Avalanche Breakdown in Lightning Strikes

Jan 11, 2006
Winnipeg, Manitoba
Maybe it's an odd question, but it occurred to me because I'm working on the lightning research part of my book.

I can remember discussions about avalanche breakdown in semiconductors during my lectures in electronics, where electrons collide with each other and create a domino effect that creates massive increases in current (often with specatular results:eek:).

Does the same phenomenon occur with air in a lightning channel?

ARe you referring to Relativistic Runaway Electron Avalanche? One of the possible mechanisms for lightning initiation involves incoming cosmic rays (extremely fast-moving particles ejected from stars, etc). Collisions of these cosmic rays with atmospheric particles can lead to RREA (electron showers). Perhaps some of the more knowledgeable lightning folks on this board can explain further (I'm off to lunch at Chile's). If not, I'll look through my notes from Atmos. Electrodynamics and try to explain it further. This topic can get quite complex very quickly (especially with the dozens and dozens of elementary particles that can be involved).
As I understand it, the avalanche process is the acceleration of electrons in a strong (with respect to the medium) electric field that propagates through a chain reaction. This process is characteristic of corona discharges and in the formation of the stepped leader. The primary discharge occurs in the ionized channel and maintains itself through thermal ionization so long as the air is hot enough. The main lightning channel stays ionized longer than the source and sink areas, so you may get multiple avalanche-discharge cycles down the same main channel(s) until the wide area electric potential is reduced.

While technically an initiator is required to start the avalanche, there're I think enough ions floating around in the high ambient electric field of a thunderstorm that cosmic ray initiation is irrelevant. Thinking out loud, perhaps this is more a factor in the highly rarefied atmosphere of sprites, etc.

Disclaimer: not an atmospheric physicist in any way, shape, or form....
The issue at stake is the way by which lightning initiates. Observed electric field intensities in thunderstorms are several times smaller than required for dielectric breakdown (the point at which electrons and other particles are accelerated in the presence of the very strong electric field to the point that electrons flood from a source to a sink). In other words, the E-field intensities that have been observed by numerous balloon launches, rockets, and aircraft observations through storms have never been strong enough to cause breakdown in dry or moist air. While the breakdown field strengths are ~3000 kV/m and 1000 kV/m for dry and moist air, respectively, the majority of observations have shown peak electric field intensities of 300-400 kV/m (though some have been closer to 700 kV/m). Corona discharge associated with oddly-shaped precipitation and cloud particles leads to a reduction in the breakdown field intensity in moist air compared to dry air, but it is still well above most observations).

So, this leaves a couple possibilities. First, the atmosphere may, through macroscale charge seperation, reach breakdown field intensities, but those areas just haven't been sampled. In other words, the balloons / observations may have missed the highest E-field areas, meaning that there are "pockets" that have electric field intensities beyond that of the breakdown intensity. This is a possibility, but the vast majority of observations have recorded E-field intensities several times smaller than that required for breakdown (so it isn't like the observations are just a little weaker than is required for dielectric breakdown).

A second mechanism, mentioned previously, involves cosmic rays that result in RREA. We know, from surface observations, that the Earth is constantly being bombarded by extremely high-energy electrons (and other elementary particles). The theory goes that these high energy particles collide with atmospheric molecules/atoms at such energy that electrons are "shot off" the atmospheric atoms, which then collide with other atoms and molecules, knocking off more electrons (this is the so called electron shower, IIRC, and it's also where many more types elementary particles get involved, such as muons and tau leptons). Although this process may not directly initiate lightning (afterall, it would imply that we should frequently see lightning well away from storms), it is quite possible that RREA locally augments the electric field to the point of breakdown. One interesting piece of evidence that supports this possible mechanism is the observation of Bremsstrahlung X-rays in and near thunderstorms. These types of x-rays are the result of extremely rapid accelerations in high-energy particles (e.g. high-energy electrons rapidly changing directions), a possibility that fits very well with the cosmic ray / RREA theory.

