# What Determines the Zig Zag Pattern of Lightning?

• Jeff Rosenbury
In summary, the experts in the conversation discussed the predictability of lightning and how its path is determined by various factors such as air conductivity, fractal nature of the discharge channel, and wind movement. They also mentioned the use of lightning rods and other methods of lightning protection, but expressed a desire for a more effective solution. The formation of the leader and the complexity of the electromagnetic fields generated by lightning were also discussed. The issue of lightning damaging electronics on sailboats was brought up and the experts debated whether the current completely leaves the conductor or finds an additional parallel route. They also mentioned previous experiences with lightning strikes and the difficulty of finding a sufficient conductor for the high voltage and current of lightning.
Jeff Rosenbury
In a discussion about lightning destroying electronics on sailboats the issue of controlling lightning came up. One member (Baluncore) proposed that lightning is completely predictable. I've been taught that lightning is somewhat random.

To me, random implies non-linear equations with sensitive dependence on initial conditions (the butterfly effect). Yet the equations lightning seems to follow are linear as far as I can tell. A puzzling note is the tendency of lightning to shoot halfway down a conductor, then jump off it and through the hull of a fiberglass boats rather than following the wire to water on the side of a boat.

So am I missing some non-linearity, or is it possible to model lightning well enough to prevent deaths analytically rather than just statistically? (The way lightning protection works now seems to be to throw enough lightning rods up to statistically catch most of it. I'd like to think there's a better way.)

Do lightning charge carriers gain enough momentum to keep going despite the fairly strong forces bending them? In air, it would seem they might gain relativistic speeds (with millions of volts potential), but in copper? It doesn't seem likely. And if lightning wants to go straight so badly, why does it zig zag across the sky?

The conductivity of air is not uniform, and the lightning takes the path of least resistance. I don't think the exact reason is settled - one possibility is that these are paths of air ionized by cosmic rays.

You can get relativistic electrons from lightning, although usually only the very largest flashes. More typically, electrons gain energy from the electric field, and then lose it to the atmosphere by ionization etc. You end up with a lot of electrons with a little energy rather than a few electrons with a lot. The rare cases with a substantial component of fast electrons is believed to be responsible for x-ray emission observed from some strokes.

Jeff Rosenbury and dlgoff
I believe the path lightning takes depends mostly on the ionization path through the air, called the leader. The formation of the leader is not well understood, but I several ideas have been proposed, including slight variations in the makeup of the air, cosmic rays, or the enhancement of the electric field by ice crystals.

Jeff Rosenbury
have a look at this video, particularly from around 2:45 to end

it clearly shows the stepped non-direct path taken

Jeff Rosenbury
Abstract from http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=902309&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D902309

Usually the electric and magnetic fields associated with lightning have been computed by assuming the lightning current to be contained in a straight vertical channel of negligible cross section above a flat perfectly conducting plane. Such a model, which does not take into account that real lightning is characterized by tortuosity and branching, is not able to justify the fine structure of the fields radiated by lightning discharges whose time-domain behavior exhibits a jagged shape with remarkable spectral content in several bands of practical interest. In this work the effect of channel tortuosity and branching is investigated by adopting a suitable numerical technique. The discharge channel has been regarded as a fractal antenna whose associated EM field has been evaluated by superimposing the contribution of the single line radiators composing the whole channel. Such a field has been compared with that generated by a simple dipole antenna in order to study the influence of the fractal nature of the channel on the generated EM fields. The relationship between the fractal dimension of the discharge channel and the fractal dimension of the generated time domain EM fields has been considered and the influence played on such a relationship by the distance between EM source and observation point has also been studied by analyzing the fields evaluated at far and close distances

Jeff Rosenbury
So does the fractal path cause the complex EM field, or does the complex EM field cause the fractal path?

Lightning happens when electrical potential difference between two points becomes sufficient that a discharge is going to occur.
The exact form (shape) of the discharge is not predictable because there are numerous factors which affect the "easiest path" for the electron flow,
and the discharge is not instantaneous, it can take a few seconds and in few seconds, the easiest path can change quite a lot.

I don't think it ads clarity to the answer, but rocket triggered lightning experiments make it graphic. Below is a still picture and a video. The still picture shows that what appears to be a single strike is actually multiple strikes all occurring in a fraction of a second. Wind moves the ionized air molecules between the strikes. The video shows what they called an anomaly in the path.

TheDemx27, Jeff Rosenbury, davenn and 1 other person
Drakkith said:
I believe the path lightning takes depends mostly on the ionization path through the air, called the leader. The formation of the leader is not well understood, but I several ideas have been proposed, including slight variations in the makeup of the air, cosmic rays, or the enhancement of the electric field by ice crystals.
That would explain the path above the lightning rod, but it wouldn't explain why lightning frequently leaves a heavy copper conductor to blast through a highly insulated fiberglass hull.

Jeff Rosenbury said:
why lightning frequently leaves a heavy copper conductor to blast through a highly insulated fiberglass hull.
Does the current leave the conductor completely, or does it merely find an additional parallel route?

Jeff Rosenbury
Jeff Rosenbury said:
That would explain the path above the lightning rod, but it wouldn't explain why lightning frequently leaves a heavy copper conductor to blast through a highly insulated fiberglass hull.
When a trillion billion electrons are trying to go somewhere in a hurry, what conductor is going to be sufficient?

A.T. said:
Does the current leave the conductor completely, or does it merely find an additional parallel route?

usually the second part ... I have seen damage done further down the cable by the significant currents still traveling within the cable

there was another thread recently on lightning where I stated experience with lightning strikes on phone cables etc
the high voltage and current of lightning strikes does not like to go around corners. Any significant bends in the cable or lightning rod grounding strap and the lightning would jump from the corner of that bend out to anything else remotely earthed that it could find other cables, metal brackets etc

I saw this effect many times in my years in the telecoms industry both on multi core cables or on radio comms poles with mast top lightning rod and large copper strap down the mast to the Earth mat

The #1 installation rule was ... NEVER do sharp bends. If you have to do bends, make sure they are
rounded and as wide angle as possibleregards
Dave

Last edited:
Tom_K said:
When a trillion billion electrons are trying to go somewhere in a hurry, what conductor is going to be sufficient?
If the charge carriers are ions or something, the momentum might be a factor. But for electrons, I don't see it.

The effect would almost need to be an electromagnetic effect to be as strong as it is at its distance scale. Perhaps the self inductance in a straight leg of conductor is strong enough to generate enough voltage at the turn to break down air? But that still doesn't explain why it goes straight instead of some random angle. Hmm...

I make a prediction. I foresee math in my future.

anorlunda said:
I don't think it ads clarity to the answer, but rocket triggered lightning experiments make it graphic. Below is a still picture and a video. The still picture shows that what appears to be a single strike is actually multiple strikes all occurring in a fraction of a second. Wind moves the ionized air molecules between the strikes. The video shows what they called an anomaly in the path.
]

The video shows that the lightning starts with the rocket's wire, but by the end, it finds its own path. This seems strange to me. Clearly there are better electrical paths than a nice, straight line of copper ions (which seems counter-intuitive). It makes me wonder it the spectral content causes the fractal pattern rather than the other way around? But what would be the mechanism?

Jeff Rosenbury said:
The video shows that the lightning starts with the rocket's wire, but by the end, it finds its own path. This seems strange to me. Clearly there are better electrical paths than a nice, straight line of copper ions (which seems counter-intuitive). It makes me wonder it the spectral content causes the fractal pattern rather than the other way around? But what would be the mechanism?

The wire is vaporized on the first stroke. Then the wind moves the ions horizontally. Later strokes follow the displaced ionized path. I don't think you can think of it as a single event.

Even if the wire doesn't vaporize, it could ionize the air near the wire surface. Then wind moves the air while the rolling boat moves the wire before the next stroke.

Also, there is often a lot of vertical wind shear near the ground.

Complexities everywhere you look. Your question is very difficult to answer.

davenn
anorlunda said:
The wire is vaporized on the first stroke. Then the wind moves the ions horizontally. Later strokes follow the displaced ionized path. I don't think you can think of it as a single event.

Even if the wire doesn't vaporize, it could ionize the air near the wire surface. Then wind moves the air while the rolling boat moves the wire before the next stroke.

Also, there is often a lot of vertical wind shear near the ground.

Complexities everywhere you look. Your question is very difficult to answer.

I don't doubt that even weak lightning would vaporize the sort of wire an inexpensive rocket could carry.

What you say about the wind is true, yet at the very end, the lightning seems to finds a wildly different path. Perhaps this is due to the wind destroying the path, but perhaps not. I suspect that is the most likely explanation. Though natural lightning seems to occur in stages, not nearly as many I think. The multiple events are possibly due to the rocket continuing into new cloud layers.

