{
  "id": "coordinated-universal-time",
  "title": "Coordinated Universal Time",
  "lang": "en",
  "source_lang": "en",
  "summary": "Coordinated Universal Time is the international reference time scale combining atomic timekeeping with controlled alignment to Earth’s rotation.",
  "category": "science",
  "subcategory": "physics",
  "type": "concept",
  "aliases": [
    "UTC",
    "Coordinated Universal Time scale"
  ],
  "generated_at": "2026-11-02",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/coordinated-universal-time/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/coordinated-universal-time/",
      "json": "https://aiwiki.page/zh/coordinated-universal-time.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Coordinated Universal Time, abbreviated UTC, is the international reference time scale used to coordinate civil time, communications, navigation, and scientific measurements. It is derived from an international ensemble of atomic clocks and advances in units of the SI second. Under the system introduced in 1972, occasional leap seconds keep it close to time determined by the rotation of Earth. Most national civil times are expressed as offsets from UTC, although the legal definition and dissemination of time remain national responsibilities.\n\nUTC reconciles two different requirements: a highly uniform measure of elapsed time and a clock that remains approximately aligned with the astronomical day. These requirements cannot be met indefinitely by a single unadjusted scale because Earth’s rotation is irregular. UTC therefore belongs both to the precision science of metrology and to the practical infrastructure of everyday timekeeping. Its construction, international dissemination, representation in software, and planned changes to leap-second policy are distinct but closely connected aspects of its operation."
    },
    {
      "heading": "Definition and relationship to other time scales",
      "text": "A time scale is a system for assigning numerical labels to successive instants. UTC provides a sequence of dates, hours, minutes, and seconds whose rate is based on atomic frequency standards. Its defining relationships involve two other scales: International Atomic Time, or TAI, and Universal Time, especially the form designated UT1.\n\nTAI is the continuous atomic reference calculated by the International Bureau of Weights and Measures, usually known by its French initials BIPM. Since the beginning of 1972, UTC and TAI have used seconds of the same duration, but their readings differ by an integer number of seconds. If that difference is denoted by \\(\\Delta_{\\mathrm{AT}}\\), the relationship is\n\n\\[\\mathrm{UTC}=\\mathrm{TAI}-\\Delta_{\\mathrm{AT}}.\\]\n\nThe expression describes the offset between corresponding readings; handling calendar labels during an inserted second requires additional care. On 1 January 1972, \\(\\Delta_{\\mathrm{AT}}\\) was 10 seconds. The leap second at the end of 2016 increased it to 37 seconds from 1 January 2017. The initial 10-second difference was inherited from earlier timekeeping arrangements rather than produced by ten leap seconds.\n\nUT1 measures Earth’s rotational angle in a form expressed as time. It is not uniform in the same sense as atomic time: changes in the rotation rate alter the duration of a UT1 day when measured in SI seconds. Under the leap-second system established in 1972, UTC is maintained within 0.9 second of UT1. The difference is conventionally written\n\n\\[DUT1=\\mathrm{UT1}-\\mathrm{UTC}.\\]\n\nUTC is consequently neither a purely rotational scale nor an uninterrupted atomic scale. Its seconds are atomic, while adjustments to its second count preserve a limited correspondence with rotation.\n\nUTC is also distinct from Greenwich Mean Time, or GMT. Historically, GMT referred to mean solar time at Greenwich, with conventions that changed over time. In contemporary civil usage, GMT often names a zero-offset time zone and is treated as equivalent to UTC at ordinary clock precision. In precise astronomical and metrological work, however, the terms are not interchangeable without specifying the intended definition.\n\nThe abbreviation UTC is internationally standardized and is used across languages. It does not follow the normal word order of either the English “Coordinated Universal Time” or the French “temps universel coordonné.” Its form avoids making either language’s initials the international designation and follows the established family of Universal Time abbreviations, including UT0, UT1, and UT2."
