The Physics of Auroras: Why the Sky Glows Green at the Poles
The northern lights are not reflected sunlight. They're atoms in the upper atmosphere being smacked by solar wind particles, then glowing as they de-excite. The physics involves magnetism, plasma, and quantum energy levels.
Table of Contents
Not What the Vikings Thought
For most of human history, auroras were mysteries. The Norse thought they were reflections off the armour of the Valkyries. The Finnish word for aurora — revontulet — means “fox fires,” from a legend about a fox sweeping snow into the sky with its tail. Chinese records describe them as celestial dragons.
The real explanation is, honestly, almost as dramatic as the myths. Auroras happen when the Sun throws a billion tonnes of charged particles at the Earth, those particles get caught in Earth’s magnetic field, spiral down toward the poles along field lines, slam into atmospheric atoms at thousands of kilometres per second, and those atoms glow as their excited electrons drop back to lower energy levels.
It’s a process that spans 150 million kilometres — from the Sun’s surface to Earth’s upper atmosphere — and involves plasma physics, magnetohydrodynamics, and quantum atomic transitions. The fact that the end result is one of the most beautiful things you can see with your naked eyes is a coincidence that physics doesn’t owe us, but I’m glad it delivered.
The Solar Wind
The story starts at the Sun. The Sun’s corona — its outer atmosphere — is so hot (1–3 million kelvin) that it can’t be gravitationally bound. Particles at the top of the corona have enough thermal energy to escape into interplanetary space. They stream outward in all directions as the solar wind — a continuous flow of mostly protons and electrons, moving at 400–800 km/s.
The solar wind is tenuous by terrestrial standards — roughly 5 particles per cubic centimetre at Earth’s orbit, compared to 2.5 × 10¹⁹ in a cubic centimetre of air at sea level. But what it lacks in density it makes up in speed and extent. The solar wind carries about 10⁹ kg of material per second and fills the entire solar system out to the heliopause (about 120 AU), where it meets the interstellar medium.
During solar storms — coronal mass ejections (CMEs) — the Sun ejects billions of tonnes of plasma in a single event. These clouds of magnetised plasma travel at 500–3,000 km/s and take 1–3 days to reach Earth. When a CME hits Earth’s magnetosphere, the result is a geomagnetic storm — and auroras can flare dramatically, expanding from their usual polar positions toward much lower latitudes.
The Magnetosphere: Earth’s Shield
Earth has a global magnetic field generated by convection currents in its liquid iron outer core — a self-sustaining dynamo. This field extends far into space, forming the magnetosphere — a magnetic bubble roughly 65,000 km toward the Sun and stretched into a long tail millions of kilometres downwind.
The magnetosphere deflects most of the solar wind around the Earth, like a rock in a stream. Without it, the solar wind would gradually strip away the atmosphere (this is what happened to Mars, which lost its global magnetic field about 4 billion years ago and subsequently lost most of its atmosphere).
But the deflection isn’t perfect. When the solar wind’s magnetic field is oriented opposite to Earth’s field (southward in the northern hemisphere), the fields can reconnect — merge — in a process called magnetic reconnection. This opens temporary “cracks” in the magnetosphere and allows solar wind particles to enter, flowing along field lines into the polar regions.
The particles that enter are channeled — funneled — toward the magnetic poles by the converging field lines. They spiral around the field lines (because charged particles follow helical paths in magnetic fields, governed by the Lorentz force) and accelerate as they approach the more intense field near the poles. By the time they reach the upper atmosphere at 100–300 km altitude, they’re moving at up to 70 million metres per second — a significant fraction of the speed of light.
What Happens When Particles Hit the Atmosphere
At altitudes of 80–300 km, the incoming electrons (and some protons) collide with atmospheric molecules — primarily nitrogen (N₂) and oxygen (O₂ and atomic O). These collisions transfer energy to the atmospheric atoms, exciting their electrons to higher energy levels.
