№ 164 · earth science

An aurora's red and green are the same oxygen at different heights

The red and the green in an aurora come from the same atom: oxygen. What decides which colour you see is how long that atom has to wait before it can glow, and whether the air around it is thin enough to leave it alone for that long. So an aurora's colours work like a height chart.

Light from a falling electron

An aurora starts when electrons from the space around Earth stream down along the magnetic field into the upper atmosphere. They crash into atoms and molecules of air and leave them excited, holding a little extra energy. When an excited particle settles back down, it gives that energy off as light of one exact colour. That is why, as Nanjo and colleagues put it in a 2026 paper, an aurora's spectrum is a set of separate lines rather than a smooth rainbow.

Barthelemy and colleagues, who built a small satellite camera to photograph auroras, list the brightest visible lines. There is green at 557.7 nanometres and red at 630 nanometres, both from atomic oxygen, and blue at 427.8 nanometres from nitrogen molecules that have lost an electron (N₂⁺).

Predict first

Aurora curtains are often green with a red top. What makes the red?

Check your answer

The last one. Both the green line and the red line come from atomic oxygen. The red needs an atom to wait about 110 seconds before it glows, and only air above about 200 km is thin enough to leave it alone that long (Nanjo and colleagues, 2026).

The waiting game

Not every excited atom glows at once. Nitrogen's blue light is prompt: Whiter and colleagues give the 427.8 nm line a lifetime of about 270 billionths of a second, far too short for a collision to get in first. Oxygen is slower. The atom that makes the green line waits about 0.7 seconds on average, and the atom that makes the red line waits about 110 seconds.

A wait that long is risky. If the excited atom bumps into another particle first, the collision can carry its energy away and no light comes out. Physicists call this quenching. Lower down, the air is denser and collisions come faster, so the slow lines are erased first.

Nanjo and colleagues give the heights. Above about 200 km the red oxygen line takes over. The green line usually peaks around 110 to 120 km. Below about 95 km even green is quenched, and the prompt light from nitrogen shows instead. Hence the rule of thumb they give: red high up, green in the middle, pinkish at the bottom. In a large study of all-sky camera images from northern Finland and Sweden, taken over seven winters from 2000 to 2007, Whiter and colleagues found that the green and the blue both typically peak at about 114 km.

Interactive One column of aurora, from 70 to 420 km up. Slide the electron energy to choose how deep the incoming electrons reach, or pick a preset. Turn collisions off to see what the sky would show if no excited atom were ever knocked out of its glow. Every control works with Tab, Enter and the arrow keys.

electron energy2.0 keV
dark brightest red survives above ~200 km green quenched below ~95 km — red 630.0 nm, oxygen (waits ~110 s) — green 557.7 nm, oxygen (waits ~0.7 s) — blue 427.8 nm, N₂⁺ (prompt) -- deep red bands, N₂ (prompt) - - energy the electrons deposit
Move the slider. Slow electrons stop high up, fast ones reach down toward 100 km.

Electron energy chooses the height

How deep the electrons get depends on how fast they arrive. Slow electrons give up their energy high in the thin air, so they light the red oxygen line. Fast ones punch down to the dense air near 100 km, where they light green and, at the bottom edge, nitrogen's deep red bands. Nanjo and colleagues tie the red oxygen lines to low-energy electrons at high altitude, and the nitrogen red bands to high-energy electrons lower down. Whiter and colleagues treat green and blue that peak above about 110 km as the mark of low-energy electrons, under about 4 keV.

Why auroras far from the poles look red

During big storms, auroras seen from countries far from the poles, such as Japan, are typically red, as Kataoka and colleagues note. On 23 April 2023 a camera at Hanle in Ladakh, India, recorded a red glow low on the northern horizon, 5 to 10 degrees up. Vichare and colleagues traced it to slow electrons, under 100 eV, falling around 54° north. These lit red oxygen at roughly 700 to 950 km, high enough to be seen over the curve of the Earth from India. The aurora itself never reached India's latitudes. Only its top was visible.

The giant storm of 10 to 11 May 2024 gave Japan something different: magenta. Kataoka and colleagues used hundreds of photographs sent in by the public to work out where the light came from: about 1,000 km up. Over the region where it shone, the air above 350 km stayed in sunlight all night. Their explanation is red oxygen light mixed with blue light from N₂⁺, much of it ions scattering sunlight. The aurora was mostly too faint for the naked eye. Phone and camera sensors caught the colour.

In short

An aurora is atoms giving back energy that falling electrons gave them. Oxygen makes both the green and the red, but the red needs about 110 seconds of peace, which only the thin air above about 200 km provides. Green survives down to about 95 km, and below that nitrogen's prompt light paints the lower edge pink. Read from the bottom up, the colours tell you how high the light is, and how hard the electrons hit.

Where this comes from

  1. Low-Latitude Auroras: Insights from 23 April 2023 Solar Storm reuse permitted with attribution
    Geeta Vichare, Ankush Bhaskar, Rahul Rawat, Virendra Yadav, Wageesh Mishra, Dorje Angchuk, Anand Kumar Singh · arXiv:2405.08821 · 2024
    arxiv.org/abs/2405.08821
  2. AMICal Sat: A sparse RGB imager on board a 2U cubesat to study the aurora reuse permitted with attribution
    Mathieu Barthelemy, Elisa Robert, Vladimir Kalegaev, Vincent Grennerat, Thierry Sequies, Guillaume Bourdarot, Etienne Le Coarer, Jean-Jacques Correia, Patrick Rabou · arXiv:2201.06973 · 2022
    arxiv.org/abs/2201.06973
  3. Calibration and observations of an AI-triggered commercial digital camera for auroral studies (Earth, Planets and Space 78, 112) linked only, not reproduced
    Sota Nanjo, Urban Brändström, Satonori Nozawa, Tetsuya Kawabata, Keisuke Hosokawa · 2026
    link.springer.com/article/10.1186/s40623-026-02451-6
  4. The altitude of green OI 557.7 nm and blue N2+ 427.8 nm aurora (Annales Geophysicae 41, 1–12) linked only, not reproduced
    Daniel K. Whiter, Noora Partamies, Björn Gustavsson, Kirsti Kauristie · 2023
    angeo.copernicus.org/articles/41/1/2023/
  5. Extended magenta aurora as revealed by citizen science (Scientific Reports 14, 25849) linked only, not reproduced
    Ryuho Kataoka, Sachin Alexander Reddy, Shinya Nakano, Joshua Pettit, Yuki Nakamura · 2024
    www.nature.com/articles/s41598-024-75184-9