    },
    {
      "heading": "From astronomical time to the modern UTC system",
      "text": "For centuries, timekeeping depended on observations of the Sun and stars. Apparent solar time follows the actual Sun, whereas mean solar time smooths the annual variations caused by Earth’s orbital eccentricity and the inclination of its rotational axis. Mechanical clocks could approximate mean solar time, but their long-term rate ultimately required comparison with astronomical observations. The 1884 International Meridian Conference selected the Greenwich meridian as the initial meridian for longitude and proposed a universal day beginning at mean midnight there. This provided an important geographical convention, not an atomic time standard.\n\nDuring the twentieth century, improved clocks exposed irregularities in Earth’s rotation. Quartz oscillators could maintain a steadier short-term frequency than the rotating Earth, and astronomical observations showed that the length of the day was not constant. The second was therefore redefined in terms of the tropical year associated with 1900, producing the ephemeris second. This separated the unit of duration from the variable rotational day, although realizing the unit still depended on lengthy astronomical work.\n\nAtomic frequency standards supplied a more reproducible basis. In 1955, Louis Essen and Jack Parry operated a practical caesium atomic frequency standard at the United Kingdom’s National Physical Laboratory. Comparisons with astronomical time helped establish the frequency subsequently used in the atomic definition of the second. In 1967, the General Conference on Weights and Measures adopted a definition based on 9,192,631,770 periods of radiation corresponding to the transition between two hyperfine levels of the ground state of caesium-133.\n\nInternational coordination of radio time signals developed alongside these changes. A coordinated system was operating by 1960, and the name Coordinated Universal Time was adopted in 1967. Before 1972, however, UTC did not follow the present integer-leap-second design. Its rate could be deliberately offset from atomic frequency, and fractional-second steps were applied to maintain approximate agreement with rotational time. This history matters when interpreting old observations: UTC before 1972 cannot be reconstructed simply by subtracting a modern-style integer offset from TAI.\n\nThe modern system began on 1 January 1972. Frequency offsets were abandoned, UTC seconds were made equal in duration to TAI seconds, and whole-second adjustments replaced the earlier fractional changes. The first positive leap second was inserted at the end of 30 June 1972. Responsibility for the international atomic scale passed into the BIPM’s time work, while Earth-rotation monitoring and leap-second decisions became associated with the International Earth Rotation and Reference Systems Service, established in 1987 and operational from 1988. The resulting division of labor separates atomic frequency measurement from the observation of planetary rotation."
    },
    {
      "heading": "Atomic standards and the calculation of UTC",
      "text": "UTC is not the reading of a single master clock. The BIPM calculates it from measurements contributed by timing laboratories around the world. These laboratories operate atomic clocks, including commercial caesium clocks and hydrogen masers, and compare their local time scales through satellite and other transfer techniques. The ensemble approach reduces dependence on any one instrument and permits failures, unusual drifts, and measurement errors to be identified.\n\nThe physical foundation is the relationship between an atomic transition and a reproducible frequency. In a caesium standard, microwave radiation interacts with atoms, and a control system adjusts an oscillator toward the transition frequency. Primary caesium fountain standards send laser-cooled atoms through a microwave interaction region, allowing careful measurement of frequency shifts. Corrections are required for effects such as thermal radiation, magnetic fields, atomic collisions, and the motion of the atoms. The definition of the second is exact; the practical realization of that definition has a measurement uncertainty.\n\nThe BIPM first forms a free atomic ensemble scale, EAL, from participating clock data. Clocks receive weights according to their evaluated performance, subject to safeguards intended to prevent excessive dependence on a small number of devices. EAL provides strong continuity and stability, but a stable ensemble can still have a small frequency offset from the SI second. Evaluations from primary and secondary frequency standards are therefore used to establish the scale interval of TAI. UTC is then derived by applying the prescribed integer-second offset.\n\nParticipating laboratories maintain real-time approximations written UTC(\\(k\\)), where \\(k\\) identifies the laboratory. Examples include UTC(NIST) in the United States and UTC(USNO) at the United States Naval Observatory. These are physical signals available locally and through dissemination services. UTC itself is calculated retrospectively: the BIPM’s monthly Circular T publishes differences between UTC and the contributing laboratory scales at specified epochs. A rapid product, UTCr, provides more frequent information supporting operational steering.\n\nThis distinction is central to metrological traceability. A laboratory or user does not obtain perfect UTC merely by possessing an accurate clock. A documented chain of comparisons must connect the local signal to UTC, with uncertainties assigned to clock behavior, signal transfer, equipment delays, and processing. A continuously broadcast national signal can be an excellent approximation to UTC while still differing from the subsequently calculated international scale by several nanoseconds or more. Those small differences are immaterial for ordinary clocks but important in precision navigation, telecommunications, and fundamental measurements."
