Imagine a world with a dead core, like Mars, but covered in water. This world has managed to hold on to its atmosphere and quite a bit of its water as well. I'm sure the prodigious life there would be grateful if the planet could also hold on to its magnetic field.

This core has been dead for quite some time. In fact, enough time has passed for its surface terrain to be largely smoothed out from erosive processes like tropical downpours and waves.

enter image description here

We know that even on Mars, which has a mostly dead core that it still has localized magnetic fields. And Callisto has an induced magnetic field from Jupiter created by a subsurface ocean.


Can a world with a dead core still retain a respectable magnetic field?

  • $\begingroup$ Earth have molten metallic core acts like bar magnet, gas giants such as Jupiter use liquid hydrogen plus it spin fast, this process is known as magnetic dynamo. Mercury is the weirdest since it's small size means core has cooled and it isn't gas planet for magnetic dynamo to take place. You'll be surprised there are more of such planets exists out there. $\endgroup$
    – user6760
    Aug 13, 2015 at 10:22

3 Answers 3



The planet needs to have fluid motion of its core in order to produce a magnetic field. The magnetic field is produced via motion in the dynamo theory. If the planet core is dead, as in solid, then it won't be generating a magnetic field.

The only idea I could think of for the planet to have a magnetic field is if iron the core cooled in the presence of a very strong magnetic field. Then the core would be a permanent magnet. However, what could create such a field and how long it could be maintained are unknown to me.

  • $\begingroup$ So in effect the core would be like the solid magnets we use everyday? Does this process to turn the core into a "rare earth magnet" always require the presence of another field to be present? $\endgroup$ Aug 12, 2015 at 20:10
  • 1
    $\begingroup$ @JoshBelmont Yes and yes. It requires a field to orient the core properly. Otherwise a fluid core would simply orient to cancel all its magnetic fields. It's the lowest energy state. $\endgroup$
    – Samuel
    Aug 12, 2015 at 20:17


But it depends upon how you look at it.

For terrestrial or rocky planets with iron cores, Samuel's answer is correct. However, there are other types of bodies for which this might not be true.

Gas giants

For gas giants, the metallic hydrogen at the core works as your conductive fluid and can maintain a strong magnetic field.

Whether you consider this a world with a "dead core" is up to you.

Ice giants

Ice giants cores are likely composed of silicates and metals. Its mantle is composed of various ices. Neither contribute to the generation of a magnetic field.

However, these bodies do have strong magnetic fields.

The mechanism for magnetic field generation isn't well understood but it is thought that a salty liquid water outer mantle works as the rotating conductive fluid and it generates the magnetic field. In this case, the core is dead, yet the planet still possesses a strong magnetic field.


enter image description here

Some think that Ceres' water ice layer may include a water liquid layer. In which case, a planet like this could generate a magnetic field like that of ice giants.

Such a planet would contain a solid silicate - metal core, liquid water inner mantle, icy outer mantle, and rocky crust.

  • $\begingroup$ Interesting, so in your opinion is it possible for an earth sized terrestrial planet to generate a magnetic field from a salty liquid layer below the surface? Or would it be too hot? $\endgroup$ Aug 12, 2015 at 22:47
  • $\begingroup$ We did recently find that the earth has subsurface oceans of water locked in rock: phys.org/news/2014-06-evidence-oceans-deep-earth.html $\endgroup$ Aug 12, 2015 at 22:53
  • 1
    $\begingroup$ I'm uncommitted. I would suspect it would be too hot but maybe not. Imagine planet formation in which an Earth sized body formed with a composition like Ceres. It then migrates closer to the Sun and becomes inhabitable by man. Humans come by and colonize. Now you have a flat featureless body, dead core, with icy upper mantle and liquid inner mantle. This is unstable and over geologic time since the water will melt and this will become a water world. In the meantime, it should meet your requirements. $\endgroup$
    – Jim2B
    Aug 12, 2015 at 23:25
  • $\begingroup$ It seems the other properties of the OP's planet (like Mars, with an atmosphere, quite a bit of water, and prodigious life) are not compatible with your solutions. Yeah? $\endgroup$
    – Samuel
    Aug 13, 2015 at 23:38
  • $\begingroup$ @Samuel, Even when there's a definitive yes/no, I like to provide some details and/or counter examples. I enjoy the discussion as much as the questions & answers. $\endgroup$
    – Jim2B
    Mar 15, 2016 at 1:27

Yes, a planetary magnetosphere can be induced without a core being involved:


The paper's abstract reads:

Induced magnetospheres occur around planetary bodies that are electrically conducting or have substantial ionospheres, and are exposed to a time-varying external magnetic field. They can also occur where a flowing plasma encounters a mass-loading region in which ions are added to the flow. In this introduction to the subject we examine induced magnetospheres of the former type. The solar wind interaction with Venus is used to illustrate the induced magnetosphere that results from the solar wind interaction with an ionosphere.

  • $\begingroup$ With an induced magnetic field, would it still be going strong if the external source of the field were removed? $\endgroup$ Aug 25, 2015 at 5:24
  • $\begingroup$ @JoshBelmont No, that's why it's called "induced". $\endgroup$
    – Hackworth
    Aug 25, 2015 at 6:14
  • $\begingroup$ If induced at the time of solidification you end up with a permanent bar magnet. $\endgroup$
    – Joshua
    Aug 28, 2015 at 17:52
  • $\begingroup$ @Joshua The induction happens only in the ionosphere or magnetosphere, not in the planet itself. There is nothing to solidify. $\endgroup$
    – Hackworth
    Aug 28, 2015 at 21:07

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .