GB3MBA · 50.408 MHz · forward scatter
Radio receivers around the UK listen to a single beacon they cannot normally hear. When a meteor burns up it leaves a trail of ionised air that reflects the signal for a fraction of a second. Everything below is counted from those reflections.
The window
The Earth ploughs forward along its orbit, so the morning side sweeps up meteors head-on while the evening side has to be caught from behind. The rate should therefore peak in the small hours and bottom out in the early evening. It does.
The same picture for each receiver on its own. They ought to agree: they are all within about 500 km of one another and they are all looking at the same sky, so they should get busy at roughly the same time of day. Where one disagrees, that usually means something local is drowning out the meteors — interference rather than a different sky.
A receiver marked uncertain has not been recording long enough, or cleanly enough, for us to say when its busiest hour really is. It is not a criticism of the receiver — a few more clear nights usually settles it.
| Site | Nights | Hours recorded | Echoes counted | Per hour | Clock locked |
|---|
It is a fair thing to ask. A blip on a radio could be an aircraft, a passing car, or a noisy phone charger. So we checked our detections against cameras that were watching the same sky.
The Global Meteor Network films meteors from the ground and publishes where each one was and how high. A radio receiver can only hear a meteor sitting in one particular spot relative to the beacon and itself — rather like a mirror, which only bounces sunlight into your eye from one angle. So we asked a simple question: when a camera saw a meteor in the right spot for one of our receivers, was that receiver more likely to have heard it?
It was — about 1.6 times more likely. And it behaved exactly as it should, which is the part that convinces us.
Meteors that turn up on their own, from random directions, showed the effect most strongly: 2.2 times. Meteors belonging to a shower all travel in the same direction, which makes the “right spot” much less meaningful for them — so the effect should mostly disappear for those. It did: 1.25 times. We expected that before we looked, and getting a prediction right is far better evidence than a single number on its own.
This check uses none of our own clocks or settings — the cameras belong to somebody else entirely. It comes from 47,643 pairings of a filmed meteor with one of our receivers, 1,323 of which we also heard, over the winter of 2025–26. It is a one-off result and does not change from night to night.
A meteor leaves a thin tube of electrically charged air behind it, and that is what reflects the beacon to us. The grain of rock itself is long gone by the time we hear anything, and it is far too small and far too fast for us to catch this way. So when this page says “echoes”, it means the trails.
We have 31 occasions where three or more receivers heard the same meteor — two of them on four receivers — and every one is backed by a camera that filmed that meteor at that moment. Repeat the same search eleven minutes later and it finds almost nothing, so these are not coincidences of timing. The network does do the thing it was built to do.
The prize is the step after that. Three listening posts ought to be enough to reconstruct the meteor's path through the sky — where it came from, how fast it was going, and where any surviving piece landed. Our attempts at that are still landing about 250 km away from where the cameras say the meteor actually was, which is not good enough to publish as a path.
The difficulty is what we are hearing. The trail hangs in the air more or less still, so all three receivers hear much the same steady note, and a steady note carries very little information about motion. The meteor itself would be far more useful: it moves so fast that its note audibly slides, and that slide is what the maths needs.
We can hear those, and we have caught a few on two receivers at once. But they are faint and brief, and any one receiver picks up only about one in twenty of them. At that rate, catching the same one on three receivers is something we would expect to happen roughly once in every hundred and twenty — and we have only found about forty so far. So the honest answer is that we have not yet looked at enough of them, not that it cannot happen.
Two things would change that. Hearing more of them in the first place, which means getting better at telling a real one from interference. And more receivers — every extra one is another chance at each meteor, and each hears its own patch of sky, so where a receiver sits matters as much as how many there are.
Two things would move this on more than anything else, and neither needs any expertise in meteors.
Every picture a person marks up teaches us to tell a real echo from interference, which is the single biggest limit on everything above. It takes a couple of minutes and no prior knowledge. Have a go →
If you are somewhere reasonably quiet for radio and can spare a small aerial, a corner of a shelf and a domestic internet connection, you could become one of the listening posts. Where a receiver sits matters more than how many we have — the right location fills a gap in the sky that none of the present ones can see. We would genuinely like to hear from you; there is a contact box at the bottom of the main page.