A thunderstorm can be considered an electric generator in which positive and negative charges are separated and stored in different regions of the cloud. As the charges are separated, the field will grow between opposite charges in the same cloud or in different clouds, or between a cloud and the ground. When the breakdown field strength is exceeded, an electric discharge will take place, sometimes in the form of a lightning flash. As further charges are separated, the process is repeated but normally at different places.
The mechanism of a thunderstorm is, in principle, very simple. But in spite of this and the many years of thunderstorm research, a detailed knowledge of the processes responsible for the charge separation and the discharges is still not at hand. There are, however, a series of known processes that actually do take place in thunderstorms and that may cause charges to separate.
To evaluate if a given process may play a role in the electrification of a thundercloud, it is necessary to determine not only if the process is able to separate charges and distribute the polarities in a way corresponding to what is measured in actual clouds but also if the rate of charge separation can explain the time variations observed in, for instance, the intervals between lightning flashes.
Direct observations from aircraft as well as radar measurements have shown that a thundercloud consists of one or more active cells with high vertical velocities. Rain and hailstones, as well as lightning activity, are formed in the cells. A particular cell may have a total lifetime of about an hour, whereas the precipitation and lightning activity may last only 15–20 minutes. A fully developed cell may have a horizontal extension of 2–10 km and a height of some 10 km, with the base of the cloud at maybe 3 km above ground (see Figure 1). The vertical velocities are typically in the order of 5–10 m·s–1. In special cases, however, velocities of 30–40 m·s–1 have been measured.
The temperature in a thundercloud varies from about 0°C at the base of the cloud to about –40°C in the upper layers. At the top of the cloud is a region with predominantly positive charge, whereas the negative-charge region is located at the base. Often, a smaller region with positive charge is also found at the base.
A thundercloud can be considered a vertical dipole (often double), and a lightning discharge can partly be described as a change in the dipole moment. The magnitude of this change is in the order of 100 C·km, corresponding to a charge of 20–30 C. As an average thunderstorm produces a lightning flash every 10–20 seconds, a charging current of about 1–2 A is required. The first lightning flash is normally registered about 20 minutes after the precipitation has started. Assuming that the precipitation is necessary for the electrical activity, a rough estimate of the charge that has to be separated in a volume of about 50 km3 in about 20 minutes (i.e., with a rate of about 1 C/(km3·min)) would be approximately ±1000 C. It should be mentioned that some thunderstorm specialists claim that, instead of precipitation, convective activity is the prerequisite for charge separation.
Turning to the actual mechanisms of charging, several processes may be responsible for the charge separation. It has been shown that regions with temperatures below 0°C are often the seat for the most active charging processes. However, it has also been shown that considerable charge separation may take place in warm clouds.
In the thundercloud, there are big ice particles in the form of hailstones formed by glazing, that is, the freezing of subcooled water by contact with a cold body. The ice formed is amorphous and glassy and has the same content of impurities as the cloud particles from which it is formed. The glazing process can take place only in the temperature interval from 0° to –15°C. At lower temperatures, the water vapor freezes out and forms pure ice crystals. In addition, the hailstones will have a somewhat higher temperature than the ice crystals, partly because of the latent heat released by the freezing of the water. If a hailstone and an ice crystal happen to come into frictional contact with each other, the more conductive, impure hailstone will become positive and the pure ice crystal will become negative. Calculations and measurements (laboratory and field) both seem to show that this process might give charging rates in the order of 10 C/(km3·min), which should be sufficient to cause lightning activity.
A similar process takes place when subcooled water drops hit ice particles. The drops will partly freeze onto the ice particles in a glazing process. Because of the released latent heat, part of the water will stay liquid, moving away with a positive charge and leaving the heavier ice particles negatively charged. The collisions between ice particles and subcooled water drops may also result in the expulsion of light ice splinters, again leaving the ice particles with a negative charge. The charging rate of these processes seems to be about 1 C/(km3·min), the estimated critical value for lightning activity.
Other possible charging processes are inductive charging in an inhomogeneous field of solid or liquid particles that come in contact with each other and are subsequently separated by gravity, and thermoelectric charging by asymmetric friction or other causes. A common feature of these theories is that they all assume the presence of precipitation elements and that the processes typically take place in a temperature region below 0°C.
