The Van Allen Belts

An invisible storm circling the Earth

Picture a doughnut of radiation surrounding the planet — not a thin ring like Saturn's, but a fat, three-dimensional torus woven into Earth's magnetic field, and a second, larger one nested beyond it. Inside those zones, protons and electrons hurtle along curving magnetic field lines at a large fraction of the speed of light, trapped because the field forms a kind of bottle they cannot easily escape. An astronaut sitting still in the heart of the inner belt would absorb a hazardous radiation dose within hours. Yet from the ground you would never know the belts were there: they emit almost no light, and they are a near-perfect vacuum by everyday standards. The "radiation" is not heat or glow — it is the sheer kinetic energy of individual particles.

The belts exist because Earth has a global, roughly dipolar magnetic field and a steady supply of charged particles — from cosmic rays smashing the upper atmosphere, from the solar wind, and from the ionosphere. A charged particle moving through a magnetic field feels the Lorentz force, which curls its path into a tight helix wound around a field line. The field is stronger near the magnetic poles, where the lines converge, and that convergence acts as a mirror that turns the particle around before it reaches the atmosphere. Caught between two mirrors, the particle bounces back and forth, pole to pole, essentially forever — or at least until a plasma wave nudges it loose. Multiply that by trillions of particles and you get the Van Allen belts.

The physics: gyrate, bounce, drift

Every trapped particle performs three superimposed periodic motions, each tied to a conserved adiabatic invariant — a quantity that stays constant as long as the field changes slowly compared with the motion in question.

1 · Gyration. The Lorentz force F = qv × B bends the particle's velocity into a circle perpendicular to the field. The gyroradius (Larmor radius) and gyrofrequency are

r_g = γ m v⊥ / (|q| B)        gyroradius
ω_c = |q| B / (γ m)           gyrofrequency (angular)

The first adiabatic invariant is the magnetic moment μ = p⊥² / (2 m B), conserved as the particle moves into stronger or weaker field. For a 1 MeV electron in Earth's equatorial field of ~5 × 10⁻⁷ T at L = 4, the gyroradius is of order ten kilometres and the relativistic gyrofrequency is a few kHz — the particle spirals a few thousand times every second.

2 · Bounce. Because μ is conserved and B rises toward the poles, the perpendicular momentum p⊥ grows at the expense of the parallel momentum until the parallel motion reverses. That turning point is the mirror point. The particle bounces between conjugate mirror points in the two hemispheres. The bounce period for a 1 MeV electron at L = 4 is roughly 0.1–1 second. The relevant invariant is the second (longitudinal) invariant J = ∮ p∥ ds taken over a full bounce.

3 · Drift. The gradient and curvature of the dipole field make the gyrocircle drift slowly in longitude — and crucially, ions and electrons drift in opposite directions. This azimuthal drift carries a particle all the way around the planet in minutes to hours, and the net charge separation is the equatorial ring current that a ground magnetometer measures as the Dst index during storms. The third invariant is the magnetic flux Φ enclosed by the drift orbit.

The whole hierarchy is nested: gyration (μs–ms) ≪ bounce (0.1–1 s) ≪ drift (minutes–hours). A particle that stays on one L-shell keeps all three invariants; when a fast magnetic disturbance violates the third invariant, the particle is transported radially inward or outward — the basic mechanism of radial diffusion that powers the outer belt.

The key numbers

The belts are organised in the radial direction by the McIlwain L-parameter — the equatorial crossing distance of a field line, measured in Earth radii (1 R⊕ = 6,371 km).

Region	L-shell	Altitude (equator)	Dominant particles	Top energy
Inner belt	~1.2 – 2.5	~1,000 – 6,000 km	Protons (+ some e⁻)	~700 MeV protons
Slot region	~2 – 3	~6,000 – 13,000 km	Depleted	—
Outer belt	~3 – 7	~13,000 – 60,000 km	Electrons	~few – 10 MeV e⁻
Third belt (transient)	~3.0 – 3.5	~13,000 – 16,000 km	Ultra-relativistic e⁻	> 2 MeV e⁻

Other anchor figures worth keeping in mind:

• Field strength. Earth's surface field is ~0.25–0.65 G (2.5–6.5 × 10⁻⁵ T); the dipole moment is ~8 × 10²² A·m². At the magnetic equator the field falls off as B ∝ 1/L³, so with B₀ ≈ 31,000 nT it is ~485 nT at L = 4.
• Inner-belt protons. Flux of >10 MeV protons peaks near L ≈ 1.5 at ~10⁴–10⁵ cm⁻² s⁻¹. These protons are remarkably stable — the inner belt barely changes from year to year.
• Outer-belt electrons. Fluxes of >1 MeV electrons can swing by three to four orders of magnitude in a day during a geomagnetic storm. The belt can be wiped out in hours and rebuilt in hours.
• South Atlantic Anomaly (SAA). Because the dipole is offset ~500 km from Earth's centre and tilted ~11°, the inner belt dips to ~200 km altitude over the South Atlantic. The ISS (orbiting at ~400 km, L ≈ 1.2–1.4) passes through the SAA and astronauts report light flashes from particle hits on the retina.

