I have some ideas for how to synchronize a galactic time system given the constraints of modern physics. However, my understanding of relativity is pretty limited so I was hoping for feedback or corrections if what I'm suggesting doesn't work, or if there are better ways of doing it. Thanks!

The Setting/Problem

In my setting, there is an established galactic community, but it is constrained by the limits of currently-understood physics. This means that travel between solar systems takes dozens of years for close ones, and requires massive nuclear-propelled starships to send even a small payload. The result of this is that almost all galactic trade, interaction, and commerce is digital, with comm-buoys set up orbiting most stars to relay messages between empires. Transmitting a message across the galaxy will still take 100,000 years, but this is far faster/cheaper than conventional travel.

Under this system, most everyday timekeeping is kept local (planet rotations, etc), due to relativity making time pass at different speeds in different places. But, there is a desire to have some sort of standard that people can share & convert from. So someone can say "the war lasted 10 [GALACTIC TIME UNITS]", and we can convert it to earth years ourselves.

My solution:

Here is the system I came up with:

  1. For most everyday people, the time-units they deal with are entirely local to their setting, so people on Earth still use seconds, days, years, etc.

  2. People on ships traveling at relativistic speeds will use the time system of their birth world, but will localize it and count up using onboard clocks from the moment their ship was first commissioned (with that being their Year 0). Ships will also manage a local day-night cycle like the planet their species first evolved on. This lets people get on with their everyday lives without having to think about it.

  3. For scientific measurements, there is a galactic unit of time based off the half-life of a known atom. This should be the same anywhere (right?), albeit relative to the speed at which time is passing in that place. So a scientist can describe an experiment in these units, but it's not useful for knowing how long ago for you the Glubglub War ended since time on Glubglug passes differently.

The Hard Part:

  1. The tricky part comes in when syncing up a common understanding of the passage of time. What I want is for the galaxy to have determined one singular "location" where time passes canonically, and everyone else converts to/from that for their local needs. This way you can say "this happened at 147 Galactic Standard Time" and everyone immediately knows that that means. The problem is, what to synchronize off of, and is said synchronization possible? There is a good answer here, which is similar to the above, but is skirts around that issue.

    One idea I had was to have everyone synchronize themselves off a commonly observable stellar phenomenon. Galactic rotation seems too slow and imprecise, but if we can theoretically measure it with any kind of precision it would be perfect. (It's probably OK for local clocks to "drift" a bit, as long as they can eventually resynchronize so the error doesn't compound).

    Is there some star or black hole in our galaxy that sends out a burst of light or radiation at set times which everyone can observe and measure? If so then it would be perfect as everyone can just check their clocks against said star.

    Another idea is to have one of the founding members of this Galactic Community just set their home planet as Galactic Standard Time, and make it's gravity/relativistic time speed known to everyone. This means that once someone knows how fast time is passing at said planet, they can convert it to their local time. The problem here is everyone needs to maintain 2 clocks, one for their local time, and one that tries to track the time on a planet where time passes at a different speed. The question then becomes, do you always know how fast time is passing for you, to allow for said conversion?

Finally, is there some other system I'm not thinking of which would do the trick?

Thanks, I appreciate the help/input!

  • 2
    $\begingroup$ I'm tempted to mark this a duplicate of my question, How to calculate the time on planet B for an event that occurred on planet A?, but you're making a proposition where I was asking for one. I suspect, based on the answers for my question, that you'll run into the same problems - and the same results. Read Ash's answer carefully. Cheers. $\endgroup$
    – JBH
    Commented May 19, 2021 at 1:58
  • 1
    $\begingroup$ Also similar? worldbuilding.stackexchange.com/questions/154482/…, worldbuilding.stackexchange.com/questions/106030/… $\endgroup$
    – Willk
    Commented May 19, 2021 at 2:12
  • $\begingroup$ It is somewhat unclear what the question is asking. A common unit of time is not hard to define -- the SI definition of a second is just fine. On the other hand, synchronizing clocks over the entire galaxy is much trickier because of the relativity of simultaneity; that is, observers in two different frames of reference will not agree that two arbitrary events are simultaneous of not, and if not, which event came first and which came second. $\endgroup$
    – AlexP
    Commented May 19, 2021 at 2:14
  • 1
    $\begingroup$ Regarding number 4, try en.wikipedia.org/wiki/Pulsar_clock . These pulses would be visible even from a space vessel travelling at relativistic speeds. There are probably enough pulsars in the galaxy that one could construct a timekeeping system that keeps track of the passage of time in the inertial frame of the galactic core, which serves as a close enough approximation to a universal reference. $\endgroup$ Commented May 19, 2021 at 2:21
  • $\begingroup$ The synchronisation element seems, in essence, a scaled-up version of the calculations that are routinely done in real life to synchronise clocks such as for GPS and co-ordinated universal time (UTC). Multiple clocks transmit information about their own local time signal, how this synchronises with the reference time signal, and their exact position and uncertainty. At least 4 such signals allow for improving or tracking your own position in spacetime. In essence, we'd mainly need to disseminate metadata about correction factors to easily-observed pulsars as @GrumpyYoungMan suggests. $\endgroup$
    – Steve
    Commented May 19, 2021 at 8:48

4 Answers 4


It's a great question, and I think the answer is no - you wouldn't be able to synchronize based on an event like that because of the precision required in determining when it happened. The issue actually comes down to determining where that event occurred.

Let's say civilizations on planet A and planet B try to synchronize based on a supernova, which occurs at a time $\tau$ and lies a distance $d_A$ from planet A and a distance $d_B$ from planet B. They observe the supernova at times $t_A$ and $t_B$, respectively. Light takes a finite time to propagate through space, and so - not accounting for any delays because of travel through the interstellar medium - civilization A knows that the supernova occurred at $\tau=t_A-d_Ac$, and civilization B knows that the supernova occurred at $\tau=t_B-d_Bc$. Therefore, civilization A can calculate $\tau$ if they know $d_A$, and likewise for civilization B. They can then synchronize their clocks, right?

Here's the issue: This assumes that the distances are known to the requisite precision. In reality, this is quite difficult to do. The Uncertainties of distance measurements to stars are often in the range of ~10-20%. Given that the supernova is (ideally!) hundreds of light-years away from both planets at the minimum, the civilizations may have measurement errors on the order of 10-20 light-years, meaning their clocks could be off from one another by 20-40 years. That's not great!

As an example, we don't have the distance to Betelgeuse - a luminous, important star - pinned down very well, with errors in the area of 25% or more in some cases (see e.g. Harper et al. 2008). The thing that's even more striking is that Betelgeuse can be observed continuously, and has been for decades - and yet, for various reasons, we still can't determine its location to a high precision. A one-off event like a supernova really doesn't given you the option of having more measurements, because the remnant is likely dim and difficult to observe at any wavelength.

A true galactic civilization, of course, will have to deal with a galaxy roughly 100,000 light-years across. This means that we're dealing with distance likely of several tens of thousands of light-years. Sure, technology has likely gotten much better than ours (and I very much envy those astronomers), but to have initial synchronization errors on the order of a year, you'd need distance measurements accurate to 0.01%, and that seems quite difficult. For example, say the event occurred near the center of the Milky Way. We don't even know that distance well; it's around 25,000 light-years, but many of the measurement errors are around 1,000 light-years (Malkin 2013)!

You could also ask about whether time dilation will be an issue, due to both the gravitational field of the Milky Way and the motion of stars within it. We can do those calculations, and determine that the fractional difference in time between the inner regions and an observer at infinity is $\Delta=7\times10^{-6}$ due to gravity and the difference between an observer orbiting with the Sun and an observer at infinity $3\times10^{-7}$ due to special relativity$^{\dagger}$. Those are both at least 5 orders of magnitude lower than the discrepancies we'd be looking at due to measurement uncertainties - and they wouldn't change substantially if we compared any two star systems.

