# Assuming...

...intelligent civilisations exist throughout the entirety of the Milkyway galaxy, and these civilisations wish for some kind of common time reference, a GTC — Galactic Time Coordinated — if you like.

Let us say they want a precision of this GTC to be +/- 1 second — as per the Système Internationale — no matter if you are in the Centaurus Arm or the Pegasus Arm; whether you are at the galactic core or the Outer Halo.

"Space[,] is big. Really big. You just won't believe how vastly, hugely, mind-bogglinly big it is!"

# What...

...apart from flying atomic clocks around and synching them to each other, can be used as a galaxy-spanning time reference, to synchronize clocks throughout the galaxy?

• In principle, this cannot be done due to relativity. Two observers standing next to each other may synchronize their clocks and see them ticking at the same rate. When one of them starts moving fast enough, both will see the other's clock ticking slower. It is not possible to use one "universal clock" in all reference frames - the duration of time between two events is not universal, but rather depends on who's observing it. There is no absolute meaning to "the same time" for all observers. Feb 14 at 19:06
• @MichaelK comment @-pings only work if the user has been involved with the post. Generally has commented, edited, or close voted. For more information see How do comment @-replies work?
– VLAZ
Feb 14 at 19:22
• I once asked How to calculate the time on planet B for an event that happened on planet A?. It's worth reading because the basic problem is it might not be possible to have a universal time reference. The problem is that everyone is observing the "reference" from their own frame of reference, which is whomping difficult to reconcile with someone else's frame. Note that my link might constitute a duplicate question reference, but I'll let others decide that.
– JBH
Feb 14 at 19:53
• even flying atomic clocks around to synch up won't work for the same reason. as soon as the clock moves its no longer in synch.
– John
Feb 14 at 21:08
• Time don't flow at the same rate, remember we need to tweak GPS satellites because the gravitational pull is slightly lesser up there than on ground also it is moving relatively fast. But you can still coordinate time with others just that one of you need to do some calibrations and in my example it is the GPS. Feb 15 at 0:33

galaxy-spanning time reference

# Nothing

These kinds of things work on Earth because anyone's relative speed to anyone else does not reach a considerable fraction of the speed of causality ($$c$$). Therefore, among ourselves, we have an illusion of simultaneity.

Once relativistic speeds come into play, this illusion is dispelled.

The suggestion to use pulsars seems good, but only works for a single planet or star system at best. You catch one of the Pioneer probes midflight and parse its message, you can math out where and "when" in the galaxy you had a place where the pulsars aligned like that.

But for any two or more places in the galaxy that are very distant from each other, and have a high velocity compared to each other, you get a very curious couple of relativistic effects. The first, more obvious one is that each sees the other experimenting dilated time. This is the stuff of mainstream sci-fi, so I am not going to elaborate further on this - I'll just remind everyone that this makes the synchronization of clocks a pain.

You might then think that this is not a problem, since the pulsars provide the tick and tock for the galactic clocks. Nope. These distant points in the galaxy will each have a different set of velocities relative to the pulsars, so they will measure their periods - and the ratios among those periods - differently. You say pulsar A has a period that is twice as long as pulsar B, I say it's the other way around.

Now let's make things worse, because science ruins everything. Once you have observers far enough and moving fast enough, the second effect I wanted to mention kicks in. Except for causally connected events, these observers will not agree on the order of the events they see.

Let's say that Alice is orbiting Sagitarius A* really close, while Bob is on Earth. They both use pulsars around the galaxy for timekeeping. Furthermore, the globular clusters M2 and Omega Centauri each elect their respective presidents for the next galactic term.

Here comes the fun part: due to relativity, Alice and Bob do not agree on the periods of pulsars in relation to one another, so they do not agree on the time that should have passed for each other since their last sync. Worse, they also do not agree about which president got elected first. It's not just about getting the signal from one place before getting the signal from another. Even if you adjust for time dilation on each point, you will still have different parts of the galaxy disagreeing about the orders of disconnected events happening far away.

TL;DR even if you manage to negotiate a system sync between two distant points, from then on they won't be able to agree on how much time has elapsed, nor the specific "galactic" dates and times that anything happened, nor even in which sequence. This renders any galactic date and time system useless.

I see this answer has gathered a lot of criticism, specially regarding the confusing nature of the impossibility of having a galactic calendar. I hope to clarify this point with a few words from Roger Penrose. This is a discussion that can be found in the book The Emperor's New Mind, about a hypothetical scenario in which two people on the street have been observing events going on in the Andromeda galaxy. This is an exxageration, as much as Schroedinger's cat is the exxageration of a particle, but it is very illustrative. Emphasis are mine:

Two people pass each other on the street; and according to one of the two people, an Andromedean space fleet has already set off on its journey, while to the other, the decision as to whether or not the journey will actually take place has not yet been made. How can there still be some uncertainty as to the outcome of that decision? If to either person the decision has already been made, then surely there cannot be any uncertainty. The launching of the space fleet is an inevitability. In fact neither of the people can yet know of the launching of the space fleet. They can know only later, when telescopic observations from Earth reveal that the fleet is indeed on its way. Then they can hark back to that chance encounter, and come to the conclusion that at that time, according to one of them, the decision lay in the uncertain future, while to the other, it lay in the certain past.

