NavList:

Tony, you wrote:
"I know there is a certain temperature above which a ferromagnetic substance looses its' magnetic properties ("demagnetises"). I cannot imagine how a "molten iron" could keep being a ferromagnetic body."

That's right.The temperature is above the Curie Point. So it has to be the "other thing". And that other thing is electromagnetism caused by currents running through that molten conducting fluid. And what generates the electric currents? Magnetic fields, of course! The magnetic fields generate electric currents, and the electic currents re-generate the magnetic fields. It's a self-sustaining system where the energy is provided by convection as the planet slowly cools over billions of years. There's a good, basic summary here: Wikipedia: Dynamo theory.

It's important to recognize the geophysicists have only primitive models of the Earth's dynamo. In many ways, we understand the magnetic field of the Sun (and the associated sunspots and 11-year sunspot cycle) generated tens of thousands of miles deep within the Sun better than we understand the generation of the Earth's magnetic field a mere thousand miles below our feet.

Most planetary magnetic fields are a bit cockeyed. The Earth's field is tilted about 10° relative to the rotation axis. For remote sensing of other planets, this is useful because the magnetic field co-rotates (presumably!) with the planet's deep interior, but being out of line we can detect radio signals as the field wobbles about. And that works well... except in one case: the planet Saturn. Until quite recently, the daily rotation rate of Saturn could only be estimated roughly since its magnetic field happens to be almost perfectly aligned with the rotation axis. Just a few weeks ago, there was some press coverage of a new method which has finally locked down the length of the Saturnian day without using its magnetic field.

Note: I said it "has to be" the other thing since it's too hot for ferromagnetism in the Earth's mantle and core. But that's not necessarily true. It's worth noting that ferromagnetism is a bulk thermodynamic property, and like other phases, it depends on pressure, as well as temperature. There's no evidence for this in relation to the Earth's magnetic field, but is is theoretically possible to have ferromagnetic states of compounds (any compounds, containing iron or not) that can exist stably at extremely high pressures. In short, there could be a permanent magnetization of some exotic substance deep in the Earth. But the default assumptiion today is that the field is due to a dynamo, generated by electric currents in the molten mantle and core.

Frank Reed