Parmi les manifestations les plus mystérieuses de l'électricité atmosphérique est le phénomène de foudre en boule, ou Kugelblitz. Un boule lumineuse soit :

  1. apparaissant après un éclair d'un nuage vers le sol et restant près du sol, soit
  2. d'abord vue en l'air, descendant d'un nuage ou arising from no obvious cause, thereafter remaining aloft until it vanishes.

Les collisions avec un appareil ont provoqué des dommages vérifiés, indiquant que la foudre en boule n'est pas restreinte au niveau du sol.

La plupart des témoins indiquent que la foudre en boule est clairement visible le jour bien que non aussi brillante qu'un éclair ordinaire. Quelques 85 % des observateurs s'accordent sur le fait que la taille et la brillance de la boule restent globalement constantes tout au long de la période d'observation et qu'aucun changement n'intervient même jusqu'à sa disparition. Une minorité signale des changements de brillance et de couleur juste avant que la boule disparaissent. Les couleurs rouge, orange et jaune sont les plus courantes, mais la plupart des autres couleurs sont vues occasionnellement. Certains chercheurs pensent que les Kugelblitz bleu ou bleu-blanc sont associés à une plus grande énergie, bien qu'il n'y ait aucune base statistique pour une telle assertion. Les diamètres de Kugelblitz rapportés varient de 5 à 80 cm avec une moyenne d'environ 30 cm. Une étude répertorie 3 [complexions] de foudre en boule :

  1. une apparence solide avec une surface mâte ou réfléchissante, ou un coeur solide avec une enveloppe translucide
  2. une structure rotative, suggérant des mouvements internes
  3. une structure avec une apparence de feu.

Le dernier type semble plus commun. Près de 1/3 des témoins détectent des mouvemenst internes ou une rotation de la boule elle-même, bien que cela puisse dépendre de la distance de l'observateur.

Une majorité de témoins rapportent le mouvement de la boule comme étant lent (environ 2 m/s) et horizontal, sans guidage apparente par le vent ou par le sol. 1 observateur sur 6 signale des vitesses dépassant 25 m/s. Plusieurs signalements indiquent bien un guidage par les lignes téléphoniques ou électriques et par des objets au sol. Une odeur de brimstone (soufre brûlant) est souvent rapportée par les observateurs proches, en particulier au moment de l'affaiblissement.

La durée moyenne de la foudre en boule en grossièrement de 4 s, avec 10 % signalant plus de 30 s. La détermination de la durée de vie est difficile car :

  1. le temps subjectif durant un événement excitant est souvent erroné, et
  2. peu d'observateurs voient une boule du moment où elle est créée jusqu'au moment où elle disparaît.

Dans tous les cas, un canal d'éclair ordinaire pouvant rester électriquement conducteur pendant seulement 0,1 s, une durée de 10 s est de 2 ordres de magnitude au-delà de ce qui est attendu.

Il n'y a pas longtemps, une discussion scientifique considérable s'ensuivi sur la question de savoir si la foudre en boule était un phénomène réel. Les scientifiques pensaient que la foudre en boule pouvait être :

  1. l'image rétinienne persistante d'un éclair,
  2. une décharge intense de point coronaire près de la cible d'un éclair sous un nuage orageux,
  3. un matériau brûlant ou incandescent projeté depuis le point d'impact d'un [bolt] d'éclair.

Today most researchers believe that Kugelblitz is a genuine electrical effect. A recent survey indicates that ball lightning may be extremely commonplace, but that the observer must be relatively close to the ball to be able to see it. Kugelblitz is probably invisible or indistinguishable in daylight at distances greater than 40 meters, which would explain why it is incorrectly believed to be a rare phenomenon.

The median distance between an observer outdoors and ball lightning is 30 meters. Sometimes ball lightning floats through buildings. The median distance between indoor observers and ball lightning is only 3 meters. The reported distance of the observer seems to be closely correlated with the reported size of the ball. A more distant observer is

  1. less likely to notice luminous balls of small diameter, and
  2. more likely to misjudge the diameter.

The second difficulty is somewhat mitigated since in most cases of ball lightning terrestrial landmarks can be used for reference in estimating distances and sizes. On the other hand, estimates of the distance and size of a luminous sphere seen against the sky can be quite inaccurate.

