Aurora Borealis Colors: Why Northern Lights Are Green, Red, and Blue

Green is the color most people associate with the northern lights, and with good reason: it is by far the most common, brightest, and most easily visible auroral emission. It comes from a single, specific source: atomic oxygen at approximately 100–150 km altitude, emitting a yellow-green photon at 557.7 nm wavelength.

The physics involves what spectroscopists call a forbidden transition — an energy transition that is technically "forbidden" by the standard quantum selection rules but occurs slowly in practice. When an incoming electron collides with an oxygen atom and excites it to the ¹S excited state, the atom would ideally release that energy immediately. But the transition from ¹S to the lower state is statistically slow — it takes about 0.7 seconds on average for the atom to emit its photon. In the dense lower atmosphere, the atom would collide with another particle in far less than 0.7 seconds and dissipate the energy without emitting light. At 100–150 km altitude, however, the atmosphere is so rarefied that collisions are rare enough to allow the emission to occur. This is why green aurora only appears above a certain minimum altitude.

The intensity and brightness of the green emission is directly proportional to the flux of energetic electrons precipitating along that field line. During a strong substorm, the green emission can be bright enough to cast faint shadows on snow — an intensity comparable to a full moon. During quiet conditions it may appear as a faint, greenish glow near the horizon that is almost invisible to dark-adapted eyes but shows clearly on a long-exposure photograph.

The exact shade of green varies from a pure yellow-green to a more saturated emerald depending on the altitude of peak emission and whether other emissions are mixed in. Photographs sometimes show a brighter, purer green than the eye perceives because camera sensors have different spectral sensitivity curves than the human eye — a distinction explored further in the final section of this guide.

Red aurora is produced by the same atom as green — atomic oxygen — but at much higher altitudes, typically above 200 km. The explanation lies in a different forbidden transition: oxygen in the ¹D excited state, which emits red light at 630.0 nm and 636.4 nm wavelengths. This transition takes even longer than the green one — up to 110 seconds — meaning it only occurs in the extreme vacuum of the upper thermosphere where collision intervals exceed two minutes.

This is why red aurora is less common than green: the emission altitude must be above 200 km, which requires either a particularly energetic geomagnetic event (to heat the upper thermosphere and increase oxygen density there) or precipitating particles of relatively low energy (which are stopped at higher altitudes before reaching the 100 km green layer). Red auroras appear most often during intense geomagnetic storms (KP 6+) or in specific diffuse aurora structures called SAR arcs (Stable Auroral Red arcs), which form through different heating mechanisms.

Visually, red aurora often appears as a faint crimson or blood-red glow above and sometimes completely disconnected from the green emission below. During the most spectacular geomagnetic storms, the top of green auroral curtains is capped by a vivid red fringe, and the entire high-latitude sky can appear stained deep red — visible even from mid-latitude cities during major events. Historical descriptions of "fire in the sky" from ancient civilizations were almost certainly accounts of high-altitude red aurora during exceptional storms.

Red aurora is also seen at latitudes far from the auroral zone during major geomagnetic events (KP 8–9+). Because 200+ km altitude emissions subtend a much wider geographic angle at the observer's horizon, red aurora from the outer edge of the auroral oval reaches observers much further equatorward than the green emission below it. This is why during the famous 1989, 2003, and 2024 major storms, reports of "blood red lights in the sky" came from places as far equatorward as Spain, Texas, and New Zealand.

While oxygen atoms dominate the upper auroral emission layers, molecular nitrogen (N₂) and ionized molecular nitrogen (N₂⁺) produce a range of blue and purple emissions at lower altitudes, typically between 60 and 100 km. These emissions are not forbidden transitions but rather allowed band emissions — nitrogen molecules excited and ionized by precipitating electrons emit photons across a range of wavelengths spanning the blue to near-ultraviolet spectrum.

The most prominent nitrogen emission bands cluster around 391–470 nm in the blue-violet range. These are produced by N₂⁺ (ionized nitrogen) in an excited electronic state. Because these transitions are fast (allowed transitions), they occur readily even at the denser lower altitudes — unlike the slow forbidden oxygen transitions that require high-altitude low-density conditions. As a result, nitrogen blue emissions appear at the base and lower edges of auroral arcs, giving tall curtains a distinctive blue or purple fringe at their bottom edge.

Nitrogen also contributes red and infrared emissions (around 650–900 nm), but these are generally weaker and overwhelmed by the oxygen red emission and the general ambient infrared emission of the thermosphere. The contribution of nitrogen becomes most visually striking during very energetic events when electrons penetrate deep into the atmosphere to the 60–80 km mesosphere/lower thermosphere boundary, producing the vivid blue-purple base structures sometimes called "dune aurora" or seen at the feet of bright substorm arcs.

