Optics

The 22° Halo: Why a Bright Ring Circles the Sun

The 22° Halo is the luminous ring you sometimes catch encircling the Sun or Moon when a thin sheet of high cirrus crosses the sky. It is not a cloud effect and not a rainbow but a piece of precise geometric optics: sunlight refracting through countless hexagonal ice crystals, each acting as a tiny 60° prism. Because a prism bends light by no less than a fixed minimum amount — about 21.8° for ice — the deflected light piles up at that angle and nowhere closer, drawing a sharp-edged bright ring exactly 22° from the Sun, with a curiously dark disk of sky inside it.

  • Also called22° halo, small halo, ice-crystal halo
  • Angular radius≈ 21.8° (sharp inner edge)
  • Refracting angle60° (alternate faces of a hexagonal ice prism)
  • Refractive index of icen ≈ 1.31 (visible light)
  • Governing principleMinimum deviation of a 60° prism
  • First explained byEdme Mariotte, ~1680s; angles by Bravais, 1847

Interactive visualization

Press play, or step through manually. The visualization is yours to drive — try it before reading on.

Open visualization fullscreen ↗

Watch the 60-second explainer

A condensed visual walkthrough — narrated, captioned, under a minute.

The setup: sunlight, ice, and a 60° prism

On a day when a thin veil of cirrostratus whitens the sky, sunlight is passing through millions of tiny hexagonal ice crystals — pencil-shaped columns and flat plates a few tenths of a millimetre across. A hexagonal prism has six long side faces; any two alternate faces meet, when extended, at a 60° apex, so a ray that enters one side face and leaves the second-next face is refracted exactly as if it had crossed a 60° glass prism. Ice bends light with refractive index n ≈ 1.31, and the fixed geometry of that bending is what paints a luminous ring at one precise angle around the Sun.

Minimum deviation: why the light piles up at 22°

A prism does not bend every ray by the same amount. The total deviation D — the angle between the incoming and outgoing ray — depends on how the ray strikes the crystal, and it has a minimum. For a prism of apex A and index n, that minimum satisfies

n = sin((A + D_min)/2) / sin(A/2)

With A = 60° and n = 1.31 this gives D_min ≈ 21.8°. The key is what happens near that minimum: as a crystal tumbles, its orientation sweeps the incidence angle through a wide range, but D changes very slowly close to its minimum. A huge span of orientations therefore funnels light into a narrow band of deviation right at 21.8° — an optical pile-up called a caustic. And crucially, no orientation deviates light by less than 21.8°, so no refracted sunlight lands inside that angle. The result is a ring with a sharp bright inner edge, a darker disk of sky inside it, and a gentle fade outward.

A worked example: where 21.8° comes from

Follow the symmetric ray — the one at minimum deviation, which crosses the crystal parallel to its base:

  1. By symmetry the ray refracts equally at both faces, so inside the ice it makes an angle r = A/2 = 30° with each face's normal.
  2. Snell's law at the entry face: sin i = n · sin r = 1.31 × sin 30° = 0.655, so the incidence angle i = 40.9°.
  3. Each face bends the ray by i − r = 40.9° − 30° = 10.9°.
  4. Two faces double it: D_min = 2 × 10.9° ≈ 21.8°.

That single number — set only by the 60° geometry of ice and its refractive index — is the radius of the halo. Rays that strike more steeply or more shallowly are all thrown outside 21.8°, which is why the ring is not a thin line but a band that brightens abruptly at its inner rim.

Color, width, and the sharp inner edge

Because n depends on wavelength, the minimum-deviation angle differs slightly for each colour. Ice refracts blue (n ≈ 1.317) a touch more than red (n ≈ 1.307), so the red minimum sits near 21.6° and the violet near 22.4°. Red is deviated least and forms the inner edge of the ring, closest to the Sun — the reverse of a rainbow's order. The whole coloured spread is under 1°, and the finite Sun (0.53° across) smears each colour by about half a degree, so the tints overlap and mostly wash out. The eye sees a largely white ring with a distinct reddish inner rim and a bluish-white outer glow — never the vivid banding of a true rainbow.

Why a full ring? Random orientation vs. sundogs

A closed ring requires the crystals to be randomly oriented — tumbling every which way — so every rotational orientation is represented and light emerges at 21.8° all the way around the Sun. This is the condition people most often overlook. If instead the crystals are flat plates settling with their broad faces horizontal (as larger, slower crystals do in still air), the 60° geometry is sampled only in a horizontal plane. The light then concentrates into two bright, often coloured patches at the same altitude as the Sun and about 22° to either side — sundogs (parhelia), not a ring. Same prism, same 22°; different crystal orientation, completely different picture. Horizontal-axis columns instead build the bright upper tangent arc perched on top of the halo.

History, folklore, and where you see it

Halos are among the oldest recorded sky phenomena — Aristotle noted them, and a famous 1629 display over Rome prompted Descartes to write about halos in his 1637 Les Météores. The correct ice-crystal refraction explanation is usually credited to Edme Mariotte in the 1680s; the full geometry, including the precise 21°50′ radius and the 46° halo, was worked out by Auguste Bravais around 1847. Halos are far more common than rainbows — seen on the order of a hundred days a year in many climates, whenever high cirrus drifts across the Sun or Moon (a lunar halo is the identical effect, merely dimmer). The saying "ring around the Moon means rain" even has real physics behind it: the cirrostratus that makes halos often runs ahead of an approaching warm front, so a halo genuinely can precede a change in weather — though far from reliably.

How the 22° halo compares with the primary rainbow and its larger, rarer cousin, the 46° halo.
Feature22° Halo42° Rainbow (primary)46° Halo
Refracting particleHexagonal ice crystalsSpherical water dropsHexagonal ice crystals
Light pathRefraction only, 60° prismRefraction + 1 internal reflectionRefraction only, 90° prism
Angular radius≈ 21.8°≈ 42°≈ 46°
Where in the skyRing centered on the SunArc opposite the SunLarger ring on the Sun
Color orderRed on the inner edgeRed on the outer edgeRed on the inner edge
How often seenCommon (~100 days/yr)Common after showersRare and faint

Frequently asked questions

Why is the sky darker inside the halo than outside?

Because a 60° ice prism cannot deviate sunlight by less than its minimum of about 21.8°. No crystal orientation sends refracted light into the disk inside the ring, so that region stays dark, while the deviation piles up as a caustic right at 21.8° to make the bright inner edge.

Is the 22° halo the same as a rainbow?

No. A rainbow comes from spherical water drops using refraction plus one internal reflection, appears ~42° from the point opposite the Sun, and shows red on its outer edge. A halo comes from hexagonal ice crystals using refraction alone, encircles the Sun itself at ~22°, and shows red on its inner edge.

Why exactly 22°?

It is the minimum deviation of a 60° prism made of ice (n ≈ 1.31): from n = sin((A+D_min)/2)/sin(A/2) with A = 60°, you get D_min ≈ 21.8°. The 60° comes from the geometry of alternate faces on a hexagonal crystal.

How can I measure 22° in the sky?

Stretch your arm out and spread your hand: the span from thumb tip to little-finger tip is roughly 20–25° for most people. The gap between the Sun and the ring is about that one outstretched hand-span.

Why is the halo mostly white with only a faint red rim?

Ice disperses colours only weakly, spreading the ring's tints over less than 1°. The Sun's own 0.53° width and the range of crystal orientations blur the colours together everywhere except the sharp inner edge, where red survives as a reddish rim.

Does a halo predict rain?

Sometimes, loosely. Halos form in cirrostratus, the high thin cloud that frequently precedes an advancing warm front. So a halo can be an early sign of incoming weather, but it is a weak forecaster, not a rule.