How we know what is inside Earth

The only instrument that reaches

Earthquakes send elastic waves through the entire planet. Seismologists read the recordings of the ground motion those waves produce, and from them reconstruct the deep interior. A seismograph works by keeping a heavy mass still while the ground beneath it moves; the difference between the two is the record.^[1,2]

That is the whole basis. Nothing in the deep interior has been observed directly. What is recorded is motion at the surface; everything below is reconstruction.^[1]

An absence that proved something

Two kinds of wave travel through the body of the Earth. Compressional waves pass through solids and liquids alike. Shear waves cannot travel through a liquid at all, because a liquid cannot be sheared.^[1]

Shear waves from an earthquake stop appearing at seismographs beyond roughly 103 degrees from the epicentre. This shadow zone is a hole in the data — a band of the Earth’s surface that receives no direct arrivals of that wave type.^[1,2]

From that absence follows the conclusion that the outer core is liquid. Nobody observed a liquid core; they observed a missing signal and reasoned about what could remove it.^[1]

An inference that was argued over

The inner core is a harder case, and its history shows what inference costs. Shear energy does pass through the innermost core, so it is taken to be solid. But when the modern argument was set out in 1971, its authors were explicit that solidity had originally been inferred from an assumption about composition — and that if the observed rise in compressional velocity reflected a compositional change instead, the inner core might be liquid.^[1]

The observations that would have settled it directly — body-wave phases crossing the inner core as shear waves — had not been reliably recorded. The conclusion was reached by other means, and the paper said so.^[1]

Images that are models

Modern tomography reconstructs three-dimensional variations in wave speed from very large numbers of complete waveforms. The results look like pictures of the mantle. They are reconstructions with finite resolution.^[1]

The limits are not incidental. Purely thermal mantle plumes are predicted to have tails only a few hundred kilometres wide — narrow enough that standard tomography would struggle to see them at all. A feature missing from a tomographic image may simply be smaller than the method can resolve.^[1]

Other lines of evidence

Seismology is not alone. The geomagnetic field, separated into its core, crustal and external contributions, points to a conducting fluid outer core, and the core-generated part alone accounts for more than 95 per cent of the field strength measured at the surface.^[1,2]

Antineutrinos escaping from radioactive decay carry information about deep heat production that nothing else can reach — though the potassium contribution lies below what those detectors can see, and had to come from elsewhere.^[1,2]

Commonly misunderstood