I agree the complexities are immense. But if it were easy, someone would have done it already. Where's the fun in that? This is an important scientific question, at least from economic and human perspectives even if it's unlikely to lead to new insights on how the universe works on a grand scale. The standard model likely applies, but we don't know how yet. It's a pleasant puzzle to chew on with real world benefits.

It's possible, even likely there are non-linear qualities that will make more than statistical prediction impossible, but we don't know that yet. The basic equations all look linear.

Yes , yes I loved one of the points earlier , remember that copper itself has a resistance which of course will of course will increase greatly with an increase in temperature due to the increased energy from the lightning , hence , at some point , theoretically the air once more becomes a better conductor
A.T. said:
Does the current leave the conductor completely, or does it merely find an additional parallel route?

Jeff Rosenbury
Jeff Rosenbury said:
In a discussion about lightning destroying electronics on sailboats the issue of controlling lightning came up. One member (Baluncore) proposed that lightning is completely predictable. I've been taught that lightning is somewhat random.

To me, random implies non-linear equations with sensitive dependence on initial conditions (the butterfly effect). Yet the equations lightning seems to follow are linear as far as I can tell. A puzzling note is the tendency of lightning to shoot halfway down a conductor, then jump off it and through the hull of a fiberglass boats rather than following the wire to water on the side of a boat.

So am I missing some non-linearity, or is it possible to model lightning well enough to prevent deaths analytically rather than just statistically? (The way lightning protection works now seems to be to throw enough lightning rods up to statistically catch most of it. I'd like to think there's a better way.)

Do lightning charge carriers gain enough momentum to keep going despite the fairly strong forces bending them? In air, it would seem they might gain relativistic speeds (with millions of volts potential), but in copper? It doesn't seem likely. And if lightning wants to go straight so badly, why does it zig zag across the sky?
Maybe it really is not unpredictable but very difficult to predict , like the possibility that a man can fly is possible just not very possible

Jeff Rosenbury
Nile Anderson said:
Yes , yes I loved one of the points earlier , remember that copper itself has a resistance which of course will of course will increase greatly with an increase in temperature due to the increased energy from the lightning , hence , at some point , theoretically the air once more becomes a better conductor

I would strongly suggest to you that the vaporisation of the copper wire happens long before that

Dave

Jeff Rosenbury
Nile Anderson said:
Yes , yes I loved one of the points earlier , remember that copper itself has a resistance which of course will of course will increase greatly with an increase in temperature...
I was rather thinking about the increase in charge density in the copper limiting the current.

A.T. said:
I was rather thinking about the increase in charge density in the copper limiting the current.
That's worse than it sounds because the skin effect will limit the charges (at least the moving ones) to the surface of the conductor. (The Skin Effect)

However, since conductors regularly vaporize in lightning strikes, we can be confident that the resistance doesn't rise enough to kick the current out. Once the copper ionizes, it's at least as conductive as the ionized air is. (I think, could I be wrong on this?)

The pinch effect should limit the ionized path. Perhaps it then pinches the ionization trail until it can no longer support the current do to the skin effect?

But this doesn't explain the main case where the cable doesn't melt the copper, yet jumps off anyway. And it does so in the same gauge wire it's happy to follow in a sraight line.

Lightning definitely doesn't like sharp changes in conductor direction, yet seems to like such changes while in the sky. It's a puzzle.

Jeff Rosenbury said:
Lightning definitely doesn't like sharp changes in conductor direction, yet seems to like such changes while in the sky. It's a puzzle.

you have to remember those zigzags seen in the lightning path in the sky are over many metres probably 10's of metres quite different to a sharp bend in a cable of less than a couple of metres

no real puzzle there

Did you watch the video I posted much earlier in the thread ( the slo mo one) showing how the zigzag path forms ?

Dave

davenn said:
you have to remember those zigzags seen in the lightning path in the sky are over many metres probably 10's of metres quite different to a sharp bend in a cable of less than a couple of metres

no real puzzle there

Did you watch the video I posted much earlier in the thread ( the slo mo one) showing how the zigzag path forms ?

Dave
I did watch the video. It was informative. It showed the mechanism of the direction change, but not a clear reason. The assumption that the current is choosing the low resistance path is belied by the currents choice to leave the low resistance copper to jump what should be a leaderless path through fiberglass.

