OpenWorldLab

Loading interactive view…

Plasma and space

The Van Allen belts and the aurora

Earth’s field traps charged particles in two belts, found by James Van Allen’s Geiger counters on Explorer 1 in 1958. A trapped particle does three things at once. It gyrates round a field line. It slides along the line and is turned back where the field grows strong near the poles, so it bounces between the hemispheres. And it drifts slowly round the planet. Particles whose turning points lie below the top of the atmosphere are lost there, and light the aurora.

Before you look

Predict first: Protons and electrons both drift round the planet. Which way do they go?

What you are looking at

Left, the meridian plane through the particle (the plane holding Earth’s axis and the particle), turning with it: the field lines, the last bounce and a half of the path folded into the plane, and white ticks where theory puts the turning points. Right, the equator seen from above the north pole: the gyration loops and the drift, with a dot each time the guiding centre crosses the equator.

The start is a 10 MeV proton on the shell L = 3 (the field line three Earth radii out at the equator) with a pitch angle of 30° to the field. It goes round in 103 s, near real time; slower particles are speeded up so that a drift takes about a minute.

The physics

The magnetic moment μ=p⊥2/2mB\mu = p_\perp^2/2mB stays constant, so a particle with equatorial pitch angle α\alpha turns back where B=Beq/sin⁡2αB = B_\text{eq}/\sin^2\alpha. On a dipole line that is the latitude

sin⁡2α=cos⁡6λm1+3sin⁡2λm\sin^2\alpha = \frac{\cos^6\lambda_m}{\sqrt{1 + 3\sin^2\lambda_m}}

Integrating along the line gives the bounce and drift periods, close to the fits

Tb≈4LREv (1.30−0.56sin⁡α),Td≈2π ∣q∣B0RE23L pv (0.35+0.15sin⁡α)T_b \approx \frac{4LR_E}{v}\,(1.30 - 0.56\sin\alpha), \qquad T_d \approx \frac{2\pi\,|q|B_0R_E^2}{3L\,pv\,(0.35 + 0.15\sin\alpha)}

where B0B_0 = 30.7 µT is the field at the equator on the ground. The field line meets the ground at cos⁡2λ=1/L\cos^2\lambda = 1/L; a particle that would turn back below it, inside the loss cone, arrives there.

Going deeper

Vector calculus and Maxwell’s equations

Each motion keeps an adiabatic invariant: the magnetic moment for the gyration, J=∮p∥ dsJ = \oint p_\parallel\,ds for the bounce, and the magnetic flux inside the drift shell for the drift. Each holds while its motion is much faster than the field changes as the particle sees it.

In a vacuum field the gradient and curvature drifts add to vd=(γm/qB3)(v∥2+v⊥2/2) B×∇B\mathbf v_d = (\gamma m/qB^3)(v_\parallel^2 + v_\perp^2/2)\,\mathbf B \times \nabla B. Averaged over a bounce, this gives the drift period above.

What to measure

Mirror latitude, bounce and drift periods, and the loss cone

  1. The checks follow 10 MeV protons at 30° and 60° on L = 3, and 20 MeV electrons on L = 4, for 20 bounces. The guiding centre’s turning latitude matches the formula within 1%; the bounce and drift periods match the integrals along the line within 1%, and the fits within 3% and 2%. Protons drift west and electrons east.
  2. Protons started just inside the loss cone reach the ground within half a degree of cos⁡2λ=1/L\cos^2\lambda = 1/L; just outside it, they bounce on.

What the model leaves out

  • The solar wind squashes the real field, so beyond about L = 5 a dipole is a poor guide.
  • Electric fields, the ring current’s own field, and collisions with the thin upper air, which slowly scatter particles into the loss cone.
  • The top of the atmosphere is taken to be the ground; aurorae glow 100–300 km up.

Things to try

  • Set the pitch angle below the loss cone (8.4° at L = 3): each particle comes down near 55° latitude, north and south in turn.
  • Switch to an electron: it drifts the other way, and its gyration is too small to see.
  • Raise the energy: the bounce speeds up with the speed, the drift with the energy.

References

  • M. Walt (1994), Introduction to Geomagnetically Trapped Radiation, Cambridge University Press.
  • D. A. Hamlin, R. Karplus, R. C. Vik and K. M. Watson (1961), Mirror and azimuthal drift frequencies for geomagnetically trapped particles, Journal of Geophysical Research 66, 1.

More experiments, and how the lab works, are in the lab’s overview.