EDIT: Apparently the accepted physics has changed, and we now use the convection model for generating planetary magnetic field, not the dynamo model. Next thing you know people will be telling me Pluto is no longer a planet. (end edit)

Obviously, my first thought was to hit Mars with Ceres, on PPV the mission would pay for itself, sadly the Boffins would never let that happen.

Using ion drives with (H2O or bi-products) as reaction mass, move Ceres to Mars orbit (they are already in nearby orbits) to act as a moon, to restart the magnetic field on mars, so an atmosphere could be maintained indefinitely. Ceres could also act as a resupply base, due to low escape velocity and abundant water (assuming any survives the trip) Mars has abundant water in any case.

I need help with calculating how much thrust for how long would be needed. I know time is relevant. I'd like a time frame within my remaining lifetime say 30 years. Even just some rough numbers and a picture of how it would work would be nice. Then we can have a goal for how many and how powerful our ion engines would need to be.

Ceres is ~1/3 water so that's our maximum reaction mass. And I know it's going to take a bunch of Ion drives using current tech.

I'm new at orbital mechanics, so which direction to fire the ions in order to drop Ceres to a lower orbit and speed it up? Eventually it would need to achieve near Mars orbit for capture. Google doesn't have any nifty info graphics on lowering orbits, everything covers getting into orbit from earth or transferring to a higher orbit.

Ceres has a much smaller ratio of mass- Mars:Ceres than our own Earth:moon, so Ceres would need to initially orbit very close to Mars in order to get the dynamo spinning again. Later on Ceres could be moved farther out again once the proper field strength is achieved.

Ceres is spinning rather fast (~9 hour day), so ion drives would need to be placed around the planet and fire as the dwarf planet rotates. Placing them at the equator on gimbals would allow full control for orbital corrections.

Orbital inclination would need to be adjusted as well, most likely.

I feel transferring Ceres to act as a moon for Mars is the best solution, forget about using nukes to spin up the core, or giant magnets... give it a moon and forget about it for the next billion years.

That last problem is there may be a liquid ocean deep under the surface, so too much thrust would shift the core, the net force would need to be balanced against that, (if there is a liquid layer), so as not to cause catastrophic imbalance.

I know it's a big problem, but the reward would be a maintenance free magnetic field on Mars, leading to a rich, possibly breathable, atmosphere (with proper mixture).

  • $\begingroup$ Around 10km deep below the surface, the Martian pressure and temperature is not very far from being breathable by us. But I see more realistic a large cloud of orientable solar sails on various orbit around the Mars. They could serve both as weather controllers and as defense from the solar wind. It would require much lesser effort, than this Ceres transfer (although the cost would be still astronomical). Oxygen could be extracted from the soil in large plants. Water is there enough on the Mars, there is like on an earthian $\endgroup$
    – peterh
    Oct 10, 2018 at 0:37
  • $\begingroup$ desert, but a planetary sized tunnel system could optimize it. $\endgroup$
    – peterh
    Oct 10, 2018 at 0:42
  • $\begingroup$ For a sustainable ecosystem, surface living is a must. You could have forests underground in giant caves with light piped in via fiber optic but underground living is a temporary solution and requires eternal upkeep. $\endgroup$
    – Aerothorn
    Oct 10, 2018 at 6:54
  • $\begingroup$ Using locally mined H2O as a reaction mass, allows us to save money by only needing to move the ion drives and mining gear there, that's a massive cost saver. In theory landing a single ion drive and miner there could eventually change the orbit enough for a mars capture. Time is the limiting factor. Do we want it there in a decade or a century? I picked Ceres because the delta V is lowest of all the possible transfers, but it's mass is (hopefully) high enough to restart differential rotation. $\endgroup$
    – Aerothorn
    Oct 10, 2018 at 7:02
  • 2
    $\begingroup$ I see sheer lack of sense of scale. You can't just throw around bodies with this kind of mass at any reasonable timescale. $\endgroup$ Jan 23, 2022 at 13:30

3 Answers 3


We're a long way away from being able to do this.

Is it theoretically possible? Let's see if we have enough delta-V: Mass is 1021 kg, with 1/3 available as propellant.

Using the Rocket Equation: $$ \Delta v = I_{sp} \ g \ ln\left(\frac{m_i}{m_f}\right) $$

Isp = 104

? = 105 * ln (1021/7*1020)

? = 35,000

Delta-V to get from Ceres to Mars is in the region of 2 km/s, so purely based on this it'd be doable.