When the excited electrons fall back to their ground state, they release the energy as photons — light. The specific wavelength of light depends on the atom and the energy transition involved. This is where the colours come from.
Green (557.7 nm): The most common aurora colour. Produced by atomic oxygen at 100–300 km altitude. The excited state involved (the ¹S state) has a radiative lifetime of about 0.7 seconds — it’s a “forbidden” transition, meaning it’s slow by quantum standards. The atom has to avoid being bumped by another particle for almost a second before it can emit. This works at high altitudes where the atmosphere is thin. Lower down, collisions quench the emission.
Red (630.0 nm): Also atomic oxygen, but a different transition (¹D state) with a much longer lifetime — about 110 seconds. This only works at very high altitudes (above 250 km) where collisions are extremely rare and the atom has nearly two minutes of undisturbed time to radiate. Red auroras form the upper edges of strong displays.
Blue and violet (391–470 nm): Produced by ionised nitrogen molecules (N₂⁺). These transitions are fast (allowed transitions), so they occur at lower altitudes where the atmosphere is denser. Blue and violet auroras appear at the lower edges of bright displays and during strong storms.
Pink: A mix of green oxygen emission and blue/red nitrogen emission, seen at the lowest altitudes during intense storms.
The colour profile of an aurora — green band, red top, blue-violet fringe — is a vertical map of the atmosphere. Each colour corresponds to a specific altitude range, determined by the interplay between collision rates and radiative lifetimes. You’re literally looking at quantum atomic physics painted across 200 km of sky.
The Auroral Oval
Auroras don’t occur randomly across the sky. They’re concentrated in two roughly circular bands — the auroral ovals — centered on the magnetic poles (not the geographic poles, which are offset by about 11°). The northern oval typically sits at about 65–75° magnetic latitude, arcing over northern Scandinavia, Iceland, northern Canada, Alaska, and Siberia. The southern oval mirrors it around Antarctica.
The ovals aren’t static. During geomagnetic storms, they expand toward the equator as more solar wind particles penetrate the magnetosphere. During the strongest storms, the ovals can reach 45° latitude or lower — bringing auroras to places like London, Paris, or New York, where they’re normally never seen.
The ovals are also not uniform in brightness. The brightest region is typically the midnight sector (the side of the oval facing away from the Sun), where particles precipitating from the magnetotail produce the most intense displays. The noon sector is usually dimmer, with diffuse “dayside auroras” produced by direct solar wind interaction.
If you want to see an aurora, your best bet is to be under the midnight sector of the oval during a geomagnetic storm — which, practically speaking, means being in northern Scandinavia, Iceland, northern Canada, or Alaska between September and March, on a dark, clear night, ideally during solar maximum (the peak of the Sun’s 11-year activity cycle).
Substorms: The Auroral Breakup
The most spectacular auroral displays are caused by magnetospheric substorms — sudden releases of energy stored in Earth’s magnetotail.
Here’s what happens. The solar wind stretches Earth’s magnetic field on the nightside into a long tail. Energy from the solar wind is stored in this stretched field — like pulling a rubber band. When the stored energy exceeds a threshold, the field lines in the tail reconnect explosively. This releases the stored magnetic energy, accelerating particles down toward the poles in a burst.
The result is an auroral substorm — a dramatic brightening and rapid expansion of the aurora, often with fast-moving curtains, corona formations (rays converging overhead), and pulsating patches. A substorm can go from calm to spectacular in minutes, and a single night can see multiple substorms.
The particle energy during a substorm increases dramatically — electrons are accelerated to 10–100 keV, sometimes higher. This drives the aurora to lower altitudes and produces brighter, more colourful displays with more blue and violet (from nitrogen emission at lower altitudes where the energetic particles penetrate deeper).
Space Weather
Auroras are beautiful, but they’re the visible symptom of something that can be dangerous: geomagnetic storms.