Once initiation occurs, there are considerably fewer questions about the lightning process, though we still don't really know why some CGs are extremely branched / tortuous while others are relatively straight (dart-like), or why some strikes contain numerous return strokes while some only have one or two return strokes (i.e. some "pulse" numerous times while some don't). In addition, there's considerably question as to why some storms exhibit inverted polarity structure. In this regard, across the country, one of the most common areas for these inverted polarity storms is the high Plains (i.e. the high Plains sees a much higher-than-"normal" proportion of inverted polarity than the vast majority of the rest of the CONUS for reasons that are not clear attm).

By no means am I an expert in this, but that's my understanding of the initiation question.
Last edited by a moderator:
I'll say that's all "over my head", as it were ;).... But I like the hydrometeor streamer and ice/water interface hypotheses much better than the cosmic ray hypothesis. Among other things, I have trouble understanding how the observation of x-rays in thunderstorms is an argument for cosmic rays. Since cosmic rays are uniformly distributed, one would expect the x-rays from their interaction with the atmosphere would be also. The stepped-leader avalanche process on the other hand seems to have plenty of sufficiently energetic electrons to make plenty of x-rays. See http://www.aip.org/pnu/2005/split/722-3.html.
David, I think the cosmic ray theory started from the lack of observations showing anything close to the electric fields required for breakdown. With commonly observed field strengths 2-3+ times smaller than would be required for even moist air breakdown, there arose a need to find another mechanism to initiate lightning. The observed electric fields in storms just aren't strong enough to initiate lightning (at least with the observations we have), even after considering corona discharge and other effects.

I think the Brehmsstrahlung Xrays aren't observed away from thunderstorms because of the relative lack of strong "clear air" electron sources/resevoirs (obviously, there are many free electrons in the atmosphere, but the electron/charge resevoirs in strong storms are orders of magnitude larger/stronger). In storms, as you know, vast "resevoirs" of charge are created by various processes, and these provide many more particles which the cosmic rays can hit. The problem then becomes a statistical one, with a certain number of incoming cosmic rays having a certain probability of creating RREAs. If you throw in cosmic rays in the presence of already strong electric fields common in storms, you may find a "hybrid" initiation mode that relies on cosmic rays to start the avalanche and the storm-scale electric fields to continue the avalanche. In any case, the lightning bolt itself certainly isn't the cosmic ray; the cosmic rays are only needed to initiate the avalanche process, after which time the strong electric fields in the storms continue it to ground (well, to an incoming streamer, or to an opposite-charge resevoir in the case of cloud-to-cloud lightning). Away from storms, cosmic rays may collide with atmospheric particles, but the fair-weather electric field is much too weak to support relativistic runaway electron avalanches. So, the Br. xrays may indeed be produced sometimes by the stepped leader, but that still doesn't solve the question of how the leaders originate in the presence of electric fields that are too weak initiate them. Note, once they get going, the electric fields may be strong enough to sustain them, but you need to get them going first.

LOL There's a lot of discovery to be had in lightning research, that's for sure. That's my 2 cents... I'll sit back and see if any others want to chime in (I don't want to hog the thread, and I know we have other members -- like David -- that are well-versed in lightning mechanism(s)).
Last edited by a moderator:
I am very skeptical of any EF measurements that suggest a storm doesn't generate high enough electric fields for lightning. The term 'initiate' is being tossed around rather ambiguously - to 'initiate' or 'trigger' a lightning discharge, you must already have an ambient electric field high enough for electrical breakdown of the air gap between charge areas. If the electric field is not high enough for breakdown, 'cosmic rays' nor ony other external source is going to somehow boost or supplement the EF to allow a lightning discharge to occur.

It has been well-documented, for instance, that lightning initiates in the presence of high EF off of physical objects (aircraft, television towers, wire-trailing rockets). The in-cloud initiation mechanism for the majority of discharges is simply not known yet. While I am open to all theories, the cosmic ray theory IMO has too many problems.
Wow, had no idea my post would be cause for such active debate, but I'm sure thankful for everybody's input on this one, as I've learned a lot from all the responses.:)

To say that I'm not biased where the theories are concerned wouldn't be honest. I've always maintained that charge separation (intracloud) or electrostatic induction (CG) could be enough for initiation all on their own, provided the electrical potential is large enough, and all other factors have been satisfied (ie: humidity, suspended particulate, etc.). Would it be unfair to discuss only one theory?

