromagnetic pulse (measured in V/m). To visualize the whole scene (rather than just the 70 km to 100 km region), you must imagine a spherically-shaped electromagnetic pulse expanding at the speed of light from the lightning location. The pulse reaches the region shown in Figure 2 only after 0.2 thousandths of a second (ms). The red contour lines in the upper panel show the strength and location of light that is emitted by excited nitrogen molecules. The entire duration of the simulation is only about 1 thousandth of a second. You may now see why the flash of light at any one point only lasts 50 millionths of a second, while the entire elve lasts almost 1 thousandth of a second: the elve consists of a rapidly expanding ring of light emissions.

The bottom panel shows the same two-dimensional cross-section of space, but now the colour scale indicates ionization enhancement. As well as exciting nitrogen to produce light, the energized electrons can ionize other neutral molecules, thus increasing the total number of electrons (and ions) available in the lower ionosphere. The colour scale shows the modified electron density as a fraction of the ambient electron density. The electron density is nearly doubled in places. These density enhancements are thought to persist for a good fraction of a minute after an elve.

Figure 2 (Click for MPEG video). Simulation of electromagnetic pulse propagation, absorption, and excitation of emissions in the top panel, and resultant ionization enhancement in the bottom panel.

How do scientists observe or measure elves?

Very occasionally, elves are bright enough to be recorded by special image-intensified video equipment. The most remarkable such case apart from that shown in Figure 1, as of 2000, is the image below. The scale is not marked, but the view is looking up from an aircraft over Europe, and the elve is likely >400 km in diameter. The central "hole" is due to the dipole radiation pattern of the vertical lightning current.

Elve over
Europe, viewed from an aircraft. Figure 3: An unusual image of an "elve" captured by M.J. Taylor and L.C. Gardner of the Space Dynamics Lab., Utah State University, during the night of the Leonids storm at 02:10:00 UT, 1998. Click on the image for a full size version.



The vast majority of elves, however, have been measured from the ground, most effectively with an array photometer -- that is, a number (typically ~10) of very fast optical detectors looking at adjacent patches of the sky. We called our such array the "Flye's Eye". Figure 4 shows the typical geometry involved in ground observation of elves. For more detailed interpretation, you may ask me, or refer to some primary literature.

Geometry of  elves when viewed by a ground observer.
Figure 4: Geometry of ground-based observation of elves.

How do elves relate to sprites?

Below, for comparison with Figure 2, is a similar simulation in which the electric field is dominated not by the electromagnetic pulse, but by the "quasi-static" electric field of a lightning stroke. This field results from the huge excess of charge left behind in a cloud after a lightning current removes either positive or negative charge to ground. In this case, the location and dynamics of emissions and ionization changes is quite different, and the phenomenon is known as a "sprite." A lightning stroke causing sprites has a very large amount of total charge moved, while one causing elves has a very large instantaneous current.

Figure 5 (Click for MPEG video). Simulation of a sprite halo electric field, optical emissions, and ionization enhancement.

A sprite may include not just the sprite halo, modeled above, but also a filamentary (corona streamer) discharge which may propagate to much lower altitudes, and typically endures for several to many thousandths of a second.

Why were elves and sprite halos confused?

The answer to this question is dealt with in great detail by a paper published in January, 2001. Video instruments with 17 or 33 ms time resolution are poorly suited to resolving sprite halos, which typically last ~1 ms, and are extremely poorly suited to resolving elves. For years, any diffuse "halo" seen in video data atop a filamentary sprite was assumed to be the signature of an elve, even though it had a horizontal spatial scale blatantly incompatible with an elve. In late 1999, this mistake became clear during analysis of some higher speed (0.3 ms resolution) video data. Now we know that halos often occur without the development of filamentary sprite structure (and vice versa). By the way, in my observations, any lightning discharge which causes a bright sprite halo has also been impulsive enough to cause an elve.

Do elves move faster than the speed of light?

Nothing with mass or energy moves faster than the speed of light. However, the size of the luminous ring in an elve begins by growing over 100 km in radius in a small fraction of a millisecond; thus an elve can be said to expand faster than the speed of light. For more detail, see
What causes elves?

Do elves occur during the day, too?

No. During the day, the lower boundary of the ionosphere is much lower than at night. As a result, the lightning electromagnetic pulse is absorbed at a lower altitude. Because the neutral atmosphere is also much denser at these altitudes, electrons collide with neutral molecules too frequently for the electric field to be able to accelerate (energize) them enough to excite optical emissions or to cause enhanced ionization.

Are any scientific publications (primary literature) about elves available online?

The figures and explanations provided above are derived from the original publications reporting these new discoveries to the scientific community. More detail on the history of elves, as well as on almost all of these topics, is available in my PhD dissertation, which is available for online viewing in
PDF format. Additionally, most of the published papers focusing on experimental observations of elves came from our group at Stanford and are also available online.

Have elves been reported by online popular press?

Yes. Newspapers around the world have taken an interest in elves and sprites. In addition, a number of
popular science articles were published in electronic format.

Who else is studying elves?

A number of groups, mostly in the U.S. and Japan, contributed to our understanding of elves in the 1990s. Searching on the web for "elves" and "lightning" will find references to materials from some of these efforts, but will also lead you to lots of poorly informed sources. Let me know if you want to have your elves-related site listed here. Feel free to pose science questions regarding elves to me by email.


Last modified: Sun Jan 3 13:00:53 PST 2001


Copyright 2001, Christopher Barrington-Leigh / Space Science Lab, Berkeley / cpbl@