This is an elaboration of the Rietdijk-Putnam argument. Again, emphasis mine:

If special relativity is true, then each observer will have their own plane of simultaneity, which contains a unique set of events that constitutes the observer's present moment. Observers moving at different relative velocities have different planes of simultaneity, and hence different sets of events that are present. Each observer considers their set of present events to be a three-dimensional universe, but even the slightest movement of the head or offset in distance between observers can cause the three-dimensional universes to have differing content.

The argument itself, as well as its criticisms (also seen in the wiki above) are focused on the amount of dimensions the Universe has (not taking into accoint String Theory, which was starting at around the time this argument was being published). Both the pro and con sides concede that the timing of faraway events will be different for different observers, and thus two observers using a distant object for timekeeping will never agree on timestamps. This, in my opinion, is what renders the idea of a single galactic calendar a moot point. You end up with a calendar of a priviledged point in the galaxy, not a calendar that is meaningful for the whole galaxy.

• I am not going to give this one the accepted because you have not shown that the issues you bring up makes the problem unsolvable. Harder to grasp and understand? Yes. But unsolvable? No, you have not shown that. And when you say "science ruins everything" and then do not present any hypothesis, test, result and/or review... instead just reling on the mere word "science" as some sort of appeal to authority... no. Feb 17 at 0:30
• @MichaelK the crux of the issue is this: Except for causally connected events, these observers will not agree on the order of the events they see.. It is an effect of Relativity as theorized by Einstein. His theory has withstood challenges from minds much greater than mine for decades. It's not that I think that this is something that can be solvable or unsolvable, it's just that it does not make sense. What you ask makes as much sense as FTL without time travel, or for Galilean transformations to be valid here. I can't state it in simpler terms than these. Feb 17 at 1:09
• @MichaelK for more explanation, probably you would need to go to Physics SE. The Square-Cube Law here is trying to make the answer understandable to laymen, while keeping the accuracy of the claims as much as possible. Feb 17 at 10:47
• All of these effects can be measured and corrected, and if a civilisation has the tech to reach the stars in the first place they would very likely not have much of a problem doing so. Sure, there's no uniquely defined notion of simultaneity that's the same for all observers, but that doesn't matter - you just decide to agree on some specific foliation of space-time and calculate your time relative to that. Feb 17 at 12:53
• Maybe a story could have The Empire(tm) use these flawed techniques. But it just doesn't care about the problems you pointed out. This would then lead to ham-handed tyranny upon which the plot hinges. Feb 18 at 4:36

## Use Pulsars AND Novas

MichaelK already pointed out the benefit of pulsars, but there is a second component to the problem that needs to be addressed. Just because everyone is counting the same number of tics per minute does not mean that everyone is recording the same timestamp on the same tick. For that, you need to exchange information on events that do not happen at a regular interval.

Novas are somewhat common events (happening about 10 times per year in the Milky Way alone), and like Pulsars, they are energetic enough for anyone in the galaxy to be able to detect them. By sending astronomical data from one planet's reference point about when and where a few Novas were recorded, another planet could use its relative distance and time delay to those same events to sync their clocks.

Granted, 2 planets 50,000 light years apart will probably not have enough astronomical data to share to sync with each other (unless it is a >50,000 year old space faring civilization), but if you assume any two inhabited planets are only 10s of light years apart, (and you have FTL communication) then you can daisy chain your records across many worlds and bring them all into sync in a relatively short amount of time.

## Why time dilation is non-issue

While an atomic clock located on Earth tics at a different speed than a clock located on Kepler-22b, the rate at which a pulsar pulses is based on its own reference frame which makes time dilation a self solving problem.

Imagine you get on a ship and start cruising along at 0.95C at an angle that is parallel to the pulsar. To you, the pulsar will appear to start pulsing ~3.2 times faster, but your clock will stay in sync with UTC because this lets you count time from its reference point, not your own.

The tricky part here only comes into play when you have to consider how far you are from the pulsar. If you were to fly directly at the pulsar at 0.95C, you would not only experience time dilation, but also a Doppler effect meaning that you would actually see the pulsar increase to a frequency of 6.24 times its normal rate. This second displacement WOULD have to calculated for. This would create a sort of Time Zone system based on how far you are from the pulsar. So if you wanted to fly a certain distance closer to the pulsar, you would have to overtake X number of pulses; so, your UTC time zones will be +/- however much closer or farther from the pulsar you need to fly.

While this technically means that planets themselves will not quite be in fixed time zones either, the speed of a planet compared to the speed of light is very small, especially if you pick an extra-galactic pulsar that is more or less in line with our galaxy's axis of rotation. This will reduce your level of error to being about as big of a deal as the leap seconds that we have to deal with in the time system we already use. It's a thing the smart people can adjust for, and everyone else would just go about their day without even noticing.

• You wouldn't actually need to synch, you can create "timezones" where a time is counted as "42492 megapulses of pulsar P45 after Nova N563-b". Feb 15 at 13:57
• @Borgh To establish "timezones", you still need to sync your underlying time system. Imagine your reference Nova happened 1000 years ago. One planet is 300 LY away and was not advanced enough to record it when the light hit them, one is 950 LY away and recorded it just fine, and another is 1100 LY away and have not recorded it yet; so, all planets can't use the same nova as a reference point. So each planetary pair will need to choose a correlator that may not work for other planets making it much more complicated than just UTC-7. Feb 15 at 14:43
• Yes, your UTC system may all be synced to "42492 megapulses of pulsar P45 after Nova N563-b", but every planet in the galaxy will be a different distance to N563-b. So, the timezone located at UTC-1212.534 may not have ever actually experienced the N563-b Nova . Feb 15 at 14:46
• Timezones have very defined spatial systems, that is one of their main points. Feb 15 at 14:49
• Like dendrochronology can be used to synchronize events that happened in disparate parts of the world in our historical record, based on climactic events that spanned the globe to within a year, then major astronomical events in the historical record, as witnessed by a civilization that has unimaginably accurate and refined resolution sensing technology, can be used to synch galaxy-wide events to the second. Like our calendar is synched everywhere on earth to a supposed event in human history, the passing of time can be referenced to a particular cosmological event. Feb 17 at 23:21

## Galactic Coordinate Time

Or TCG (from the French “Temps-coodonnée galactique”), which you will have to define. No, not GTC: that would be Galactic Coordinated Time, and “coordinated” time does not make much sense in this context.

You want a coordinate time instead. This is a relativistic notion. The basic idea is that you define a coordinate system where each event (each point in space-time) is identified by four coordinates: (t, x, y, z). The first coordinate (t) is the time scale you are looking for. In the context of special relativity, this is pretty easy to get: take a notional observer located at the center of mass of our galaxy, who does not rotate relative to distant quasars, and use his frame of reference to build your coordinate system. In general relativity it is a bit more involved, because you have the time dilation due to the galaxy's gravity well, and the frame dragging due to its rotation. You then have to instead use a notional observer co-moving with our galaxy, but outside it's gravity well. This is exactly how Barycentric Coordinate Time has been defined in our solar system, within a fully-relativistic framework. The same thing can be done at the scale of the galaxy.

Note that this time scale will differ from the proper time of any individual star system: this is the cost of having something “universal”. However, any civilization advanced enough to do interstellar travel will easily have the know-how needed to convert between TCG and their local time scales.

# UTC

Step 1, convince everyone that using good old earth UTC is the way to go (this is the hard part. Then again, tradition dies hard..).

Then send messages to everyone "it is now precisely 18:59:55 UTC on earth, and one second is defined by ΔνCs = 9192631770Hz on earth".

Everyone that receives this message then first adds their distance from earth in lightseconds to that time and sets it to their clock, and if they want to be precise they also set the time dilation setting on their clock based on their relative speed to earth and depth of their gravity well.

Done.

Now, you will have some relativity weirdness, like some fast observer may claim that time X on Earth happend before and/or after time X on Sirius or something.

Just ignore those.

You will find that events theoretically effected by this never influence each other. Any seeming paradox will resolve itself once messages (or messengers) about an event have actually been exchanged.

Anyways most things in the milky way are not fast enough for any of this to matter much anyway.

Maybe your citizens even want to rather go with leap seconds than use different length seconds in different locations.

• Somebody gets to pick, +1. Everywhere else is too far away to matter because you'll never meet them anyway unless you have FTL. And if you have FTL and you can't tell time... then something went wrong there. Feb 16 at 7:03
• Please, do yourself a favor and use TAI, not UTC. Leap seconds cause way too much trouble and their raison d'être (sync up UTC with mean solar time on Earth) isn't even relevant other than on Earth. So, you'd just invite trouble caused by a solution to a problem that you don't even have. In fact, UTC is set to abandon leap seconds within the next 10–15 years or so for precisely that reason. Feb 17 at 22:03
• The Milky Way is some 87,000 light years across. That is a long interval to wait for the 'time click' tone. Feb 18 at 17:57
• @JustinThymetheSecond Not actually, the question asked for science-based, so I assume no FTL. No FTL means the time signal will arrive even before the settlers get there (and I assume it will be refreshed periodically). Feb 19 at 8:53
• The term 'science-based' no longer applies to FTL communications when one talks about settings in the near future, The only significant question remaining in science is the potential band width Feb 19 at 13:58

Tl/Dr: We make corrections as we always do. They are relativistic corrections, but they're just corrections. Barycentric Coordinate Time (TCB) and Geocentric Coordinate Time (TCG) demonstrate precisely how we get around the nightmare of relativity

We have fought this battle before, on Earth. So we can see some of the important parts you need already built out. Unfortunately, some of it will require sending atomic clocks around, or at least information sent by radio waves.