    },
    {
      "heading": "Earth’s rotation and the leap-second mechanism",
      "text": "Earth’s rotation changes because angular momentum is exchanged among its solid interior, atmosphere, oceans, and other components. Winds, ocean currents, seasonal mass redistribution, and interactions involving the core contribute variations over different timescales. Tidal interactions, especially with the Moon, produce a long-term tendency toward slower rotation. Shorter-term fluctuations can oppose that tendency, so the length of the day does not increase smoothly from one year to the next.\n\nIf the rotational day were always exactly 86,400 SI seconds, UT1 and a suitably aligned atomic scale would remain synchronized without corrections. In reality, a small daily difference accumulates. A day longer than 86,400 SI seconds by 1 millisecond produces approximately 0.365 second of accumulated displacement over a year if that excess persists. The relevant quantity is the integrated departure of rotational time from atomic time, not simply whether Earth is described as rotating “fast” or “slow” in a particular month.\n\nA leap second changes the UTC count to limit this displacement. A positive leap second is inserted at the end of a UTC month. The sequence of labels is\n\n\\[23{:}59{:}59,\\quad 23{:}59{:}60,\\quad 00{:}00{:}00.\\]\n\nThe affected UTC day contains 86,401 seconds. A negative leap second would omit the final ordinary second, producing the sequence \\(23{:}59{:}58\\), followed by \\(00{:}00{:}00\\), and a day of 86,399 seconds. The system permits both signs, although the adjustments introduced from 1972 through the end of 2016 were all positive. Leap seconds are unrelated to leap days: the former address rotation relative to atomic time, while the latter reconcile the calendar with the seasonal year.\n\nThe International Earth Rotation and Reference Systems Service, or IERS, determines when an adjustment is required and announces it in Bulletin C. The preferred dates are the ends of June and December, with other month ends available under the governing rules. Announcements normally provide several months of notice. The correction occurs at the same physical instant worldwide; a local clock with a positive offset from UTC may therefore encounter the event on the following local calendar date.\n\nThere is no fixed schedule comparable to the four-year pattern of calendar leap years. Future Earth rotation can be predicted only with limited accuracy, so leap-second dates cannot be assigned indefinitely in advance. Broadcast DUT1 information and more precise Earth-orientation products allow users to recover rotational time from UTC when needed."
    },
    {
      "heading": "Dissemination and synchronization",
      "text": "UTC becomes practically useful through signals that connect international timekeeping to local clocks. Traditional dissemination uses terrestrial radio transmitters. Services such as WWV and WWVB in the United States, DCF77 in Germany, and MSF in the United Kingdom transmit frequency or time information related to national realizations of UTC. Their different frequencies and modulation schemes support different ranges and receivers. Reception accuracy depends not only on the source clock but also on propagation delay, interference, and the receiver’s decoding process.\n\nSatellite navigation systems provide another major route. A global navigation satellite system broadcasts precise timing information because positioning depends on measuring signal travel time. A timing receiver with a surveyed antenna position can use those signals to estimate local clock offset. Antenna cables, receiver electronics, satellite orbit and clock errors, and atmospheric delays all affect the result. Specialized timing installations calibrate these delays rather than treating the received timestamp as exact.\n\nNational laboratories also compare clocks directly through satellite links. Common-view techniques compare observations of the same satellite from two sites, canceling some shared errors. Two-way satellite time and frequency transfer exchanges signals in both directions so that parts of the propagation delay cancel. Optical-fiber links offer much higher performance over suitable routes and are important for comparing advanced frequency standards. These metrological links serve a different purpose from ordinary public time broadcasts: they help construct and verify the reference scale itself.\n\nOn the Internet, the Network Time Protocol, or NTP, synchronizes computer clocks through timestamped packet exchanges. Its hierarchy places directly referenced servers near the top, with other servers and clients at progressively greater logical distance. NTP estimates clock offset and network delay, filters measurements, and adjusts local clock behavior. Variable routing and unequal outward and return delays limit accuracy. A server’s stated hierarchy level is therefore not a guarantee of measurement quality.\n\nThe Precision Time Protocol, standardized in IEEE 1588, supports closer synchronization in controlled networks, particularly when hardware records timestamps near the network interface. Telecommunications, industrial control, and measurement systems can achieve substantially tighter coordination than ordinary wide-area NTP installations. Performance nevertheless depends on network architecture, oscillator quality, calibration, and treatment of asymmetric delays.\n\nAccuracy, stability, resolution, and availability remain separate properties. A device may display nanoseconds while being wrong by milliseconds, or maintain an exceptionally stable frequency while showing the wrong date. Reliable UTC dissemination requires both a credible reference and a characterized transfer path. Authentication and resilient source selection additionally address deliberate interference, which ordinary precision specifications alone do not cover."