The same currents that produce auroras also induce electric currents in the ground — geomagnetically induced currents (GICs). These flow through any long conductor connected to the ground: power lines, pipelines, railway tracks, telecommunications cables. In power grids, GICs can saturate transformer cores, causing overheating and, in extreme cases, permanent damage.
The most famous example is the March 1989 geomagnetic storm, which knocked out the Hydro-Québec power grid in 90 seconds, leaving 6 million people without electricity for up to 9 hours. The Carrington Event of 1859 was far more powerful — telegraph systems worldwide failed, and operators reported receiving shocks and seeing sparks from their equipment.
A Carrington-scale event today would be catastrophic. Modern power grids are far more extensive and interconnected, satellite systems would be disrupted, GPS accuracy would degrade, and high-frequency radio communications would be blacked out for hours. Studies estimate the economic damage from a modern Carrington-class storm at $1–2 trillion.
The physics of forecasting such events — predicting when the Sun will erupt, what direction the CME will travel, and how the magnetosphere will respond — is the domain of space weather research. We’re getting better at it, but predictions are still imprecise. We typically get 1–3 days of warning after a CME is detected leaving the Sun. Sometimes less.
Looking Up
There’s something about auroras that short-circuits the usual detachment of physics. You can know exactly what’s happening — solar wind, magnetospheric reconnection, electron precipitation, forbidden line emission from excited atomic oxygen — and it still takes your breath away when you see it.
The green light is oxygen atoms radiating at 557.7 nanometres. The curtains are shaped by Earth’s magnetic field lines. The flickering is substorm dynamics playing out in real time. Everything has an explanation. None of the explanations make it less astonishing.
That’s unusual in physics, honestly. Most phenomena become less impressive the better you understand them. Auroras go the other direction. The more you know about what’s happening — that the Sun has flung a billion tonnes of plasma across 150 million kilometres, that Earth’s magnetic field is catching it and threading it down toward the poles, that individual atoms a hundred kilometres above your head are flipping between quantum states and emitting photons that land in your retinas — the more remarkable the whole thing seems.
Frequently Asked Questions
Why are auroras usually green?
The most common aurora colour — green — comes from oxygen atoms at altitudes of about 100–300 km. When a solar wind electron collides with an oxygen atom, it can excite one of oxygen's electrons to a higher energy level. When this electron falls back down, it emits a photon at 557.7 nm — green light. This particular transition is 'forbidden' in quantum mechanics jargon, meaning it happens slowly (the excited state lasts about 0.7 seconds). At higher altitudes, where the atmosphere is thinner and collisions are less frequent, the oxygen atom has time to emit before being de-excited by another collision. At lower altitudes, collisions quench the emission before it can happen. The green glow sits in a sweet spot of altitude where the excited oxygen lives long enough to radiate.
Can you see auroras from the equator?
Rarely, but yes — during extremely strong geomagnetic storms. In normal conditions, auroras are confined to the auroral ovals, roughly 65–75° magnetic latitude. But during severe solar storms (like the Carrington Event of 1859), the auroral oval expands dramatically toward the equator. The Carrington Event produced auroras visible in Cuba, Hawaii, and Colombia. More recently, the May 2024 geomagnetic storm (caused by an X5.8 solar flare) produced aurora sightings as far south as northern Mexico and the Mediterranean. These events are rare — perhaps once or twice per solar cycle — but they do happen.
Do other planets have auroras?
Yes. Any planet with a magnetic field and an atmosphere can produce auroras. Jupiter has spectacular auroras at both poles, powered largely by charged particles from its volcanic moon Io rather than the solar wind. Saturn, Uranus, and Neptune all show auroral emissions. Mars, which has only remnant crustal magnetic fields (no global dipole), has patchy auroras that occur over regions of strong crustal magnetism. Even some moons — Ganymede, which has its own magnetic field — exhibit auroral glows. The physics is the same everywhere: charged particles follow magnetic field lines into the atmosphere, excite atoms and molecules, and the atoms emit light as they de-excite.