Adam Hibberd
I have been studying some very unusual celestial bodies, the ‘dark comets’ which are particular sorts of asteroids exhibiting strange inexplicable forces whilst showing absolutely no observable comae or tails as we might see in true comets. These two phenomena are evident signs that outgassing of volatile materials is taking place from the object’s surface, in turn causing a rocket effect, so can we discount this as happening for these dark comets? Absence of osbervational evidence doesn’t necessarily mean absence in reality.
What I’m finding is a convincing pattern emerging, the evidence is accumulating, but let’s put that aside for the moment and talk about the particular dark comet designated ‘1998 KY26’.
The Japanese Aersopace Exploration Agency (JAXA) is sending a mission to this object, the Hayabusa2 spacecraft, arriving in July of 2031 – but what on Earth will it find?
I have found sound and compelling evidence that this so-called ‘dark comet’, hitherto thought to be a natural object and which has several unusual characteristics, may in fact be a defunct Russian Mars probe lost on its way to the red planet way back in 1988.
Let’s discuss what we have from the historical record.
Phobos 1 was launched on 7th July of 1988.
We know that in late August 1988 the Russians uploaded a faulty command to the Phobos 1 probe, on the way to Mars, causing the spacecraft to lose attitude control on 2nd September 1988.
The Russians must then have lost the orbit of the spacecraft, since they requested the European Southern Observatory (ESO) attempt observe the probe based on its last known course. ESO attempted a sighting 20 days after the loss of the probe, but it was not seen at its predicted coordinates leading them to ask ‘had the probe experienced a faulty rocket fire?’
Through my research I discovered that a rocket fire around loss of mission followed by a second much later on, within the thrust envelope of the probe, could just account for the difference in orbits of the two objects – 1998 KY26 and Phobos 1 – opening up the stunning possibility that they could actually be one-and-the-same object.
But there are further pieces of evidence which add weight to this theory. For instance 1998 KY26 is very small at only 11 metres in diameter, just the length of Phobos 1 from one tip of a solar panel to the other.
1998 KY26 is also unusually reflective for an asteroid which agrees with this object being shiny and man-made.
Note also that HUGE fluctuations in apparent magnitude indicate it is spinning and highly elongated. That fits too – the Phobos 1 probe had a high aspect ratio because it was much narrower than its maximum span of 11 metres.
So will the Hayabusa2 spacecraft find the Phobos 1 probe in 2031? We shall have to wait and see.
Go here for the preprint.

For those of you interested, the software I used to reconstruct the possible trajectory that allowed Phobos 1 to morph it into 1998 KY26 was written by me in the ‘C’ programming language and exploiting the REBOUND software library.
REBOUND allows integration of the Phobos 1 trajectory forwards in time using numerical integration techniques and assuming the presence of several gravitating bodies, the key ones in this case being the Sun, Earth and Mars of course (for completeness I modelled all the Solar System planets).
Inputs to the software included the DeltaVs themselves – i.e. the directions and magnitudes of the two velocity increments (or ‘impulses’) and the respective times of their delivery. Output from the software was the displacement between 1998 KY26 and the Phobos 1 probe. This ‘displacement’ was NOT simply the positional discrepancy between the two objects, but a highly significant statistical deviation, based on the known uncertainty in the orbit of 1998 KY26, and taking into account position AND velocity (the so-called ‘Mahalanobis Distance’).
A second software package, NOMAD, as well as a CMA-ES python solver (CMA-ES = Covariance Matrix Adaptation – Evolutionary Strategy) were used to find the combination of inputs leading to the minimum output.
Reassuringly, the total velocity impulse from the solution found by NOMAD and CMA-ES, turned out to be 1.9 km/s, which is just within the Delta-V envelope of the Phobos 1 probe. This probe possessed a thrust module for Mars Orbital Insertion with an impressive momentum-change capability.
Although the solution DeltaV total discovered was 1.9km/s, I am still convinced there is a LOWER DeltaV solution out there yet to be found, I am determined to find it when I have the time.
Look below for Tony Dunn’s simulation of the theoretically feasible trajectory followed by Phobos 1 to make it ‘become’ 1998 KY26.