Nor does the change itself seem to be over a large volume of space. The turns look sharpish.

The distance of each segment does seem largish. Still the distance from the cloud to the ground is only a few thousand feet. It's easy to misjudge vertical distances.

http://ieeexplore.ieee.org/xpl/abstractAuthors.jsp?tp=&arnumber=902309&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D902309 examined the fractal nature of the current. It occurs to me the straight paths may act as antenna segments which discharge energy depending on spectral content. Thus there may be a limited set of possible segment lengths depending on what the spectrum of the lightning is at any particular point. But the mechanism isn't clear to me.

Speculating:

Accelerating charges emit photons. But due to the phase (gauge?) relationship with the length of the current path certain energy/frequency photons can't be emitted (or at least are much less probable). This would cause the charge density mobility (current) to clump into some speeds/frequencies. This might create some sort of back electromagnetic force which would cause a build up of charge at a sharp turn in the conductor (allowing flashover through the fiberglass). However I know of no mechanism for such a back EMF. It could be a quantum effect, or it could just naturally fall out of the equations in what to me is a counter intuitive manner. Or it could be bovine fecal material (BFM). My bet is on the BFM, but if I'm wrong, this could be an explanation.

I simplify the jaggedness of lightning down to the repulsion of electrons in the drift. Imagine: +[ ]- you have 2 electrons at the plus that are heading in generally the same direction towards the minus. They repulse branching out from the drift, but then one path is always favorable, hypothetically. So now the next pair of electrons to follow tend towards the favorable path ahead of it, branching perpendicular to the guiding path, giving electricity it's 90 degree kick every step.

jerromyjon said:
I simplify the jaggedness of lightning down to the repulsion of electrons in the drift. Imagine: +[ ]- you have 2 electrons at the plus that are heading in generally the same direction towards the minus. They repulse branching out from the drift, but then one path is always favorable, hypothetically. So now the next pair of electrons to follow tend towards the favorable path ahead of it, branching perpendicular to the guiding path, giving electricity it's 90 degree kick every step.
The pinch effect seems to prevent that. It maintains current primarily in one path. Of course this works only up to a point, so lightning does occasionally split. But it clearly doesn't take all paths, only a few, and often only one. (At least for the vast majority of the current.)

I do have another contender pre-hypothesis: It has been postulated that marine epoxy might contain copper dust as an anti-foulent. At high voltage levels this might make it behave as a supercapacitor. I have no idea if either of these conditions are valid. If they both are true, the hull could act as a short to the high frequency components, allowing a pinch to form and drive the current through the hull.

That's the other side of the story, conductors provide a stable "channel" for the electrons to organize into an efficient "stream" around a wire. The problem with wires is they don't always lead exactly where the current wants to go, so at some point the easiest path veers too far off course and it has to redirect.

Jeff Rosenbury said:
The pinch effect seems to prevent that.

not something I have heard about before as relating to lightning (electrical discharges in general)

will have to read up more

Jeff Rosenbury said:
so lightning does occasionally split. But it clearly doesn't take all paths, only a few, and often only one. (At least for the vast majority of the current.)

yeah it does split ... often
a pic from my own collection about 500m to 1km away from me ...

the final one or 2 paths taken are really controlled by the stepped leader coming down from the cloud finally meeting the streamer coming up from the ground
there can be a number of streamers but only one may connect. As in this pic. you can see 2 streamers one from a power pole on the left and another from the
tree. Amazingly that streamer from the tree didn't connect with the stepped leader. There must have been another streamer from the tree that reached a bit higher
NOT MY PIC !

cheers
Dave

jerromyjon said:
That's the other side of the story, conductors provide a stable "channel" for the electrons to organize into an efficient "stream" around a wire. The problem with wires is they don't always lead exactly where the current wants to go, so at some point the easiest path veers too far off course and it has to redirect.

Clearly the current finds the easiest path. But just as clearly the easiest path isn't always the lowest resistance path. So the question is, what causes the impedance to be so different from the resistance?

Jeff Rosenbury said:
what causes the impedance to be so different from the resistance?
The impedance simplifies to resistance in direct current, from what I've read. I'm still trying to fit the whole puzzle together.
Jeff Rosenbury said:
allowing a pinch to form and drive the current through the hull.
I think the pinch you were referring to is caused by the current organising so that the magnetic field helps to condense the stream, meaning the current precedes the pinch not the pinch precedes the current.

