The Science of Discworld Revised Edition

deviation from an ideal sphere (strictly, spheroid, because of the flattening of the poles) is about one-third of a per cent – about as irregular as the shallow indentations you find on a basketball, which add to its grip. Our home planet, give or take a bit of squashing, is remarkably round and surprisingly smooth. Gravity made it that way, and it keeps it that way – except that some small but interesting movements in the mantle and the crust add a few wrinkles.
    How do we know all this? Mainly because of earthquakes. When an earthquake hits, the whole Earth rings like a bell hit by a hammer. Shockwaves, vibrations emitted by the earthquake, travel through the Earth. They are deflected by transition zones between different kinds of material, such as that between core and mantle, or lower and upper mantle. They bounce off the Earth’s crust and head back down again. There are several kinds of wave, and they travel with different speeds. So the short sharp shock of an earthquake gives rise to a very complex pattern of waves. When the waves hit the surface they can be detected and recorded, and recordings made in different places can be compared. Working backwards from these recorded signals, it is possible to deduce a certain amount about the underground geography of our planet.
    One consequence of the Earth’s internal structure is a magnetic field. A compass needle points roughly north. The standard ‘lie-to-children’ is that the Earth is a giant magnet. Let’s unpack the next layer of explanation.
    The Earth’s magnetic field has long been something of a puzzle since magnets are seldom made out of rock, but once you realize that the Earth has a whopping great lump of iron inside it, everything makes much more sense. The iron doesn’t form a ‘permanent’ magnet, like the ones you inexplicably buy to stick plastic pigs and teddy bears on the fridge door; it’s more like a dynamo. In fact it’s called the geomagnetic dynamo. The iron in the core is, as we’ve said, mostly molten, except for a slightly lumpy solid bit in the middle. The liquid part is still heating up – the old explanation of this was that radioactive elements are denser than most of the rest of the Earth, and therefore sank to the middle where they became trapped, and their radioactive energy is showing up as heat. The current theory is quite different: the molten part of the core is heating up because the solid part is cooling down. The liquid iron that is in contact with the solid core is itself slowly solidifying, and when it does so it loses heat. That heat has to go somewhere, and it can’t just waft away unnoticed as warm air because everything is thousands of miles underground. So it goes into the molten part of the core and heats it up.
    You’re probably wondering how the part that is in contact with the solid core can simultaneously be getting cooler, so that it solidifies too,
and
be getting hotter as a result of that solidification, but what happens is that the hot iron moves away as soon as it’s been warmed up. For an analogy, think about a hot air balloon. When you heat air, it rises: the reason is that air expands when it gets hot, so becomes less dense, and less dense things float on top of denser things. A balloon traps the hot air in a huge cloth bag, usually brightly coloured and emblazoned with adverts for banks and estate agents, and floats up along with the air. Now hot iron rises, just as hot air does, and that takes the newly heated iron away from the solid core. It heads upwards, cooling slowly as it does so, and when it gets to the top it cools down – comparatively speaking – and starts to sink again. The result is that the Earth’s core circulates up and down, being heated at the bottom and cooling at the top. It can’t all go up at the same time, so in some regions it’s heading up, and in others it’s heading back down again. This kind of heat-driven circulation is called convection.
    According to physicists, a moving fluid can develop a magnetic field provided three conditions hold. First, the fluid must be able to conduct electricity – which iron can do fine. Secondly, there has to be at least a tiny magnetic field present to begin with – and there are good reasons to suppose that the Earth had a bit of personal magnetism, even early on. Thirdly, something has to
twist
the fluid, distorting that initial magnetic field – and for the Earth this twisting happens by way of Coriolis