This tweet says in part:

Earth's potential new minimoon, 2020 SO may be the Surveyor 2 Centaur rocket body, launched in September 1966. Integrating backwards shows 2020 SO2 to also be orbiting Earth in September 1966.

and links to this Orbit Simulator setup which exhibits said behavior. I guess that you have to speed it up to move forward a half-century but slow down once 2020 starts:

In fact this is the third pass of that simulated orbit through Earth's neighborhood (association with the Sun-Earth Lagrange points) since the 1960's.

Suppose there was support to find out if 2020 SO is the rocket body or not, and if so, do some in situ studies of long term exposure of (1960's) aerospace materials to deep space and perhaps bring back a few samples to see what the cat brought home.


  1. Could such a mission be thrown together while it's still "in the area", for of the order of $200 million? It would need appropriate maneuverability to get up close and take images or other measurements (e.g. a stripped-down electron microscope)
  2. This seems harder; if it had a mechanism to remove a chunk sample, could it then remain in Earth's neighborhood without needing a big delta-v so that a later recovery mission could return the sample back to Earth? Or could it be given a small nudge that puts it into a longer-lasting orbit, something to buy a few years for a better-planned mission to then go and measure and perhaps recover a sample?
  • 1
    $\begingroup$ Does "in the area" constrain launch date or the rendezvous date of the mission? $\endgroup$ Dec 4, 2020 at 16:16
  • $\begingroup$ @SE-stopfiringthegoodguys yes it sure does! Without propulsive intervention it only stays cis-lunar for a few months. $\endgroup$
    – uhoh
    Dec 4, 2020 at 21:19
  • 1
    $\begingroup$ The Discovery Program seems relevant, "a series of small-scale spacecraft designed to proceed from development to flight in under three years for a cost of less than $150 million". (though 13/14 missions exceeded that cost goal, most by a significant amount). $\endgroup$ Dec 5, 2020 at 20:49

1 Answer 1


Generating the JPL HORIZONS ephemerides, the two close encounters (Dec 1, 2020 and Feb 2, 2021), are way to close in time to us to throw anything together. Especially since it has only been observed since September 2020. Even if some rocket stands ready on the launchpad somewhere, we don't have the craft.

That leaves us with heliocentric phasing.

From the velocity vectors of the second close encounter, 2020 SO should have around ~100m/s of Vinf, putting it into a new heliocentric orbit with a shorter orbital period than the Earth.

One orbit after leaving the Earth, the target will arrive back at the same point. The Earth will not be there though, missing by over a week. But this is the point in the orbit we will want to launch a "catch-up probe" into an even shorter orbit, over several orbits undoing the drift the target has built up.

This makes the available development time a whole number of years. Using the 3 year development to flight goal of the Discovery Program as a baseline, it seems unrealistic to be able to cut enough corners to get it down to 2 years.

So there's 3 years of drift to catch up to when our probe finally launches. For phasing of very close orbits, the phasing orbit period can be approximated as being linearly dependent on the initial Vinf. A two-year rendezvous is then only a matter of a 250m/s Vinf, which is a single digit delta-v cost on top of an Earth escape due to the Oberth effect, and another 150m/s delta-v deep space burn when arriving at the target. Less than 10% of the probe's mass has to be hydrazine tank to achieve that.

That's a not too wishful 5 year mission plan to achieve the basic goals of part 1).

The relatively simple phasing manoeuvre also makes things pretty flexible: Each extra year of development means 2/3 year of extra coasting required for the same delta-v cost. Not a bad deal. Relatively minor increases in the delta-v budget can further improve the ratio, making the development time dominant. Taking your time to do things properly is then an option.

Part 2) I have more doubts about. Five years down the line, after rendezvous, the drift relative to the Earth has grown substantially. A ~200m/s burn to initiate reverse phasing means the whole sample return has grown into a decade long mission, with considerably more complications than a rushed 3 year development can likely accommodate. Bringing the entire Centaur stage back to park seem even more unrealistic. To achieve that within two or three decades, a substantial amount of hypergolics (at least 200kg) has to accelerated into a heliocentric location.

  • 1
    $\begingroup$ This is my initial thoughts after seeing the trajectory data, making the somewhat optimistic about your proposed mission. I know from experience that spending bounties without getting any answer is frustrating, so here's the outline (feel free to request parts expanded!). In the mean time, Hop David catching an asteroid is relevant reading. $\endgroup$ Dec 9, 2020 at 8:19
  • $\begingroup$ Thanks! This is appreciated :-) comment on the first sentence only (will read further soon): You don't think that if someone asked SpaceX to squeeze in a launch for $100 million providing them the opportunity to demonstrate an emergency on-demand F9 launch they wouldn't go for it, or just that there would't be anything to use as payload, or that it would need a Falcon Heavy to get there in the next months? $\endgroup$
    – uhoh
    Dec 9, 2020 at 13:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.