Nighttime radio reach

Why can a distant broadcast station arrive clearly after dark yet vanish completely by day?

Tune an AM receiver at midday and you hear the local stations. Tune the same receiver at midnight and the band fills with distant ones — another country, hundreds of miles away, arriving clearly enough to listen to. By breakfast they are gone.

Nothing has moved. The transmitter is in the same place at the same power, the receiver is on the same table. The only thing that changed is the sun. What is remarkable is not merely that reception varies, but that an entire institution — how stations are licensed, how much power they may run, which ones exist at all — has been built around this daily change.

The physical half of this is covered elsewhere in the collection and I will use it rather than re-derive it. Two companion accounts establish the mechanism: sunlight ionises the lowest region of the ionosphere, that region absorbs medium-frequency signals rather than reflecting them, and it recombines and effectively disappears after dark, leaving the higher reflecting layers to return signals to earth far away. A related account derives why a plasma has a cutoff frequency at all, and why it differs between day and night.

Take those as given. What follows from them is a broadcasting problem, and that is what this piece is about.

Start with what the two conditions mean for a station. By day it has a groundwave service area: signal following the earth's surface, reliable, bounded, a few tens of miles. That is its audience, and it is stable. At night it keeps the groundwave and acquires a skywave — energy that went upward, bounced, and came down hundreds of miles away.

Now notice the asymmetry that creates the entire regulatory problem. The skywave is a benefit to the station that produced it and a hazard to every distant station on the same frequency, because it arrives in their groundwave area as interference. And it is mutual: at night, every station on a channel is spraying signal into every other station's territory.

Follow the arithmetic of the band. There are only so many channels, and far more stations than channels, so many stations must share each frequency. By day that works, because absorption confines everyone to their own patch. At night, sharing breaks, because the patches overlap. So a rule is needed, and the rule cannot be a technical fix — the propagation is not adjustable.

What emerged instead was an allocation scheme, and its features all follow from the above. Some channels are reserved for a single dominant station permitted to exploit the skywave, with others on that frequency required to protect it. Stations must reduce power at sunset, switch to a directional antenna pattern that puts a null toward the station they must protect, or leave the air entirely until sunrise. And because sunset moves through the year, the licence conditions are tied to actual local sunrise and sunset rather than to clock times — so a station's operating schedule shifts week by week.

Then the consequence people actually experience. A daytime-only station simply stops broadcasting; a listener at its edge loses it not because conditions worsened but because it was ordered off. And the distant station arriving instead is not a curiosity — it is often precisely the station whose protection required the local one to shut down.

So the answer to the question has two layers, and the second is the one usually missed. The distant signal arrives because absorption ceased. The local one vanishes because a regulator made it. Both are consequences of the same daily cycle, and a listener cannot tell them apart by ear.

The station did not get weaker at night, and this is the most common misreading. People describe distant stations "overpowering" the local one. Usually the local one has reduced power or signed off under licence conditions, and the distant one has gained a skywave it did not have at noon. Both sides of the change are happening at once, and attributing it to a single cause gets the mechanism wrong.

This is specific to the medium-frequency band, and neighbouring bands behave differently for reasons worth keeping separate. FM broadcasting sits far above the frequencies these layers reflect, so it is essentially line-of-sight and shows none of this. Shortwave sits in the range that reflects readily by day and night, which is why it was the long-distance broadcast band; its behaviour varies with the solar cycle rather than simply with the clock. So "radio travels further at night" is a claim about one band, not about radio.

The regulatory arrangement is historically contingent, not a physical necessity. Which channels were made exclusive, and to whom, was decided by particular administrations at particular times, and reflects the commercial and political interests of the era as much as any propagation calculation. Different jurisdictions solved the same physics differently — in the number of protected channels, in how strictly protection is enforced, and in whether small local stations were permitted to continue at reduced power or required to close. Treating any one country's scheme as the natural consequence of the physics would be a mistake.

The scheme is decaying, and for reasons unrelated to propagation. Rising urban electrical noise has degraded medium-wave reception generally; audiences have moved to FM, digital and streaming; and some administrations have relaxed night-time protection because the assets being protected are worth less than they were. So the elaborate apparatus is now maintained for a shrinking purpose — an institution outliving the conditions that justified it.

The falsification test. If the mechanism is loss of absorption rather than any change at the transmitter, then a distant station's arrival should track local sunset along the propagation path rather than clock time, should move through the year with the seasons, and should be observable while the transmitter's own logs show constant power. If distant reception improved at a fixed clock hour regardless of season and latitude, the account would be wrong and something about transmission schedules or receiver conditions would have to explain it.