A second defined by an atom
Since 1967, the second, the basic unit of time in the International System of Units, has been defined not by any astronomical motion but by a single number: 9,192,631,770 cycles of the microwave radiation that a caesium-133 atom absorbs and emits when it flips between two specific hyperfine energy states. The caesium atomic clock is the device built to count that many cycles reliably, using the atom itself as the pendulum rather than any mechanical or electronic oscillator, because every caesium-133 atom in the universe flips between these two states at exactly the same frequency, unlike a spring or a crystal, which drifts with temperature, age and manufacturing variation. The claim behind the clock is not merely that it keeps good time, but that it keeps a kind of time that does not depend on which specific piece of equipment, or which laboratory, produced it.
From Essen’s 1955 clock to caesium fountains
Louis Essen and Jack Parry built the first working caesium clock at the National Physical Laboratory in the United Kingdom in 1955, following theoretical groundwork on atomic beam resonance from Isidor Isaac Rabi in the 1930s. The technique tunes microwave radiation until it matches the frequency that maximises the fraction of caesium atoms flipping between energy states, using that peak as the reference. The 1967 international redefinition of the second formalised this method by fixing the caesium-133 transition frequency at exactly 9,192,631,770 hertz, later refined in 1997 to specify an atom at rest and at a temperature of absolute zero, in effect removing small residual sources of error from the definition itself. Later designs, known as caesium fountain clocks, improved on the original by tossing a cloud of laser-cooled atoms upward through a microwave cavity, letting them interact with the field for longer and yielding uncertainties around one part in ten thousand trillion.
The clock inside every GPS satellite
The caesium standard has held up as the practical backbone of modern timekeeping for decades. Caesium clocks, at various levels of precision, sit inside GPS satellites alongside rubidium clocks, and the accuracy of a satellite-based position depends directly on the accuracy of the clock aboard it, since a timing error of a single nanosecond corresponds to roughly thirty centimetres of positional error once multiplied by the speed of light. Rival navigation systems, Europe’s Galileo, Russia’s GLONASS and China’s BeiDou, all rely on the same basic approach, combining atomic clocks of different types to reach timing accuracies in the tens of nanoseconds. None of the underlying physics has needed revision since caesium was adopted as the reference; refinements have improved precision without overturning the method.
Overtaken by optical clocks
What has changed is that caesium is no longer the most precise atomic clock available, only the one still legally defining the second. Optical clocks, which use transitions at far higher frequencies than caesium’s microwave transition, have overtaken it by orders of magnitude: strontium and aluminium-ion optical clocks have demonstrated uncertainties down toward one part in a hundred quintillion, versus roughly one part in ten thousand trillion for the best caesium fountain clocks. This gap is why metrologists expect the second itself to be redefined around 2030 to 2034 using an optical transition instead of caesium’s microwave one, once optical clocks can be compared reliably enough across different laboratories and countries to serve as a shared international standard, a coordination problem that is proving harder to solve than building the clocks themselves.
A clock precise enough to test relativity
Precise, shared timekeeping is infrastructure that most people never see but constantly depend on. GPS and its rival satellite navigation systems require atomic clocks not as a refinement but as a precondition, since positioning is fundamentally a timing calculation converted into distance by the speed of light. The same clocks have also become sensitive scientific instruments in their own right: the newest optical clocks are precise enough to detect the general-relativistic effect where time runs measurably slower deeper in a gravitational field, over height differences as small as a millimetre in laboratory demonstrations. A technology developed to fix the definition of a second has ended up serving simultaneously as the basis for global navigation and as a working test of Einstein’s general relativity.
Infrastructure nobody sees
This is a solid, unglamorous piece of physics infrastructure, and worth understanding precisely because its consequences are so widely felt while its mechanism is rarely explained. The material rewards attention to the difference between the caesium standard, still the legal definition of the second, and the optical clocks that have already surpassed it in raw precision, since that gap is the live edge of the field right now rather than settled history. Readers should not expect drama: the story is one of steady refinement from 1955 onward, with no dramatic overturning of the basic method, only tighter numbers. For anyone who wants to understand what a GPS receiver, a satellite, or a physics experiment testing relativity actually relies on underneath, this is time well spent.