Jumping from this question about terraforming Mars, how would one mitigate the main obstacle, namely Mars's lack of a magnetic field? Are there any hypothetical processes that might restart Mars's magnetic field (assuming it had one in the past)?

  • 7
    $\begingroup$ Perhaps we could hit it with a big enough rock.. $\endgroup$
    – agweber
    Commented Oct 15, 2013 at 0:01
  • 13
    $\begingroup$ @agweber - That's certainly what geologists and astrophysicists think got our own core going (and gave us the Moon). Honestly, when you hear about how much had to have gone right in Earth's early childhood to support life, and how much of it seemingly went very wrong, it's a wonder we're here at all. $\endgroup$
    – KeithS
    Commented Oct 15, 2013 at 0:12
  • 9
    $\begingroup$ "terraforming Mars, how would one mitigate the main obstacle.." I'd say the main obstacle in terraforming Mars is the lower gravity, not the atmosphere (or lack of). And as far as the atmosphere being stripped from Mars goes, I heard it took in excess of 100 Million years. If the inhabitants cannot manage to direct some more comets and other objects down to Mars to replenish it in that time, it seems they would never have had the tech. to leave Earth in the first place. $\endgroup$ Commented Oct 15, 2013 at 18:36
  • 3
    $\begingroup$ If our nuclear waste were enough to melt Mars' core, it'd be hot enough to make Earth uninhabitable too. In reality, we'd need 10^18 times more nuclear waste than we have for @Exchangemaster 's idea to work. $\endgroup$
    – Hobbes
    Commented Sep 6, 2014 at 8:17
  • 4
    $\begingroup$ You don't need to. If you have the ability to terraform Mars in times scales of interest to a civilization (e.g. 10's, even 100's of thousands of years, if you're very patient), then the timescales of the loss of the atmosphere you built are irrelevant to you, which are 100's of millions of years. You can easily replenish as needed. Though your civilization will likely be long gone by the time it is needed. $\endgroup$
    – Mark Adler
    Commented Oct 11, 2018 at 1:52

7 Answers 7


Since the question states that the answer can be "hypothetical":

Since the core of Mars does not have enough heat to start the convection process, we can drill a hole to the solid core and connect them to a source of electricity, and pass a huge current so it heats up ($\textrm{Heat}~=I^2*\textrm{Resistance of core}*\textrm{Time of passage of current}$) the core until the core melts because of the heat, so the convection process becomes self-sustainable.

Another way is to hit it with a huge asteroid (as agweber said in a comment)

  • 4
    $\begingroup$ Instead of re-heating the core, would it be possible to turn the cooling core into a permanent magnet somehow? Like a REALLY powerful magnetic field to align all the atoms? $\endgroup$
    – john3103
    Commented Oct 15, 2013 at 14:01
  • 2
    $\begingroup$ @ john3103 The core of Mars is mostly made of Iron and iron can be magnetized by heating but it's is possible to magnetize iron by non heating methods but they are temporary the magnetic moment decrease with increase time $\endgroup$
    – Hash
    Commented Oct 17, 2013 at 15:54
  • 1
    $\begingroup$ @Hash how could the convection process become self-sustainable? $\endgroup$ Commented Oct 11, 2016 at 16:52
  • 1
    $\begingroup$ Can we just use that electricity meant for heating to just generate some magnetic field, like series of huge electromagnets? Might this be an efficient way to generate magnetic field? $\endgroup$ Commented Oct 11, 2018 at 6:33

There are some very good ideas. This requires a multiple answer approach. This can not be resolved by one method only. I don't believe mass is the issue (if) the iron core is large enough in comparison to the over all mass of the planet. If that is the case then re-starting the mantel is within our current technology to do. We may be able to reach and initiate colonizing the planet within a decade but this would be no more then a .001% colony till we transform the planet. Hence it will take decades to do and if we don't start now then we are only delaying the inevitable. I propose the following, a combination of the proposed ideas.

  1. Go to the asteroid belt and grab a few good sized rocks. Toss them at Mars so that they impact at a tangent. Not only does this increase the green house gasses and increase the atmospheric temperature, but it also gently starts to increase the rotation of the planet. Several of these impact placed and timed correctly can do the above.

  2. A series of properly placed nuclear shape charges planted into the mantel to assist in the melting process. Not enough to cause any long lasting side effects. Just enough for part 3 to work.

  3. While grabbing rocks we build a new moon for Mars. Placing this in orbit to increase and stabilize the tidal forces on Mars. this will help heat the core and keep it molten. It will take some artificial stabilization to keep the new moon in orbit till gravity and tidal forces straighten enough to hold it in place.

  4. All of this still may not be enough to hold back the solar winds. so an artificial shield to strengthen the now occurring magnetic shielding maybe needed.


According to Wikipedia on formal definition of the dynamo theory, which itself paraphrases The Earth as a Distant Planet, Vázquez et al.:

There are three requisites for a dynamo to operate:

  1. An electrically conductive fluid medium

  2. Kinetic energy provided by planetary rotation

  3. An internal energy source to drive convective motions within the fluid.

All conditions are already met by Mars. But, if we can somehow increase the speed of its rotation, more powerful convection process will be observed due to increased kinetic energy and cause a dynamo action similar to Earth's.

This can be done in three ways:

  1. Hitting it tangentially with a large asteroid or a number of smaller asteroids.

  2. Slingshooting a bigger body of about one-fourth the mass of Mars around it, increasing its rotation due to the gravitational pull. Could be also achieved with a number of smaller asteroids, but would take longer time to sufficiently increase frequency of its rotation.