This is the journey of atomic time. Time is simple, until you have to deal with timezones... or you get down into the peculiar worlds where physics gets tough. This is where the International Astronomical Union (IAU) lives. And what they deal with is truly mind bending.

And I am sorry in advance for how long this answer is. The raw amount of detail behind our current equivalents for GCT are astonishing. There's good reasons we do the things we do... we just do a lot of things. The first half of this answer is the very generalized "what is a time system anyway, and what are they for?" That paves the way for decisions we've made with time systems in use today, and can be used to project towards a fictional GCT based on how the system is being used. And honestly, this information is spread across so many documents that having it in one place is of some use. I've broken it up into 3 sections: background, history, and solutions to your problem

## Background

What makes a good time system? In fact, what makes a time system at all? The Time Ontology in OWL captures a great deal of what we care about. It builds off of TemporalEntities, the Instant and the Interval. These are the instantaneous events and extended events that we might be interested in speaking to. From the perspective of a single observer, these Instants and Intervals have a well understood ordering.

Why start so abstract? Well, in the end, time systems exist for a purpose: to assign mathematical values to things, and that can be pretty darn generic. The Aztecs did not need our precise mathematical definition of time, but they might care whether their ritual events (an Interval that might last a day or days) was OverlappedBy a solar eclipse. As it turns out, the latest studies show that their calendar wasn't precise enough for predictions, but they did care about the order of events enough to write things down. It is these events that can be measured which matter.

To quantify these, we add two additional concepts: the time point and the duration. These are used to quantify Instants and Intervals and their relationships. Time points are associated to where Instants occur, and durations are associated with the extended span of Intervals. A duration is the span of time between two time points.

And this is where it starts to get a little quirky. As it turns out, while its easy to define a duration as the time span between two time points, practically speaking durations are easier to measure. If I'm judging a pinewood derby, I rarely know that a race started at 2024-02-18T13:45:13.125Z and ended at 2024-02-18T12:45:21.439Z (putting the cart before the horse and using ISO 8601 date-time stamp notation). That kind of precision is hard for timestamps. However, with nothing more than a stopwatch I can measure the duration associated with the race Interval is 8.314 seconds.

Why is this? Durations are repeatable. If one has an experiment which reliably takes the same duration every time, we can use it as a clock to determine how long other things take. However, a time point only happens once. We can't do any repeated measurements to quantify its time point (we can do parallel measurements, but nothing after the fact).

As a result, nearly every time system ever defined in human history first defines a duration. Whether it's the duration of the day or the duration it takes a standardized incense stick to burn, durations come first. We then add an epoch to define time points. An epoch is an Instant with a pre-defined time point associated with it. Once one has an epoch, one can calculate the time point associated with any instant as long as one can identify a series of adjacent intervals and their durations. All practical time systems invented by humans take the form of a defined duration (day, second, incense-stick, etc.) and an arbitrarily chosen epoch time point. The duration can be re-measured and re-defined, but the epoch cannot be re-measured.

That last bit is important, because it opens the door for a key property of time systems. We can often define the relationship between two time systems. With such a definition, we can compute a time point in one system using another. If we know the relationship between Mountain Standard Time and Eastern Time, we can compute that 6:00 PM MST is equivalent to 8:00 ET (they are associated with the same set of Instants). You will see we leverage this property heavily in our time systems, and it will be very important for dealing with relativity for GCT.

So that's the introduction. What a long document this is turning out to be! Now let's get into the actual time systems we've used in the past.

## History

We can skip over an astonishing array of local time calendar systems. The old systems are quite interesting in that they are practical. Time was measured to support the time operations that were needed. For example, in many cases time was measured differently at night because the primary thing needing timing support was shift changes.

This all changed with the invention of the chronometer. For the first time, a clock could be used to accurately determine longitude for ships at sea. But the clocks were not as immaculately precise as our current atomic ones. If left free-cycling, they'd soon drift uselessly out of date. So they were tied to observations of something rigid and unyielding: the rotation of the Earth. The first universal time system, Greenwich Mean Time(GMT), was built to this need. GMT defined an epoch of "noon at the Royal Observatory in Greenwich" and all clocks were set from it.

And I already lied. It actually wasn't set to noon. Noon drifts over the course of the year due to effects like the elliptical orbit of the Earth around the sun. GMT is Greenwich mean time because its associated with the average(mean) noon over the year, and was actually measured from studying the angular distance between the moon and certain stars at night. So we see a common pattern that will emerge over and over: using mathematics to compute a time system whose events are not exactly experienced anywhere.

