Planetary Science

Earth's Magnetic Shield: How a Molten Core Keeps the Solar Wind at Bay

Spin a ball of liquid iron the size of Mars at the center of a planet, keep it churning at roughly 5,000 K, and you generate a magnetic bubble that deflects a million-tonne-per-second river of solar particles screaming past at 400 km/s. That bubble — the magnetosphere — stretches about 10 Earth radii toward the Sun and trails a tail past the Moon's orbit. Switch the core off and, over tens of millions of years, the atmosphere would slowly bleed into space, as it almost certainly did on Mars. Everything about the field, from compass needles to the aurora, traces back to that hidden, spinning dynamo.

  • Field strength at surface25,000–65,000 nT (0.25–0.65 gauss)
  • Source regionLiquid iron outer core, 2,890–5,150 km deep
  • Outer-core temperature~4,000 K (top) to ~5,700 K (bottom)
  • Magnetopause distance~10 R⊕ (~64,000 km) sunward
  • Last full reversalBrunhes–Matuyama, ~780,000 years ago
  • Solar-wind speed~400 km/s (up to ~800 km/s in gusts)
  • First global modelWilliam Gilbert, 1600 (De Magnete)

Interactive visualization

Press play, or step through manually. The visualization is yours to drive — try it before reading on.

Open visualization fullscreen ↗

Watch the 60-second explainer

A condensed visual walkthrough — narrated, captioned, under a minute.

What you'd actually see — and feel

You can't see Earth's magnetic field, but you meet its effects constantly. A compass needle swings to align with field lines that emerge near the geographic south pole, arc around the planet, and plunge back in near the north — which is why the north-seeking end of your compass points to what physicists call a magnetic south pole. The field is a tilted dipole: the magnetic axis sits about 11° off the spin axis, and the north magnetic pole has been sprinting across the Canadian Arctic toward Siberia at roughly 50–60 km per year at its recent peak, a pace that has since begun to slow.

The most spectacular visible signature is the aurora. Charged particles funneled down field lines slam into oxygen and nitrogen 100–300 km up, lighting curtains of green (557.7 nm oxygen) and red. Auroras ring both magnetic poles in ovals precisely because that's where the shield's field lines dip down into the atmosphere.

Numbers to anchor the scale:

  • Surface field strength ranges from about 25,000 nT near the equator (over the South Atlantic Anomaly, even lower) to roughly 65,000 nT near the poles — far weaker than a fridge magnet (~5,000,000 nT) at its face.
  • The magnetosphere it inflates reaches a magnetopause about 10 Earth radii (≈64,000 km) toward the Sun, and streams into a magnetotail that extends hundreds of Earth radii downwind, well past the Moon.
  • Trapped particles form the Van Allen radiation belts, discovered by James Van Allen's team with Explorer 1 in 1958, doughnut-shaped zones satellites must design around.

The mechanism: a self-sustaining geodynamo

Earth's field is not a giant buried bar magnet. Iron loses its permanent magnetism above its Curie temperature (~1,043 K for pure iron), and the core is many thousands of degrees hotter than that. Instead, the field is regenerated in real time by a geodynamo — a natural electromagnetic generator running in the liquid outer core, a shell of molten iron-nickel alloy between about 2,890 and 5,150 km depth.

Three ingredients make it work:

  • An electrically conducting fluid. Molten iron conducts current superbly.
  • Convection. The outer core is heated from below by the hotter inner core and by the slow release of primordial and radioactive heat. Hot iron rises, cools, and sinks. Crucially, as the solid inner core freezes out at the bottom (growing ~1 mm per year), it releases light elements like sulfur and oxygen that buoyantly stir the fluid — compositional convection that may power the dynamo more efficiently than heat alone.
  • Rotation. Earth's spin twists rising and sinking columns via the Coriolis effect into helical rolls aligned roughly with the rotation axis. That organized swirl is what lets a chaotic soup produce a coherent, mostly-dipolar field.

The physics is a feedback loop: moving conductive fluid across existing field lines induces electric currents; those currents generate magnetic field; the field acts back on the moving fluid. When the geometry is right, the loop amplifies and sustains itself against ohmic decay. Left alone with no dynamo, Earth's field would decay away in only tens of thousands of years — so the fact that rocks record a field billions of years old proves the dynamo has run almost continuously. The whole engine draws only a modest slice of Earth's heat budget; estimates put the dynamo's power at a few hundred gigawatts to ~1 terawatt, tiny beside the ~44 TW flowing out of Earth's interior in total.

The shield in action: deflecting the solar wind

The Sun continuously boils off a supersonic plasma — the solar wind — that reaches Earth at typically ~400 km/s (gusting to ~800 km/s during high-speed streams) and densities of a few particles per cubic centimeter. It carries the Sun's own magnetic field frozen into it. When this wind hits Earth's magnetosphere it can't easily flow through; instead it drapes around the obstacle.