  3. Giving Mars a new moon of about one-fourth its mass close enough to increase friction due to tidal forces (tidal friction), increasing the temperature inside its core.

Most people tend to think that the first option is the easiest. But such asteroid impact could also cause Mars to leave its present orbit, destabilizing orbits of terrestrial planets like Earth, in lower orbits around the Sun.

I'm in favour of the second option, as we will be capable of capturing and re-orbiting asteroids in the coming decades.

  • $\begingroup$ A Mars day -- sol -- is approximately 24 hours, 39 minutes, 35 seconds long. If the Martian rotation is increased, the sol is shortened. This actually brings the duration of the sol closer to the Earth day of 24 hours. Is there an equation that relates the time difference between Earth day and a solar day of a planet and energy of the convection? $\endgroup$
    – Kav
    Commented Jul 12, 2021 at 10:49

Yes, this is purely speculation (+: (apart from the content available from web-references)

Wikipedia writes to say

Mars is the fourth planet from the Sun and the second smallest planet in the Solar System. Named after the Roman god of war, it is often described as the "Red Planet" because the iron oxide prevalent on its surface gives it a reddish appearance

Iron Oxide is commonly known as rust; typically caused when iron well, oxidizes. So depending upon the oxidation it may/may not be ferromagnetic. That's one part of it.

The other part is that all of the Solar System is steeped in Sol's magnetic field

"The sun's magnetic field extends all the way to the edge of the solar system," explains Opher.

Assuming the iron-oxide exists in adequate quantity on the surface of Mars (which it should, seeing as we're able to discern the colour way out here!) here's what I'd do

  • Determine just how ferro-magnetic the iron oxide out there is
  • Come up with a mechanism to align all that iron oxide in a circular band around the Martian surface pole-to-pole
    • Let the repeated caresses of Sol's magnetic field magnetize the band (That's how we magnetize razors as kids anyway - move the magnet above the razor blade again & again in the same direction)
    • This has the advantage that rust being a semi-conductor will hopefully prevent demagnetization as the band spins away.
  • If necessary
    • Reduce the iron-oxide so it is adequately ferromagnetic for this scheme to work
    • Drill into the the Core, and connect the band there (Then mix this scheme with Hash's answer)

Again (+: this is hypothetical

  • 1
    $\begingroup$ Most of the rust on mars is Red Rust, and Red Rust is not magnetic. #Busted. Also while the sun does send it's magnetic effect throughout the solar system, I really doubt that it's actually a large enough effect. $\endgroup$
    – Civilian
    Commented Oct 3, 2016 at 6:30

Mars's has an exceedingly low flux magnetic field generated by its core because there is very little convection of its conductive core material. This is likely because the natural nuclear fission process in the core has stalled. When planets form they become molten, which causes the constituent materials to differentiate. Lighter materials e.g. silicates gravitate to the surface, heavy materials like iron sink to the core. However, the ultra dense materials, like uranium, end up at the absolute centre of the planet where they form a georeactor, which is a naturally moderated breeder fission reactor.


The heat from this georeactor drives the convection which creates the magnetosphere. However, on Mars the georeactor seems to have stalled, presumable because it has run out of nuclear fuel and suitable fertile breeder materials.

So to answer you question, in order to kick start Mars Magnetic field you would need to inject a relatively large quantity of fissionable and fertile materials into the core. Convection is restarted along with volcanism and plate tectonic motion. Fortunately the pyroclastic gases produce an atmosphere and, when they condense, an ocean. I'd say shake and bake terraforming except I believe someone already has that trademarked...

  • $\begingroup$ do you mean the Martian core can become a dumping site of nuclear waste from nuclear fission for electricity? $\endgroup$
    – Kav
    Commented Jul 12, 2021 at 11:27

As fun as it might be to start hitting Mars with big rocks, there are serious concerns with:

  1. Orbital debris post bombardment.

  2. Destruction of any possible resident life forms that have not been discovered.

  3. Existing human habitation. Although Eminent Domain would surely apply.

  4. The environmentalists would have a fit.

Less catastrophic methods would obviously be preferable.



Apparently the accepted physics has changed, and we now mainly use the convection model for generating planetary magnetic field, not the dynamo model. Differential crust to core rotation, as well as salt water circulation must have an additional effect, however.

One could drill down deep, close to the mantle, in places the crust is much thinner on mars New Gravity Map Suggests Mars Has a Porous Crust. Once close enough to the mantel, fill with radioactive material and let it go critical, in theory it would melt it's way into the mantel where it could eventually make it's way to the core (and/or erupt in volcanoes on the surface if your math fails?). If done enough times, the radioactive material could restart convection. Add a Ceres moon to be sure we get it right. Try to basically emulate the earth moon:system on a smaller scale, because it works.

Ceres is the closest Dwarf Planet to Mars, at times its orbit is quite close. Use equatorial placed Ion engines on gimbals and locally obtained H2O as reaction mass. Allow Mars to capture Ceres in a close orbit, which may help induce a magnetic field. Ceres can be moved to optimum orbital distance at will using the same Ion engines. Matter from Ceres can be transferred to Mars via same ion drive for heating of polar caps and generation of initial atmosphere.

  • $\begingroup$ Geodynamos and convection aren't competing models, they are completely different kinds of concepts. Geodynamos are driven by convection. $\endgroup$ Commented Apr 26, 2020 at 20:18

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