In one report, a red lightning ball the size of a large orange fell into a rain barrel which contained about 18 liters of water. The water boiled for a few minutes and was too hot to touch even after 20 minutes. Assuming

  1. that the water temperature was initially 20°C,
  2. that 1 liter of water evaporated, and
  3. that 17 liters were raised to 90°C,

one needs roughly 8x106 joules of energy (equivalent to 2 kg of TNT). For a ball 10 cm in diameter (the size of a large orange), the energy density is then 5x109 joule/m3. But if all the air in a volume were singly-ionized, the energy density would be only 1.6x108 joule/m3. Both the energy content and the energy density of ball lightning as derived from the singular rain barrel observation seem incompatible with the non-explosive character of most Kugelblitz. Although many lightning balls emit a loud explosive (or implosive) noise upon decay, effects characteristic of the release of energies of the order of 2 kg of TNT have rarely been reported (understandably if the observer was within 3 meters) . Moreover, explosive or implosive decays have been noted indoors with no apparent heat or damage to nearby ceramic objects. Nevertheless, there are enough well-documented cases of extremely high energy Kugelblitz to make the water barrel report very believable. Probably there is a wide range of possible energies for a lightning ball, with the vast majority of Kugelblitz possessing energy densities less than that of singly-ionized air. The minimum possible energy of a lightning ball is that required to illumine a sphere about 25 cm in diameter with the brightness of a fluorescent lamp. With 10% efficiency, this means a source of 250 watts for 4 sec., or about 1000 joules of energy. We can only conclude with certainty that the energy of a lightning ball lies somewhere between 103 and 107 joules.

Theoretical efforts have focused on the energy estimate of the rain barrel observation. To maintain a fully-ionized, perhaps doubly-ionized mass of air requires either

  1. a large amount of energy concentrated in a small volume and shielded from the surrounding air by a remarkably stable envelope, or
  2. a continuous energy flow into a small volume, presumably by focusing power from the environment.

Theories which attempt to bottle fully-ionized plasma by magnetic fields or magnetovortex rings are faced with severe stability problems. There is no known way to contain plasma in the atmosphere for as long as a few seconds. Moreover, a fully-ionized plasma ball would be hotter and probably less dense than the surrounding air, so that it would tend to rise rather than descend or move horizontally. Chemical combustion theories cannot explain the high energy content or the remarkable antics of the ball. Nuclear reactions would require an electric potential of at least 106 volts between the center and surface of the ball, and a mean free path for the ions as long as the potential gap. This situation seems unlikely, and faces similar problems of stability.

Theories which depend on an outside source of energy such as microwaves or concentrated d-c fields cannot explain how ball lightning can survive indoors.

If energies as high as several megajoules are not required, we can try other hypotheses. One suggestion is that the lightning ball is a miniature thundercloud of dust particles, with a very efficient charge separation process. Continuous low energy lightning flashes are illuminating the cloud. Another idea is that a small amount of hydrocarbon, less than that required for combustion, is suddenly subjected to strong electric fields. The hydrocarbons become ionized and form more complex hydrocarbon molecules which clump together. Eventually there is enough combustible material in the center to allow a burning core. If the concentration of hydrocarbon decreases, the ball disappears if the concentration increases, the ball ignites explosively. (This represents the swamp gas theory for ball lightning).

Much depends on a reliable energy estimate for the Kugelblitz. If the energy is as high as indicated by the water barrel report, we have a real dilemma. At present no mechanism has been proposed for Kugelblitz which can successfully explain all the different types of reports. Probably several completely different processes can produce luminescent spheres in the atmosphere.

We conclude this section with summaries of several eyewitness reports of Kugelblitz.

The first few cases concern aircraft.

  1. A commercial airliner (LI-2) was struck by ball lightning on 12 August 1956 while flying in the lower Tambosk region of the USSR. Before being struck, the aircraft had been flying at 3.3 km altitude through a slowly moving cold front which contained dense thunderclouds. During a penetration of one thundercloud, where the air temperature was about -3°C, the crew saw a rapidly approaching dark red almost orange fireball 25 to 30 cm in diameter to the front and left of the aircraft. At a distance of not more than 30 to 40 cm in front of the nose, the ball swerved and collided with a blade of the left propeller, exploded in a blinding white flash, and left a flaming tail along the left side of the fuselage. The sound of the explosion was loud enough to be heard over the noise of the engine. No substantial damage could be found. One of the left propeller blades had a small fused area 4 cm along the blade and less than 1 cm in depth. Around the damaged region was a small area of soot, which was easily wiped off.
  2. In 1952, a T-33 jet trainer was flying near Moody AFB in Georgia. Because of a thunderstorm, the pilot was told to proceed to Mobile, Ala. As the T-33 rolled out onto a westerly heading at 4 km altitude, it collided with a "big orange ball of fire" that hit the nose head-on. The jolt was such that the student pilot believed there had been a midair collision with another aircraft. The low frequency radio compass no longer functioned, and they had to receive radio guidance to another base. On examination of the aircraft, they did not find a single mark or hole. The only damage was to the radio compass unit in the nose of the T-33 which was practically melted inside and was rendered useless. After the radio compass was replaced, everything functioned normally.
  1. Another pilot distinguishes ball lightning from balls of St. Elmo's fire, and states that he has only seen "true" ball lightning near severe thunderstorms associated with squall lines, mountainous terrain, and significant cloud-to-cloud lightning. He defines "true" ball lightning as having the following characteristics:
    1. diameters between 15 and 30 meters,
    2. never originates outside the main thunderstorm cloud,
    3. generates from a single point and expands in exactly the same manner as the fireball of an atomic explosion, but with a longer lifetime,
    4. earphones detect soft sibilant hiss, easily distinguishable from crash static, which gradually increases in loudness concurrent with the growth of the ball, then rapidly decreases in loudness after peak brightness,
    5. no apparent thunder.