The encounter builds distinct structures:

  • A bow shock forms ~3 R⊕ upstream of the magnetopause, where the supersonic wind abruptly slows, heats, and deflects — the plasma analog of the shock cone ahead of a supersonic jet.
  • The magnetopause is the boundary where the magnetosphere's outward magnetic pressure balances the wind's inward ram pressure. Its distance breathes with solar activity: nominally ~10 R⊕, it can be crushed inside geosynchronous orbit (6.6 R⊕) during severe storms.
  • Field lines dragged downstream form the vast magnetotail, storing energy that reconnection events later dump into the polar atmosphere — driving auroral substorms and geomagnetic storms.

The protection is real but selective. The field is superb at deflecting the bulk low-to-moderate-energy solar wind, sparing the upper atmosphere from steady erosion. It is far less effective against the highest-energy galactic cosmic rays (protons at GeV energies and up), which the atmosphere's ~10 tonnes of air per square meter ultimately absorbs. So think of the magnetic field and the atmosphere as a two-layer defense: the field turns aside the charged solar plasma; the air soaks up the hard radiation that punches through.

Reversals, excursions, and a wandering field

The most startling fact about the shield is that it is not stable in orientation. Over geologic time the whole dipole flips — north and south magnetic poles swap — in events called geomagnetic reversals. When new ocean crust erupts at mid-ocean ridges and cools past the Curie point, it locks in the field direction of the moment, producing symmetric magnetic 'stripes' on the seafloor. Reading those stripes in the 1960s (Vine, Matthews, and Morley, 1963) clinched the theory of plate tectonics.

Key facts about reversals:

  • They are irregular. The average over the last 20 million years is roughly one every 200,000–300,000 years, but the intervals are wildly uneven — one stretch, the Cretaceous Normal Superchron, ran without a reversal for ~37 million years.
  • The last full reversal, the Brunhes–Matuyama, occurred about 780,000 years ago. The most recent brief flip-and-back, the Laschamp excursion, happened ~41,000 years ago and lasted only a few centuries, during which surface field strength dropped to perhaps a quarter of normal.
  • A reversal is not instantaneous. The transition takes on the order of 1,000–10,000 years, during which the clean dipole weakens and breaks into a messier multi-pole pattern — the shield gets patchy but doesn't vanish.

Today's field is, in fact, weakening: the dipole has lost roughly 9% of its strength since precise measurements began in 1840. The most dramatic soft spot is the South Atlantic Anomaly over Brazil and the South Atlantic, where the inner Van Allen belt dips close enough to the surface (~200 km) that satellites, including the Hubble Space Telescope, routinely power down sensitive electronics while passing through. This weakening is sometimes cited as a possible sign of an approaching reversal, but paleomagnetic records show the field has been this weak before without flipping — so it is not a reliable countdown.

Common misconceptions — and what a reversal would (not) do

The magnetic shield attracts more myth than almost any other Earth-science topic. A few corrections worth having straight:

  • 'A reversal would kill everything.' Reversals have happened hundreds of times, and the fossil record shows no reliable mass-extinction signature tied to them. Our ancestors lived through the Laschamp excursion 41,000 years ago. During a transition the field weakens and gets messy, likely increasing radiation dose at high altitudes and expanding the auroras toward the equator — a headache for satellites and power grids, not an apocalypse.
  • 'The magnetic poles are where the field lines are vertical, and they match the geographic poles.' No — the magnetic poles are offset by ~11° from the geographic poles, they wander independently, and they are not even exactly antipodal to each other because the field isn't a perfect dipole.
  • 'The molten core is molten because of pressure/friction from spinning.' The outer core is liquid because of temperature exceeding the melting point at that pressure; the inner core, though even hotter, is solid because the immense pressure (~330–360 GPa) raises iron's melting point above the local temperature. Spin doesn't melt it — spin organizes the convection into a dynamo.
  • 'The field protects us from all space radiation.' It deflects charged solar-wind plasma effectively but is far less able to stop the most energetic galactic cosmic rays; the atmosphere does the heavy lifting there. Astronauts beyond low Earth orbit — outside the magnetosphere entirely — lose that protection, a central problem for crewed Mars missions.

The single most consequential real risk isn't a reversal but a severe geomagnetic storm from a large coronal mass ejection. The 1859 Carrington Event set telegraph lines sparking; a comparable strike on today's grid could induce currents that damage high-voltage transformers over continent-wide areas. That's a space-weather hazard riding on top of the shield, not a failure of the core.

