Normally, a radio wave launched from an antenna bends (is refracted) towards the Earth. The wave also spreads as it propagates toward the distant receiver and if the path is obstructed, only a small amount of signal energy can be captured beyond the obstruction. Propagation is normally by both refraction and diffraction. See the page on the normal troposphere for more on refraction.
Duct set-up
Typically, energy spreading upwards is lost to space. But sometimes an upper limit is forced on that propagation by the set-up of a layer which reflects energy along the path, rather like the ionosphere does but lower. The result is a surface-based duct. Sometimes too, a lower limit is forced on the wave reflected from above by another lower layer. The result then is that the wave is trapped in an elevated duct.
Ducts are formed by a change in the atmosphere’s refractive index from the normal 50% of time of -40N/km to -157N/km or below. This is the inverse beam-bending case where the obstruction of the Earth’s bulge into the radio path disappears, and the stations are effectively ‘looking at one another’.
If a duct is formed, the loss approaches the Free Space Loss. This means that very long-distance paths that were previously inoperable (through excessive diffraction loss over FSL) become operable.
Duct frequency of occurrence
Ducts at VHF and UHF are rare around the British Isles. Favourable propagation more often comes about here by the more-frequently occurring slightly reduced refractivity to around -100N/km represented by a k of 2.75 exceeded for 5% of time. This reduces the diffraction loss to something allowing 500-700km, rather than the longer 1,000km plus paths of ducting.
Ducting is typically caused by a warm moist air mass when it slides off the land onto the sea. The air mass cools rapidly, moisture condenses out (since cold air can hold less moisture) and the refractive index drops. This really only occurs over the North Sea in summer. There are three centres of frequent ducting, one of which is off the coast of Senegal and other West African states where much of the empirical work on ducting was done.