Moon Time

Lunar time has two halves, and the calendar is the one you can see with your own eyes. Measured against the stars, the Moon goes once around Earth in 27.32166 days, and it spins once on its axis in the same time, which is why the same face always points our way. But a day in the everyday sense, from one sunrise to the next, takes longer: 29.53059 days on average, the same as the interval from one new moon to the next. The extra 2.21 days come from the trip around the Sun. Earth and Moon move along their orbit together, so after one full turn against the stars the Moon has to turn a little further before the Sun is back in the same place in its sky. Astronomers call the first period the sidereal month and the second the synodic month, or lunation, and even the lunation is only an average: NASA's eclipse tables show individual ones running up to about seven hours longer or shorter.

On the ground near the equator that works out at roughly fourteen and three quarter Earth days of sunlight followed by the same again of darkness. The Sun crawls: it moves about 12 degrees a day, half a degree an hour, so it takes more than a week to climb from the horizon to noon. Bernhard Hufenbach of ESA's Moonlight team put the problem for future crews plainly in 2023: in the equatorial region each day is 29.5 days long, including freezing fortnight-long lunar nights. From the near side, Earth hangs in roughly the same patch of sky all month, rocking a little as the Moon librates; from the far side it never rises at all, though the far side gets just as much sunlight. Near the poles the pattern changes. The Moon's spin axis is tilted only about 1.5 degrees, so the polar Sun skims along the horizon, and some crater floors near both poles are permanently shadowed.

The night is the hard part. NASA gives lunar surface temperatures of about 127 °C in full Sun and about minus 173 °C in darkness, and the first night reading from China's Chang'e 4 lander, in January 2019, bottomed out at minus 190 °C. For a solar-powered spacecraft, sunset means two weeks with no power at those temperatures. Mission designers have three ways to deal with it: finish the job within one lunar day, carry a heat source that does not depend on the Sun, or switch everything off and hope.

Most landers take the first route. India's Chandrayaan-3 lander and rover were specified by ISRO for a mission life of one lunar day, about 14 Earth days. Intuitive Machines' Odysseus touched down on 22 February 2024, worked for seven Earth days and fell silent once the Sun set; the company listened for it from 20 March, when it calculated that enough light might be reaching the solar arrays again, and on 23 March concluded that the lander had permanently faded. Firefly's Blue Ghost 1 shows what planning around the clock looks like when it goes right. It landed in Mare Crisium on 2 March 2025, worked through 346 hours of daylight, photographed the sunset from the surface and kept operating for just over five hours into the night before the mission ended on 16 March.

The second route is older than most people assume. The Soviet Lunokhod 1 rover drove on solar power by day and parked at night with a polonium-210 radioisotope heater keeping it warm; it was meant to work for three lunar days and worked for eleven, until operations ended on 4 October 1971. China followed the same principle four decades later. According to a 2020 review by engineers at the China Institute of Atomic Energy, China's use of plutonium-238 in space began with Chang'e 3, which carried heater units imported from Russia: six were bought in 2009, three rated at 120 watts of heat, one at 8 and two at 4. Chang'e 4 carried plutonium-238 heat sources and a plutonium-238 generator, both developed with Russia, and the generator powered the thermometer that logged that first minus 190 °C night. When the Sun came back to Von Kármán crater, the Yutu-2 rover woke itself at about 20:00 Beijing time on 29 January 2019 and the lander followed at 20:39 the next evening.

Nuclear heat produced the longest records on the Moon. In September 2024 Chinese state television reported that Yutu-2 had finished its 71st lunar day and driven 1,613 metres against a design life of three months, and Zuo Wei, a deputy chief designer of the Chang'e 4 ground application system, said both the Chang'e 3 and Chang'e 4 landers were still active. Those were the most recent official figures available when this page was checked.

The third route produced the strangest result of recent years. Japan's SLIM landed on 20 January 2024 within about 10 metres of its target, but one of its two main engines most likely lost thrust near the end of the descent and it came to rest on its nose, with its solar cells facing west. JAXA had expected surface operations to last only a few days, and power returned only on 28 January, once the Sun had moved round. The lander was never designed to survive a lunar night. It answered Earth after its first one anyway, in a window from 25 February to 1 March, and again from 27 to 30 March and from 23 to 29 April. JAXA's own announcement describes those three rounds of operation after crossing the night as something the mission had never planned, and its December 2024 briefing says it is difficult to determine how SLIM survived. Attempts to reach it in later lunar afternoons failed, and operations were formally ended on 23 August 2024.