  • you've disassembled Ceres to get all the water out, so you're left with a pile of loose rubble instead of a planetoid.
  • converting 3*1020 kg of water into ions will take ungodly amounts of energy, good luck generating that in 30 years.
  • ion engines don't use water, they use inert heavy gases, so Isp will drop
  • you'd have to convert a large fraction of Ceres' mass into ion engines to get enough thrust.
  • the biggest industries in the world (e.g. coal mining) move 1012 kg per year. If you moved that entire industry to Ceres, they'd need 10 million years to extract all the water.
  • to move all that equipment to Ceres, we'd need to scale up our rocket industry by a factor of 103, assuming your timetable of 30 years and a need to move 109 kg of equipment.
  • $\begingroup$ "you'd have to convert a large fraction of Ceres' mass into ion engines to get enough thrust." - Is... is there even enough surface area on the surface of Ceres to put the thrusters? How many ion engines is that? Upwards of many tens or hundreds of million? $\endgroup$ Oct 11, 2018 at 19:31
  • $\begingroup$ Great answer by the way though, answers feasibility for sure. $\endgroup$ Oct 11, 2018 at 20:03
  • $\begingroup$ Ceres might not need to be disassembled or mined in a conventional manner, if liquid oceans exists, pipe to surface, electrolysis to obtain ions for thrust, expel ion. Biggest problem is crust wrinkling as mass leaves (balloon deflating). Not as efficient as heavy ions with present tech, but maybe still viable.. And planetary mass will decrease as mass is expelled obviously. $\endgroup$
    – Aerothorn
    Oct 12, 2018 at 6:58
  • $\begingroup$ Even if all that water were conveniently available, you'd have to process it (pump it out, ionize). For a Fermi estimate like this one, I'm going to put our capacity to do that at 10<sup>12</sup> kg per year as well, and the answer won't change. $\endgroup$
    – Hobbes
    Oct 12, 2018 at 7:35

For example, from the Moon, the characteristic time of the atmosphere loss would be some ten millions of years*. Mars is further from the Sun, and has a higher gravity. This would be enough for us. The problem is that there is no way to give that atmosphere to it.

From the Ceres, and from small asteroids it the popular belief that they are small. And yes, they are small - compared to the planets. In fact, the mass of the Ceres is $\approx 10^{21}$ kg ref. It is around the mass of the oceans of the Earthref.

I didn't find a reference, what would be the timeline of the build up / build down of the planetary magnetic field due to the tidal effect of a moon, but it is probably not instantenous.

If the Mars has a magnetic field, it still won't have an atmosphere. To have an atmosphere, from somewhere that gas should be transferred there.

  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – called2voyage
    Oct 11, 2018 at 14:41
  • $\begingroup$ Why not convert the CO2 ice and water ice that is already abundant near the surface? With lower pressure would not a higher O2 content be preferable? IF that is not enough, we are back to bombardment by comets/high water content planetoids $\endgroup$
    – Aerothorn
    Dec 13, 2018 at 17:09
  • $\begingroup$ @Aerothorn It would work well. The main problem is that politicians are focused to the human society of the today, all of their goals and games are on the Earth. Even the few space exploration what we (Humanity) have done, had in fact earthly goals (to win the Cold War & to show to the people of the USA that still they are the best country). I think it could be done easily, simply big, nuclear or solar-powered machines should melt the soil and electrolyze it. $\endgroup$
    – peterh
    Dec 13, 2018 at 17:31
  • $\begingroup$ @Aerothorn Most of it are different metal oxides, so we would have a lot of iron and aluminium, a lot of glass (silicon dioxide can't be electrolyzed) - on a planet having a fresh oxygen athmosphere. What I see harder problem: there is very few hydrogen and nitrogen on both the Moon and the Mars. Not even in the soil. So both will be desert, there will be no seas - a planetary-level irrigation system will be needed. Nitrogen-containing minerals will need to be mined and processed to fertilizers. $\endgroup$
    – peterh
    Dec 13, 2018 at 17:34
  • $\begingroup$ Water is abundant and there is enough water to create an atmosphere are two vastly different things $\endgroup$
    – Antzi
    Sep 12, 2020 at 16:28

Ceres orbital speed is 17km/sec

To make it intercept, you need to do a 5km/sec burn according to http://www.projectrho.com/public_html/rocket/appmissiontable.php

Ceres is 9x1020 kg

In other words you need to make Ceres lose 1x1030J

Of course, we should reduce that number because you’re using the mass of Ceres itself as reaction mass. Since you propose using up to ⅓ of the mass, and I’m too lazy to calculate let’s say 1029J

In 2013 the world used 520 J of energy

In other words, all the energy used in the world is not enough to move Ceres in 10.000 years... and we didn’t even begin talking about transporting that energy there, efficiency losses, ...

No, deorbiting Ceres is impossible.

  • $\begingroup$ @uhoh I guesstimated it would take 10km/sec to lower the périgée enough for an intercept, I don’t know if there would be a smarter orbit or if the number is correct $\endgroup$
    – Antzi
    Sep 13, 2020 at 2:55
  • $\begingroup$ Edited because I used km/h instead of km/sec $\endgroup$
    – Antzi
    Sep 13, 2020 at 2:59
  • $\begingroup$ oh, got it, yes that makes sense. That was my Sunday morning pre-coffee brain... $\endgroup$
    – uhoh
    Sep 13, 2020 at 3:36
  • $\begingroup$ For an object with a circular orbit at 2.77 AU the initial speed is about 17.9 km/s. For an elliptical orbit that went between 2.77 and 1.0 AU the semimajor axis is 1.88 AU. That orbit would have an apopapsis speed of 13.0 km/s. I think if Ceres slowed to 7 km/s it would reach 0.2 AU at periapsis? $\endgroup$
    – uhoh
    Sep 13, 2020 at 3:54
  • 1
    $\begingroup$ @ChristopherJamesHuff so it's even more impossible :D $\endgroup$
    – Antzi
    Sep 14, 2020 at 2:17

Your Answer

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

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