The leap second is headed for retirement. The International Bureau of Weights and Measures, known by its French acronym BIPM, says it will stop issuing leap seconds for Coordinated Universal Time in May 2027, ending a system that has periodically inserted an extra second to keep atomic time close to the Earth’s uneven rotation.
For most people, the change may pass unnoticed. But Coordinated Universal Time, or UTC, sits beneath a vast amount of modern technical infrastructure. Databases, software interfaces and other systems use it to label and coordinate events. The seemingly small act of adding a second has repeatedly created problems for systems that must agree precisely about the order and timing of events.
Why leap seconds were introduced
Leap seconds were introduced in 1972 to address a mismatch between two ways of measuring time. Atomic clocks provide an extremely stable time scale. Earth’s rotation, on the other hand, is not perfectly regular and is slowing overall, in large part because of tidal friction associated with the Moon. Conditions involving the planet’s molten core, oceans and atmosphere also affect its rotation rate.
UTC is based on atomic clocks, while Universal Time, or UT1, is tied to astronomical observations of Earth’s rotation. Leap seconds were designed to prevent those time scales from drifting too far apart. Rather than adding time on a regular schedule like leap years, the BIPM compares the output of hundreds of atomic clocks with astronomical time and determines when an adjustment is needed.
Thirty-two leap seconds have been added since the system began in 1972. The most recent was inserted in 2016. Each addition was intended to keep the clock used by digital infrastructure close to the rotation-based measure that connects timekeeping to the position of the Earth.
A one-second problem for computing systems
That alignment came at a cost. A leap second is unusual enough that many systems were not designed to represent or process it cleanly. UTC is used widely in application programming interfaces, databases and networked services, where a duplicated or unexpected second can create ambiguity or failures if every part of a system does not handle the event in exactly the same way.
Some major technology companies have adopted a workaround known as a time smear. Instead of recognizing one extra second at a single instant, the system spreads the adjustment across a longer period, such as a day, in tiny increments. The approach reduces the abrupt edge case for that company’s own systems, but it also illustrates the lack of a single operational response across the broader technology ecosystem.
Past disruptions have made the problem less theoretical. The article notes that Qantas experienced a booking-system problem during the 2012 leap second. The details differ from one implementation to another, but the larger lesson is straightforward: a correction too small to notice in everyday life can be difficult for interconnected digital systems to execute safely.
The looming concern over a negative leap second
The decision also removes the prospect of a different and potentially more difficult kind of adjustment: a negative leap second. A positive leap second adds one second to UTC. A negative leap second would instead require the system to omit a second, something the current arrangement has never had to do.
The BIPM’s move is therefore more than a symbolic cleanup of an obscure standard. It changes the future relationship between the atomic time foundation of UTC and the astronomy-based measure of Earth’s rotation. UTC will continue to be the time scale central to global technical coordination, but it will no longer receive these periodic one-second corrections after the planned 2027 cutoff.
What will change in 2027
The immediate operational change is specific: beginning in May 2027, the BIPM will no longer issue leap seconds for UTC. That ends a practice that has existed since 1972 and has produced 32 additions.
For systems administrators and software developers, the change should reduce one recurring class of timekeeping exception. It does not make time simple. Technology will still need to account for time zones, clock synchronization, daylight-saving rules where they apply, historical timestamps and the distinction between different time scales. But it removes the rare, irregular insertion that has demanded special handling from systems built on UTC.
For people outside timekeeping and computing, the most visible impact may be the absence of an event they rarely knew was occurring. Midnight will continue to arrive, atomic clocks will continue to provide highly stable measurements, and UTC will remain the common reference used around the world. The difference is that the time scale will no longer periodically pause for a one-second correction intended to track the planet’s rotation.
A shift in what UTC is optimized to do
The history of the leap second reflects a compromise between two goals: retaining a close connection to the astronomical day while providing a uniform time scale for increasingly complex technology. In 1972, the adjustment offered a way to preserve both. Decades later, the scale and interdependence of digital systems have made the compromise harder to maintain.
Ending leap seconds does not erase the physical difference between atomic time and Earth rotation. It instead recognizes that UTC’s role as a dependable technical reference has become the more pressing concern. The practical effect is a cleaner, more predictable system for the technologies that rely on coordinated timestamps, even as the planet continues to turn at its own subtly variable pace.
- The BIPM plans to stop issuing UTC leap seconds in May 2027.
- Thirty-two leap seconds have been added since 1972; the latest was in 2016.
- The change removes a recurring source of complexity for UTC-based digital systems.
This article is based on reporting by Wired. Read the original article.
Originally published on wired.com
