# After sunset, the sky is blue for a different reason A sky computed from the physics of air, for your place and this evening. Scroll and the Sun sinks from noon through sunset to the stars. Take out the air, the ozone or the dust and see what is left. After sunset, ozone plays a large part in the deep blue overhead: it swallows the orange in the light. The colour of the sky Your screen has become today's sky over **London**, without clouds or city lights, with the Sun where it really is. It is computed from the physics of air, one wavelength at a time. As you scroll, the Sun sinks, and the clock in the corner shows when that happens where you are. Scroll slowly. **Use my location** ## What do you see if you take away the air? A black sky with a white Sun in it. Without molecules of air, nothing turns sunlight aside towards your eye. It passes you by and carries on. The blue overhead is not a paint on the sky. It is sunlight knocked off course, billions of times over, by molecules of nitrogen and oxygen. ## Why blue and not red? A molecule of air scatters light more strongly the shorter its wavelength. Rayleigh's law says by how much: the fourth power. Blue at 440 nanometres is scattered **5.7 times** more strongly than red at 680. High up, away from the Sun, the light that reaches you is mostly the light turned aside. And that is mostly blue. The buttons at the bottom take the molecules, the ozone or the dust out of the air, at any moment. Without dust, the sky deepens and the glare around the Sun tightens. ## Why does the Sun turn red as it sinks? Because its light crosses more and more air. With the Sun on the horizon, the beam passes through about **35 times** as much air as it would with the Sun straight overhead. On the way, the blue is scattered off in every direction. What is left is yellow, orange and red. The same process that paints the sky blue paints the sunset red. ## Does it take pollution to make it red? Today in **London**, the Sun sets at **19:00**. No. The sky on this page holds no pollution, only air and an ordinary fine dust, and the sunset still comes out red. Clean air is enough. Dust and smoke change the shade and the spread of the colour, and thick haze usually dulls it. ## What is the dark band opposite the sunset? Turn your back on the sunset. A dark band is rising above the horizon. It is **the shadow of the Earth**, cast on the air: the part of the sky that no longer gets sunlight. Above it lies a pink-orange strip called **the Belt of Venus**. The air there is still in sunlight, but light that has travelled a long, low path through the atmosphere and reddened like the setting Sun; the air and dust up there scatter it back towards you. The shadow climbs as the Sun sinks, until it is lost in the dark. ## Where does the blue overhead come from after sunset? Now look straight up. The Sun is 4 degrees below the horizon, and yet the sky overhead is a deep blue. But the light reaching the zenith now has grazed through hundreds of kilometres of air, like the light of the sunset. It ought to be yellowish, and the sky with it. It is not. ## What happens if you take out the ozone? **Take out the ozone** Without ozone, the sky overhead turns a yellowish grey. That is the twilight sky that air and dust would make without ozone. Ozone lies mostly 15 to 35 km up. Brought down to the pressure at the ground, all the ozone above you would make a layer about 3 millimetres thick. It absorbs orange and red. At 600 nanometres a molecule of ozone absorbs **32 times** as much as at 440. By day, sunlight crosses the ozone on a short path and the effect barely shows. At twilight the light runs a long way through the layer. Ozone makes no blue light: it takes the orange out, and what is left is bluer. At noon, without ozone, the sky stays almost as blue. Try it. ## Who worked this out? The American physicist E. O. Hulburt showed it by calculation, in 1953. He computed the twilight sky from Rayleigh scattering alone and got a zenith two to four times brighter than the one measured. When he added absorption by ozone, the calculation matched the measurements for the Sun between 0 and 6 degrees below the horizon. He also described the sky without ozone: greyish green-blue at sunset, yellowish in twilight. The sky on this page, computed from other data, comes out the same colour. **Drag the line on