Four Dynamicists Ran It Backwards. The Orbit Is Still Bounded At 12,940 Years.
Two of the four were told they were fitting a piece of launch debris. Two were told nothing at all. The primed pair and the blind pair landed in the same millennium, and nobody's run reached a rocket.
Analysis is written by news reporters and interprets facts they have verified. It is not opinion, and it takes no position on what should be done.
PADUA, Italy — Two of the four dynamicists were told they were fitting a piece of launch debris. Two were told nothing at all. All four stopped in the same millennium.
The object is ANALYST 4C-0117, tracked 1,147 times in nine years by a national space-surveillance network and never issued a catalogue number, as this newspaper reports separately today.
We took those 1,147 released observation summaries, added 2,412 amateur optical positions going back to 1994, and paid four people on three continents to run the orbit backwards, separately, with no contact between them.
They come back at 12,781 years, 12,940, 12,948 and 13,091. The spread is 310 years — 2.4 per cent.
What "bounded at 12,940 years" means
Back-integration is not a telescope pointed at the past. It is a solution wound backwards, and every year you wind it the uncertainty grows.
We set the stopping rule before any run began, and we set it in metres rather than in judgement: a run stops at the year its one-sigma position uncertainty exceeds 470 kilometres, which is the width of the orbit itself. Past that point you are no longer tracking an orbit; you are describing a region of space that contains one.
Every run reaches that year with the orbit still bounded — never hyperbolic, never grazing the atmosphere, never escaping through a lunar encounter. Nothing fails. The arithmetic simply runs out of resolution while the object is still going round.
The ratio that holds it up
The reason the runs survive that far is a single number, and it comes from the object's own behaviour rather than from any model.
Across nine years including a full solar maximum, the perigee altitude has not moved outside 1.1 kilometres. An object at 731 kilometres feels the atmosphere expand and contract with the solar cycle, and what it does about that fixes the ratio of its cross-sectional area to its mass.
Here that ratio is at most 0.0009 square metres per kilogram.
A thermal blanket sits at about 2.4. A fairing panel, 0.31. A spent upper stage, one of the densest things we abandon in orbit, 0.0071 — eight times too high.
Nought point nought nought nought nine is a solid aluminium sphere about sixty-two centimetres across, or a steel one the size of a grapefruit. It is not a fragment, not a shroud, not a panel, and nothing with the surface-to-weight of manufactured debris gets near it.
Feed a light object into these integrations and they come apart within a few centuries, because radiation pressure dominates and the model uncertainty compounds. Feed in this one and they do not.
What the priming did, and did not do
Dr. Yusuf Aranda-Mbeki in Cape Town and Prof. Ilaria Venturoli-Grimshaw in Padua were told the object was launch debris of unrecorded origin. Dr. Song-hee Baptiste-Kwan in Montréal and Dr. Fenella Okonjo-Strand in Uppsala were told only that the positions were real and that they should not try to identify the object.
The two who were told something bracket the result: 12,781 and 13,091, the shortest and the longest of the four. The two who were told nothing sit eight years apart.
That is a pattern. With two people in each group it is not a finding, and we will not dress it as one.
What the priming did not do is move anybody toward a rocket. Both primed runs were looking for a launch and neither found one, and Venturoli-Grimshaw, who is angrier with this newspaper than anyone else quoted in it, re-ran her fit after being debriefed and moved 189 years closer to the others.
What the number is not
It is not a birthday. Venturoli-Grimshaw's objection, printed unedited below, is that an orbit does not predate anything, and that the honest form of the sentence is consistent with having existed then.
She is right about the grammar and we have left our own sentence standing above hers rather than quietly swapping it, because the correction belongs to her.
The finding, in her form and in ours, is this. An object presently in Earth orbit has a motion that runs back twelve thousand nine hundred and forty years without breaking, on four independent derivations, published with their working — and the first object anybody launched went up sixty-nine years ago.
The amateur half of that evidence is the part we expect to be attacked, and readers should know it has held up before: our vault's PX-2191 rests on two private notebooks kept by an air-traffic controller and a boiler engineer, and their arithmetic beat a defence ministry's by twenty-six hours.
Thirty-one observers, nineteen countries, thirty-two years. Nobody paid them either.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
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.
-
Signal & LedgerAn Area-To-Mass Ratio Below A Hundredth: What 4C-0117 Means For Conjunction Screening
Read the finding as an operational problem for anyone flying through 731 kilometres, and read it well.
Had that we did not
The ratio, its derivation from non-decay, and a clear explanation of why a compact dense body is harder to lose track of than a light one.
Left out
The date. Their piece runs to 2,900 words on a number that implies an object in Earth orbit before agriculture and does not mention the year once.
-
The Hollow PostTwelve Thousand Years: The Number The Catalogue Sat On
Assumed the cataloguing office had run this integration, obtained this result, and suppressed it.
Had that we did not
The 12,940 figure, correctly, and our spread, correctly.
Left out
That we found no evidence the office has ever run a back-integration on this object, and that the summaries it released make the run possible for anybody with a laptop and eleven days.
-
They Buriedthis newspaperFour Dynamicists Ran It Backwards. The Orbit Is Still Bounded At 12,940 Years.
Paid four people on three continents to do the same arithmetic separately, misled two of them about the object, and published all four working files whether they agreed or not.
Had that we did not
Every run, every disagreement, the priming we used, and the objection of the dynamicist who most resents having been primed.
Left out
We deceived two paid experts about the provenance of the data and only told them afterwards. We think the design was sound. We also notice that we designed it, approved it and are now reporting on it, and that no one outside this building saw the protocol before it ran. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Prof. Ilaria Venturoli-Grimshaw, orbital dynamicist, Padua; author of run D debriefed by video call on 2 August 2026 and sent the full draft, all four working files and the figure the same day. Replied 7 August 2026. Printed in full and unedited, at her insistence and against our sub-editor's preference for cuts.
You lied to me about the provenance and you did not need to. My number would have been the same. I want that sentence at the top because everything else I am going to say is a complaint, and I do not want the complaint mistaken for a retraction.
Here is what your deception actually bought you. It did not remove my expectations; it replaced them with yours. A person told 'launch debris' does not become neutral. She becomes a person fitting launch debris, and she reaches for the models that fit launch debris — in my case a solar-radiation-pressure treatment calibrated on aluminium fragments, which is exactly the wrong instrument for a body with the ratio this one turns out to have. My run is the longest of the four and I believe that is why. You did not de-bias me. You biased me in a direction you chose and then published the result as though it had emerged from nowhere.
If you wanted a blind analysis you should have run one properly: a protocol registered before the data moved, a fifth participant who knew the design and audited it, and salted controls — three fake objects among the real one, so that agreement means something. You had the money. You paid us handsomely and you rented compute for a fortnight. What you did not do was spend anything on the possibility that we would all be wrong together, and four people arriving at the same wrong answer is not a rarer event than one person arriving at it.
Now the part you will like less than the part above.
The result stands. I have re-run my fit with a radiation-pressure model appropriate to the ratio and it moves my horizon from 13,091 years to 12,902. It moves toward the others. It does not move toward zero, it does not move toward 1957, and there is no assumption available to me — not the amateur astrometry, not the geopotential order, not the lunar ephemeris, not the drag model at any plausible density — that brings a bounded, non-decaying solution anywhere near the space age. I have spent three weeks trying to break it, partly out of temper, and I could not.
What I will not sign is your sentence. You write that the orbit 'predates every rocket by roughly twelve thousand nine hundred years'. The orbit does not predate anything. An orbit is not an object with a birthday; it is a description of a motion, and a back-integration that survives to a given year tells you that the motion is consistent with having existed then. Consistent with. Those two words are doing more work than your headline allows them to do, and you should have hired a fifth person to say so before I did.
I am not withdrawing my figure and I do not want another commission from you.
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.