The joy of lightning

On July 24, 2006, I had the honor of almost being struck by lightning. That was not the interesting part, though.

The truly interesting thing was the metallic click I heard just before. I was doing measurements on lightning for my PhD thesis in Finland at a meteorological observatory. The equipment was in a detached outbuilding to minimize radio disturbances. The storm was overhead, and I was having as much fun as a nerd with an oscilloscope can have. Excellent data.

Taking a small break, I went to the door to enjoy the sights. Immediately, lightning struck an electric pole on the other side of the road.

I have never been more frightened. The shock wave from so close is truly mind-boggling, in the sense of boggling your brain. The hairs on your body rise up and the skin goes into goosebumps. For some reason, you feel nauseous for a long time afterwards, as well as a metallic taste in the mouth. I suspect the latter is due to an adrenaline rush. All of this is standard operating procedure (SOP) for near lightning strikes.

What was not SOP was something I did not expect. And could not immediately explain. At the same exact time the lightning hit the pole, and significantly BEFORE the shock wave, I heard a loud metallic click. Very much like the crack an electrified fence makes when touched (try it).

Sound takes half a second to travel that distance, so there seemed to be no logical explanation for this. A hypersonic pressure wave? Perhaps, but nothing like that has been reported, and it does not satisfactorily explain why the intermediate time was so quiet. It took detailed study of lightning processes to understand what was possibly going on.

One thing to understand is that a lightning stroke does not in fact travel all the way to the ground. Rather, it travels down from the cloud as a stepped leader. Just before it hits the ground, there is actually an upward stepped leader which travels from the ground up and connects with the downward leader. The real current only starts to flow and the big kabuum is heard when the connection occurs.

There can be several such attempted upward leaders, which do not manage to join the downward leader and hence die out.  This is well seen in Figure 1 (photograph by by Antti Tiihonen). Multiple downward leaders try to reach the ground. Two of them actually do; this kind of forking is quite common. Several attempted upward leaders jump up from the treetops. One clear one is seen next to the right-hand branch; there are weak upward leaders near the left-hand image as well but they are very poorly visible.

Figure 1: Downward and upward leaders.

The best hypothesis was that an upward attempted leader was the cause of the click. An attempted leader is after all a spark that can meters to tens of meters in height, and could create a sonic boom of it own. Such a noise simply had not been recorded before.

Figure 2 shows a map of the area. There were two possibile sources: a 20-meter metal observation tower about 100 meters away, or the hut itself. There were also some grounded buildings nearby. Looking at the time dfferences, the most logical choice was that the upward leader was driven from the hut. In other words, if the downward leader had happened to connect with “my” upward leader, the flash would have hit the hut and, with any luck, me.

The hut was well grounded as it had been used for storage of hydrogen for weather sounding balloons. It probably would not have been damaged badly. Nevertheless, the idea was not attractive. I was standing at the door of the hut, so the lightning could have taken any path. Almost certainly, there would have been hearing loss.

Figure 2: Map of the situation. The instrumented tower and the observation point are almost equidistant from the strike, so the tower cannot have been the source of the click. The source almost certainly was the hut itself.

The anecdote remained just that: an interesting occurrence that no one seemed to have replicated. Some amateur storm chasers did report various kinds of crackles before close flashes. (Professional lightning researchers, at least those who want long careers, try to avoid close flashes). My hunch was plausible — but unverifiable.

However, in 2009 I noticed a paper which suggested my explanation was correct. The phenomenon had been observed accidentally by Lu and Walden-Newman (2009). An attempted leader occurred near their equipment, and they got both (partial) video, electric field, and audio signals from it, ensuring that the interpretation is correct. Figure 3 shows the audible “click” quite clearly about half a second before the krakabuuuum of the actual lightning flash.

Figure 3: The first known actual recording of such a “click”, by Lu and Walden-Newman (2009).

There was not enough data for me to even consider publishing the event (no video, no audio, only a badly distorted electric field recording) so that the authors are the first ones to have measured this effect in any real scientific sense. It is nevertheless a satisfying feeling to have observed something, come up with a hypothesis, and then later found proof that suggests the hypothesis was correct.

Moral of the story? Perhaps it is always worthwhile to stay observant for anomalies, whatever the situation? And take notes. I collected what information I had, and wrote it up rather quickly. There is an old science saying: “If you didn’t write it down, it didn’t happen”.  It’s a good motto for life in general.

Writing this post, I feel nostalgic for my research days. Science is cool, especially lightning science.

