How lightning discharges are generated

From thunderstorm electrification to leader attachment and the return stroke: a technical, sequential view of the physical phenomenon that produces lightning.

Lightning is a process, not an instant

A lightning discharge is the visible manifestation of an electrical sequence that develops inside a thunderstorm and, in some cases, between the cloud and the Earth. The phenomenon begins long before the flash: charges separate, the electric field grows, ionized regions appear and channels develop until they may connect and carry a large current.

In cumulonimbus clouds, strong updrafts and downdrafts keep supercooled water droplets, ice crystals and graupel in constant interaction. Collisions among these particles favor charge transfer and separation. The real distribution is dynamic and complex, but simplified models commonly describe predominantly negative charge in middle or lower regions and positive charge aloft.

Three broad discharge families

Not all electrical activity in a thunderstorm reaches the ground. Three broad families are useful for understanding the phenomenon:

• Intra-cloud: discharge occurs between regions at different potential within the same cloud.
• Inter-cloud: an electrical channel develops between different clouds or cloud regions.
• Cloud-to-ground: an electrical connection is established between a charged cloud region and the Earth.

Cloud-to-ground flashes are especially relevant to external lightning protection design, but they are only one part of a much broader three-dimensional atmospheric phenomenon.

1. Preliminary breakdown: dielectric failure begins

Charge separation increases potential differences inside the storm and between cloud and ground. As the electric field intensifies, air no longer behaves as an ideal insulator. Local ionization begins and small conductive regions can form.

This preliminary stage should not be imagined as a complete lightning flash, but as the beginning of dielectric breakdown that prepares the medium for more organized channels.

2. Stepped leader: the descending channel explores space

In a typical negative cloud-to-ground flash, a stepped leader may propagate from the cloud toward the ground in successive steps. The channel contains ionized plasma and branches while exploring multiple possible paths.

The leader does not know its final attachment point in advance. Its path evolves with the electric field around the cloud, terrain and structures. This is why lightning cannot be reduced to the idea that it simply seeks the tallest object or the path of least resistance.

3. Upward leader: the surface participates

As the descending leader approaches the surface, the electric field becomes particularly intense around tips, edges, elevated structures and geometric discontinuities. Upward streamers or leaders may develop from some of these points.

Several upward channels may compete at the same time. The final attachment point is determined by the connection process between one upward channel and a branch of the descending leader.

4. Attachment: the connection is established

When the distance between a descending channel and an upward streamer becomes sufficiently small, the intervening air completes its ionization and attachment occurs. A continuous conducting path between cloud and ground is then established.

This moment is fundamental to protection engineering: the strike point is not determined solely by object height, but by local electric-field geometry and by which upward channel successfully connects.

5. Return stroke: the main flash appears

Once the conducting path exists, a return stroke occurs. A large current travels through the channel and creates the intense luminous flash that we normally identify as lightning. Rapid heating of the air causes violent expansion that is later perceived as thunder.

The return stroke is therefore a later stage in the sequence. Before the bright flash, a complex evolution of charge, leaders and streamers has already taken place.

6. Dart leader: the channel may be reused

The first return stroke does not necessarily end the event. The previously ionized channel may remain favorable to another discharge for a short time. If additional charge remains available in the cloud, a new leader—often called a dart leader—may travel approximately along the existing channel.

This subsequent process usually has fewer branches because it takes advantage of part of the ionized path created by the earlier discharge.

7. Subsequent strokes: why lightning can flicker

A new dart leader can trigger another return stroke. The sequence may repeat several times while adequate electrical conditions remain in the cloud. This is why a single visible lightning flash may contain multiple strokes and appear to flicker.

The expression “a lightning bolt” can hide this complexity: one visible event may actually be a sequence of several electrical processes following a similar path.

Polarity: cloud-to-ground flashes are not all the same

Negative cloud-to-ground flashes are the most common: a negatively charged stepped leader descends from the cloud and eventually connects with an oppositely charged upward channel. Positive cloud-to-ground flashes also occur and can behave differently, including a greater likelihood of sustained continuing current.

For engineering, polarity matters because waveform, transferred charge, current duration and multiplicity are not identical in every event. Lightning protection standards therefore do not represent lightning with a single current number; multiple current parameters are used to evaluate thermal, mechanical, electromagnetic and surge effects.

