The Aurora: How the Solar Wind Paints the Sky

The power supply: solar wind and magnetic reconnection

The Sun continuously blows off its corona as the solar wind — a fully ionized plasma of protons and electrons streaming past Earth at roughly 400 km/s (fast streams reach 800 km/s), with a typical density of about 5 particles per cm³. A solar-wind proton at 400 km/s carries ½m_p v² ≈ 830 eV of bulk kinetic energy, but the wind's real currency is the magnetic field it drags along, the interplanetary magnetic field (IMF), at a few nanotesla.

Earth's dipole field (about 50 μT at the surface, falling as 1/r³) carves a cavity in this wind — the magnetosphere — with its dayside boundary, the magnetopause, standing off at roughly 10 Earth radii ≈ 64,000 km. Coupling switches on when the IMF points southward, antiparallel to Earth's field at the nose. There the two field topologies annihilate and reconnect: magnetic reconnection splices solar-wind field lines onto terrestrial ones, and the solar wind's flow then drags magnetic flux over the poles and piles it into the long magnetotail downstream. Energy accumulates in the stretched tail like a wound spring until a second reconnection event in the tail releases it explosively — a substorm — hurling plasma back toward Earth and lighting the auroral oval within minutes.

The guiding law: charged motion in a magnetic field

Everything downstream of the power supply is governed by one equation, the Lorentz force:

• F = q(E + v × B) — the force on a charge q moving at velocity v through electric field E and magnetic field B.

With no electric field, the v × B term makes a charge spiral around a field line rather than crossing it. Setting the magnetic force equal to the centripetal requirement, qvB = m v²/r, gives the gyroradius r = mv/(qB) and the field-independent-of-speed cyclotron frequency:

• ω_c = qB/m, or f_c = qB/(2πm).

For an electron in Earth's ~50 μT field, f_c ≈ 1.4 MHz and a 100 eV electron's gyroradius is under a metre. The particle is therefore tied to its field line, sliding freely along it while orbiting tightly across it. This is why the aurora traces the geometry of the magnetic field with such fidelity: the glowing curtains are field lines made visible, and the fine vertical striations are individual flux tubes each ~100 m to 1 km wide.

A second consequence is the magnetic mirror. As a spiralling particle moves toward the pole into stronger B, conservation of the magnetic moment μ = ½mv⊥²/B forces perpendicular energy to grow at the expense of parallel motion, and the particle can be reflected. Particles inside the loss cone — those with too little perpendicular velocity — instead penetrate to the atmosphere and precipitate. It is precisely these loss-cone electrons that make the aurora.

The accelerator: field-aligned currents and the auroral potential

Reconnection alone leaves electrons with hundreds of eV — enough to reach the atmosphere but too diffuse for bright discrete arcs. The sharp, structured curtains require an extra parallel electric field that accelerates electrons downward along B, typically through a potential drop of 1–10 kV at altitudes of 4,000–12,000 km. This is the auroral acceleration region, mapped in detail by NASA's FAST satellite in the late 1990s.

The circuit is closed by field-aligned (Birkeland) currents, named for Kristian Birkeland, who proposed them in 1908. Solar-wind-driven flow in the outer magnetosphere acts as a generator; current flows down one set of field lines, horizontally through the conducting ionosphere, and back up another, carrying on the order of 10⁶ A total. Where the current demand exceeds what the tenuous plasma can supply, the plasma responds by building a parallel electric field to accelerate charge carriers — the same physics as a vacuum diode running short of emission. An electron dropped through a 5 kV drop arrives with 5 keV of energy, travelling at about 0.13c ≈ 4×10⁷ m/s.

• Energy gained: ΔE = qΔΦ = e × 5000 V = 5 keV.
• Higher potential ⇒ faster electrons ⇒ deeper penetration ⇒ brighter, lower, greener aurora.

The paintbrush: collisional excitation and forbidden lines

When a keV electron plunges into the thermosphere it undergoes a cascade of inelastic collisions, ionizing and exciting atmospheric atoms and losing ~35 eV per ionization. Each excited atom later relaxes by emitting a photon whose colour is set by the quantum energy-level spacing, ΔE = hν = hc/λ. The dominant emitter is atomic oxygen, and its two famous lines are forbidden transitions — electric-dipole-forbidden, so the excited states are metastable and long-lived:

• Green 557.7 nm: the O(¹S → ¹D) transition, ΔE = 2.22 eV, upper-state lifetime ≈ 0.7 s.
• Red 630.0 nm: the O(¹D → ³P) transition, ΔE = 1.97 eV, upper-state lifetime ≈ 110 s.

