What we actually know about the stuff at the bottom of the planet
The inner core sits somewhere between 5,200 and 5,700 degrees Celsius, give or take a few hundred depending on which paper you trust this month. That is roughly the same temperature as the surface of the Sun. It sounds absurd when you say it out loud because the idea that the center of a rock is as hot as a star has always bothered people who think about it too much. The reason it stays solid anyway comes down to pressure, not some special property of iron. At around 330 gigapascals, the melting point of iron gets pushed way up past the actual temperature, so you end up with a solid ball of crystalline iron-nickel sitting under half the planet's weight. Figuring out that number was not straightforward and nobody gets credit for just measuring it directly because obviously you cannot drill that deep. The whole thing relies on seismic data. You take the travel times and attenuation of different wave types — specifically PKIKS and PKiKP reflections off the inner core boundary — and you feed them into models that invert for temperature based on known equations of state for iron at high pressure. The problem is that every lab experiment measuring iron's melting curve gives slightly different results, and those differences cascade into massive uncertainty bands on the final number. I spent a chunk of my grad school time trying to reconcile thermal conductivity values for hexagonal close-packed iron under core conditions. One group's data said the core was cooler, another said hotter, and the spread was something like 800 degrees. I ended up running a combined inversion that used both thermal and electrical conductivity constraints from resistivity experiments on iron alloys with light element impurities. The workaround was basically admitting that no single measurement would settle it and treating the inner core temperature as a probability distribution rather than a fixed value. You publish the range, not the point estimate, and you move on.
Here is something most people do not expect: the inner core might actually be cooling slowly right now, which sounds fine until you realize that cooling is what keeps the geodynamo running. If the core cools too much, convection in the outer core slows down and the magnetic field weakens. We have already seen the South Atlantic Anomaly expand and the field strength drop about nine percent over the last century. That is not immediate doom but it is a reminder that this hot solid ball is the engine under the hood and it is not infinitely fueled. Another counter-intuitive thing is that the inner core may be rotating at a slightly different speed than the rest of the planet. Some studies have suggested it is spinning a fraction of a degree faster per year, though other groups have pushed back hard on that. The mechanism would involve magnetic coupling between the liquid outer core and the solid inner core, along with gravitational interactions with the mantle. It is still debated. What is less debated is that the temperature gradient between the inner core boundary and the outer core is steep enough to drive compositional convection as the inner core grows and releases light elements like sulfur or oxygen into the surrounding liquid. That release of light material is probably more important for driving the dynamo than thermal convection alone. If you are looking for a single number to throw around in conversation, 5,400 degrees Celsius is a reasonable middle ground. But the honest answer is that we are working with estimates that could shift by several hundred degrees when better experiments come out, and that is normal for this kind of science. The inner core is not something you can touch or sample directly, so every conclusion is one step removed from actual observation. We build models, test them against seismic waves, and keep refining. It is slow, messy work but it is the only way we have to know anything about a place we will never reach.
