Sounds of the Aurora

by Brian Dunning
July 15, 2025

Sponsor an episode!

The lights of the aurora are one of our planet's most exquisitely beautiful sights. The sounds of the aurora, however, are something that might be new to many of us. Those who live in aurora country and see them frequently are all too familiar with the crackling and popping sounds they've either heard themselves, or know people who have. But to the rest of us, this seems like a very dubious claim. How can the gentle interaction of charged particles with a trillion trillion trillion air molecules in the ionosphere — with each individual interaction being unspeakably weak — generate noise that we can hear down on the ground? It seems an impossibility; and it may indeed be so. But people are hearing something, and today we're going to shine the light of science until our skeptical eye falls upon the explanation.

Aurorae occur mainly in the range of 100-300 km above the Earth (occasionally they extend a bit lower or a bit higher). So we might be inclined to wonder whether it's even possible for any sound to propagate at that altitude? As they say: In space, no one can hear you scream. And it turns out that that's the right question to ask. At a certain altitude, atmospheric pressure is low enough that acoustic waves can no longer propagate within the audible range. As we move up through the atmosphere, the distance between air molecules increases, and higher frequencies with shorter wavelengths can no longer be supported. Go even higher, molecules separate even farther, and longer wavelength audio frequencies disappear. Finally, at about 160 km up, no acoustic wavelengths within the range of human hearing are possible, and we've entered what's called the anacoustic zone.

Thus, the great majority of aurorae take place in the anacoustic zone. Even if they did produce sound, it would only be at the lowest altitudes; and even then would be infrasound below the range of human hearing. It's estimated that a tone of 20-100 Hz might be possible at an altitude of 100 km, but here's the rub: that's assuming you're up there with it to hear it. But you're not; you're 100 km away. We can calculate that for that tone to be audible at ground level, it would need to be played at 130 dB.

And that brings us headlong into our final brick wall. Aurorae happen when charged particles interact with the atoms in atmospheric gases. These particles knock electrons into higher energy states, which then collapse back down into their lower energy states, releasing photons. These are really low energy interactions. To actually move a molecule of atmospheric gas would require far higher energy; and to move many millions of them to create even the tiniest sound would require even more.

So we're left with two strikes against the physical possibility of some people being able to hear aurorae: First, aurorae don't have enough energy to make any sound; and second, even if they could, it would be far too low frequency and low volume to be audible.

So is that the end of our exploration into this question?

No. Because it turns out that there is a substantial body of evidence of sounds being detected during aurora events. Not only can some people hear them, but instruments can too. So we've got ourselves a nice little quandary. All of our atmospheric physics shows that aurorae can't make any sound; and yet people are hearing them. A quandary indeed.

So this is where the scientific method comes in. In broad strokes, the scientific method is a way to find an explanation for an observation. So let's start with that observation. Who is hearing these aurora sounds, what exactly are they hearing? Because the very first thing we have to do is establish that there are sounds being heard. As Hyman's Categorical Imperative warns: Do not try to explain something until you are sure there is something to be explained.

Most (but not all) of these observations are anecdotal. People who live where aurorae are common report that they sometimes hear faint popping, hissing, buzzing, or crackling sounds — and they say they're coming from the sky. So the first question I wonder about is whether these sounds actually are coming from the sky, or from some other earthbound source. Let's run through some options:

• The sounds are illusory or imaginary. No doubt some are, but reports are numerous enough that this is almost certainly insufficient to explain them all — or even a significant percentage of them.
  
  
• The sounds are not audio but radio, and are being picked up the same way some people can hear radio. This is not an acceptable explanation. We went into the science of the microwave auditory effect in episode #761. This would be orders of magnitude too weak to produce the high frequency thermal expansion and contraction required to produce sound.
  
  
• The sounds are coming from the tips of the trees. Although St. Elmo's Fire does produce the kinds of noises described, it requires a far stronger electrical field than the rain of charged particles responsible for aurorae can produce. It is only seen during much higher current events like thunderstorms.
  
  
• The sounds could be mechanical noises from frost and ice, unrelated to the aurorae. This is always a possibility, and would need to be tested for and excluded. But it does seem improbable that people who live in these conditions would single out the aurora sounds as something heard only during these geomagnetic storms.
  
  
• The sounds could be coming from anything and anywhere else. No doubt many people have heard sounds during aurora events that had nothing to do with the aurora, but they made the connection because they were primed to. These have to be excluded from the observational data set, if possible.

Many of us may not have even heard of this phenomenon before, and much of the reason for that may be that it occupies an odd vacuum between two otherwise-unrelated scientific fields: acoustics and atmospheric physics. Consequently there are very few people studying it. If you search online, one of these few people is British astrophysicist Carolin Crawford.

Crawford stresses that trying to figure out this problem is more of a hobby for her, and she has come up with a conjecture that she is the first to admit is highly speculative. In her scenario, the noise accompanies the aurora, but is not generated by it. These geomagnetic storms cause minute fluctuations in the Earth's magnetosphere. If there are ferromagnetic materials nearby, for example a metal fence, it might react to these changes in the magnetic field and vibrate in such a way as to cause the noise. You've experienced this before if you've ever had an MRI. Those loud banging sounds are coming from the physical structure of the machine itself being pushed and pulled by the insanely strong electromagnets inside it.

However, the fluctuations of the magnetosphere during a geomagnetic storm are far, far smaller and slower than those in an MRI machine. The Earth's magnetic field is also much too weak to move things about like a metal fence. The needle of a compass is about all it can manage.

Crawford acknowledges all of these flaws in her conjecture, and if you want to know more, she will point you instead to the man doing most of the research and publishing in the field. He is a Finnish professor emeritus of acoustics, Unto Laine. Dr. Laine had gotten some early recordings of these aurora sounds back in 2004. Here is what he got:

He refers to these sounds as "clapping." But he wanted to make sure where they were coming from. So Laine and his colleagues set up arrays of microphones allowing them to triangulate the source of any sounds that might be recorded. If the sound was coming from somewhere on the ground, it was determined to be unrelated and was excluded; but if it was coming from the sky, it was recorded and characterized.

In September 2011, near the village of Fiskars in southern Finland, they collected the first recordings of aurora sounds using this triangulation technique to precisely locate the origin of the sounds. And where was it? Were the clapping sounds coming from 100 km up in the sky? Nope; try 1500 times closer. Laine's claps originated in the air only some 60-70 meters up.

After much work, Laine formed a hypothesis which he presented in 2016. It depends on local weather conditions: a clear, calm night with very low temperatures. Under such conditions, an inversion layer is likely to form: warm air sitting above colder air, forming a sort of lid that prevents vertical air movement. If there is a breeze, the inversion layer is quickly dissipated and these conditions will not occur. Atmospheric data from the Finnish Meteorological Institute confirmed that these layers existed at the times and places Laine acquired his audio recordings — and the inversion layer was at that crucial height of 60-70 meters.

So Laine's Inversion Layer Hypothesis checks three interesting boxes that we discussed before:

1. The sounds are not coming from the aurorae, as eyewitnesses have long believed. Rather they are coming from the same source that causes the aurorae: a rain of charged particles during geomagnetic storms.
   
   
2. The very low energy of an aurora event, insufficient to produce sound, is not a problem in this scenario. The energy is coming instead from electrical charges that have been accumulating in the inversion layer throughout the day and night.
   
   
3. The great distance of the aurora — 100 km away — is not a problem because the sound is coming from so much closer, only 60-70 m, requiring far less energy.

You might wonder why we don't see visible sparks or tiny lightning bolts, if Laine's explanation for the sounds is correct. The reason we don't see anything is that visible sparks like lightning are arc discharges, whereas these would be corona discharges. But unlike St. Elmo's fire, where the corona discharge is concentrated around a physical point so much that it becomes visible as a faint blue glow, these discharges are distributed throughout the inversion layer and lack sufficient energy and current density to produce a visible display.

And unfortunately, that's where we have to leave this story. We do not have any definitive proof. What we have is a series of ideas that fit well together and provide a strong candidate explanation for aurora sounds, but no proof that's how they're being generated. Laine's Inversion Layer Hypothesis remains a hypothesis, though it is unquestionably the leading explanation at this time. Interestingly, if true, it would mean we would no longer be able to call these sounds aurora sounds. They would be geomagnetic storm sounds. Their manifestation correlates with the aurora, but is not caused by it.

And I've still never gotten to either see or hear a decent aurora.

By Brian Dunning
Follow @BrianDunning