8 thoughts on “The joy of lightning”

  1. I talked with Sami and searched Internet for this. Direct effect of hearing an electromagnetic pulse directly in the brain is called Microwave auditory effect or Frey effect:

    I couldn’t found a good evidence that a lightning could cause this. Most references in the Internet quote the Wikipedia article above, which has earlier read something like:

    “Some scientists state that electrophonic effects may also be caused by lightning strikes, very bright auroras[1], and earthquakes.”

    Now this text that was there at least some time year 2010 has been deleted, probably because lack of a citation. There are some other speculations of this connection:

    “I’ve heard a similar phenomenon on several occasions. I suspect that you actually “heard” the pulse of EMP as the downward and upward leaders connected for the powerful lightning return stroke. This pulse traveled at the speed of light, and so led the sound waves from the event. Workers near high power pulsed RF and RADAR sources often “hear” these as click or pops. This is called the “microwave auditory phenomenon”, “microwave hearing”, or the “Frey Effect”. The exact mechanism is not completely understood, but is thought to be either a thermal-acoustic pulse that is induced within your inner ear or it may be detected directly by your brain itself.” –


    “Another possibility, and the one I suspect, is that you actually detected the electromagnetic impulse from the lightning strike itself. If so, it may be related to a phenomenon known as the “microwave auditory phenomenon” (also known as the “microwave hearing” or “Frey Effect”). This occurs when humans (and some animals) are subjected to short impulses of microwave radiation. The exact mechanism(s) are not completely understood. Some theories invoke thermal-acoustic pulse, which is induced within the water in your soft tissues and then detected by your inner ear. Other research implies that there may be more direct detection processes going on within the brain itself. For a pulsed microwave signal, the “sound” that is heard is a click coincident with the leading edge of the pulse. Changing the parameters of the pulse can change the sound, and audio modulated words can also be detected… sounding sort of like a Cylon (robot-like).

    There is no doubt that the above effects are real and repeatable for microwave and radar pulses… perhaps for lightning RF impulses as well??” –

    But can lightning cause this? Definitely worth researching I think.

    About Cooray’s theory: Detection of EMP by eyes or ears is much easier than direct cortical stimulation because sensory organs have good “amplification circuits”. Direct cortical stimulation requires quite high magnetic field gradients so I think it’s more likely that flash or sound of the lightning causes sensory overload and epilepsy-like effects.

    1. Cool stuff! My suspicion is that there are multiple phenomena, some of which may be present at a particular time and some not. (Re Cooray: I omitted to mention that their theory works if the lightning flash is seen, if not directly then at least indirectly. Since multiplicities can be as high as 10, there is enough flicker to trigger epilepsy. OTOH this is also a reasonable explanation for why ball lightning is seen so extremely rarely: you have to have a person who is susceptible to such epilepsy, and a high-multiplicity flash, both of which are relatively rare. It hasn’t been proven in any way though, so other explanations are certainly possible).

      At the moment there is not enough data to prove anything either way. In fact, IMHO the way to proceed at this stage would be to just collect (anecdotal) data without trying to make any assumptions at all. With say a hundred well-documented case studies, one might start to have enough information to at least start to see patterns. But it’s crucial that the collectors be completely neutral (of course everyone has a pet theory, but a competent person is able to avoid letting that interfere).

      This is too big an effort for one person, but if we could find say half a dozen people to jointly find and collate the information (and search on the Internet like you started) then we might start to get somewhere. I can volunteer to be one of the six, but can we find others? :-)

  2. Just one thing to add: the answer! Hasn’t this been out before? If not, then why?

    I had faint memories of MRI devices and their pulses being heard by the patients. While doing searches to confirm this, I remembered what the teacher in radio tech classes told: you can hear the radar pulses if you happen to cross the path of the radar. That’s it! High energy EMP doing something in the brain cells, or at the inner ear. Either way, the coupling is electromagnetic, not acoustic.

    Now I also remember being told that it’s possible to stimulate the visual cortex magnetically, so that phosphenes are perceived. Here’s a link to one abstract of a study “Magnetic stimulation of visual cortex: factors influencing the perception of phosphenes.”

    When it’s possible to create false visual signals, surely producing an audible click must be easy. Just add power ;)

    1. Interesting concept. While it might be possible to test experimentally, it might be difficult to find volunteers… (I would volunteer without hesitation of course).