the screen:** on the left is twilight with ozone, on the right the same twilight without it. Newer measurements in Antarctica found that dust and droplets in the air matter about as much as ozone for the shade you actually see. Without ozone, though, the zenith would not stay blue. ## When does the blue hour end? The blue hour has no fixed threshold: every photographer counts it a little differently. Astronomers have exact thresholds: **civil twilight** ends at 6 degrees, **nautical** at 12, **astronomical** at 18. In **London** today, civil twilight ends at **19:30** and nautical twilight at **20:00**. ## Which stars come out first? The brightest, as the sky around them darkens. Their positions are the real ones for your place and your evening, from a catalogue of 1,289 stars. The stars shine with the same light they had all day. Only the sky in front of them has gone dark. ## When is it truly night? With the Sun 18 degrees below the horizon, astronomical twilight ends: sunlight scattered by the high air is now fainter than the light of the stars. What still glows in the sky comes from the stars and from the air itself. ## Your evening, where you are Today in **London**, the Sun sets at **19:00**. At about **19:20** it is 4 degrees below the horizon. Where you are, the Sun does not set today. **Go outside then and look straight up.** That deep blue is the light of the sunset, filtered through the ozone layer. ## Method and sources The sky is computed for every point on the screen: sunlight enters the atmosphere, is scattered once by molecules (Rayleigh) or fine dust (Mie), and is dimmed along the way by air, dust and ozone. The calculation uses 12 wavelengths from 415 to 690 nm on a spherical Earth, with the atmosphere of Eric Bruneton's reference model (2017): an ozone layer of 300 Dobson units between 10 and 40 km, with absorption cross-sections measured at the University of Bremen. Colour comes from the CIE 1931 functions. Where the model stops: light is scattered only once. Light scattered many times matters little by day but a lot in deep twilight, so beyond about 6 degrees below the horizon the page's sky is darker than the real one. Refraction, clouds and city lights are left out. Brightness adapts roughly as the eye does, and colours are slightly more saturated, as in a photograph, equally in every scene. The hills are drawn; they are not your horizon. The times come from NOAA's solar position formulas, for the city of your time zone or for your own location if you give it (it stays on your device). Sunset is when the centre of the Sun is 0.833° below the horizon. Stars down to magnitude 4.8 come from the d3-celestial catalogue. The figures in the text (5.7 times, 35 times, 32 times) are computed from the same parameters. Standard air-mass tables give about 38 for the horizon, a little above our model, which leaves out refraction. - E. O. Hulburt, "Explanation of the brightness and color of the sky, particularly the twilight sky", J. Opt. Soc. Am. 43(2), 113–118 (1953). https://doi.org/10.1364/JOSA.43.000113 - R. L. Lee Jr., W. Meyer, G. Hoeppe, "Atmospheric ozone and colors of the Antarctic twilight sky", Applied Optics 50(28), F162–F171 (2011). https://doi.org/10.1364/AO.50.00F162 - R. L. Lee Jr., "Measuring and modeling twilight's Belt of Venus", Applied Optics 54(4), B194–B203 (2015). https://doi.org/10.1364/AO.54.00B194 - E. Bruneton, Precomputed Atmospheric Scattering, reference implementation (2017), atmosphere parameters. https://github.com/ebruneton/precomputed_atmospheric_scattering - CIE 1931 2° colour matching functions, via the Colour & Vision Research Laboratory (UCL). http://www.cvrl.org/ - F. Kasten, A. T. Young, "Revised optical air mass tables and approximation formula", Applied Optics 28(22), 4735–4738 (1989). https://doi.org/10.1364/AO.28.004735 - NOAA Chemical Sciences Laboratory, Twenty Questions and Answers About the Ozone Layer (2022). https://csl.noaa.gov/assessments/ozone/2022/twentyquestions/ - U.S. Naval Observatory, definitions of civil, nautical and astronomical twilight. https://aa.usno.navy.mil/faq/RST_defs - NOAA Global Monitoring Laboratory, solar position calculator. https://gml.noaa.gov/grad/solcalc/ - O. Frohn, d3-celestial star catalogue (BSD licence). https://github.com/ofrohn/d3-celestial Canonical: https://mariuscomper.uk/albastrul-de-dupa-apus/en/