The combination of green oxygen (100+ km), blue-purple nitrogen (80–100 km), and red oxygen (200+ km) gives well-developed auroral curtains a characteristic vertical color gradient: red at top, green through the middle, blue-purple at the base. This layered structure is sometimes visible to the naked eye during bright displays but is captured most faithfully by the camera — another reason aurora photography can reveal structure invisible in naked-eye viewing.

White and pink aurora represent special cases where multiple emission sources blend into mixed-color light rather than displaying distinct spectral bands.

White aurora appears when multiple emission components — green, red, and blue — are present simultaneously in roughly equal proportions and at overlapping altitudes. The eye's inability to spectrally separate these emissions perceives the mixture as white or whitish-gray. White aurora is most commonly seen during particularly intense substorms when all atmospheric layers from 80 to 200+ km are simultaneously emitting, or in auroral structures so bright that they overwhelm the color receptors in dark-adapted human vision. (The cone cells responsible for color perception require a certain light level to operate; very bright aurora may paradoxically appear whiter because it saturates the color-sensitive cones.)

Pink aurora — sometimes described as salmon or magenta — results from a specific combination of the red oxygen emission (at the top of an arc) and the blue nitrogen emission (at the bottom of the same arc). When these are superimposed spatially — either through a very tall arc where the emission columns overlap, or through the optical mixing of light from different altitude layers in the same line of sight — the result is a warm pink or magenta that can be quite vivid. Pink often appears at the base of intense substorm arcs or along the lower edge of tall curtains during the peak of strong geomagnetic storms.

There are also specialized phenomena that produce unusual colors:

SAR arcs (Stable Auroral Red) appear as diffuse, wide bands of deep red light at lower latitudes than typical aurora, caused by thermal excitation of oxygen in the plasmasphere during geomagnetic recovery phases rather than by direct particle precipitation. They are often so faint as to be invisible to the naked eye but brilliantly red in long-exposure photography.

Beyond aurora itself, the STEVE (Strong Thermal Emission Velocity Enhancement) phenomenon produces a distinctive purple or mauve streak of light at sub-auroral latitudes. Discovered by citizen scientists and studied intensively since 2017, STEVE is not technically aurora (it is caused by a different ionospheric current system) but frequently accompanies aurora displays and appears most prominently in long-exposure aurora photographs.

One of the most common questions from first-time aurora viewers is: "Why do my photos look so much more colorful than what I see with my eyes?" The answer lies in fundamental differences between human visual physiology and digital camera sensors.

Human eyes contain two types of photoreceptors: cones (for color vision) and rods (for low-light black-and-white vision). At very low light levels — like a dark arctic night — the eye shifts to predominantly rod-based vision through a process called dark adaptation. Rods are monochromatic: they detect light intensity but not color. The green aurora at 557.7 nm falls near the peak sensitivity of the rods, so a moderately bright green display appears greenish to most people — but the subtler colors (red, blue, purple) may appear as gray or near-black because they stimulate the rods less efficiently.

A digital camera sensor, by contrast, behaves like a long-duration integrator — it accumulates light over the entire exposure time (typically 4–30 seconds for aurora photography). Even a very faint red aurora that the eye cannot detect will register clearly on a 20-second exposure at ISO 3200, because the camera adds up all the photons that arrive during those 20 seconds. The result is that red tones, blue nitrogen fringes, and subtle pink edges are captured by the camera but missed entirely by the eye.

Additionally, camera sensors have relatively flat spectral sensitivity across the visible spectrum — they respond to red and blue light with similar efficiency to green. Human cones are not: the three cone types have overlapping sensitivities but are centered on different wavelengths, and the overall system is most sensitive to yellow-green light (around 555 nm, very close to the aurora green line at 557.7 nm). Red wavelengths stimulate a narrower range of cones less strongly, making the eye relatively less sensitive to red light in dim conditions.

The practical implication: during a moderate aurora display, your naked eye may see a greenish glow with occasional movement. Your camera on the same scene may reveal a complex, colorful structure of green curtains with red tops, purple-blue bases, and subtle STEVE features you missed entirely. This is not the camera "lying" — it is capturing physical reality that your visual system cannot resolve in darkness. Understanding this difference is one of the reasons amateur aurora photography has become so popular: the camera reveals layers of the phenomenon invisible in real time.