History and how we know all this

Humans navigated by lodestone compasses for a millennium before anyone understood why they worked. The Chinese described magnetic direction-finding by the 11th century; European mariners adopted the compass by the 12th–13th. The conceptual breakthrough came in 1600, when the English physician William Gilbert published De Magnete and argued, from experiments with a magnetized sphere he called a terrella, that Earth itself is a giant magnet — the first correct global picture.

The modern chain of understanding:

  • Carl Friedrich Gauss, in the 1830s–40s, put the field on mathematical footing, using spherical harmonics to prove that the source is overwhelmingly internal to Earth, and organized the first global observatory network.
  • Seismology revealed the layered interior: Richard Oldham (1906) found evidence of a distinct core, Beno Gutenberg (1913) fixed the core–mantle boundary near 2,890 km, and Inge Lehmann (1936) discovered the solid inner core inside the liquid outer core — from how seismic waves bend and where S-waves (which can't cross liquid) fail to arrive.
  • The geodynamo idea matured through the 20th century (notably Walter Elsasser and Edward Bullard in the 1940s–50s), and only in recent decades have supercomputer simulations — the first self-consistent one by Glatzmaier and Roberts in 1995 — reproduced a dipole field that spontaneously reverses, matching the geologic record.
  • Spacecraft mapped the shield in space: Explorer 1 (1958) found the Van Allen belts, and the ESA Swarm trio (launched 2013) now tracks the field's fine structure and its ongoing weakening from orbit.

What makes this a genuinely astronomical story is that the same test — does a rocky world have a churning conductive core? — sorts the planets. Earth and Mercury (surprisingly, given its size) have active dynamos; Mars and the Moon have only fossilized crustal magnetism from dynamos that died billions of years ago. The magnetic shield, in other words, is a readout of a planet's thermal life story, written in iron.

Earth vs. Mars: why one kept its shield and one lost it
PropertyEarthMars
Global magnetic field todayYes, ~25,000–65,000 nT at surfaceNo global field; only crustal 'fossil' patches
Molten, convecting coreYes — liquid iron outer coreCore largely frozen/stagnant; dynamo died ~4 billion years ago
Atmospheric pressure at surface~1,013 hPa (1 atm)~6 hPa (~0.6% of Earth's)
Standoff against solar windMagnetopause ~10 R⊕ upstreamSolar wind reaches near the surface; ions stripped away
Radius6,371 km (1 R⊕)3,390 km (~0.53 R⊕)

Frequently asked questions

If the core is solid iron in the middle, where does the magnetism come from?

From the liquid layer around it. The solid inner core can't generate the field (iron above its Curie temperature isn't magnetic anyway). The field is produced by convecting, electrically conducting molten iron in the outer core (2,890–5,150 km deep), whose motion — organized by Earth's rotation — acts as a self-sustaining dynamo. The inner core helps by releasing heat and buoyant light elements as it freezes, driving that convection.

How strong is Earth's magnetic field, really?

Weak, in everyday terms. At the surface it ranges from about 25,000 nanotesla near the equator to about 65,000 nanotesla near the poles — roughly 0.25 to 0.65 gauss. A common refrigerator magnet is on the order of 50–100 gauss at its face, over a hundred times stronger locally. Earth's field wins only through sheer size, filling a magnetosphere tens of thousands of kilometers across.

Would we be in danger if the poles reversed?

Not the way movies suggest. Reversals have happened hundreds of times with no reliable link to mass extinctions, and humans survived the Laschamp excursion ~41,000 years ago. During a transition (lasting roughly 1,000–10,000 years) the field weakens and becomes multi-polar, likely raising radiation exposure at high altitudes and disrupting satellites, navigation, and possibly power grids — serious for technology, not an extinction event.

Is the field really weakening, and does that mean a reversal is coming?

The dipole has weakened about 9% since 1840, and the South Atlantic Anomaly is spreading. But paleomagnetic records show the field has dropped this low before and then recovered without flipping. So the weakening is real and worth monitoring — especially for satellites over the South Atlantic — but it is not a reliable sign that a reversal is imminent.

Why did Mars lose its magnetic field and its atmosphere?

Mars is smaller (radius 3,390 km, ~0.53 R⊕), so it cooled faster and its core convection shut down, ending its dynamo around 4 billion years ago — leaving only patchy crustal magnetism. With no global shield, the solar wind gradually stripped ions from the upper atmosphere, a process NASA's MAVEN mission has measured directly. Today Mars's surface pressure is under 1% of Earth's.

Could a solar storm actually get past the shield and cause damage on the ground?

Yes — indirectly. A large coronal mass ejection can compress the magnetopause and inject energy that reaches the ground as rapidly changing magnetic fields, which induce currents in long conductors like power lines and pipelines. The 1859 Carrington Event sparked telegraph wires; a comparable modern storm could damage high-voltage transformers across a continent. The shield deflects the steady solar wind well, but violent space weather still couples through it.