Saturn's Decagon Drifts 0.04 Seconds A Year. The Hexagon Drifts The Same.
The new ten-sided wave at the south pole and the hexagon 120,000 kilometres away are keeping one clock — 10h 39m 24s — that matches neither pole's cloud rotation, neither radio period, and nothing else on the planet.
A months-long reporting project. Documents cited below are held in The Vault and available to readers.
TUCSON, United States — Saturn's south pole has a ten-sided wave in it, and it is keeping the same time as the six-sided one at the north.
Both patterns turn on a period of 10 hours, 39 minutes and 24 seconds. Both have moved by about four hundredths of a second a year for as long as anyone has been able to measure them.
Nothing else on that planet keeps that time. Nothing else on that planet keeps time at all.
What was found this week, and what was already known
A persistent ten-sided wave in the cloud tops above Saturn's south pole was described this week and is genuinely new. It sits at about 87 degrees south, it has held its shape through four years of imaging, and it is the first polar polygon anyone has resolved outside the northern hemisphere.
The northern hexagon has been known since Voyager flew past in November 1980. It is roughly 30,000 kilometres across, it has six sides that stay six sides, and it has outlived every explanation offered for why it should not.
The two are 120,000 kilometres apart, through the whole body of the planet.
Why we went to the raw frames
The published discovery images are pretty and they are useless for this. A vertex fit needs the frame, the exposure time, the pointing and the reduction pipeline, and press imagery has been through a pipeline chosen to make it look like something.
So we bought the raw archive and paid two people to fit a rotation period to the polygon vertices independently, neither told that the other existed.
The method is unglamorous. Find each vertex in each frame. Convert to planetographic longitude. Plot longitude against time. Fit a line. The slope is a rotation rate and the reciprocal is a period.
The number
Six fits, at six epochs, across forty-five years.
The hexagon: 10h 39m 22.4s in 1980, 23.4s in 2007, 24.0s in 2016, 24.2s this August.
The decagon, which has only two usable epochs: 24.0s in March 2019 and 24.3s this August.
The whole spread is 1.9 seconds. Over the same forty-five years Saturn's radio period — the closest thing the planet has ever had to a published rotation rate — moved by 402 seconds, and moved by different amounts at each pole, which is the reason nobody quotes it any more. The equatorial jet moved by more than three thousand.
The thing a fluid does, and the thing it does not do
A polygon in a rotating fluid is ordinary. A shear layer between a fast inner vortex and a slower outer flow rolls up into a wave with a whole number of sides, and the number depends on how hard the shear is driven. It happens in a tank of water on a turntable and it happens on a gas giant.
None of that is in dispute here and it never has been.
What a jet-driven polygon does not do is keep a period. It should wander with the jet that makes it, and Saturn's jets wander — seasonally, and by a great deal more than two seconds.
Two of them, at opposite ends of the planet, are not wandering, and they are not wandering by the same amount.
What we could not fill in
Between March 2019 and January 2024 nobody imaged the southern polar region at anything like the cadence a vertex fit requires.
That is not a gap in the record because someone removed something. It is a gap because the observing time went elsewhere, which is the ordinary and boring reason for almost every gap in every archive this newspaper has ever worked in.
It still means the decagon could have lost a side and grown one back while we were not watching, and this desk would have no way of telling.
The 1980 problem
Our own drift figure — the 0.04 seconds a year in the headline — depends on the earliest measurement, and the earliest measurement is the worst one.
Voyager gave four hours and eleven minutes of usable imaging. The fit that comes out of it carries an error of plus or minus 4.1 seconds, which is four times the entire spread of the five later fits put together.
One of our two processors would throw it out. If he is right, the honest statement of this story shrinks to the two polygons agree with each other and have not moved since 2007, which is still the finding, and is a smaller finding than our headline.
We have kept the 1980 epoch, marked it on the plate, and said this here rather than in a footnote.
What this desk takes from it
That the interesting question about Saturn is no longer how many sides a polar wave has.
It is why two of them, in different hemispheres, in different seasons, driven by different jets, are being told the same time by something neither of them can see — and why the only candidate for the thing doing the telling has never been observed either.
Both reductions are published with this story, along with every frame we used and every frame we threw away. If a reader can break the lock, this newspaper will print that on the front page.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
We ignored the published images and went to the raw frames, then paid two processors who did not know of each other to fit a rotation period to the polygon vertices at every epoch where the cadence allows it. Both fits, both rejection lists and the frame identifiers are published with this story so that a third party can break them.
Who we spoke to
- Dr. Ottilie Marchetti-Banda, Planetary imaging specialist, Padua. Commissioned to re-reduce the raw polar frames from archive; her working published in full June–August 2026 Returned 10h 39m 24.2s for the hexagon and 24.3s for the decagon. Says the agreement surprised her and that she re-ran it three times before sending it.
- Dr. Kwabena Lindqvist-Achebe, Second processor, Cape Town. Given the same frames and the same brief without sight of the first result, and without being told there was one July–August 2026 Returned 24.1s and 24.4s. Agrees on the lock. Disagrees that the 1980 epoch should be in the fit at all.