According to another theory, the strong vertical air movements in the thunderclouds may play a vital role in the charging process. The theory maintains that, in the lower atmosphere, the air has a small excess of positive ions and that a cloud formed in this region will therefore initially be positively charged and attract negative ions from the upper atmosphere. The negative ions may be caught by a downdraft and on their way attach themselves to droplets or other elements of precipitation. In this way, a negative charge is formed at the cloud base, and the resulting field may be strong enough to cause corona discharges at sharp points on the ground. Positive ions will be attracted to the cloud but may be caught by an outside updraft and carried to the top of the cloud, increasing the field so that more negative ions are attracted from the upper atmosphere, thereby amplifying the charging process until lightning activity starts.
It is not possible to point to one of these theories (or others) as being mainly responsible for the charge separation in thunderclouds. Probably the charging is a result of more than one and not always the same processes.
Current Balance in the Atmosphere
It has already been mentioned that the fair-weather current and the charge brought to ground by elements of precipitation account for a positive current density of about 4·10–12 A·m–2. To evaluate whether this can be balanced by the effect of thunderstorms, the contributions from lightning discharges and from the current induced by the fields below and above the thunderstorms must be looked at separately.
As far as the lightning discharges are concerned, it is estimated that about 1800 thunderstorms are active at any one moment. Each of these storms produces about 60 lightning flashes per hour, each carrying about 20 C, with approximately 80% of the flashes bringing negative charge to the ground. This will then correspond to a total current of about –1.3·106 C/hr, or –360 A, which corresponds to–0.7·10–12 A·m–2.
The current caused by the fields below and above the thunderclouds can be estimated in two different ways. The current below a thundercloud has been measured at –10–8 A·m–2. With about 1800 simultaneously active storms, each with an approximate horizontal dimension of 50–60 km2, this corresponds to an average current density for the earth as a whole of –2.3·10–12 A·m–2. It should be mentioned that it is extremely difficult with any degree of accuracy to measure the vertical current density during a thunderstorm. In another method, the current above the thunderclouds is measured to have an average of about –0.8 A per thunderstorm. Using the same assumptions as in the previous argument, this corresponds to an average current density of about –2.8·10–12 A·m–2.
The contributions from the various processes can be summarized as shown in Figure 2. It appears that about –0.7·10–12 A·m–2 is lacking in order to make the current budget balance. It should, however, be kept in mind that the calculation of the two negative contributions is based on very uncertain estimates of the number of active thunderstorms and the average lightning rate. It has, for instance, been suggested that the number of thunderstorms active at any one moment might be closer to 3000 than the figure of 1800 used above, but this higher figure could also include a number of weaker systems with lower lightning rates and possibly weaker fields. Despite the uncertainties, the estimated values of the atmospheric electric parameters seem to fit the circuit reasonably well.
Although lightning discharges contribute rather modestly to the current balance in the atmosphere, these same discharges have such violent (direct and indirect) effects and properties that a short survey of this phenomenon would be appropriate.
A lightning discharge is a transient current of high intensity spanning several kilometers. Lightning may be produced by sandstorms and snowstorms or by erupting volcanoes, but the most common cause is the activity in cumulonimbus clouds (thunderclouds). Although the most common type of lightning discharge takes place entirely within the cloud (intracloud strokes), the discussion will be limited to the discharges between a cloud and the ground.
Any cloud-to-ground discharge is made up of a series of partial discharges separated in time by 40–50 milliseconds and lasting about 200 milliseconds for the total flash. Figure 3 shows some of the characteristic features of a cloud-to-ground stroke as it would appear on a streak-camera photograph.
The Stepped Leader. Each lightning discharge starts with a predischarge or leader that propagates from the cloud to the ground in weakly luminous steps. The predischarge may start as a local discharge between the N and p regions in the base of the thundercloud (see Figure 1), converting part of the negative charge bound in the base into highly mobile electrons, which will be carried to ground in a negatively charged column. The column appears to move in luminous steps about 50 m in length, with a time between steps of about 50 microseconds, during which time the intensity of the steps is too weak to be observed.
The predischarge moves with a velocity of about 1.5·105 m·s–1 (>500,000 km/hr), and because the base of a thundercloud is typically at an altitude of 3 km, it takes about 20 milliseconds for the predischarge to reach the ground. The negative charge in a predischarge is about 5 C, and the average current is therefore approximately 100 A. The diameter of the discharge has been measured photographically to be between 1 and 10 m, but it is assumed that the actual charge transport takes place in a narrow core surrounded by a luminous corona sheath, which is what is actually observed.
The Main Stroke. When the predischarge brings the negative charge at high potential close to the ground, the field strength at ground level may be high enough to cause ionization and make the discharge move from the ground to the leader. When the two discharges meet, the leader is effectively grounded and its conductive channel will support a very luminous main or return stroke.
