The White House just released the "Policy on Celestial Time Standardization in Support of the National Cislunar Science and Technology Strategy" (https://www.whitehouse.gov/wp-content/uploads/2024/04/Celestial-Time-Standardization-Policy.pdf).

According to this, Coordinated Lunar Time (LTC) would be defined based on "an ensemble of clocks on the Moon", so that "users in Cislunar space can reference a time standard defined near the gravitational environment in which they operate".

It sounds like this will indeed be an independent Moon time; it only requires "traceability" to the Earth-based UTC, so that that LTC can be converted to UTC (within specific tolerances, using distributed offsets) for operations interacting with Earth.

@dp9000 If I understand correctly, traceability means that it's not independent.

It's useful to have a bunch of atomic clocks on the moon even if they're not independent from Earth, because they can tell you how much time has passed in between timing pulses you might get from Earth, and can keep time well enough if the synchronisation cuts out for whatever reason. But if it's traceable to UTC, then if I understand right, that means they're being synced to UTC back on Earth and slewed to correct for errors.

This would be similar to how you can get a quartz oscillator on Earth and attach it to a GPS receiver, and adjust the temperature of the quartz crystal until it's ticking at the same rate (65536 oscillations per GPS second, or whatever). You've now got your own timing standard - the quartz oscillator - that you might use to keep all your lab equipment's timing in sync. But it's not independent, it's still based on GPS time.

Similarly, GPS satellites in Earth orbit are not on the geoid, and contain atomic clocks ticking at a different rate than on the surface for the same reason that clocks on the moon tick at a different rate. They define "GPS time", which, nonetheless aims to track time at the geoid. GPS time is synced to UTC (well - more like TIA with some leap seconds grandfathered in) regularly, so although it is a separate time standard to UTC, it's not independent.

It sounds like this is how the US proposal would suggest things should work on the moon - atomic clocks defining a local time standard, but one that is ultimately still synced to UTC and slewed to correct for errors.

It's possible I've misunderstood something, it's a bit confusing to think about. But that's my impression at the moment.

@chrisjbillington one way the above could be wrong - they could define LTC to be synced not with UTC, but with UTC times a fixed scaling factor, chosen to match the gravitational time dilation factor between the Earth's geoid and a hypothetical lunar geoid (or a surface of a specific fixed gravitational potential - not quite the same thing I think given local gravity changes slightly throughout the moon's orbit).

That way SI seconds and LTC seconds would actually have the same duration (if you're on the lunar geoid).

That seems like the less likely option though. And it would still be referenced to UTC, just with a scaling factor.

(Edit: re-reading the memorandum with this question on mind, it does kind of sound like this might be what they're going for! They mention wanting to not break systems that assume SI seconds. I take it back that this is the less likely option.)

Perhaps it's not 100% clear what this market means by "independent" vs "referenced to Earth". If the seconds go by in LTC at a different rate than in UTC, but it's by definition exactly 58 microseconds extra per day, and constantly being compared with UTC to make sure that's the case, is that "independent moon time", or "referenced to Earth"? I'd assume the latter, but perhaps others have different thoughts.

@chrisjbillington The way I understand it, LTC as proposed would be defined by clocks on the Moon, analogous to but independent from UTC/TAI. These lunar clocks would be synchronized only between each other and not with Earth clocks, so LTC would be allowed to drift away from UTC in a not accurately predictable way.

I think the whole point is that PNT (position, navigation , timing) applications require higher precision than what can be attained by synchronizing individual lunar clocks from Earth; they need an independent time standard that does not have its precision "ruined" by Earth synchronization.

Traceability would just mean that the difference between UTC and LTC is continuously measured to make conversion possible, for applications that need it.

@chrisjbillington So my thinking when creating the market was there were two possiblities:

1. Lunar clocks would regularly check in and sync with Earth clocks reading UTC (or similar). Timekeeping would be entirely referenced to the Earth, and there would be no independent lunar time.

2. An array of atomic clocks would be put on the moon to define an LTC, in much the same way UTC is defined on Earth. This LTC would be allowed to drift from UTC as suggested by @dp9000. There would be no reference to an Earth standard time; lunar time would be entirely independent.

It's clear now that there is at least one more possibility/set of possibilities:

3. An array of atomic clocks is put on the moon to define an LTC, but this LTC is forced to sync up somehow with Earth UTC. (This could be either via a scaling factor to make the duration of the Moon second match that of the Earth second, or via regular offsets like leap seconds, or some other way.)

I can't see how possibility 3 is consistent with the plain meaning of an "independent system" of lunar time, since it is clearly regularly "referenced" (using the plain meaning of the word) to Earth time. So, given the way I worded the question, I'm inclined to say that possibility 3 will cause the market to resolve "referenced to an Earth standard time".

As an aside, whatever system is chosen, it will clearly be necessary to be able to convert from Earth time to lunar time. But I don't think the mere existence of a conversion factor bears either way on the market.

@dp9000 you can do both - have an array of clocks that sync with each other, and have the whole thing steered to keep it synced with earth. You get better short-term timing precision by having local clocks (and more than one of them), which is necessary for good PNT. And you get better long-term timekeeping by having it sync with Earth.

It's the same concept as the disciplined quartz oscillator I mentioned above. Useful to have a local oscillator providing short-term timing precision (GPS pulses don't come often enough), and even better if you have more than one syncing with each other. But you can still keep them in sync on longer timescales with an external reference, and then you get both short-term precision and long-term accuracy.

I think you're right about traceability though - it can be established with after-the-fact measurements, whether you're syncing or not.

And now that I think about it, Earth's geoid where UTC/TIA is defined isn't a gravitational isopotential either*, so it doesn't make sense to define another time standard at some other isopotential via a fixed scaling factor. You're going to have to measure the difference, you can't define it.

My bad! I think I was totally wrong.

(* it is at a given moment in time, but is not fixed over time - "mean sea level"? Sea levels change! Edit: that's not even correct, it's an isopotential of "geopotential" which is not just gravity!)

Is there the same market about Mars?

@bessarabov it's not quite the same question, but this is a related one:

@chrisjbillington Ooh, very cool.

Clocks tick at different rates on Earth too, depending on location and altitude, and they tick at different rates at different locations on the Moon or in lunar orbit. Not to mention that they tick at different rates as you travel to the moon on a spacecraft.

Wherever you are on Earth, unless you happen to be on the geoid, you can't naively use the tick rate of a local atomic clock to track UTC/TIA anyway. You have to correct for gravitational effects in any case. This is already how it works.

Clocks ticking at a different rate on the moon is not a new problem, it is just the same problem as on Earth to a greater extent, and I suspect will not be a reason to depart from UTC/TIA.

They will simply use different correction factors depending on location to estimate UTC/TIA, if they use local atomic clocks, just like we do here on Earth.

@chrisjbillington I don't really get the different clock rate thing.

I guess I don't understand gravity well.

(No need for anyone to explain, I *do* (sorta) understand it. It's just a stupid pun.)

@chrisjbillington Wait, did you read the whole article? The point is that for precise operations like navigation and subcentimeter measurements, we rely on UTC + a geoid to manage the variations in gravity across the earth’s surface, thus as more separate entities operate on the moon doing precise operations in parallel, developing a moon-centric spatial model plus moon-time makes sense, to keep everyone on the moon in sync.