$^{\dagger}$Time dilation due to a potential difference $\Delta\Phi$ can be written as $$\Delta t_r=\Delta t_{\infty}\sqrt{1-\frac{2\Delta\Phi}{c^2}}$$ for bodies at radii $r$ and $\infty$. From special relativity, we get that a star moving at a speed $v$ experiences $$\Delta t_v=\Delta t_{\infty}\sqrt{1-v^2/c^2}$$ For a typical galactic potential and a stellar speed typical of the Sun, you can check that you get the results I listed above.

  • $\begingroup$ Is it necessary to synchronise based on a single event? The clocks of different areas can be resynced by observing the arrival of light from an extragalactic supernova (chosen for having a useful duration). Everyone gets to observe that light sooner or later, logging it in the interstellar transmission-stream allows recipients to calculate the transmission-delay from any source-system. Such beacons appearing from different angles in the sky, the standard timekeeper would become a notional observer located somewhere near the galactic centre with time-dilation of someone at "infinite" distance. $\endgroup$
    – Anon
    Commented May 19, 2021 at 2:59
  • 1
    $\begingroup$ @Anon Having it be based on an extragalactic object is an interesting possible solution. Running the numbers, to have a difference in travel time (using my formula from here) of ~1 year from the center to the outskirts would require a source about 400 million light-years away, which is certainly reasonable for detecting a supernova or GRB - but the two observers would still observe it happening a year apart. $\endgroup$
    – HDE 226868
    Commented May 19, 2021 at 3:10
  • $\begingroup$ True. Going back to the OP: Duration might be standardised by a hypothetical observer keeping "the beat" with a (say) caesium clock at "infinite distance" and beaming that "beat" to galactic centre at the galactic-time-unit (events "lasted 42 GTU"). A hypothetical clock located there provides a common reference point which nobody actually uses, with Year Zero being defined arbitrarily (dates are said as "at 147 GST"). Observing extragalactic events lets you sync these clocks more easily. (Round-trip signals also give you that.) $\endgroup$
    – Anon
    Commented May 19, 2021 at 3:43
  • $\begingroup$ From my understanding the huge error bars in interstellar distance come from the fact that we are trying to measure them all from earth, so from a single point. One can measure the angle between two different stars as seen from earth to a much better accuracy. Once you can make these angle measurements from several distinct stars you can compute distances to the same accuracy using trigonometry. $\endgroup$
    – quarague
    Commented May 19, 2021 at 12:10
  • $\begingroup$ @quarague You'd still need to know some distances between certain objects, though, subject to the same issues - and given that the OP seems to want to obey relativity, it would take quite a long time to travel to places to make the measurements and communicate the information to the home planet. $\endgroup$
    – HDE 226868
    Commented May 19, 2021 at 12:46

First I agree with the points made that this time synchronization would not be very important in a galactic "civilization" with such limited interaction. Also that time dilation would have a slight but not huge affect on planets in different solar systems because they're all in similar gravitational fields and all moving at far less than 1% the speed of light relative to each other.

Rather than half life of an element I would look to the oscillation period of a neutral hydrogen atom. The Hydrogen Line. This is what NASA used on the Golden Record on the Voyager Probe to indicate a recognizable unit of time to any aliens that might intercept it.

The easiest way to keep time between worlds would be to put atomic clocks on all the starships to keep time from the time they leave earth. These remain accurate to within one second after running for 300,000 years. However, assuming these starships travel at any notable fraction of the speed of light (anywhere near 1% or more) you'll worry about time dilation. If it was my galactic empire I would come up with some technology to receive the radio signal from some quasar, presumably the one with the strongest signal on average across the Milky Way. Quasars aren't found in the Milky Way but they would be detectable throughout it. Everyone would receive the same radio pulses from the same quasar but, for instance, if you were traveling at 10% the speed of light on a starship, you would observe the frequency to be lower than would observers living on a planet. Maybe time 0 is the time the first starship left earth and you count galactic time by the number of oscillations of the radio frequency of the selected pulsar since then.

This would be different periods of perceived time to different people and planets, but it would be a reliable galactic standard. This would be useful for galactic historical records like you mentioned but of course it wouldn't be relevant to the average citizen or even scientist on any given planet. More for record keeping and communications protocols than anything else I would think.

Also should be noted that there are other bodies in space that give off frequencies in the radio spectrum any many others. But, my understanding is that quasars are very powerful and would be reliable for this purpose. Google tells me that they tend to last 100-1000 million years. On some timelines it could take around this long for us to actually populate the galaxy but I would also think that that's plenty of time for us to come up with an artificial radio pulse powerful enough for everyone in the galaxy to pick up on. Once that is constructed the galaxy could be instructed to switch to this artificial frequency after some determined (upcoming) oscillation number of the pulsar. Alternatively, there may be another type of body with similar useful properties that also lives longer. It guess it all depends on how long it's been since first launch in your world.

P.S. If you haven't seen Sharkee's video on how we could feasibly populate the galaxy with reasonable limits like these you should check it out here. One of my all time favorite YouTube videos.

Edit: if you picked a quasar that is "off to the side" of the milky way, the close side would receive the pulse 52000 years before the far side. If you chose a quasar that is "above" the disk galaxy, then everyone would receive the pulse at closer to the same time in a sort of radio wall passing through the whole galaxy at once. Alternatively, if you picked an off to the side one, the receivers could just account for the difference because it would be well known where they are in relation to the rest of the galaxy and the pulsar.

  • $\begingroup$ A note about quasars: I don't think they really pulse with any regularity; there may be outbursts of activity, but not usually anything cyclic. $\endgroup$
    – HDE 226868
    Commented May 19, 2021 at 12:40

Frame Challenge:

If you are limited to light speed transmission you aren't going to have a 'galactic civilization', you might have clusters of star systems in (very slow) communication but they are each going to use time systems that make sense to them. Most likely the time system of the source system of the local diaspora.

The point is, what possible information could civilization A on one far end of the Milky Way have to say that would be of any interest to civilization B 100k LY away (other than perhaps 'we were here'), but that doesn't require any sort of agreed-upon system of time units).

  • $\begingroup$ So exchange is limited, true, but this is kind of the whole idea behind my setting. Since technological advancement has mostly ground to a halt due to running into the hard limits of physics, things in this galaxy progress a lot more "slowly". Wars can last tens of thousands of years, or more, as one civilization sends a ship to the outer edge of an enemy system and hides, capturing asteroids to build an automated attack fleet (sending a large force is impractical). Moreover, arts and entertainment are regularly transmitted and traded between systems, despite the long transmission time. $\endgroup$ Commented May 19, 2021 at 23:32

There's not really going to be a huge difference in times between solar systems within the galaxy; the galactic gravitational field is not that strong, and relative velocities between stars are only of the order of hundreds of km/s, which won't produce vast time dilation effects. Keeping clocks synchronized to the millisecond will be hard, but keeping them synchronized to within a few seconds should be do-able.

This is in some sense a vastly scaled up version of the problems encountered in timekeeping even here on Earth. The consensus seems to be to use a standard reference ellipsoid (the world geodesic system) and synchronize clocks to this, adjusting clock rates for local conditions like height. There's a lot of finicky calculations to do this: modern atomic clocks are in fact sensitive enough to detect differences in the passage of time due to height differences of less than a meter, so even here on Earth relativity must be taken into account.

I would imagine that your galactic civilization could come up with some similar "standard" galactic reference surface (perhaps a disk or sphere centered at some point near the center of the galaxy). Correcting local time to match the notional standard time will be harder than it is on Earth, since broadcasting time will be slow. But everyone could use a set of reference pulsars as a starting point.

EDIT: in order to fulfil the "hard science" requirement: the formula for time dilation due to relative velocity $v$ is $t' = t * sqrt(1-(v^2)/(c^2))$. Gravitational time dilation is proportional to the gravitational potential, and in fact is equal to the relativistic time dilation of the escape velocity from the potential; escape velocity from the galaxy is about 550 km/s, so this gives an upper bound of 53 seconds per year for gravitational time dilation within the Milky Way. Relative velocities of stars within the Milky Way are clearly less than the escape velocity, so special relativistic (non-gravitational) time dilation will be less than this.

A good reference for time keeping in the GPS, which discusses some of the subtleties, is Relativity in the Global Positioning System


Not the answer you're looking for? Browse other questions tagged .