Why the sky is blue — and sunsets red.
Air molecules scatter short wavelengths far more than long ones (I ∝ 1/λ⁴) — enough to paint the whole daytime sky blue, and, once sunlight has to cross far more air at a low sun angle, to scatter away nearly all the blue before it reaches your eye directly, leaving red and orange behind.
Wavelength λ
λ⁻⁴ ∝ I
380 nm · violetshorter λ scatters much more750 nm · red
Relative scattering
× vs 550 nm
midday direct beam survives
sunset direct beam survives
Try
Named for Lord Rayleigh, who worked out the 1/λ⁴ law in the 1870s — see Rayleigh scattering. Using 450 nm and 650 nm as representative blue and red wavelengths, blue scatters about (650/450)⁴ ≈ 4.35× more than red at those specific wavelengths — a different pair of representative wavelengths gives a different number (400 nm vs. 700 nm gives (700/400)⁴ ≈ 9.38×, close to the "about 5.5×" figure sometimes quoted elsewhere), so there's no single "the" ratio without stating which wavelengths went in. The two atmosphere path lengths above (airmass ≈1.15 at 60° elevation, ≈37.9 at the horizon) come from the empirical Kasten–Young formula, computed live, not looked up; the sun and sky colors are a simplified composite built by weighting each sample wavelength's color by how much of it survives that path (Beer–Lambert, using a real sea-level reference optical depth of ≈0.1 at 550 nm), assuming an evenly-bright input spectrum rather than the sun's true irradiance curve — illustrative, not a radiative-transfer simulation. This model only holds because air is mostly N₂/O₂ molecules much smaller than a wavelength of visible light, which is exactly the regime Rayleigh scattering requires; larger particles like water droplets, dust, or pollution scatter all visible wavelengths roughly equally (Mie scattering) which is why clouds and haze look white or grey rather than blue.