The dynamo theory, which explains how Earth generates its magnetic field through convection of liquid iron in the outer core, is supported by multiple independent lines of evidence. These range from seismic readings confirming the outer core is liquid, to magnetic reversal records preserved in ocean floor rocks, to computer simulations that reproduce the field’s behavior, and even laboratory experiments using liquid metal. Together, they build a compelling case that no permanent magnet sits at Earth’s center. Instead, a self-sustaining electromagnetic engine powers the geomagnetic field.
Seismic Proof That the Outer Core Is Liquid
The dynamo theory requires a layer of electrically conducting fluid inside Earth, and seismic waves confirm exactly that. When earthquakes send energy through the planet, one type of wave (called an S-wave) cannot travel through liquids. S-waves vanish when they hit the outer core boundary at about 2,900 km depth, then reappear beyond the inner core. This tells geophysicists the outer core is fluid while the inner core is solid.
Compressional waves (P-waves) do pass through the outer core, but they slow down and bend sharply at the boundary, consistent with a dramatic change from solid rock to liquid metal. Recent seismic analysis has detected localized regions where P-wave speed increases by about 1 to 1.5%, which researchers attribute to pockets of lighter elements moving through the liquid iron at speeds around 40 km per year. That kind of fast fluid movement is exactly the type of flow the dynamo theory predicts is necessary to generate and maintain a magnetic field.
Magnetic Reversals Recorded in Rock
A permanent magnet can’t flip its poles, but a fluid dynamo can. One of the strongest pieces of evidence for the theory is that Earth’s magnetic field has reversed its polarity hundreds of times throughout geologic history. When lava cools or sediments settle, iron-bearing minerals lock in the direction of the ambient magnetic field at that moment. By studying these rocks, scientists can read a timeline of magnetic orientation stretching back billions of years.
The reversals are irregular. They happen as often as every 10,000 years or as rarely as every 50 million years, with no predictable cycle. The last full reversal occurred about 780,000 years ago. Reversals typically unfold over hundreds to thousands of years, though at least one may have happened in as little as a single year. This erratic, non-periodic pattern is exactly what you’d expect from a chaotic fluid system rather than a static source.
The most visually striking evidence comes from the ocean floor. As new crust forms at mid-ocean ridges, it preserves the magnetic field’s direction like a tape recorder. The result is a symmetric pattern of alternating magnetic stripes on either side of the ridge. These stripes not only confirm that reversals happen but also helped establish the theory of plate tectonics itself.
The Energy Source Driving Convection
A dynamo needs energy to keep running, and two processes in the core provide it. The first is thermal convection: the core is losing heat to the mantle above, and that cooling drives hot fluid upward. The second, more powerful source kicked in when the solid inner core began crystallizing, roughly one billion years ago. As iron solidifies onto the inner core, it releases latent heat and expels lighter elements like sulfur, silicon, and oxygen into the surrounding liquid. These lighter materials rise buoyantly, stirring the outer core and generating the electric currents that sustain the magnetic field.
Energy budget calculations estimate that 6 to 10 terawatts of heat currently flow from the core into the mantle. Before the inner core existed, the dynamo ran on thermal convection alone, which is less efficient. Maintaining the field back then required either a much higher heat flow (14 to 24 terawatts) or the field was simply weaker. Paleomagnetic intensity measurements from rocks older than one billion years suggest the field strength was between one-quarter and the full value of today’s field, which fits neatly with the idea that inner core crystallization supercharged the dynamo.
Computer Simulations That Reproduce the Field
In 1995, Gary Glatzmaier and Paul Roberts ran a landmark three-dimensional computer simulation of the geodynamo. Their model solved the equations governing fluid flow, heat transfer, and electromagnetism in a rotating spherical shell meant to represent the outer core. The simulation maintained a self-sustaining magnetic field for over 40,000 simulated years, and it spontaneously produced a reversal of the magnetic poles without any external trigger.
The simulated reversal shared key features with real reversals seen in the paleomagnetic record, including a temporary weakening of the field before the poles flipped. The model also included a finitely conducting solid inner core, which played a stabilizing role during the reversal. This was a turning point for the theory: it showed that the basic physics of convecting liquid metal in a rotating shell can, on its own, generate a magnetic field that looks and behaves like Earth’s.
Laboratory Experiments With Liquid Metal
If the theory is correct, a fast-moving conducting fluid should be able to generate its own magnetic field without any external magnets. In 2006, a French research team demonstrated exactly this with the VKS experiment, which used rapidly stirred liquid sodium in a cylindrical vessel. The experiment produced a self-sustaining magnetic field purely from the motion of the liquid metal.
More remarkably, the experiment reproduced several behaviors seen in Earth’s field. By adjusting the flow speed and stirring pattern within a fairly small range of parameters, the researchers observed stationary dynamo states, periodic oscillations, intermittent bursts of magnetic activity, and random reversals of the field’s polarity. Despite extremely strong turbulent fluctuations in the flow, the magnetic field’s large-scale behavior remained governed by just a few interacting modes. This showed that a turbulent fluid dynamo doesn’t produce random noise. It produces organized, recognizable magnetic behavior, just as the theory predicts.
Comparative Evidence From Other Planets
The dynamo theory also predicts that any planet with a convecting, electrically conducting fluid interior should generate a magnetic field, and planets without one should not. The solar system lines up with this prediction remarkably well.
Earth and Mercury both have active global magnetic fields and are thought to have liquid, convecting outer cores. Mercury’s field is much weaker, consistent with the idea that its smaller core has cooled more and may be nearing the end of its dynamo phase. Mars and the Moon, by contrast, have no global magnetic fields today. Both are smaller bodies that lost their internal heat faster, shutting down core convection long ago. Yet Mars shows patches of strong magnetization in its oldest crustal rocks, evidence that it once had an active dynamo billions of years ago before its core solidified too much to sustain one.
The Sun provides another test case. Its magnetic field flips roughly every 11 years as part of the sunspot cycle, driven by the flow of hot, ionized gas. Differential rotation stretches magnetic field lines and wraps them around the Sun, while the Sun’s spin twists them into loops. This solar dynamo operates on the same basic principle as Earth’s: moving conducting fluid generates and regenerates magnetic fields. The Sun’s 22-year full magnetic cycle, the equator-ward drift of sunspots, and the reversal of polar fields near each cycle’s peak all match predictions from dynamo models.