The stepped leader in numbers

For a typical negative cloud-to-ground flash, the descending leader progresses in steps on the order of tens of metres. The U.S. National Weather Service describes typical steps of roughly 50 m and an overall leader speed on the order of 200,000 mph—about 90 km/s—while emphasizing that actual speeds vary considerably.

Near the surface, the process becomes local. Upward streamers intensify from structures and terrain features as the descending leader approaches. There is no universal attachment distance: geometry, polarity, electric-field distribution and the surrounding environment all influence the final connection.

Return stroke: current, speed and channel temperature

Once a conductive path is established, the return stroke propagates extremely rapidly. NOAA/NSSL reports roughly 60,000 miles per second, while NWS educational material gives an equivalent order of magnitude near 200 million miles per hour.

The current-carrying core is physically narrow: NSSL describes approximately 1–2 inches in diameter, surrounded by a wider region of charged particles. The air in and around the channel can be heated to about 50,000 °F (approximately 28,000 °C) in an extremely short time, producing the shock wave that becomes thunder.

Peak current is not a fixed universal value. Engineering practice commonly deals with return-stroke peaks in the tens of kiloamperes for many events, while protection levels and component tests must account for a much broader statistical distribution and severe events.

One flash may contain many strokes

A visible flash may contain one return stroke or a sequence of repeated strokes. NSSL notes that a negative cloud-to-ground flash may contain as many as about 20 return strokes. Dart leaders can reuse a previously ionized channel, and some flashes also contain continuing current.

Continuing current is important because thermal and ignition effects depend not only on peak current but also on how long charge continues to flow through the channel.

Electrical parameters that matter to engineers

Asking only “how many amperes are in lightning?” is insufficient. Different damage mechanisms depend on different parameters:

• Peak current: affects potential rise, electromagnetic forces and component duty.
• Current steepness, di/dt: strongly influences induced voltage and electromagnetic coupling.
• Transferred charge: relates to thermal effects and melting at attachment points.
• Specific energy/current integral: contributes to thermal and mechanical stress.
• Duration and continuing current: important for heating and ignition.
• Stroke multiplicity: subjects the same path to repeated stress during one flash.

IEC 62305 treats these parameters separately because a protection system must address both direct strike effects and conducted/induced effects on internal systems.

From direct strike to electromagnetic impulse

Damage is not confined to the physical strike point. Lightning can simultaneously cause ground-potential rise, hazardous potential differences, inductive coupling into cable loops, conducted surges on power/control/telecommunication lines, intense electromagnetic fields around down conductors and internal arcing.

This is why a complete lightning protection system is more than an air terminal: interception, down conductors, bonding, grounding and surge protection must operate as one coordinated architecture.

What an observer actually sees

The human eye mainly perceives the luminous channel produced during return strokes, while the stages that determine its path begin earlier and are much less visible. The descending leader, upward streamers and attachment form the electrical architecture that makes the main flash possible.

This distinction matters because lightning protection does not attempt to stop the atmospheric phenomenon. It aims, within the limits of engineering, to control where a discharge may be intercepted and how its current can be conducted through designed paths that reduce risk to people, structures and electrical systems.

From the physical phenomenon to a protection system

Understanding this sequence explains why lightning protection must be treated as a complete architecture:

• Air termination: provides designed interception points.
• Down conductors: establish defined current paths.
• Grounding: supports current dispersion into the Earth.
• Bonding: reduces hazardous potential differences between conductive elements.
• Internal protection: limits the effects of transient overvoltages on electrical and electronic equipment.

None of these layers should be analyzed in isolation. Lightning is transient, three-dimensional and variable, so design begins with risk assessment and uses geometric and electrical criteria defined by the applicable standards.

The next level: interception geometry

Once descending leaders, upward channels and attachment are understood, geometric design methods such as the rolling sphere become more intuitive. The method uses an imaginary sphere, whose radius depends on the selected protection level, to identify which points of a structure may remain exposed to an approaching downward channel.

The rolling sphere does not explain how to “attract” lightning. It is a design tool used to identify potential interception points and locate the air-termination system appropriately. That criterion will be developed in a separate technical Insight.

Conclusion

Lightning is not a spark that suddenly appears between cloud and ground. It is the result of a sequence: thunderstorm electrification, charge separation, preliminary breakdown, stepped-leader development, upward channels, attachment, return stroke and possible subsequent strokes.

Understanding that sequence changes the engineering perspective: instead of trying to “prevent lightning,” the objective becomes designing a coordinated architecture to intercept, conduct, dissipate and limit its effects according to the real risk of each installation.