      Cortical stimulation is quite possible, in fact there is an interesting theory (by prof Cooray et al in Uppsala) that ball lighting could actually be (sometimes) an effect caused by the stimulation of frontal lobes by EM fields from lightning (especially high-multiplicity flashes). The descriptions of ball lightning often are similar to descriptions given by people with frontal lobe epilepsy, and a high-multiplicity flash could trigger such epilepsy in prone people (just as a blinking light can). As far as I know it’s still all speculation, but it makes eminent sense. [I have my own theory on ball lightning, but it’s not as credible as Cooray’s].

      So yes, direct cortical or auditory stimulation cannot be ruled out. It’s simply that I prefer Occam’s razor, and the upward leader hypothesis assumes fewer unknown physical processes. So I’m still sticking with the present hypothesis.

      It would be great to zap people to test this, but it’s difficult to find funding for unethical human experiments that have no real purpose. :-)

  3. Thanks for your comments. It already got difficult to keep track on them :) I should create a separate document for precise argumenting. Anyway, I’ll try to be brief. I wrote the word “echo”, and with that I mean anything that differs from the direct path of the audio signal. Be it echoing or reverberation. Note that reverberation can be so short you cannot hear as a timely event, but it “bends” the spectrum of the audio. Our auricles do that also, they obstruct and reflect the audio, and that causes spikes and notches to the spectrum, which helps us to locate the audio source in 3D. None of this was heard in my outdoor click observation. The sound was purely and distinctively in “mono”. I know the click is short, but not too short to be noticed as the basic process: timing and level difference between right and left ear.

    About electronics:

    0. “However, if they both heard it and it was recorded on audio, it would seem unlikely that it was caused at the least in the audio circuitry. (Something that intense would have broken the equipment).”

    I was thinking about the microphone preamp. It deals with micro to millivolt range and very low currents. Surely it’s possible to pick up EMP through the circuit without breaking anything. In case of EMP (sorry for this very unprecise expression), there will be no high currents in the circuitry, because of the high impedance of the amplifier, design (short leads) and the shielding, which attenuates the EMP well to keep the voltages low. Electronic fields can also be picked up by the microphone itself. Try that by going near a CRT of a TV set with your camcorder. If you don’t have a CRT at home, I can do it for you :) The computer monitor tubes are different, that’s why a real old-fashioned TV is required. Of course, now I need to verify this! I surely remember dynamic microphones producing buzz when brought close to a TV CRT, but how about the condenser mikes more generally used. We’ll see… or hear :)

    1. Psychoacoustics: this may vary from person to person, but the ear is much faster than the eye, and the spatial information decoding is “built-in”, we don’t think about it. We first hear the sound, and later on start thinking about what caused it. Then the memories begin to fade. But first we hear it, psychoacoustics (in this non-realtime sense) kick in later.

    2. We don’t have metallic roof (well, we do have metallic outlining of the roof and the water collection system) nor antenna, not even galvanic cable TV. What is left, is the electrics, coming from underground. The indoor click is interesting, because the structures of the house attenuate high frequencies very well. If it was coming from outside of the house, it could not have sounded the way it did. And if it came from outside at that intensity, it would have the reverberation from the ground and surrounding buildings. And I have heard sparkovers from the wall outlets for example. They’re very different, focused on lower end of the audible spectrum and being softer on the attack. Surely they may vary according to the voltage for starters.

    And for the attempted upward leaders: I don’t see how they could “always” (with respect to the number of observations) produce just one sharp click. I’ve heard the true crackling of a close strike, it’s like pending drum sticks into pieces. By the way, the actual drum stick pending heard on an AC/DC record sounds like something is generating strong electric sparks – very nasty sound :D
    I’ve heard from somewhere that the crackling comes from the leaders. It could fit. They don’t have the energy to generate pressure waves like the discharge channel does. The energy (on the audio frequencies) is on the mid to upper frequencies – depending on how thick walls you have ;)

    Footnote. The indoor click I heard was caused by a strike, that was captured by the Mikkelin Ursa sky camera. I also know the discharge location within a triangle based on two other observations from the other corners of the triangle (+ added location accuracy by the photo, which I can not find at the moment…). Not that this makes any proof to my argumentation, but it is just nice bonus information for myself :)

    1. Maybe we should start an open-science project on this. :-) I can set up a Google Docs page after midsummer, and we can try to get e.g. observations from Ursans. Everything is anecdotal, but we need to start somewhere (and of course we should urge all storm chasers to be on the lookout for this).