    He considers smaller luminous balls seen near his aircraft to be St. Elmo's fire. If Kugelblitz within clouds can be as large as is estimated by this pilot, then ground-based observations reflect only weak manifestations of the phenomenon.

  2. In Klass's book there is a remarkable photograph taken by an RCAF pilot in 1956, which seems to confirm the above observations. The pilot was flying westward at 11 km altitude over the foothills of the Canadian Rockies near Macleod, Alberta, through what he describes as the most intense thunderstorm he ever saw in North America. Cloud pillars extended above 12 km. The sun was setting behind the mountains and was obscured from view. The ground was dark. Through a break in the clouds he observed a bright stationary light with sharply defined edges "like a shiny silver dollar." The light was nestled deep within the thunderstorm, suspended above some cumulus reported at 4 km altitude. The object remained in view for 45 seconds as he flew across the cloud break. The diameter of the light is estimated to be at least 15 to 30 meters.

The following case is indicative of high-energy ball lightning.

  1. At , following a moderate rainstorm at Roche-fort-sur-Nier (France), an extremely brilliant flash of lightning branched into three directions. At the first impact point, a witness described a ball 15 to 20 cm in diameter and 2.5 meters above the ground which passed only 4 meters in front of him. He felt a breeze of air at the same time. The globe climbed an iron cable which it melted and pulverized, producing smoke in the process. The electrical conduits of an adjoining house were burned and the meter was damaged. The observer, who was installing a gas pipe, received a shock. At the second impact point several workers saw a globe also 15 to 20 cm in diameter touch the top of a crane. There ensued a great explosive noise accompanied by a blue spark as large as an arm which flew 40 meters and struck the forehead of a dock worker, knocking him to the ground. A dozen shovelers working 10 to 50 meters from the crane received shocks and were knocked over, one being thrown 60 cm into the air. The shovels were torn from their hands and thrown 3 or 4 meters away. No smoke or odor was perceived. At the crane, current flowed along the electric cable, boiled the circuit breaker board and the windings of the crane's electric motor. The chief electrician received a violent shock and was unable to free his hands from the controls. At the third impact point, a ball of fire as large as two fists hit a lightning rod and descended along the conductor to the ground, disappearing behind a building. Two workers saw a ball of fire roll very rapidly along the ground.
  2. A Hanovre (Allemagne) lors d'un orage en , a fireball the size of an egg came through the window, left a burnt spot near the ceiling, traveled down the curtain, and disappeared in the floor. No burnt marks were found in the floor or curtains, but the ceiling had a slightly charred mark the size of a penny.

Cases like these are not unusual. Ball lightning has been known to cut wires and cables, to kill or burn animals and people, to set fire to beds and barns, to chase people, to explode in chimneys, and to ooze through keyholes and cracks in the floor. It has even been reported in the passenger compartment of a DC-3 aircraft. Moreover, lightning conductors are not always able to dissipate the energy of Kugelblitz. In St. Petersburg, Fla., during the summer of 1951 an elderly woman was found burned to death in an armchair near an open window. Above one meter, there were indications of intense heat - melted candles, cracked mirror, etc. A temperature of 1400°C would have been needed to produce such effects. But below one meter there was only one small burned spot on the rug and the melted plastic cover of an electric outlet. A fuse had blown, stopping a clock in the early morning hours. Since lightning is common near St. Petersburg, this case has all the marks of Kugelblitz.

  1. Le , Diane de France, fille illégitime de Henri II, alors le Dauphin, épouse Francois de Montmorency. La nuit de leur mariage, une flamme oscillante entre dans leur chambre par la fenêtre, se déplace de coin en coin, puis finalement sur le lit nuptial, où il brûle les cheveux de Diane et and night attire. Elle ne leur fit aucun autre mal, mais on peut imaginer leur frayeur.