One day, alien pranksters decide to stop all the clocks on Earth at the same time. Watches, computers, atomic clocks, microwave ovens, etc. At the same time, they scramble the memories of people so some people think it’s 2013, some think it’s 1998 and all other dates evenly distributed, and 1/7 of people think it’s each day of the week.
How can we objectively determine the correct time and date? I’m thinking year and season should be easy, time to the hour can be done, can we get back to the exact time we were on before?
We know the exact moment of eclipses and with some error for atmospherics, we know the times of sunset/rise for any date and place. I think we could certainly get back to the second.
Dates and times are simply conventions of society.
We call this year 2013 because somebody long ago decided year such and such was year 0. Somebody (maybe the same person, maybe somebody else) said each year will start on day so and so. The time at which each day starts was born of a similar process.
I am compelled to note that 2013 is not relative to any “year 0.” The year 1 in our calendar is preceded by the year 1 Before (Common Era or nominal Christ).
As long as the written records of these conventions persist (just the living memories are scrambled), then we can reboot the clocks with a high degree of accuracy.
I’m confident that they definitely can do this. I don’t know the calculations personally, but here’s what I do know:
We all know that centuries ago, astronomy advanced to the point where some slippage in the calendar was noticed, and they decided to fix it by inserting an extra day between February and March, and to do this on a fairly regular basis, 99 times in every 400 years. This is known as a leap day.
What many people don’t realize is that in recent years, they’ve noticed some additional slippage, of a much smaller and finer nature, and they fix it by adding another second every now and then. This is known as a leap second and has been added to the calendar 25 times since 1972.
It seems logical to me, that the question of whether or not to add a leap second is essential identical to the OP. In both cases, they look at the stars and whatever else, and ask themselves if the clocks need to be adjusted.
we can probably get reasonable precision by looking at written logs, such as ship’s logs or legislature journals.
We can get within a year of the correct date by visiting a cemetery in a large-enough city.
We can determine the month by finding an advanced sundial such as this one and monitoring it for a few days.
We can get within a day of the correct date by monitoring sunrise and sunset for a few months (until the next equinox or solstice).
To those who believe they are in 1999, ask them where the World Trace Center towers are and why there are no Concordes on the tarmac at any airports. Now, of course, if the aliens can make us forget historical events, it’s more complicated…
Ask yourself how we set our current clocks. How do we know when to add a leap second? The answer is that we track the stars with better precision than our clocks are set to. A transit telescope can measure time to a very small fraction of a second. A set of known stars passes directly overhead. We know the location of these stars to astounding accuracy, and measure their transit to essentially the limit of the resolution of the telescope. This allows us to know the time to sufficient precision that we are now starting to argue about whether the Earth’s rotation period is the right definition of the length of a day. The pivotal definition of time was based upon Airy’s transit telescope at Greenwich.
We have an enormous amount of data under our current system. Without knowing how to tie the post-event time into the pre-event time, that data has much less value. Hence the OP is a good question.
Assuming astronomers have their current capabilities, I’d guess the error could be reduced well under a second, to milliseconds or maybe much better. We have to know time very accurately to point telescopes, and we have knowledge of many pulsar’s periods. Pulsar periods usually vary smoothly, although they “slip” now and then, so whether we’d be able to use a given pulsar would depend on whether it had slipped or not between the pre- and post-event measurements. The longer time span, the fewer pulsars available to use. With only one or two pulsars, their relative periods would repeat with some longer period. With more pulsars, their total period would get longer, allowing you to synch them to other, coarser time estimates.
The OP specified “alien pranksters decide to stop all the clocks on Earth at the same time.” …bolding mine.
So…couldn’t we just call up to the International Space Station and ask them what time it is?
Similarily, with all of our own extra-terrestrial satellites and probes, we could resynchronize our date and time.
Thinking about things a bit more - there have been a few Sci-Fi stories, usually involving a seriously long relativistic journey returning to the solar system, and having to work out when it is they have returned. The answer being, as standingwave points out, to look at the positions of the planets and deduce the date. Not as trivial as all that, and given enough (geological) time, possibly sufficiently chaotic as not to be viable. But within reasonable constraints the position of the planets provides a unique and reasonably clear indication of the date.
The GPS satellites all contain atomic clocks. So, even if we ignore Sparky812’s precise reading of the OP, and allow them to be wiped too, they will still be cheerfully orbiting in the same orbits as before. Thus we re-establish communication, reboot them, and use the existing earthbound infrastructure to work out what orbits they are in. (We have to do this normally, as whenever a new one in inserted into orbit, we can’t loft it with the precision needed, we have to measure it once up there, and work out its ephemeris. Clearly new ones can now use the existing constellation, but there has to be a way of starting from scratch.) We should be able to iteratively add satellites into the constellation, and as we do, we can re-establish the time required by the on-board clocks, and eventually rebuild the constellation into a fully working system. At this point we will know the time to within less than a nanosecond relative to the time prior to our alien friend’s interruption. That is better than a transit telescope, but not enough better that (almost) anyone would actually care.
Easier than that, just turn on your GPS. The GPS satellites will have the correct time down to the millisecond for a few days before their clocks start to drift.
Position of planets, sun, and moon should easily get you down the day.
With a little more work using the moon or postions of asteroids you should be able to get down to hours and with more work minutes or even a few seconds.
Once you are down to seconds you should be able to piggyback of off slower pulsars then use those to work you way down to millisecond pulsars.
And these pulsars obviously have periods of milliseconds, so my WAG is you could get down to about 1/100 of a millisecond (or perhaps much better).
So, we would finally know what time it was back down to about .000001
Not sure there is anything in astronomy you could use to get back down to knowing what atomic clock " second" it was.
I’m kind of curious about the extent of this memory-scrambling. If we were really starting from scratch, for example, would we return to a 24-hour day, 60-minute hour and 60-second minute?
If we had a huge infrastructure of non-functioning clocks and watches, could we end up with a new 12-hour day (assuming we forget that the clock is supposed to go around twice) with the hours twice as long as previously?
I thought we measured leap seconds not instantly, but over the course of several years. Not, “hey, the time now is 0.001 seconds slower than last year”, “hey, the time now is 10 seconds slower than 400 years ago”. Can we measure astronomical objects with such precision? Something like Gravity Probe B?
Ok great. That scenario was just a way to ask: what is our most precise method of measuring time that isn’t specifically for measuring time?