In the space digest for September 24th, there was a line you could easily scroll past: Curiosity has counted 5000 Martian sols (5137 Earth days), just a little short of Opportunity's record. I caught on not to the anniversary, but to the word "sols" — because behind it hides the entire essence: a Martian day is 39 minutes and 35.244 seconds longer than an Earth day, and this isn't a footnote from Wikipedia, but a physical quantity around which twenty years of human and machine work has been built.
The oddity is that Mars is usually described through rockets, launch windows, and regolith. But here we're talking about a metric that costs nothing, waits for no one, and has never received a common standard. For twenty years, humanity has lived on Mars by homemade clocks — each mission with its own.
I'll start with the arithmetic, because everything else depends on it. The mean Martian solar day equals 88,775.244 seconds, meaning 24 hours 39 minutes 35.244 seconds — and this is the official figure NASA cites in technical notes for its Mars24 tool. Mars's sidereal day is shorter: 24 hours 37 minutes 22.663 seconds — versus Earth's 23 hours 56 minutes.
Then begins the part intuition can't hold. A Martian hour lasts 3699 Earth seconds — about 1.65 minutes longer than ours. This means sunrise on Mars comes nearly 40 minutes later by Earth clocks each time, and Martian clock hands lap the Earth dial roughly once every 36 and a half days. Within a month, no Martian noon arrives twice at the same Earth hour.
Same with the year. A Martian year is 668.59 sols, or 686.97 Earth days. Planetary scientists have their own era: Mars Year 1 begins with the northern vernal equinox on April 11, 1955, a zero point Todd Clancy proposed in 2000 and himself called arbitrary (though lucky with a bright marker — the great dust storm of 1956). By this count, September 30, 2026 begins Mars Year 39 — exactly one month after Curiosity counted its five thousandth sol.
Mars didn't inherit a prime meridian — it was assigned, and assigned in an amusing way. In the 1830s, astronomers trying to measure the planet's rotation latched onto a small dark spot; in 1877 they used it as the zero longitude and named it Sinus Meridiani — "Meridian Bay." After Mariner 9, they made the zero a half-kilometer crater inside this bay, which in turn they named Airy-0 — in memory of George Biddell Airy, the Astronomer Royal whose Greenwich telescope established Earth's prime meridian. Mars got its Greenwich literally by inheritance from a man who never thought about Mars.
Next — the attempt to make this a standard. GISS introduced the concept of Airy Mean Time (AMT, mean solar time at the Martian prime meridian) and even carved the planet into 15-degree time zones mimicking Earth's: Olympus, for example, lives in zone AMT-9. Not one zone ever got a name — in the tool they go by technical indices.
The irony is that AMT was never used by any mission. The reason is prosaic: the position of Airy-0 itself on the planet was imprecisely known, and by the time the MER rovers began work, this uncertainty corresponded to an error of about 20 seconds in implementing AMT. Local time for a specific point was simpler and more reliable to calculate than planetary time. By 2014 the task was put in order: zero longitude is now defined as the point exactly 47.95137° east of Viking Lander 1, accurate to about six meters. But AMT didn't become a standard people actually use.
Hence — the most interesting layer. "Sol" is the count of days since your own landing, not Martian days in general. And NASA couldn't agree even with itself for years about what to call zero.
Going through missions in sequence, you see the inconsistency accumulate. Viking Lander 1 started with Sol 0 from midnight by local true solar time at a point chosen during planning (312.5°E, though the craft landed somewhere else). MER departed from this and built "hybrid local solar time" — so that by about sol 45 of the mission, the clocks would match the actual terrain: for Spirit the difference came out to over 41 minutes, for Opportunity — over 37. Phoenix landed 5 km off target but kept the old clock — result, three and a half minutes of discrepancy. InSight left its clock 85 seconds ahead of local time. Curiosity landed "long," and the discrepancy came out to just a few seconds — simply because the miss was tiny.
And here all the irony of the topic comes out in full. The difference between Earth and Martian days — 39 minutes and change. The difference between one mission's Martian clocks and its own Mars — up to 41 minutes. Error within the system turned out larger than the deviation it was measuring.
Separately: about Chinese Zhurong, these same notes contain an honest line — information on whether CNSA uses sol counts and local solar clocks has not been released. No denial, no confirmation. Just a gap in the overall picture, and I don't know how to fill it.
Now — what makes these minutes worth talking about at all. From January 2004, the Spirit and Opportunity rover team went on Martian time literally, along with their work schedule. About 250 people, and each day they came on shift almost 40 minutes later than yesterday.
When Steven Squyres, the two rovers' principal investigator, gave an interview in 2003, he started with a joke: the long day sounds attractive, "you sleep 39 minutes longer every morning" — and then explained where the trap is. The team lives on stretched days inside a world that keeps running on a straight 24 hours. Step out of the house for coffee — and you're hit by signals from an alien rhythm. Additional nuance: Spirit landed in Gusev Crater, Opportunity in Meridiani Planum, nearly exact opposite sides of the planet with different local times, so work had to be conducted not in one Martian day, but effectively in two.
The organizers didn't arrive at this immediately, and a curious document survives — a report on strategy selection, where four schedule options were compared by weights: 400 points for "accomplish mission task," 150 for number of people, 150 for process complexity, 100 each for infrastructure load and for personnel morale. Morale in this breakdown is the cheapest of five criteria. The same report notes with a caveat: the rigid Martian time regime lasts no longer than about 30 sols, beyond that additional shifts are needed. The recommendation to crews was specific — no more than four Martian shifts in a row, then days off to catch up on sleep.