How the belts are observed

The belts cannot be seen optically, so they are measured in situ by particle detectors flown directly through them. The historical instrument was the Geiger–Müller counter; modern missions carry magnetic and electrostatic spectrometers that resolve particles by energy, pitch angle (the angle between velocity and field), and species.

• Geiger counters (1958). Explorer 1 and 3 carried simple GM tubes. The defining clue was counterintuitive: above a certain altitude the count rate dropped to zero, because the radiation was so intense it paralysed the tube. Van Allen's team realised "zero" meant "off-scale high."
• Particle telescopes and spectrometers. The Van Allen Probes' MagEIS, REPT, and RBSPICE instruments measured electrons and ions from eV to ~20 MeV with fine energy and pitch-angle resolution, mapping how the phase-space density (the relativistic, invariant-coordinate version of particle density) evolves.
• Wave instruments. Because the belts are controlled by wave–particle interactions, missions carry electric and magnetic field sensors (EMFISIS on the Van Allen Probes) to catch chorus, hiss, and electromagnetic ion-cyclotron (EMIC) waves in the act of accelerating or scattering particles.
• Phase-space-density inversion. A key diagnostic is to convert measured fluxes to phase-space density at fixed adiabatic invariants. A radial peak in phase-space density is the fingerprint of local acceleration by chorus waves; a monotonic profile points instead to inward radial diffusion. The Van Allen Probes settled a decades-old debate in favour of local acceleration dominating during many storms.

Worked example: how fast does a killer electron drift around the Earth?

Take a relativistic electron of kinetic energy E_k = 1 MeV mirroring at the equator at L = 4. The gradient–curvature drift period for an equatorially mirroring particle in a dipole is approximately

T_drift ≈ (π q B₀ R_E²) / (3 L γ m c²)        (equatorial, simplified)

where B₀ ≈ 3.1 × 10⁻⁵ T is the equatorial surface field and γmc² is the total energy. A convenient empirical form gives, for electrons,

T_drift ≈ 1.0 / [ L · (E_k in MeV) · (1 + E_k/1.02) ]  hours    (order-of-magnitude)

Plugging in L = 4, E_k = 1 MeV, the factor (1 + 1/1.02) ≈ 1.98:

T_drift ≈ 1.0 / (4 × 1 × 1.98) hr
        ≈ 0.13 hr
        ≈ 7.6 minutes

So a 1 MeV electron laps the entire planet in roughly eight minutes — far slower than its sub-second bounce and far, far slower than its microsecond gyration, confirming the gyrate ≪ bounce ≪ drift hierarchy. The opposite-sign drift of ions and electrons separates charge and drives the ring current. Note that more energetic particles drift faster (shorter period), which is why a sudden drift-resonant ULF wave preferentially energises the most relativistic electrons.

Discovery, missions, and people

The Van Allen belts were the first major scientific discovery of the Space Age — the very first thing humans found once they could put an instrument above the atmosphere.

• James A. Van Allen (1958). His Geiger counter on Explorer 1 (launched 31 January 1958, the first U.S. satellite) showed anomalous saturation. Explorer 3 (March 1958) added a tape recorder that confirmed the altitude pattern. Van Allen announced the trapped-radiation region in May 1958; Time put him on its cover in 1959.
• Mapping the outer belt (1958–59). Explorer 4, Pioneer 3, and the Soviet measurements by Sergei Vernov's group filled in the second, outer zone. The Soviets independently detected trapped radiation with Sputnik 2 and 3.
• Starfish Prime (9 July 1962). A 1.4-megaton U.S. nuclear test detonated at 400 km altitude over the Pacific injected a vast population of fission-decay electrons, creating an artificial radiation belt that lingered for years and disabled or degraded at least six satellites, including Telstar 1.
• Apollo (1968–72). Trajectories were planned to cross the belts quickly and off the most intense field lines; astronaut doses were kept to a few mSv — comparable to a few CT scans — not the lethal levels a stationary stay would deliver.
• CRRES (1990–91). The Combined Release and Radiation Effects Satellite caught a March 1991 storm forming a brand-new belt in minutes, the first clear evidence of rapid shock injection.
• Van Allen Probes / RBSP (2012–2019). NASA's twin spacecraft, flying matched, eccentric orbits through the belts, delivered the definitive modern picture: the transient third belt (Sept 2012), confirmation of local chorus-wave acceleration, the impenetrable barrier at the inner edge of the outer belt, and detailed wave–particle physics.