GMT caused some confusion. Astronomers started their day at noon, which means a nightly set of observations all occurred in the same "day." Most people started their day at midnight, so their daily activities all occurred in the same "day." This lead to our first time system with "universal" in the name, Universal Time (UT). This was really just a disambiguation. Whether GMT started at noon or midnight required context. UT always started at midnight. And, since we're about to undergo a journey through universal times, we'll skip ahead and call it UT0. The IAU accepted this standard in 1884. Please note just how recent this is. In the history of human kind, precise time is a rather recent novelty. And note that people will always choose an epoch that is convenient to them. Even to this day, astronomers think of the day as starting at noon. (You'll see tables tabulated in Julan Dates(JD) that are all ending in 0.5, because JD is an astronomical concept built on noon, but all of the metrologists tabulated their data for days built on midnight)

Now time starts to get squirely. It turns out the planet doesn't spin straight. It has a wobble called polar motion. As a result, it turns out impossible to measure the duration of previous days using the current day as a "meter stick." So in the spirit of adding more mathematical corrections to make things more precise, in 1956 the International Time Beaudreau (BIH) adopted UT1 and UT2. UT1 included polar motion correction terms to once again describe time with respect to an abstract mathematical place where the "observations" were made. UT2 included seasonal variations, intended for civil consumption, but it wasn't as big of a deal because UTC soon came forth.

Coordinated Universal Time (UTC) arrived in 1963, along with International Atomic Time (TAI) at the onset of the atomic clock era. Atomic clocks had two huge advantages over Earth based measurements. First was that you could observe the TAI second using your own laboratory hardware, independent of the Earth's motion. The second was that as we got better duration measurements, we discovered the Earth's motion was astonishingly more complex than we initially thought. Started in 1955 (with the timescale named $$T_m$$), TAI measured time using the weighted average of atomic clocks.

This is a very important thing for your story. These clocks were not run in isolation. The standard was defined by a set of atomic clocks around the world which communicated using VLF radio signals. They worked together to keep the standard in sync, even though they were at a distance from one another. This quickly proved to be the most accurate time scale humans had created, and became the standard.

And here we have UTC. It got mentioned a few times on answers here. UTC is hilarious because, other than the name Coordinated Universal Time, its actually one of the worst possible "universal" systems to base a GTC around. UTC exists because we want the Earth to be privileged in some cases. We want one day to be one "Earth" day. And so, UTC has an astonishingly complicated history of slews and steps and updates, culminating in the leap second. If all goes well, the leap second will be abolished by 2035, although it will be replaced by an equally quirky leap-minute. And that is all I have to say about the compromise that is UTC here, other than to perhaps point out that even the acronym is compromise. It took several years of debate between English speaking countries, who wanted Coordinated Universal Time to be CUT, and French speaking countries who wanted Temps Universel Coordonné to be (TUC), to arrive at UTC which was equally wrong for both languages and started with U just like UT.

Let's get back to the interesting standards that apply to your problem. But first, I wish to revisit the problem of epochs. How does one pick an epoch? If one picks it arbitrarily, there's no way to convert between two systems. So typically we pick an epoch from the previous time system, and declare them equal. TAI, under the name $$T_m$$, was started with an epoch of 1958-01-09T00:00:00.0 which coincides with that date in UT2. There's no way to recover that date if it were lost (i.e. if all of the clocks shut down at the same time).

So how about them correction terms? In the 1970s, we noticed that all of the atomic clocks weren't operating at quite the same speed. They were at different altitudes, and thus experienced relatavistic differences. In 1977, TAI began accounting for this, making yet another abstract mathematical correction. It adjusted such that all clocks would be measured as if they were at mean sea level on the Earth. And this brings us to the start of solutions to your problems.

## Solutions

At the time we began using relativistic corrections, we introduced three more time systems, Barycentric Coordinate Time (TCB), Geocentric Coordinate Time (TCG), and Terrestrial Time (TT). TCB, TCG, and TT are all relativistic corrected time systems. TCG measures time as it would be experienced by a point co-moving with the Earth, but not rotating and arbitrarily far from the Earth as to not be affected by our gravity well. TCB does the same thing, except instead of being tied to a point co-moving with the Earth, it co-moves with the barycenter (center of mass) of the solar system.

These show you the essential first step to your GTC. This is how our metrologists have defined a time system that has relativistic corrections put in place. Now per relativity, there is no truly definitive time system, but we can pick arbitrarily large systems and apply the same pattern. GTC would have a correction factor such that it measures time as an observer arbitrarily far from the galaxy but co-moving with its barycenter would experience. Using the barycenter like this does make a privileged point, but it has two advantages over other points. First off, it's fair. It really is the most democratic solution. The second appears when one starts to define reference systems to measure space. The barycenter has some unique properties that make it immune to really hard to work with relativistic effects like frame dragging. If you define time and space at the barycenter of the massive things you care about, the relativistic effects become manageable.

Note that all of these have been completely dependent on not only mathematical corrections, but on a consensus. BIPM asserts that no one measurement device can truly measure the current time in TAI, TCB, etc. It is defined as a consensus, after the fact. And, as such, the true "TAI time" of an event can only be known for events sufficiently far in the past as to have a published consensus about that time. You can approximate it really well, but it's not official until the official atomic clocks around the world have been compared and the metrologists have had time to analyze the details. The galaxy is 100,000 light years wide. This means any such consensus based time could only define exact time points for events 100,000 years ago. One would have to rely on a local time system and an approximate conversion. This isn't new. If you've ever set your watch to a site like time.gov, you've done this... just on a much smaller time scale. A question for you would be what time scale does your galactic society work on? Remember that they do need to coordinate to be a society, and that is unlikely to happen on any faster timescale, unless you add faster than light travel/communication to your world.

