What Is a Time Signal? Broadcasting the Answer to What Time It Is
By the BrainSnail editorial team. How these articles are written and checked, and how to tell us when one is wrong.
A clock is only useful if it agrees with other clocks, and no mechanism keeps perfect time by itself. The solution adopted everywhere is to measure time extremely accurately in a few laboratories and then broadcast it, so that any clock can compare itself against the transmission and correct itself.
Before radio
The need arose at sea, because finding longitude requires knowing the time at a reference meridian, so a ship's chronometer had to be set accurately before departure and checked whenever possible. Ports therefore installed visual time signals, most famously a large ball raised up a mast and dropped at a fixed moment each day, which could be seen from ships in the harbour and allowed every chronometer to be checked simultaneously. The first was installed at Portsmouth and the best known sits on the Royal Observatory at Greenwich, where it has dropped at one in the afternoon since 1833, and the practice spread to ports worldwide, with the New Year celebration in New York descending from the same idea. Guns were fired for the same purpose in several cities, and a few still are. Railways created a second and larger pressure, since a timetable is meaningless if towns keep their own local solar time, and the resulting standardisation of time across a network is what created time zones, adopted nationally in Britain by the 1840s and internationally from the 1880s.
The broadcast systems
Radio replaced light and sound, and several distinct services grew up serving different needs:
- •Long-wave and low-frequency stations transmitting a coded time signal continuously, which radio-controlled clocks and watches decode automatically, including services in Germany, Britain, the United States, Japan and China
- •Shortwave time stations broadcasting audible ticks and spoken announcements, historically the main reference for navigation and scientific work
- •The familiar broadcast pips on radio, six tones marking the hour, introduced in Britain in 1924 and generated from the national time standard
- •Satellite navigation, which is fundamentally a timing system, since each satellite carries atomic clocks and a receiver solves for position and for time simultaneously, making it the most widely used time distribution method in existence
- •Network time protocols over the internet, which synchronise computers by measuring round-trip delays to reference servers
- •Dedicated fibre and precision protocols in financial and scientific settings, where microsecond or nanosecond agreement is required and ordinary network timing is not good enough
What the time actually is
Since 1967 the second has been defined by a property of the caesium atom, specifically a fixed number of cycles of the radiation associated with a particular transition, which replaced definitions based on the Earth's rotation and orbit because atomic behaviour is far more stable than the planet. International Atomic Time is produced by combining readings from hundreds of atomic clocks in laboratories worldwide, weighted by their performance, which means the world's official time is a computed average rather than the reading of any single clock, and the definitive version is published retrospectively rather than in real time. Coordinated Universal Time is that atomic scale adjusted to stay within a set margin of the time implied by the Earth's rotation, which drifts irregularly, and the adjustment is made by inserting a leap second when needed. Newer optical clocks based on other elements are substantially more precise than caesium and a redefinition of the second is under active consideration, which would be the first change to the definition in more than half a century.
The leap second problem
The Earth's rotation is not uniform, so astronomical time and atomic time diverge, and leap seconds have been inserted to keep them aligned, with more than two dozen added since 1972. That causes real difficulty for computing, because a minute containing sixty-one seconds breaks software that assumes it cannot happen, and documented outages at major internet companies have been traced to leap second handling. Various workarounds exist, including smearing the extra second across a day by slightly slowing every clock, which keeps systems consistent internally while putting them briefly out of step with the official standard. After long disagreement between the astronomical and technological communities, an international decision was taken in 2022 to stop inserting leap seconds by 2035 and allow the two scales to drift apart, with a larger correction deferred to the distant future. A further complication has emerged because the Earth's rotation has recently sped up slightly, raising the possibility of a negative leap second, meaning a minute of fifty-nine seconds, which no system has ever been tested against.
The takeaway
No clock keeps perfect time alone, so a few laboratories measure it precisely and broadcast the result for others to correct against. Ports dropped time balls so ships could check chronometers, and railways forced the standardisation that created time zones. Official time is a weighted average of hundreds of atomic clocks, not any single one. Leap seconds keep it aligned with the Earth's uneven rotation and break software, so they are being abandoned by 2035.