Just to make certain that I'm understanding things correctly, wouldn't the runaway breakdown propagate the flow of electrons to the point where there would be a vast increase in current between the point of initiation and the contact point of the main channel (ignoring the effect of branching)?

Maybe this is an overly simplistic visualization of the process?

've always maintained that charge separation (intracloud) or electrostatic induction (CG) could be enough for initiation all on their own, provided the electrical potential is large enough, and all other factors have been satisfied (ie: humidity, suspended particulate, etc.). Would it be unfair to discuss only one theory?

{emphasis added by me}

I had a similar view as you (and the others) in that enough charge could be seperated to induce strong enough electric fields for moist air breakdown (until the atmospheric electrodynamics class I had last semester). One of the main problems with the classic "breakdown" method is that the observations thus far don't support electric fields in thunderstorms being strong enough to yield breakdown (observed E-fields in storms have been, largely, 2-3 times smaller than that required for dielectric breakdown). So, this is the question: how do you "start" a stepped leader if the electric field isn't strong enough to do so by itself? Of course, there is a possibility that the initiating regions (areas where E-field intensity is equal to the breakdown E-field) just haven't been sampled yet, implying very small areas of massive E-field fluctuation (in either time or space). In the absense of observations supporting the "lightning initiating via breakdown" theory, however, other possibilities/theories have been developed. Whether cosmic rays really affect (or initiate in the presence of storm-scale electric fields) lightning is unknown at this time, but the idea that they do has support in the observations.

So, the condition of "provided the electric potential is large enough" seems to be a "deal-breaker", since observations indicate that this may not occur.
Last edited by a moderator:
This is the first I've heard of the notion that thunderstorms don't produce high enough electric field to produce lightning by themselves, based on EF observations. Again, I am highly skeptical of these observations as it pertains to this subject. The reasons I doubt the low EF observations are twofold.

First, the EF needed to overcome dielectric strength (and start electrical breakdown) of air is a principle of physics. If the potential between two charge regions is not sufficient to break down the air between them (as is suggested by the aforementioned measurements), then breakdown cannot occur. Even if cosmic rays somehow triggered breakdown (which is possible, with a critical EF level), breakdown could not continue propagating unless the ambient EF was high enough to support it. The only way to achieve breakdown with an EF smaller than needed to overcome an air gap is to physically shorten the air gap with object(s) (water droplets, aircraft, television tower, wire-trailing rocket, etc).

Second, electric fields in a storm are highly dynamic and irregular. Regions of EF that are high enough to support lightning are localized and very transient. You'd have to find the 'needle in a haystack' where the EF was reaching its critical breakdown voltage, and take the measurement before the lightning discharge occured (thereby neutralizing the charge). In most cases that's less than a minute or two. Given those challenges, it's not too surprising that actual measurements of thunderstorm-envoronmental electric fields showed the lower voltages.

If the EF is at critical stage (where breakdown is imminent), then I could see the possibility of any number of triggers starting the breakdown process, including cosmic rays. But if the EF is smaller than needed to overcome the air's dielectric strength, breakdown's just not going to happen.

In my many years of involvement with studying and observing lightning, I see a lot of published information on the subject that is questionable, particularly (and surprisingly) in textbooks, news articles, books, magazines, documentaries, etc. Unfortunately lightning has consistently not enjoyed the same attention to accuracy as other studies of science, and sadly the classroom is not immune to the problem. I would take anything read, viewed or taught with a grain of salt.

EDIT: I would highly recommend the book entitled 'The Lightning Discharge' by Dr. Martin Uman. Much of what is discussed here is covered in great detail. Any publications by either Dr. Uman or Dr. Richard Orville (both respected pioneers in lightning research) are the most trustworthy sources when delving deep into the subject of lightning.
Last edited by a moderator: