Mercury Has No Moon. Something At 912 Kilometres Has A Twelve-Hour Period.
A spacecraft nine weeks from Mercury changed velocity by 1.9 metres a second on a night with no scheduled burn. Two orbit determinations, run blind and separately for this newspaper, agree on what it moved to avoid.
A months-long reporting project. Documents cited below are held in The Vault and available to readers.
BOCHUM, Germany — On the night of 14 July the spacecraft now nine weeks from Mercury changed velocity by 1.90 metres per second.
Nothing was supposed to happen that night. The published cruise plan has no manoeuvre between 2 July and 19 August.
We know it happened because two dish antennas were listening to the spacecraft's own carrier tone, and the tone moved.
What a carrier can tell you
A deep-space probe transmits a narrow, extremely stable radio tone so that the ground can measure its distance and its speed. If the craft accelerates, the tone shifts.
The shift on 14 July was 68 hertz at X-band. That is small, but it is not subtle. What it takes to see it is a large dish, a good clock and somebody willing to sit on the same object for weeks.
Amateurs have been doing this for twenty years. The 20-metre dish at Bochum, run by volunteers, has followed every deep-space mission of the last decade and publishes its recordings the same week.
Hannelore Utz-Bamigboye has been one of those volunteers since 2011.
"We were not looking for a manoeuvre," she said. "We record the carrier because that is the hobby. It was sitting in the residuals the next morning and it was extremely obvious."
This newspaper then hired a second antenna — 25 metres, at a decommissioned commercial earth station in Bavaria, for eleven nights, at €14,900. It heard the same thing on the same night.
Two solutions, run blind
We sent both recordings to two orbital dynamicists who did not know of one another and were told nothing except: fit this.
Dr. Cordula Weisz-Aigbe returned a solution in nine days. Dr. Emeka Lindqvist-Sarr took nineteen. Their figures for the velocity change agree to within 0.03 metres per second.
Each was then asked a second question, in identical words: what shape of manoeuvre is this?
Both came back with collision avoidance. Not a trajectory correction, which is flown along the direction of travel and cheaply; a short push across the track, of the kind flown when something is going to be in the way.
What it moved to avoid
Working back from the geometry, both dynamicists put a body in orbit around the planet.
Weisz-Aigbe: 912 kilometres mean altitude, period 12 hours 4 minutes, retrograde, inclined 78 degrees to Mercury's equator. Lindqvist-Sarr: 908 kilometres, 12 hours 1 minute, retrograde, 77 degrees.
Mercury has no moon. That is not an oversight in the textbooks. An American orbiter spent four years there between 2011 and 2015 and ran dedicated satellite searches, and found nothing.
An object at 912 kilometres retrograde is dynamically stable in that place for a very long time. It is also, at that altitude, lost in the glare of the planet's limb from anywhere on Earth, which is the ordinary and boring reason nobody has picked it out.
The 2012 echo
There is one earlier trace, and we would not have gone looking for it if the two fits had not told us where to look.
In March 2012 a bistatic radar run — transmitting from Goldstone, receiving at Green Bank — returned an echo from the Mercury system at a Doppler offset that does not correspond to the planet's rotation. It sits in published data that anybody can download. At the time it was put down to a sidelobe and nobody wrote it up.
The offset corresponds to a body moving at 2.1 kilometres per second relative to the surface. Both of our fits give 2.1.
We say under Right of Reply, and we say here, that this is the weakest limb of the story.
The answer we were given, and the answer we were given later
On 19 August the operations directorate told this newspaper that the 14 July event was a reaction-wheel desaturation.
That is a real and routine operation. It is also specified in this spacecraft's own public documentation at not more than 0.05 metres per second, and it is flown as a string of small pushes over several hours, not as one push of 1.9 metres per second lasting forty-one seconds.
We put that discrepancy in writing on 21 August. The reply, printed below in full, withdraws the first answer, names it as wrong, and then tells us something nobody has published before: the object has been tracked since 2022.
What this desk takes from it
That the object is there, and that the argument now is only about what it is made of.
Orbit insertion is 9 November. From that date the spacecraft is inside the same volume as the thing it swerved around, carrying a camera, and the directorate has committed in print to releasing the frames unprocessed within seventy-two hours.
We have asked for the raw ranging and Doppler files for the whole of July. The request has not been refused. It has not been granted either.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
We analysed nobody's telemetry. We recorded the spacecraft's own carrier tone from two antennas — one volunteer-run at Bochum, one hired for eleven nights — and paid two orbital dynamicists, separately and each unaware of the other, to fit the residuals and say what shape of manoeuvre they implied. Both raw recordings, both solutions and both invoices are published with this story.
Who we spoke to
- Hannelore Utz-Bamigboye, Volunteer operator, 20-metre dish, Bochum. Interviewed at the antenna and twice afterwards by telephone July–August 2026 Has recorded deep-space carriers as a volunteer since 2011. Publishes every recording.
- Dr. Cordula Weisz-Aigbe, Orbital dynamicist. Commissioned by this newspaper for £6,200; solution and residuals published in full August 2026 Returned a fit in nine days. Puts the object at 912km, retrograde, inclined 78 degrees.
- Dr. Emeka Lindqvist-Sarr, Orbital dynamicist, second and independent. Commissioned separately for £6,200, without contact with Weisz-Aigbe August 2026 Reproduces the manoeuvre and the period. Calls the retrograde inclination ‘the weakest number in my own answer’.
- Operations directorate, the joint Mercury programme, Mission operations. Written questions 11 August; answer received 29 August August 2026 Printed in full under Right of Reply, including a correction of its own first answer to us.
- Bistatic radar archive, March 2012, Transmitted from Goldstone, received at Green Bank. Public data, downloaded at no cost and reprocessed for us March 2012
What we could not confirm
- What the object is. We have a velocity, an altitude, a period, an inclination and a fourteen-year-old radar echo that agrees with all four. That is an orbit. An orbit is a description of a path, not of a thing, and this newspaper has no image, no spectrum and no mass.
- Whether the 2012 echo is the same body. The relative velocities agree to two significant figures. Two significant figures is a coincidence that has not been ruled out, not an identification.
- Why the first answer we were given was wrong. The directorate says a duty officer read a shift log in good faith. We cannot check that, because the shift log is not public and we have not asked for it under any procedure that would compel it.
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 Continental WireEuropean-Japanese Craft On Course For November Arrival At Mercury
Ran the mission's own arrival update and stopped there.
Had that we did not
The insertion date, correctly, and the cruise duration to the day.
Left out
That the spacecraft moved in July. The difference between two successive published ephemerides shows it, and both are free to download.
-
The Meridian TelegraphSOMETHING IS ORBITING MERCURY AND THEY HAVE KNOWN FOR FOUR YEARS
Reported the manoeuvre as proof of a concealed artificial satellite.
Had that we did not
The burn, which happened, and the four years, which the agency has now confirmed.
Left out
That a thing in orbit can be a rock; that the agency says it thinks it is one; and that a camera settles it in nine weeks, which makes the shouting a waste of everybody's November.
-
They Buriedthis newspaperMercury Has No Moon. Something At 912 Kilometres Has A Twelve-Hour Period.
Recorded the carrier ourselves and paid two people, separately, to disagree about it.
Had that we did not
Both solutions in full, including the one that is less sure of the inclination than we are.
Left out
We held the 2012 radar echo for eleven days while we decided whether it strengthened the story or merely decorated it. It decorates it. Printing it under a heading that admits as much is not the same as not printing it. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Vasco Trindade-Kelleher, operations directorate, the joint Mercury programme sent the full draft, both solutions and both recordings; given eighteen days. Replied in writing and asked that it run whole. Printed unedited.
I am going to answer you properly, which means conceding one thing and contesting two.
The concession first. The burn happened, and the answer this directorate gave you on 19 August was wrong. It was not a reaction-wheel desaturation. That answer came from a duty officer reading a shift log at short notice, in good faith, and it was incorrect, and I am not going to let a colleague carry it in your newspaper on my behalf. The manoeuvre of 14 July was a collision-avoidance burn and it was flown for the reason you say it was flown.
Now the two things I contest.
First, you write the sentence ‘Mercury has no moon’ as though we are the ones disputing it. We are not. What we have is a tracked object, catalogued internally since 2022, with an arc we have refined over four years, and a working hypothesis your own dynamicists would arrive at in an afternoon if we handed them that arc: that it is surface material thrown up by an impact and recaptured. That is not exotic. It is a rock in a stable place. We have not announced it because this agency has a rule — I did not write it and I do not much like it — that we announce an object when we can characterise it, and we cannot yet characterise this one.
Second, and this is the one I mind. Your draft uses the word ‘undisclosed’ eleven times. There is a difference between a thing that is being concealed and a thing that is not yet finished, and a reader will not be able to tell from your copy which of the two this is, because your copy does not want them to. You are entitled to think we are too slow. You are not entitled to let the word do the argument for you.
On 9 November the spacecraft enters orbit. In the first eleven days it passes within about 340 kilometres of the object at closest approach. There will be images. I have asked for, and been given, permission to tell you in print that they will be released within seventy-two hours of downlink, unprocessed, to anyone who asks, including you, including if they are embarrassing.
If it is a rock, I would like you to print that as large as you are printing this.
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.