- Prof. Imogen Farrow-Adisa, Planetary fluid dynamicist, Reykjavík. Interviewed by video call three times; read the full draft and both reductions August 2026 Accepts the measurement without reservation and rejects the inference. Her reply is printed unedited and is the best argument against this story.
- Raw polar imaging, 1980–2026, Four instruments, 11,402 frames. Pulled from public mission archives; the 2026 southern frames bought at the archive's commercial rate, $2,900 November 1980 to August 2026 Every frame used, and every frame rejected, is listed with its identifier in the file published with this story.
What we could not confirm
- What is under the cloud deck. Nothing has ever imaged below about three bars at either pole. Everything in this story is measured on the top of an atmosphere, and a period is not a picture.
- Whether the 1980 epoch belongs in the fit. It rests on four hours and eleven minutes of Voyager imaging and carries an error of ±4.1 seconds — four times the entire spread of the five later fits. One of our two processors would drop it. We have kept it and marked it, and if it goes the drift figure in our headline goes with it.
- What the south pole did between March 2019 and January 2024. Nobody imaged it at the cadence a vertex fit needs. The decagon could have lost a side and grown it back in that window and we would not know.
How Others Covered This
The same events, as reported elsewhere on the same day. We list what each outlet had that we did not, as well as what we had that they did not — including where we come off worse. Why we print this.
-
The Meridian TelegraphAstonishing Ten-Sided Storm Found Swirling Over Saturn
Ran the discovery, the press image and a scientist calling it beautiful, inside four hours.
Had that we did not
The wave, its ten sides, and the fact that it is genuinely new.
Left out
That the same team's own supplementary table gives a rotation period, and that the number in it is the northern one.
-
The Hollow PostTWO ARTIFICIAL STRUCTURES ARE BROADCASTING FROM INSIDE SATURN
Read the shared period as two machines under the cloud deck, running since before the solar system cooled.
Had that we did not
The lock. It is there and this newspaper found the same number twice.
Left out
That a shared period is a shared period and not a transmitter, that nothing has ever seen below three bars at either pole, and that the paper's own sources say so on the record.
-
They Buriedthis newspaperSaturn's Decagon Drifts 0.04 Seconds A Year. The Hexagon Drifts The Same.
Ignored the press image, bought the raw frames, and fitted a period to both poles independently.
Had that we did not
Both reductions in full, the 2019–2024 cadence gap, and the 1980 error bar that carries the whole baseline.
Left out
The hexagon's period has been in the published literature since 2007 and the southern number has been in a supplementary table since March 2019. Two spreadsheets, one subtraction. This desk needed a wire item to do it. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Prof. Imogen Farrow-Adisa, planetary fluid dynamicist Sent both reductions and the full draft; given two weeks. Asked to answer at length rather than in a quote. Printed unedited.
Your measurement is good and I am going to say so first, because I am about to spend six paragraphs disagreeing with you and I do not want that read as a dismissal.
You have shown that two polygonal jets at opposite poles of Saturn share a rotation period to within a few tenths of a second, and that the period has moved by under two seconds in forty-five years while every other clock we have on that planet has moved by minutes. I have checked your processors' working. It holds.
Now. A polygon in a rotating fluid is not exotic. Put a tank of water on a turntable, spin the bottom faster than the walls, and the shear layer rolls up into a polygon — three sides, six sides, ten sides, depending on how hard you drive it. Undergraduates do this. Saturn is a very large tank.
What is not ordinary, and what you have correctly identified, is that the pattern keeps time. A jet-driven polygon should wander with its jet, and Saturn's jets wander. Yours does not wander. That is a real result and you should be pleased with it.
But the explanation for a pattern that keeps better time than the atmosphere above it is a forcing from below, and the forcing we already have a name for is a deep-seated convective column — a rotating structure in the molecular envelope, thousands of kilometres down, imposing its own period on everything stacked above it. This is not a rescue I have invented for your benefit. It is in the literature and it predates your story by twenty years.
You have found a metronome. You have not found a machine, and the distance between those two words is the entire content of the disagreement between us.
And here is the part your standards editor will want, so I will save her the trouble of asking. Nobody has ever imaged a convective column at Saturn. The models that produce one need a boundary at depth to anchor it, and Saturn does not obviously have one — it grades from gas to liquid to whatever metallic hydrogen is without a surface anywhere. I am asking you to prefer my unobserved thing to your unobserved thing, and my only real advantage is that mine has been unobserved for longer.
Print that. I would rather your readers had it from me than found it themselves.
Published unedited under our right-of-reply guarantee.
How was this story?
We publish the result, whatever it is. Reader verdicts appear on the front page and in our newsroom metrics.
Readers' Letters 0
Printed at once under the name you give and read by the desk afterwards; anything unfit is removed, with a note saying so, and nothing else is ever deleted — only corrected. Letters that changed something in the story carry a mark saying so, and there are 5 of those across the archive.