    },
    {
      "heading": "Civil time, time zones, and date notation",
      "text": "UTC supplies a shared reference, but it does not determine local clock rules. National and regional authorities establish time zones, their offsets, and any seasonal changes. Local civil time is commonly represented as\n\n\\[\\text{local time}=\\mathrm{UTC}+\\text{applicable offset}.\\]\n\nOffsets need not be whole hours: India uses UTC+05:30, while Nepal uses UTC+05:45. A locality’s offset can also change because of daylight saving or legislation. UTC itself does not advance or retreat for daylight saving.\n\nA numeric offset and a named time zone convey different information. The offset +01:00 identifies a relationship to UTC at an instant, but it does not identify a unique jurisdiction or specify future rules. A regional identifier such as Europe/Paris refers to a history and rule set maintained in the IANA Time Zone Database. For recurring civil events, such as a meeting scheduled at 09:00 local time each week, that rule set matters. Treating the event as permanently fixed to one UTC offset can shift its local occurrence after a seasonal clock change.\n\nThe international date-and-time standard ISO 8601 provides forms such as 2024-01-15T12:00:00Z, where Z indicates a zero offset from UTC. An explicitly offset representation, such as 2024-01-15T13:00:00+01:00, identifies the same instant. The letter Z is also pronounced “Zulu” in aviation, military, and other operational settings. It is a notation for the time reference, not a separate physical scale.\n\nA complete timestamp needs more than a clock reading. The date, time scale, offset where relevant, and sufficient precision must be known. “12:30” alone cannot establish an unambiguous instant, and a local time during the repeated hour at the end of daylight saving may identify two different instants. Conversely, some local clock labels never occur when clocks move forward. UTC avoids these seasonal ambiguities, although leap-second labels create a separate exceptional case.\n\nThe legal connection between UTC and civil time differs among jurisdictions. Some laws refer directly to UTC; others refer to national standard time, Greenwich time, or an official realization maintained by a designated institution. An international recommendation does not itself rewrite national legislation. UTC’s role is to provide a common technical reference against which such legal arrangements can be expressed, not to impose a single worldwide civil clock or calendar practice."
    },
    {
      "heading": "Digital timestamps and leap-second handling",
      "text": "Digital systems often assume that every day contains exactly 86,400 seconds. That assumption is convenient for arithmetic but does not faithfully represent UTC on a leap-second day. The conflict appears in operating systems, databases, communication protocols, and file formats. Merely labeling a stored number “UTC” does not establish how the system handled an inserted second.\n\nUnix time conventionally counts seconds from an epoch at 00:00:00 UTC on 1 January 1970 while excluding leap seconds from the count. The mapping therefore does not supply a unique ordinary integer label for every UTC second at an insertion. Implementations may repeat a value, pause or step the system clock, or use another documented behavior. Differences between implementations can produce inconsistent event ordering even when their clocks agree before and after the leap.\n\nSome services use a leap smear: they distribute the one-second adjustment across a longer interval by slightly changing the effective clock rate. This avoids presenting 23:59:60 or an abrupt step to applications. However, during the smear the delivered time is not strictly UTC. Different providers can choose different intervals and mathematical profiles, so two apparently synchronized systems may disagree while each follows its own policy. A smeared source must not be mixed uncritically with an unsmeared source.\n\nCalendar time and elapsed time also require different tools. A wall clock is appropriate for recording when an event occurred, but it may be adjusted by synchronization software or administrative action. A monotonic clock is designed not to run backward and is generally the appropriate basis for measuring intervals within a running system. Its reading normally has no direct civil-date meaning. An application timing a transaction and recording its completion date may therefore need both kinds of clock.\n\nFor archives and scientific records, the time scale is part of the data. A timestamp expressed in TAI, UTC, GPS time, or a smeared scale cannot always be converted correctly without knowing the relevant offsets and conventions. Converting historical UTC to a continuous count requires the applicable leap-second history; pre-1972 records require additional information about frequency offsets and fractional steps. Future conversion can be uncertain because future leap-second decisions are not yet known.\n\nHigh-resolution timestamps also do not guarantee globally correct ordering. If two records differ by less than the combined uncertainty of their clocks and transfer paths, their apparent order may not establish their physical sequence. Distributed computing therefore supplements physical timestamps with other ordering mechanisms when correctness depends on causality rather than on approximate civil time."