Because these states live so long, they can be destroyed before radiating by a collision — collisional quenching. Below ~200 km the air is dense enough that the 110-second red state almost never survives to emit, so red aurora appears only at high, thin altitudes (200–400 km). The 0.7-second green state survives down to ~100 km, where oxygen is still abundant and electron flux peaks — hence green is the workhorse colour of the aurora. Blue and purple fringes at the bottom come from ionized molecular nitrogen (N₂⁺) at ~110 km; deep-red lower borders come from N₂ first-positive bands.

Reading the colours: altitude as a spectrometer

Because each species and state has a characteristic quenching altitude, the vertical structure of an aurora is effectively a spectrograph you can read with your eyes. The Chapman profile describes how a beam of monoenergetic electrons deposits its energy: penetration depth depends on how much column mass (integrated density) the beam must traverse, so energy maps almost one-to-one onto altitude.

• Soft precipitation (~0.5–2 keV) stops high, ~250–400 km, exciting the red 630.0 nm line — diffuse red glows and the tops of tall rays.
• Medium precipitation (~2–10 keV) stops near 100–150 km, exciting the green 557.7 nm line — the bright, sharp arcs.
• Hard precipitation (>15 keV) penetrates below 100 km, exciting N₂⁺ blue/violet emission along the lower border — the mark of an intense, active substorm.

Thus a curtain that is red at the top, green in the middle, and violet along its base is telling you the incoming electron spectrum: soft on top, hardening downward. The rapid flickering and folding (0.1–10 Hz) reflects the dynamics of the acceleration region and Alfvén waves bouncing along the field lines, whose characteristic speed v_A = B/√(μ₀ρ) reaches thousands of km/s in the magnetosphere.

Scales, storms, and technological consequences

The full auroral display balances at spectacular scales. The auroral oval is a ring 3,000–4,000 km in diameter centred on the geomagnetic pole, sitting near 67° magnetic latitude when quiet and expanding equatorward during storms. During a strong substorm the precipitating electron flux can exceed 10 erg cm⁻² s⁻¹ (10⁻² W/m²), and integrated over the oval the deposited power reaches ~10¹¹ W (100 GW). Brightness is measured in rayleighs (1 R = 10⁶ photons cm⁻² s⁻¹ into 4π); a faint arc is ~1 kR while a brilliant display exceeds 100 kR.

The same energy that paints the sky perturbs the ground. The horizontal ionospheric currents that close the Birkeland circuit — the auroral electrojets, ~10⁶ A — vary rapidly, and by Faraday's law dΦ/dt a changing magnetic field induces geomagnetically induced currents (GICs) in long conductors. The March 1989 storm collapsed the Hydro-Québec grid in 90 seconds, and the 1859 Carrington event set telegraph offices on fire. Auroral activity also heats and inflates the thermosphere, increasing satellite drag, and it disrupts HF radio and degrades GNSS positioning through ionospheric scintillation.

Subtleties and common misconceptions

Several intuitive pictures of the aurora are simply wrong, and correcting them clarifies the physics:

• It is not the solar wind hitting the atmosphere directly. The solar wind almost never reaches the ground; it deposits energy into the magnetosphere, which stores it and later accelerates a separate population of electrons already trapped in near-Earth space. The aurora is powered by the Sun but not made of solar particles.
• The lights are not hot. The emitting oxygen sits at ~1000 K but at densities so low that the total thermal content is negligible; you would feel nothing standing beneath an arc. The glow is a non-thermal, quantum line emission, not blackbody heat.
• Colour is set by physics, not altitude alone. Red is not merely 'high aurora' — it is high aurora because the 110 s metastable state can only radiate where collisions are rare. Change the quenching rate and you change the colour map.
• Aurorae occur on other planets too. Jupiter's aurorae, powered mainly by its own rotation and volcanic Io rather than the solar wind, radiate in the ultraviolet and outshine Earth's by orders of magnitude — proof that the essential ingredients are a magnetic field, a plasma source, and an atmosphere, not the Sun specifically.

The two dominant auroral emission lines of atomic oxygen — why altitude and density decide the colour

Property	Green line (¹S → ¹D)	Red line (¹D → ³P)
Wavelength	557.7 nm	630.0 nm & 636.4 nm
Photon energy	2.22 eV	1.97 eV
Excited-state lifetime	≈ 0.7 s	≈ 110 s
Emission altitude	~100–120 km	~200–400 km
Quenched below	Survives to ~95 km	Collisionally quenched below ~200 km
Requires	Faster electrons, denser air	Slower electrons, thin air