What happened in practice, the report doesn't soften. One staffer after a series of shifts walked into a wall. Another fell asleep on a freeway exit ramp. Listed symptoms include stomach upset (try figuring out what and when to eat if your shift starts at 2:30), increased conflict in families where both partners carried Martian schedules, and a distinct social mechanism: living on Martian time worked as a mark of belonging. Refusing such a schedule meant in a way admitting you weren't fully in the project.
And then there was the object that made all this arithmetic visible. JPL engineers Julie Townsend and Scott Doudrick made the rounds and got rejections: mechanical watches can't be slowed, quartz ones — only by order from ten thousand units. Watchmaker Garo Anserlian of Montrose agreed, spent about $1000 of his own money on experiments — and solved the problem by adding micro-weights to the balance wheel and reworking the gear ratio: a 21-jewel mechanical watch began measuring 3699 Earth seconds per Martian hour. The first piece came out accurate to no worse than ten seconds per Earth day — exceptionally tight for a fully mechanical machine.
On January 2, 2004, the shop was packed with scientists who came for their watches. Later they approached it more cheaply: programmable wristwatches for fifty bucks, where the second increments not through 32,768 ticks, but through 33,668.833. Rounding error — about two seconds per day, a minute per 78 sols, 8.57 minutes per Martian year. A discussion surfaced in comments: one reader took 37 minutes 22 seconds — the sidereal day — and nearly reprogrammed the whole firmware with the wrong constant, because he confused which day people working shifts actually need. It's very easy to make a mistake on this difference.
Now the main thing everything's been leading to. The problem wasn't the schedule. The problem is that human internal clocks are tuned to a different number.
The internal period of the human circadian rhythm is approximately 24.18 hours. Under dim light, to which space crews are usually confined, the system simply doesn't lock onto Martian days: 24.65 hours lies outside the entrainment range. A 2007 study (Scheer, Wright, Kronauer, Czeisler) tested this directly in a lab without temporal cues and found a workaround: with moderate lighting around 450 lux for half the waking period, participants entrained to both 24.65 and 23.5 hours. The side result proved more important than the main one: after living on Martian days, subjects' intrinsic periods significantly lengthened — the first demonstrated plasticity of internal clocks.
Practice confirmed it. Phoenix program: 19 people, 78 days on a Martian day, daily diaries, actigraphy all the way through, and hormonal rhythm measurement via urine. 87% of participants achieved a period compatible with Martian days. The price is measured in hours of sleep: in the synchronized state people slept 5.98 hours, desynchronized — 4.91. A difference of an hour plus, and it emerges precisely when the mission enters its hottest phase.
A separate study, hot on the heels of MER-2004, found one more detail: with extended shifts and shortened sleep, people exhibited a pronounced 12-hour activity rhythm — the daily cycle sort of split in half, with an overall drop in level. This is the body trying to exist when the day matches neither external nor internal clocks.
It ended predictably. In November 2012, 89 sols after landing, the Curiosity team returned to Earth time — and this was planned in advance. The condition for transition was not victory over physiology, but compression of the planning process: from 16 hours in the first weeks to 12 hours, plus dispersal of scientists to their home institutions. People stopped chasing Mars and started ordering work from it in advance.
Those who want to argue should keep in mind one caveat from internal documents. The authors of the MER report wrote directly that they couldn't assemble a coherent picture of this regime's impact: different teams implemented the schedule differently, and they didn't attempt to separate the effect of Martian days from ordinary shift work on night hours. The figures about 41 minutes, the wall, and the freeway exit are there. Proof that it's all about those 39 minutes — is not.
Now the side of the topic that made me sit down to write this. While ground teams learned to live on alien days, the question "what time is it" on Mars remained unanswered at the treaty level — but on Earth they're about to decide it. Mars has no UTC of its own. There's a set of local conventions, there's AMT which nobody uses, there's MTC — "Mars Coordinated Time" as a UTC analog but without leap seconds, meaning effectively a UT1 analog; and it too isn't used by any mission.
On Earth the exact opposite is happening: they're trying to bring time to a common denominator right now. From October 13–15, 2026, the 28th meeting of the General Conference on Weights and Measures will be held in Versailles — the very structure that defines what a second is. On the agenda are two documents worth reading at least diagonally.
Draft Resolution C introduces continuous UTC starting May 20, 2027 — the leap second as practice goes away, and the allowable discrepancy with Earth's rotation widens by 2035, so the scale can live peacefully another century. The motivation in the text is purely engineering: transport, energy, telecom, and space projects require time without future edits.
Draft Resolution D concerns another planet: it speaks of a unified international lunar time scale with traceability to UTC. A reference system already exists — LCRS with coordinate time TCL, adopted by the IAU in 2024. And the figure for which all this is being done is given there: from the perspective of an observer on the Moon, Earth clocks lag on average by 58.7 microseconds per Earth day, plus periodic variations. For navigation this is fundamental, because events simultaneous on Earth are not simultaneous on the Moon. U.S. policy separately required NASA to prepare a lunar standardization strategy no later than December 31, 2026.
And here you can see where it's all heading. Mars in twenty years of work received a set of local clocks and zero international agreements — because there was nothing to decide: spacecraft tolerated it, and people went home. As soon as infrastructure appears that must be shared and navigated by, the question "what time is it" stops being a technical detail and becomes a matter for voting. The first standard beyond Earth will be lunar — and this will happen before the first habitable module appears on Mars.