    },
    {
      "heading": "Relativity and specialized scientific time scales",
      "text": "Precision timekeeping must account for the fact that clocks do not all accumulate time at the same rate. Special relativity predicts rate differences associated with motion, and general relativity predicts differences associated with gravitational potential. These are measurable engineering effects, not merely conceptual qualifications. Without the appropriate corrections, clocks at different elevations or on orbiting satellites would gradually diverge even if their atomic mechanisms were otherwise identical.\n\nTAI’s scale interval is referred to a conventional terrestrial gravitational potential rather than to the uncorrected proper time of every contributing clock. Laboratories therefore evaluate relativistic frequency corrections when their standards contribute to international timekeeping. Near Earth’s surface, a height difference of about 1 metre corresponds to a fractional gravitational frequency difference of roughly \\(10^{-16}\\). At this level, the height and gravitational environment of a clock become part of its frequency evaluation.\n\nThe Global Positioning System illustrates the operational importance of relativistic time dilation. Relative to a reference clock on the rotating geoid, the weaker gravity at a GPS satellite’s orbit tends to make its clock run faster by about 45.7 microseconds per day, while orbital motion contributes an opposing effect of about 7.2 microseconds per day. The net difference is approximately 38.5 microseconds per day before the system’s prescribed corrections. Additional terms account for orbital eccentricity and signal propagation geometry.\n\nGPS time is a continuous system time scale without leap seconds. At its epoch on 6 January 1980 it agreed with UTC, and its nominal relationship to TAI is a constant 19-second offset. As UTC accumulates leap seconds, the GPS–UTC difference changes. Satellite navigation messages supply information needed to relate system time to UTC realizations. Other navigation systems use their own conventions, so a receiver or data processor must identify the particular scale rather than treating all satellite time as identical.\n\nAstronomy uses further scales because different calculations require different reference systems. Terrestrial Time, or TT, is a theoretical time scale for geocentric applications; its realization TT(TAI) is defined by adding 32.184 seconds to TAI. Barycentric scales are used for dynamics referred to the solar system’s center of mass. Their transformations involve relativistic modeling, not simply the addition of leap seconds.\n\nUTC remains useful for observation logs, coordination, and communication, but it is not automatically the best independent variable for precise dynamics. An astronomical observation may need UTC for its recorded civil timestamp, UT1 for Earth orientation, and a dynamical time scale for orbital calculations. These parallel uses reflect different physical questions rather than competing definitions of a single universal clock."
    },
    {
      "heading": "Governance and the planned change to leap seconds",
      "text": "UTC is maintained through cooperation among organizations with different mandates. The BIPM calculates the atomic reference scales and publishes comparisons with laboratory realizations. The IERS measures Earth orientation and announces leap-second decisions. The International Telecommunication Union, particularly its Radiocommunication Sector, develops recommendations for standard-frequency and time-signal emissions. National metrology institutes, observatories, communications operators, and legal authorities provide the operational and civil links to users.\n\nDebate over leap seconds intensified as digital infrastructure became more dependent on continuous, globally coordinated timestamps. Positive insertions have required special behavior in software that otherwise assumes a fixed number of seconds per day. Telecommunications operators, satellite systems, and large computing services have often developed internal continuous scales or adjustment procedures to manage that exceptional case. The absence of a fixed long-term leap-second schedule also complicates advance calculations and systems operating with infrequent updates.\n\nThe case for retaining a close relationship between UTC and UT1 has centered on continuity with astronomical time and on applications that use Earth orientation. Celestial navigation, telescope pointing, and some tracking systems require rotational information. Their technical needs do not disappear if UTC becomes more continuous; instead, the difference between UTC and UT1 must remain available through a separate, dependable mechanism. Changing UTC therefore involves interfaces, documentation, broadcast messages, and equipment assumptions as well as clocks.\n\nIn 2022, the 27th General Conference on Weights and Measures adopted Resolution 4 on the use and future development of UTC. It called for the maximum permitted difference between UT1 and UTC to be increased in or before 2035, with the aim of allowing UTC continuity over at least a century. The resolution did not itself prescribe a complete operational replacement for the existing leap-second rules. It requested further work on the new limit, implementation planning, and coordination with relevant international bodies.\n\nThe 2023 World Radiocommunication Conference supported the transition toward continuous UTC through the corresponding international telecommunications process. The central policy direction is to relax the present subsecond constraint, not to change the duration of the SI second. A more permissive bound would allow UTC and Earth-rotation time to diverge for much longer before any corrective action was required.\n\nImplementation still requires decisions about how UT1 information will be distributed, how legacy systems will behave, and how legal definitions and technical standards will be updated. Historical leap seconds remain part of UTC’s record regardless of future policy. Software interpreting older timestamps must therefore retain the existing adjustment history even if future UTC operates for extended periods without new one-second steps."
    }
  ],
  "links": [
    "second",
    "earth",
    "metrology",
    "international-atomic-time",
    "greenwich-mean-time",
    "sun",
    "general-conference-on-weights-and-measures",
    "atomic-clock",
    "measurement-uncertainty",
    "metrological-traceability",
    "atmosphere-of-earth",
    "ocean",
    "moon",
    "leap-second",
    "internet",
    "special-relativity",
    "general-relativity",
    "time-dilation"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}