      There’s one aspect about fizz/crackle/pop/click that I haven’t mentioned: there are basically four types of ionization that can occur when the electric field starts to rise, each with somewhat different audio characteristics.
      1. Corona discharge (St Elmo’s fire). Air is slightly ionized, but no real movement of charges. Possibly humming or fireplace-like crackle.
      2. Streamer. A “cold” discharge in which the plasma in the channel does not heat up appreciably. More or less all the sparks we observe in normal life are actually streamers. These are extremely short-lived and produce a snap.
      3. Leader. Qualitatively different from the streamer in that the plasma is heated up significantly, the channel can be quite thick (a centimeter or so), the leader can live for quite a while, and the expansion causes a more intense sound. I would tend to put my observation into this category, since the click really did not resemble any normal spark that I’ve experienced in my life.
      4. Lightning discharge channel. Kabuum. I don’t remember from memory, but I think the plasma channel thickness can be up to a meter and the corona sheath can be 30 meters. Maaany orders of magnitude bigger than anything else.

      I can’t unfortunately remember any exact figures from memory, but those discharges are in any case very different from each other. So in practice, I’m pretty sure that one reason why observations will vary is that different phenomena have been observed.

      Very complex stuff, lightning… :-)

  4. This is interesting, because I’ve heard a click or snap a few times myself and been thinking about logical causes. I’ve also written about the clicks to the URSA stormchasers list, as I recall. Did it really happen when I’m not 100% sure I _wrote_ about it? ;)

    I could say I’m not a very scientific person, but I’m into audio. I began creating music soon after I got my first computer, Commorode Amiga 500 at the end of 1989. Editing audio samples became necessary and when I couldn’t get what I wanted from the editing software I had, I wrote my own pieces of code (in BASIC) to do some things I needed. Obviously, one of the tasks was to draw the audio waveform. Later on by experimenting with simple maths I found a simple way to boost or attenuate the highest frequencies (the Nyquist frequency) in the audio signal. It was very useful for enhancing the sounds I used. That was done by calculating the difference between the adjacent samples, multiplying the result and adding to the original 8-bit signal. I was wondering, what if I calculated the difference from n > 1 samples apart? Would that target the effect to lower frequencies? I was looking for ways to create a some kind of equalizer without complex maths :) Indeed, the n had something to do with frequencies, but the usefulness of the algorithm was of random nature. I guess it would only work with pure sine wave. And I still couldn’t figure out a way to process arbitrary frequencies, just the F/2, F/3, F/4… However, this algorithm was enhanced by Teijo Kinnunen and it ended up to a music editor software called OctaMED.

    In the early 90’s, I got an analogue synthesizer, namely Roland Juno-106, and got familiar with subtractive synthesis. Soon I knew what knobs to adjust to get the sound I wanted.

    That was a bit too long story, but the purpose was to say I’ve seen the audio waveform in several contexts and know what sound comes from which shape. Also, understanding the nature of audio makes it easier to remember what you’ve heard – I think. Short click comes from a short pulse. The shorter the pulse, the higher the mean frequency. In the case of a thunder click, we’re hearing something caused by a very short event, or actually, the “fastest rising edge” of the event, presumably the transition from pre-discharge to discharge. The sound has never been a “thud” or a “knock”. The flash dies off rather slowly, so that’s what we’re not interested in here.

    Now, let’s put the prologue aside. I’ll present some of my observations trying to clarify the phenomenon (and willing to oppose your hypothesis, let’s see if I can).

    0. Firstly, before I forget, I’d like to ask how the Lu and Walden-Newman (2009) recording was proven NOT to include an electromagnetic pulse possibly induced to the microphone cabling or the circuitry inside the recorder? This is one of the basic problems in any scientific studies: how to eliminate noise. How to verify the existence of noise in recorded signal, when the recorded event is unpredictable and has unknown parameters? If the scientists themselves heard the click, and later saw it in the spectrogram, was it really the same click? That is difficult to prove.

    1. We have two ears. We can hear directions of the audio source. In my observations, I haven’t been able to pinpoint the source, not even then when I’ve seen the discharge channel(s) directly (psychological connection between eyes and ears did not occur).