If you made it this far, and compared notes, you might notice I left one time system completely undefined TT. I left it to the end to highlight a fundamental limit of your GTC which cannot be overcome using current physics. TT is Terrestrial Time, and it realized via TAI: TT = TAI + 32.184s, as both TT and TAI are measured with respect to the Earth's Geoid. Why the 32.184s? History. TT was designed to read the same as its predecessor, Terrestrial Dynamic Time (TDT), at its epoch. In turn, TDT was based on Ephemeris Time (ET), a pre-atomic timescale designed to avoid the imprecision of the solar second of UT1. It defined the second as a fraction of the tropical year 1900 (356 * 24 * 60 * 60 seconds in a year). By the time TAI was defined w.r.t UT2, ET and UT2 had drifted by over 32.18 seconds. By the time ET was revised to use atomic seconds, the current difference of 32.184 was set in stone.

This shows something important for your GTC. There's no way to re-measure an epoch for a time system. They always daisy chain from one to the next. If you needed some "calendar event" to synchronize everyone in the galaxy, you will find it hard to come by. Better to have your GTC grow off of one civilization's most accurate time system, share an epoch, and then start ticking from that point on in a galactic way. Later, compare notes just like they do with TAI.

And do remember that this is relativistic time we're dealing with. I do not measure time the way you do. I may even see events occur in a different order than you do. However, by applying corrections, we can both determine what an abstract outside observer would see (comoving with the galactic barycenter), and we can both take the other's measurements and convert the back into our local time systems.

• 'An epoch can not be recreated'. But it can be re-experienced. Consider if the 'epoch' was the actual 'Big Bang'. yes, it happened only once, and will not happen again. But we can determine WHEN it happened by observing the on-going results of it. We view the light currently arriving from it, and by using the speed of light we can back-track as to when it occurred. Then we can re-set our epoch based on this calculation. That is, the actual time the Big Bang occurred will not be lost. It can be observed and re-calculated. Feb 18 at 18:21
• @JustinThymetheSecond You have stumbled on a layer of complexity I only alluded to once in my entire answer: time systems versus their realizations. You can define a time system with an epoch of "the big bang," which then must be realized, putting numbers in where the concepts go. I mention TT is realized by TAI. TAI is the weighted average of clocks. TT is a continuation of ET using SI seconds -- SI seconds being an idealized concept. I rounded the offset to 32.184 because that's traditional. The IAU actually updates this based on our latest numbers yearly. Feb 18 at 20:44
• BIPM realizes TT with TT(BIPM22)=TAI + 32.1840277s. And this shows why measuring time from the big bang as an epoch is a problem. One can make time system that is "SI seconds since the big bang," but the realizations of it have such great uncertainty that future realizations are likely to be wildly different. BIPM is realizing TT using TAI to a factor of 10 picoseconds. The time since the big bang is estimated to within millions of years. Feb 18 at 20:48
• Th Big Bang can not be observed directly, but for instance if we can differentiate a unique event 'spike' that it caused and happened after it, we can back-track to the epoch. Then, if we 'lose' it, detect another event that came after it but was caused by it, and back-track again. The trick is, without FTL communication, how do we communicate this 'mark' event to all of those in the galaxy outbound of us? Our 'this is the mark event' would come to them AFTER the event., and certainly well after it was seen at the leading edge of the galaxy. 100,000 years later, across the galaxy, in fact. Feb 18 at 21:04
• @JustinThymetheSecond That's more of the typical consensus problem for time -- well solved. One just needs to identify a class of events that can be "spikes," and have everyone record when these spikes occur in their local time system. They then can develop a consensus as to which one should be used as an epoch. As long as your recordkeeping is good, you can eventually converge on an agreed upon system. Feb 18 at 21:08

Frame Challenge

Unless one can answer the question 'What is meant by 'the same time at two different places?', this question can not be answered. It has no criteria by which to answer it.

Let me illustrate.

Everything that follows is to be assumed to be in the same relativistic frame of reference, the same relativistic perspective, of the 'pulsar' (as my 'concession' to Relativists).

Let us say that there is a pulsar that gives out waves of massless particles at a given pulsing frequency. This wave, of course, propagates at c.

As this wave expands, it forms an ever larger expanding concentric sphere with every point around this wave equidistant from the center.

We can reasonably state that every point on this sphere experiences the same time 'click' at the same 'time'. That is, the same event (the passing of the wave) happens simultaneously around the sphere. Further, as the sphere expands, every point on the new expanded sphere experiences the same passage of time from when the sphere was in its former position. The phrase 'Hey, the wave was X distance away Y seconds ago' could be spoken at the same time anywhere on the sphere, and x and y would be the same value at every point around the sphere. It would seem to be a measure of the passage of time that the OP is asking for. Everyone around the circle sets their clock at exactly the same time - the passage of this wave, and subsequent waves define a standard interval of time around the sphere. Recall that everything happens in the same relativistic time frame.

c, of course, by definition, is a constant. Distance over time. The same distance over the same time gives the same value.

Now suppose that this wave front goes through some 'sticky substance' at various areas around the wave front that alters the progress of this wave. Some aspect of this wave reaches a given circumference at a different 'point in time' than another part of the wave. The wave front is irregularly shaped, not a perfect sphere. This 'sticky substance' has either altered time, or altered distance, or altered the value of c. Since it is the same relativistic framework, both time and distance are unaltered. The value c itself must have changed, in order to change the velocity of the wave.

So let us consider that it alters c .

Here is the question. How do you define 'the same time' around this sphere? Is it that every point on that circle has the same time, regardless if the wave front has reached it or not, or is the 'same time' a point on the wave front itself, regardless if it has reached the concentric circle? Is it 'the same time' at various places around the sphere at the time when the wave actually passes the sphere? And what about the interval of time? How do we measure it? The time for this wave to go a specified distance, even as it goes through this 'sticky substance', or the time it takes for subsequent waves to cross this given reference circumference? On what exactly do we base our Universal Galactic Time Code?

So, what is this 'sticky substance'? Can it actually change c? Is this all balderdash? Or does it exist?

Let us check the recent scientific literature.

Two papers, published in the European Physics Journal D in March, attempt to derive the speed of light from the quantum properties of space itself. Both propose somewhat different mechanisms, but the idea is that the speed of light might change as one alters assumptions about how elementary particles interact with radiation. Both treat space as something that isn't empty, but a great big soup of virtual particles that wink in and out of existence in tiny fractions of a second.

And further

The charges of all these particles are important to their model, because all of them have charges. A quantity called impedance depends on the sum of those charges. The impedance in turn depends on the permittivity of the vacuum, or how much it resists electric fields, as well as its permeability, or how well it supports magnetic fields. Light waves are made up of both an electric and magnetic wave, so changing those quantities (permittivity and permeability) will change the measured speed of light.

In other words, it is the stuff that forms 'empty space' is this 'sticky stuff'. Theoretically, it exists.

If, indeed, this expanding waveform traveling at c is not perfectly concentric, and that indeed an event (the passing of this wave front) can happen at a different 'time' at various points around this equidistant circle from the wave origin, then exactly what does it mean to have a 'Universal Galactic Time Code'? And of what practical value would it have? How could it be used to co-ordinate the passage of time in the galaxy? What exactly do we want to measure?

Second Frame Challenge

One of my favorite sci-fi authors has been writing a space opera 'war' series over maybe two decades. I have every book in the all of the series. This author is particularly stringent on sticking to science-based principles, and his stories have evolved over this period to reflect his deeper understanding of the principles involved. One of those concepts involves how to target the 'enemy' spaceship when traveling at velocities of 0.2 c relative one ship to the other. (Yes, he has wrestled over the series, and the decades, with how to handle "Helmsman, increase speed to 0.2 c", and what that velocity is relative to.) He has limited all space engagements between opposing ships to 0.2 c relative velocity because he understands that beyond this velocity, even the most advanced computer would have extreme difficulty targeting, even with energized beam weapons. Throughout the series, he makes evolving reference to issues involving 'How long for weapons to impact with the enemy'. Following his newest book, this understanding is still evolving.

So here is the frame challenge, which I hope will further refine the criteria for evaluating answers, and even the question itself.

Suppose we have two ships traveling at different velocities relative to a 'standard' framework given by three pulsars, triangulated. The velocity of each spaceship can be standardized based on its displacement over time through space relative to these triangulated pulsars, and thus a standard velocity relative to these pulsars can be determined for each ship. How this is computed is not relevant, only that the two ships can within their navigation systems determine that each is going a different velocity to the other, relative to this standard frame of reference. One is going 0.15 c and the other is going 0.25 c, relative to this standard. It is not part of this discussion to attempt to actually calculate their velocity relative to each other.

The ships are approaching each other at some angle (not head on), in a collision trajectory (or they fire a torpedo at each other, or whatever).

The weapons officer of one ship says 'On the mark, collision with enemy ship in 2 hours 3 minutes and 25.42 seconds, or X seconds UGT (Universal Galactic Time) from NOW.'

The weapons specialist on the other ship says, at exactly the same moment, 'Collison with the enemy ship in y seconds, or z seconds UGT on my mark NOW.' Coincidentally, the mark NOW happens simultaneously, to the hundredth of a second (the MARK is very precise).

The frame challenge questions (based on the assumption that the passage of time is relative to the frame of reference of each ship, and each ship is going at a different velocity relative to a standard point of reference) are:

One, is the value of y equal to 2 hours 3 minutes and 25.42 seconds?

Two, is the UGT interval x equal to 2 hours 3 minutes and 25.42 seconds?

Three, what is the relationship between x and y, the expected passing of an UGT interval on each ship? If there could be any such thing as an UGT, would x and y not have to be the same? Otherwise, what is 'universal' about it?

As mentioned in some answers, a pulsar gives a very accurately measurable rate of the passage of time that everyone can agree on. As mentioned in other answers, relativistic effects are relevant on a galactic scale. For example, the rate at which you observe a pulsar pulsing depends on your velocity relative to the pulsar.

First of all, let's observe that Galactic Time cannot pass at the same rate for everyone in the galaxy. Due to relativity, someone flying around at high speed will come back to the same point with less time having passed for them than for someone that stayed there. We want a Galactic Time that still reports the same time when they meet up again, so it must have passed for them at different rates. Therefor, you wouldn't use Galactic Time to plan your day, you would use local time for that. Instead, Galactic Time is primarily useful for planning meetings.

For Galactic Time, you'd pick a single reference pulsar. Assuming the reference pulsar is visible at all times during your journey, if you meet up with someone, go on a long journey through the galaxy separately and later meet up again in an entirely different place, you will still have counted the exact same number of pulses. Relativity can't influence that. You might have experienced different amounts of local time, but you'll have seen the same amount of Galactic Time pass.

Even if the reference pulsar isn't visible, a computer can make a reasonable estimate of its pulses based on local time and your relative trajectory. You'll also need a computer to estimate the duration of the journey to your meeting point in Galactic Time, which is a different computation from the duration in local time, but computable all the same.

As for accuracy, as long as the pulsar is visible, I'd say it's very high, much better than 1s. When it's not, it depends on how capable your computer is in making its estimation. If a reference pulsar in another galaxy is chosen, it might be visible from almost anywhere in our galaxy.

• Galactic Time could still be the preferred measurement for most purposes since time dilation would generally be way less than 1% unless you are on a newtonian relativistic spaceship. Also, if you setting uses an alcubierre warp drive, spacetime remains more-or-less flat inside of your warp bubble meaning that local time will not significantly change no matter how fast the bubble moves. Feb 15 at 16:46
• My understanding is that pulsars change (spin down) on a rapid enough time-scale that an observer on one side of the galaxy would see it rotating at a different rate than an observer on the other. I suppose you could pick a pulsar as close to the galactic core as possible, but even then the 50k years would likely matter to an observer on the galactic rim compared to one that was relatively nearby. Feb 15 at 18:18
• @SoronelHaetir Alternatively, use the spindown as a second, long term time aquisition method. Furthermore, just use multiple pulsars, and their relative and absolute ticks to calculate not just current time, but also position and speed, giving you simply Unified Galactic Coordinates. Feb 18 at 0:40

The solution is fairly mundane and simple.

First, assume that relativity is a bust, and has no significance to the real world. It was a red herring, with no importance. The theory just has too many loopholes and inconsistencies that are now so incredibly patched over to make it 'work', to conform to current evidence. Just like the early models of the solar system that had everything rotating around the Earth became so ungainly in their complexity that they had to be abandoned and replaced by a better model, not just 'updated' and 'evolved'. "Jim, it's dead. Let it go".

Second, assume Quantum Physics rules the Universe at even the cosmic scale, which there is an abundance of evidence to support. Photosynthesis, for instance. The knowledge imparted by the theories grows exponentially, unconstrained by the blather of Relativists, and finally physicists see the light. They actually now understand the truth, not myth.

Third, consider that there is a 'quantum state equation' that overs the entire Universe in one huge matrix equation. One all-encompassing universal field that blankets and defines the universe. No warping, bending, transforming, multidimensionality, or other crap that was needed to maintain the integrity (theology) of Relativity.

With these three assumptions, all of which are perfectly valid and perfectly supported by evidence, then one can posit that instantaneous communication everywhere in the Universe is a given. That there is one universal field that is so absolutely 'stiff' that any displacement in that field is immediately transmitted throughout the field, without the necessity for inertia or compression waves, or time dependent 'cause and effect' ripples to propagate through time.. An 'entanglement', if you will, of everything.

So now it is a simple matter of positing a regular oscillation in that field that can be picked up by instruments, like a clock beat, and used for a universal clock.

• For what it's worth, quantum entanglement does seem to propagate at 10,000x the speed of light in a vacuum. Also this would mean Galilean transforms work at interstellae or even intergalactic distances. Feb 18 at 18:18
• 'Does seem to propagate'. The devil is in that 'seem to' qualifier. How do you establish a 'mark' to determine the time interval? The conundrum arises out of 'retrocausality'. Is there a delta-t between the advanced and the retarded time interval? Feb 18 at 18:28
• @The Square-Cube Law "Notice that this result does not eliminate the possibility that the influence of entanglement actually is instantaneous – it merely sets a limit saying how close the influence must be to infinitely fast." Quote from the article. Feb 18 at 19:03
• @The Square-Cube Law Personally, I have absolutely no doubt that 'locality' will soon be thrown out the window. It just can't survive the evidence that is mounting up against it. Feb 18 at 20:56
• Just throwing known physics out of the window in a question that asks for science based answers does not seem like a good answer to me, sorry. Feb 19 at 17:17