Earth's belts versus other planets

Any planet with a global magnetic field and a particle source builds radiation belts. The strength scales dramatically with the magnetic moment and rotation rate.

Planet	Surface field (equator)	Peak electron energy	Notes
Earth	~0.3 G	~few – 10 MeV	Two belts + slot; SAA from offset dipole
Jupiter	~4.3 G (~14× Earth)	tens of MeV	Fiercest in the solar system; fried Galileo electronics; Io plasma torus feeds it
Saturn	~0.2 G	~MeV	Belts carved up by rings and moons absorbing particles
Uranus / Neptune	~0.2–0.1 G	~MeV	Wildly tilted, offset dipoles → exotic, time-variable belts
Mercury	~0.003 G	—	Field too weak; no stable trapped belt

Jupiter's belts are the benchmark extreme: its field is about fourteen times Earth's, it rotates every ~10 hours, and it is continuously fed by sulfur and oxygen ions from Io's volcanoes. The result is a synchrotron-emitting electron belt so intense that the Galileo orbiter absorbed a radiation dose hundreds of times the lethal human limit over its mission, and Juno's vault wraps its electronics in a centimetre of titanium.

Related phenomena and the belts' dynamics

• The slot region. The quiet gap between the belts, kept empty by plasmaspheric hiss scattering electrons into the loss cone. It is not permanent: strong storms can fill the slot for weeks before it drains again.
• The transient third belt. Discovered by the Van Allen Probes in September 2012 — the outer belt split, leaving an isolated ring of multi-MeV electrons that survived a month until a shock erased it.
• Chorus and hiss waves. Whistler-mode chorus outside the plasmasphere energises electrons to MeV energies via cyclotron resonance (the "local acceleration" mechanism); hiss inside the plasmasphere scatters them away. The belts are a tug-of-war between these waves.
• EMIC waves. Electromagnetic ion-cyclotron waves can rapidly scatter the most relativistic electrons into the atmosphere, a leading candidate for sudden "dropout" events.
• The South Atlantic Anomaly. The low-altitude dip of the inner belt over the South Atlantic, where satellites and the ISS see elevated radiation; Hubble suspends some observations while crossing it.
• Cosmic-ray albedo neutron decay (CRAND). The main source of the inner-belt protons: galactic cosmic rays hit the atmosphere, splash neutrons back upward, and those neutrons beta-decay in the trapping region into protons and electrons.
• Loss cone and aurora. Particles whose mirror point would lie below ~100 km fall into the atmosphere instead of bouncing; the angular window they occupy is the loss cone, and their precipitation lights the diffuse aurora.

Common misconceptions and subtleties

• "The belts are solid rings of radiation." They are an extremely tenuous plasma — fewer particles per cubic centimetre than a good laboratory vacuum. What makes them dangerous is energy per particle, not particle density.
• "You can't fly through them safely." You can, if you cross quickly. Apollo crews crossed in under an hour with doses of a few mSv. It is dwelling inside the belts that is hazardous; satellites that must live there are radiation-hardened.
• "The inner belt is electrons, the outer belt is protons." It is the reverse for the headline populations: the inner belt is dominated by very energetic protons, the outer belt by relativistic electrons. (Both belts contain both species at various energies.)
• "The belts are static." The outer belt is one of the most dynamic plasma systems near Earth — its multi-MeV electron flux can change by 10,000× in a day. The inner belt, by contrast, is remarkably steady.
• "A storm always boosts the belts." Storms can either fill or empty the outer belt; whether flux rises or falls depends on the competition between acceleration (chorus, radial diffusion) and loss (EMIC scattering, magnetopause shadowing, outward diffusion). Many storms cause a net dropout first, then a rebuild.
• "The belts cause the aurora." The aurora is driven mainly by fresh particles injected from the magnetotail along auroral field lines, not by the trapped belt populations — though belt particles scattered into the loss cone do contribute to the diffuse aurora.