It is obviously a complicated phenomenon but I'm determined to figure it out!

Jeff Rosenbury said:
Clearly the current finds the easiest path. But just as clearly the easiest path isn't always the lowest resistance path. So the question is, what causes the impedance to be so different from the resistance?

Path singular? Event singular? My thinking keeps leading back to the plurals of both.

As time progresses, more ions will be created, thus changing the path resistance greatly. Resistance and impedance are transient and distributed. To model this problem you need differential equations distributed in 3D space. Maybe you even need Q.E.D.That sounds impossibly difficult to me.

But there.might be hope if you narrowed the OP question. A lightning rod's wire carries the current to ground. The wire does not evaporate, nor does it have sharp bends. Think of the guy wires for 1000' TV antennas. Does the current jump out of (or in to) the wire sometimes? There may be photographic evidence. If so why?In past PF threads, Davenn said that no matter what, the protected equipment inside get fried sometimes. That suggests that anaomlous outcomes do occur. Have a look at the remarkable picture of the Washington Monument that Dave posted in this thread.

Last edited:
davenn and Jeff Rosenbury
jerromyjon said:
The impedance simplifies to resistance in direct current, from what I've read. I'm still trying to fit the whole puzzle together.

I think the pinch you were referring to is caused by the current organising so that the magnetic field helps to condense the stream, meaning the current precedes the pinch not the pinch precedes the current.

It is obviously a complicated phenomenon but I'm determined to figure it out!
Current can (temporarily at least) travel through a capacitor. Then, at least theoretically, the current could form a magnetic field which forced the current to continue after the capacitor filled. Think of it as a small inductor in series with a low voltage capacitor. The inductor could easily break down the capacitor with careless design.

Of course that assumes the hull acts as a high value capacitor, which it doesn't under normal conditions. It's an unlikely explanation; probably more BFM.

I really think I'm missing something right in front of my face, but I can't see it. I keep looking at the leaders coming out of the ground, wondering why current would be going the wrong way. I wonder if that could be + ions being drawn towards the - charged cloud?

jerromyjon said:
wondering why current would be going the wrong way.

what do you mean by that ?

jerromyjon said:
I wonder if that could be + ions being drawn towards the - charged cloud?

a good probability

and don't forget that the charge polarities can also be the other way around

Dave

davenn said:
what do you mean by that ?
I meant I was stuck in a rut only considering the electrons and not the ions. Electrons wouldn't be coming out of the ground going towards the cloud while the electrons in the cloud are going towards the ground, that wouldn't make sense.
davenn said:
and don't forget that the charge polarities can also be the other way around
That was when it dawned on me, it goes either way for either polarity. It does make much more sense now. The leaders from the ions would be less "mobile" than the electrons, wouldn't they? That would explain why the branches are much more prominent on the electron side. Now I'm curious if I can find slo-mo of lightning going from ground to cloud... I think it would look different if charges were reversed.

It appears I am correct. Now I just have to figure out how the connection is formed to start the "seek and destroy" (discharge)

Last edited:
jerromyjon said:
I meant I was stuck in a rut only considering the electrons and not the ions. Electrons wouldn't be coming out of the ground going towards the cloud while the electrons in the cloud are going towards the ground, that wouldn't make sense.

That was when it dawned on me, it goes either way for either polarity. It does make much more sense now. The leaders from the ions would be less "mobile" than the electrons, wouldn't they? That would explain why the branches are much more prominent on the electron side. Now I'm curious if I can find slo-mo of lightning going from ground to cloud... I think it would look different if charges were reversed.

It appears I am correct. Now I just have to figure out how the connection is formed to start the "seek and destroy" (discharge)

Charge mobility is different for different charges. The most mobile seem to be electrons. The next most mobile are holes I think with about ⅓ the electron's mobility. Holes exist in semiconductor material in what I assume is a quantum effect. They are unfilled electron shells which act as positive charge carriers even though the constituent atoms are locked in a matrix. I mention them to show there are other possible charge carriers than electrons or ions. (I don't know which work in plasma though.)

Ions are several thousand times more massive than electrons and should be much slower for that reason. But I suspect Anorlunda's Quantum Electrodynamics would be needed to figure out which charges are moving where.

Also, cloud charges are sometimes reversed. I seem to recall reading this is common in monsoon storms for some reason. It is infrequent (I've read 10%) in temperate storms. Anyway, I tend to discount which charge is which for this reason, though perhaps that's a mistake.