Surviving the night matters for timekeeping too. Kevin Coggins, deputy associate administrator for NASA's Space Communications and Navigation programme, explained in an interview recorded on 10 March 2026 that the atomic clocks a lunar time system needs want a stable temperature and continuous power, because every shutdown means a fresh warm-up and resynchronisation. He expects some of the clocks on the surface to be "equipped to survive the night". This is where the two meanings of Moon time meet: the slow calendar of sunlight, and the rate at which a clock actually ticks.

The clock problem comes from relativity. A clock deeper in a gravity well ticks more slowly, and so does a clock that is moving fast. A clock on the Moon sits in a much shallower gravity well than one on Earth's surface, and once the gravity and velocity effects are added up it comes out ahead. NIST physicists Neil Ashby and Bijunath Patla calculated in 2024 that a clock on the lunar surface gains 56.02 microseconds a day on one at sea level on Earth, accumulated over a lunar orbit, and ESA, NASA and NIST all quote about 56. The White House memo that launched the American effort gave a different number: to an observer on the Moon, it said, a clock on Earth appears to lose on average 58.7 microseconds per Earth day, with additional periodic variations.

The two figures do not contradict each other, because they describe different clocks. The 58.7 figure matches the average rate of Lunar Coordinate Time, the idealised time scale the International Astronomical Union defined for the Moon in 2024, against Earth's time scales. China's LTE440 ephemeris puts that rate at 1 + 6.798 × 10⁻¹⁰ relative to barycentric dynamical time, which by its IAU definition keeps pace with Earth's Terrestrial Time on average, and 6.798 × 10⁻¹⁰ of a day is 58.74 microseconds. A real clock standing on the lunar surface sits in the Moon's own gravity well and runs about 2.7 microseconds a day slower than that ideal, according to Pascale Defraigne of the Royal Observatory of Belgium and her colleagues, which brings it to about 56. The exact rate also shifts with the clock's position, on the ground or in orbit, and with periodic terms that follow the Moon's orbit. As Defraigne's group puts it, a universal difference between lunar and Earth time does not exist; there are only differences between particular clocks.

Fifty-six microseconds sounds like nothing, but light covers about 17 kilometres in that time, and satellite navigation works by turning signal travel times into distances. NASA's lunar navigation lead Cheryl Gramling put it in football fields: an observer on Earth who ignored the effect for a day would place an astronaut orbiting the Moon about 168 of them away from where the astronaut really is. Earlier missions could live with it because each one ran on its own clock. Mercury, Gemini, Apollo and the shuttle ran their timelines on mission elapsed time counted from launch, and ESA noted in 2023 that each lunar mission until then had run on its own timescale exported from Earth, kept in step through deep space antennas. NASA planned Artemis II the same way. That stops working once several landers, rovers, relay satellites and crews have to share one time and fix their positions without waiting on Earth.

The most direct political push came from Washington. On 2 April 2024 the White House Office of Science and Technology Policy directed NASA, working with the Departments of Commerce, Defense, State and Transportation, to define and implement a Coordinated Lunar Time, abbreviated LTC, with four required features: traceability to UTC, accuracy good enough for precision navigation and science, resilience to loss of contact with Earth, and scalability to places beyond the Earth-Moon system. NASA has until 31 December 2026 to hand the White House a finalized implementation strategy. In September 2024 the agency said LTC would be set by a weighted average of atomic clocks at the Moon, with exactly where still to be determined. Congress has not written any of this into law. The Celestial Time Standardization Act (H.R. 2313), which would require it, was approved by the House Science Committee on 29 April 2025, and similar provisions sit in NASA authorization bills that House and Senate committees approved in February and March 2026, but as of 22 September 2026 none had passed either chamber.

The astronomers moved four months later. In August 2024 the International Astronomical Union's General Assembly adopted Resolution II, which defines a Lunar Celestial Reference System and its time coordinate, Lunar Coordinate Time (TCL), built with the same equations the IAU adopted for Earth's geocentric system in 2000, with lunar quantities put in place of Earth's. TCL runs in SI seconds and is tied to the other coordinate times by one arbitrary instant: it reads 1977 January 1, 0 h 0 m 32.184 s when Barycentric Coordinate Time reads the same at the centre of the Moon. The resolution adds that the same formulae work for any body in the Solar System, and a January 2026 paper in Metrologia notes that whatever is adopted for the Moon will be reused for Mars. TCL is a mathematical scale that real clocks are compared against. It still has to be realised by clocks on and around the Moon, and it does not by itself say what a lunar clock display should read.