The return stroke differs in many ways from the leader stroke. The strongly ionized wave front moves with a velocity of about one-tenth of the speed of light, covering the distance from the ground to the base of the thundercloud in about 70 microseconds.
The region between the wave front and the ground is traversed by strong currents that bring the excess negative charge in the leader channel to the ground. Measurements at ground level have shown currents in the main discharge of 10–20 kA during the first few microseconds, falling off to half the initial value in 20–60 microseconds, but with currents of hundreds of amperes to flow for several more milliseconds.
During the main stroke, considerable amounts of energy are dissipated in the discharge channel, and the temperature will be higher than during the predischarge. Because the gas density cannot change instantaneously, the pressure in the channel will be higher than in the surroundings, and the channel will expand supersonically, producing a shockwave that gives rise to the sound of thunder. The shockwave phase lasts about 5–10 microseconds, during which time the gas density in the channel behind the shockwave will decrease until a state of equilibrium is reached between the channel with high temperature and low density and the surrounding air with low temperature and high density. In this state, the discharge channel has a diameter of a few centimeters.
The Dart Leader. The lightning discharge is not necessarily finished even if the current in the main discharge has decreased to zero. The main discharge has provided a conductive trail, and if extra charge from the N region is available less than some 100 milliseconds after the main stroke, this charge may move through the channel as a continuous discharge or dart leader. The dart leader appears as a luminous section of the channel about 50 m long moving toward the ground with a velocity of about 2·106 m·s–1, or about 10 times as fast as the stepped predischarge. The dart leader reaches the ground in about 1 millisecond and carries a charge of about 1 C. During this charge transfer, the ionization of the channel has increased and a new main stroke is possible. This process may be repeated (normally about 3–4 times), but much larger numbers (up to 26) of return strokes have also been observed.
The first return stroke, the actual main discharge, is strongly branched downward as in the preceding stepped leader. The subsequent return strokes, following the dart leaders, are only slightly branched. The first return stroke transfers more charge than do later return strokes, but because a more or less continuous current is flowing in the time between the two return strokes, the total charge transferred by a lightning flash is about twice that of a single-stroke flash.
Positive Strokes. Occasionally, stepped leaders bringing positive charges to the ground have been observed, but in such cases, the leader current is carried by negative charges (electrons) flowing out of the top of the leader into the positive region of the cloud and thereby charging the channel positively. The charge transferred by positive strokes may be about three times that of negative strokes, with maximum values of about 300 C. Positive strokes rarely have more than one return stroke.
Upward Leaders. If the field strength at ground level is particularly high (e.g., at very high structures or mountain tops), the ionization may start here, with the leader developing upward.
Charge Balance in Thunderstorms. As the majority of lightning flashes from a thundercloud to the ground carry a negative charge, it is expected that a thundercloud would eventually get an excess positive charge, reducing the possibility of bringing further negative charge to the ground. However, this tendency is counterbalanced both by positive discharges from the top of the cloud to the surrounding air and by the vertical current above the cloud often being greater than that below the cloud.
Effects of Lightning Discharges
This article primarily addresses the electrical phenomena and processes taking place in the atmosphere, but it does seem appropriate to briefly mention the effects on buildings, installations, and human beings brought about by the lightning discharges.
If the lightning strikes a conductor, an amount of heat approximately proportional to the charge is dissipated in a relatively small volume around the point of impact. The material may melt and be thrown around because of magnetic and pressure forces. Furthermore, the lightning current will dissipate heat and create magnetic forces on any conductor through which the charge is led to ground.
The lightning may also create overvoltages in installations and even along the ground. These overvoltages may appear in the medium itself because of resistive coupling or, in conductive surroundings, because of inductive coupling through the magnetic field or capacitive coupling through the electric field.
Finally, humans may be fatally injured and suffer brain damage by being hit directly by lightning strokes. Burns and damage to organs are rather rare because the discharge normally runs on the surface of the body, leaving lightning figures on the skin. However, nearby strokes may create overvoltages dangerous to human beings up to about 100 m from the point of impact.
It naturally follows that a relatively short review such as this one has to be rather summary. A series of relationships between the various elements have therefore been bypassed or only mentioned superficially, which may give the impression that the atmospheric electric circuit is a simpler phenomenon than is actually the case.
Niels Jonassen, MSc, DSc, worked for 40 years at the Technical University of Denmark, where he conducted classes in electromagnetism, static and atmospheric electricity, airborne radioactivity, and indoor climate. After retiring, he divided his time among the laboratory, his home, and Thailand, writing on static electricity topics and pursuing cooking classes. He passed away in 2006.