    2. I have heard it indoors as well. For the record, the last time I heard it April 27th 2012, when the first small thunderstorm of the season went over Mikkeli. This click was the loudest I’ve heard. The strike had a lot of energy as well, estimated by the rumble I heard (from the audio you can evaluate many things from the strike, was it cloud-to-ground or not, did it have multiple discharges, the direction related to the observer and “branchiness”, among other things contribute to the audio very distinctly). I was in our kitchen, where a lot of metallic objects are found, most notably the cooker hood less than 2 meters away from me. I considered it as the most probable source at first. But my son heard the click as well, and he was upstairs in the other end of the house. From that I concluded it didn’t come from the cooker hood, or any other smaller metallic object.

    3. In one case, in quiet environment outdoors (no rain, no wind, traffic etc.), the click, or snap has sounded very distinctly like it was inside my head. The discharge occurred several kilometers away. What might help the click to be heard is that the air was very humid. I’m saying that maybe the electric field shakes the water molecules in the air, and they connect to the ear drums? I read from somwhere, that the ear drum needs only move just 1/10 of the width of hydrogen atom, for audio to be observed.

    4. I came to think that I’ve never heard multiple clicks either from a single discharge.

    5. I’ve observed is that the click occurs at the exact moment of the main discharge, regardless of the distance to the strike.

    6. I’ve never heard the click echoing. Especially the #3 could have created echoes, should the click been a normal audio pulse. I was at a suitable distance from buildings, and the place was very familiar for me, so I guess I would have recogniced the echo bouncing from the walls.

    I’m sure we need this kind of non-scientific observations as well. Some of them can push the scientists into the right direction, recording the phenomenon with scientific instruments.

    1. Good stuff. :)) Some comments:

      0. It’s difficult indeed to prove. However, if they both heard it and it was recorded on audio, it would seem unlikely that it was caused at the least in the audio circuitry. (Something that intense would have broken the equipment). Also in any scientific setup great care is taken to avoid inductive currents in the equipment. It does happen of course, but again something intense enough to spark an audible noise would probably break stuff. Also, the electric field and audio correlate so well that it’s difficult to see what in the equipment would have caused it. I’ll put it this way: this click is so loud that if you hear it, your first assumption is that something has been seriously broken. So an equipment issue cannot be completely disproven, but it’s not likely.

      1. I agree with the difficulty of pinpointing. I heard a similar click earlier during the same measurements, but much weaker, and I didn’t observe the flash (or there was too much of a delay between the click and the flash to make a mental connection). And that one was impossible to pinpoint either. Could this in part be a psychoacoustic phenomenon? If you have absolutely no expectation of a click, are your ears going to be able to pick up the direction particularly well? I’m personally not very strong at pinpointing noise direction in any case, so it doesn’t surprise me that I didn’t.

      2. I remember several people reported hearing it indoors when I did a brief survey among the Ursa storm chasers in 2006. Two possibilities: a sparkover in the electrification of the house (although that should have left damage) or an upward leader from the top of your house. (Most houses here don’t have separate lightning conductors, but if the roof is grounded correctly, the likeliest location would be the top corner of the roof. Or the chimney or TV antenna, which is a less pleasant though). I would still tend to suggest the latter, as otherwise there should be damage somewhere. That would also explain why it’s so difficult to pinpoint it: you would have been actually standing under the upward leader.

      3. Possible… but there is another explanation: the horizontal extent of the downward leaders can be several kilometers or more before it actually hits the ground, and also there can be many attempted downward leaders. So it’s completely feasible that there was in fact a very strong electric field around you, possibly strong enough to cause an attempted upward leader. It doesn’t take a very strong field to produce a corona (St Elmo’s fire) but those tend to sizzle rather than click, and not very strongly. An intermediate possibility is that there was a streamer close to you, not intense enough to form into an actual upward leader. So while your explanation is not impossible, the alternative mechanism doesn’t require any mechanisms that would not be completely known already.

      4. Typically, once the downward leader attaches to the upward one, no more upward leaders are generated. And even if they were, any audio would be drowned out by the roar of the thunder. So hearing just one click is consistent with this theory, but it’s also consistent with other theories.

      5. This would support the upward leader theory. The attachment occurs at most a few microseconds before the upward leaders are formed, so to the ear they are simultaneous. This simultaneousity was very striking to me also: the click occurred at exactly the same moment the visible flash hit the pole, much before the thunder.

      6. This supports my hypothesis in #3 I think. If the click was produced locally and was not loud in absolute terms, then there would have been little or no echo.

      I don’t know if the scientists are terribly interested in this phenomenon as such, since upward leaders are a well-known phenomenon… :) The audio click is an anecdotal curiosity, but certainly interesting enough to have been published in JOLR. Hard to see any practical importance to this…

Comments are closed.