Europe came at the problem through navigation. At a meeting at ESA's ESTEC centre in the Netherlands in November 2022, the agencies involved agreed on the importance and urgency of a common, internationally accepted lunar reference time, as part of LunaNet, a shared set of standards for lunar communication and navigation. ESA's Moonlight programme and NASA's Lunar Communications Relay and Navigation System are meant to broadcast positioning signals around the Moon, and ESA has said the two should use the same timescale; a 2024 report for Moonlight introduced TCL and proposed ways to realise it with clocks in orbit and on the ground. The current LunaNet Interoperability Specification, version 5 of 29 January 2025, written and approved by NASA, ESA and JAXA, requires every provider's signals to be referred to a LunaNet Reference Time. The standard that is supposed to define that reference time is still listed in it as to be determined, and version 5 was still the edition NASA was publishing on 22 September 2026.

China's most visible contribution so far is a piece of public mathematics. In December 2025 the Chinese Academy of Sciences' Purple Mountain Observatory released LTE440, a numerical lunar time ephemeris built on the IAU's definition of TCL and on JPL's DE440 planetary ephemeris, published in Astronomy & Astrophysics and posted openly on GitHub. It converts between TCL and the Solar System's barycentric time scales; the paper claims an accuracy better than 0.15 nanoseconds before 2050, and the Academy's announcement of 22 December 2025 says the accumulated error stays under one twenty-millionth of a second even after a thousand years. The observatory's own framing, reported in January 2026, is frank: it calls the definition of lunar standard time an international hot topic that is not yet settled and has become a focus of competition among the space powers. China is also studying its own Queqiao constellation of lunar orbiters and relay spacecraft, which the lunar programme's chief designer Wu Weiren said in April 2025 would provide communication, navigation, positioning and timing: one more system that will need to agree on time with the rest.

The live argument is over what, exactly, the official lunar time should be, and it has not been settled. Defraigne and her co-authors set out three options. The first is to use TCL exactly as defined, which is what they recommend: scaling it would force a matching rescaling of distances and mass values in every Earth-Moon calculation, and the lunar equivalent of sea level that one of the scaled versions would need has not been defined. The second would scale TCL so that its average rate matches a perfect clock on the lunar surface, the way Terrestrial Time is matched to clocks at sea level on Earth, so that a good surface clock would tick the reference time directly. The third would scale it so that it never drifts away from Earth time at all, leaving only periodic wobbles. The US memo describes Lunar Time only by analogy with Terrestrial Time, set by an ensemble of clocks, and ESA noted in 2023 that it was still open whether a single organisation should keep lunar time, and whether it should run independently on the Moon or stay synchronised with Earth.

The next decision point is in France. The General Conference on Weights and Measures, the body that oversees the SI and the international time scale UTC, meets at Versailles from 13 to 15 October 2026 with a Draft Resolution D on the definition of an international lunar reference time scale and its traceability to UTC. The latest published text, version 5 of 13 July 2026, says a proliferation of reference time scales must be avoided, that thorough studies have shown TCL without any scaling would meet the scientific requirements, and that space agencies' operational choices may add requirements of their own. It recommends that if a version of TCL scaled to a lunar geoid is considered, its offset be fixed by a conventional value set by bodies such as the IAU and the International Association of Geodesy, and it asks the metrology community to work with the agencies on the realisations of a single lunar reference time scale. It does not itself adopt one, and what the conference decides will not be known until it votes. The same meeting is also being asked to make UTC continuous, with no further leap seconds, from 20 May 2027, so the Earth time that lunar time must trace back to is changing as well.

So, as of 22 September 2026: the Moon has an internationally defined coordinate time, TCL, but no operational lunar time that clocks keep and people read. This page found no published NASA strategy, the LunaNet document meant to define the reference time is still a placeholder, and when Coggins was asked in March 2026 whether NASA knew how many lunar clocks it needed and where to put them, he answered: "We don't." Proposals keep arriving, among them a May 2026 paper in PNAS by Jun Ye and colleagues suggesting an ultrastable laser cavity inside a permanently shadowed crater to anchor a lunar time standard. None of it is on the Moon yet. Meanwhile the calendar half of Moon time keeps its own schedule, and every lander that arrives still has to plan around a sunset about two weeks away.