How Far Back (or Forward) in Time Can I Use Stars to Determine the Date?

Quasi-inspired by this thread, let’s say I have a TARDIS I bought off eBay from a sketchy-looking Sontaran, and I have discovered that the navigation and sensing is hosed. So when I step out into a tropical jungle, I can’t be sure if I’m in the Cretaceous or the Jardim Botânico c. 2016. Sure, I could wait to see if the life forms that wander by look more like a Noasaurus or a drunk gold medalist, but what I’d rather do is check the stars.

So is that reasonable? Could I determine from star and visible planetary positions what year I’m in? What month? What day?

Hmm… I would think that with sufficient reference data and precision observations, you could determine quite a lot over a wide swathe of time.

The planets in the solar system behave a little like clock hands, each with its own period. Eventually, if you go far enough back or forward, those periods will change however, as the planets drift in and out. And eventually you’d come to a point where all the hands come back to a similar point, or close enough given the margins of error.

All the stars in the sky also have “proper motion” ; as the stars move through space, they seem to move across the sky very slowly. Some are extremely slow, some not quite so slow. Comparing a few of these would help place your era in longer scales.

I hope that this helps.

Even assuming that the TARDIS has good stellar records exceeding what we know (for example, locations of stars that have since burned out/exploded and are no longer visible, stars that have coalesced from current nebulas or gas), I’d have to say no, not unless you got lucky and there was a predictable event (comet, eclipse, supernova, etc.) visible. At least not with the resolution of human eyesight.

Given instruments that could tell you relatively exact positions of stellar objects and a computer to do the math/database work, I’d think, yes. Stars are very long lived objects, and they move in more or less perfectly predictable ways, at least as a population. That should be good enough to get you down to a period of no more than a couple centuries, at which point the positions of the planets would get you an exact year, and the position and phase of the moon (I assume you’ve got imaging that can give you your exact LOCATION on the earth, and can wait until night to see those stars/moon) would narrow it down as much as your imaging can take. Leave a few reflectors on the moon before you visit, and it wouldn’t surprise me if you could determine your time to the second.

Yes and no. It helps in confirming that my basic understanding is accurate, but it doesn’t add any new information. My question is: what, in numbers, is a “wide swath” of time? Armed with a telescope from Sharper Image, could I successfully tell I was in the Cretaceous period as opposed to the Jurassic Period? Could I figure out how long I had before the Cretaceous-Paleogene (K-Pg) extinction event came upon me? (Makes a difference in deciding to buy or rent). Could I narrow in down to the nearest million years? The nearest hundred thousand? The nearest week?

Or would the best estimate simple be “a really long time ago?”

Interesting how we approach the problem with very different assumptions, TimeWinder! :wink:

First of all: awesome user name/subject combo.

Secondly: yes, I can load stuff up here and now before my vacation starts. So I can buy a laptop, a good portable telescope, and a sextant if they would help.

My question is: do we, humans, have such data somewhere? Can I download it? Is there an app?

Cretaceous period started around 145 million years ago. Our solar system would have been on the other side of the Milky Way galaxy at that time. Some of the stars we see in our sky may still be visible, but there’d be no way to tell which is which. I think all you’d know is that you are at least many millions of years away from the present day.

Other galaxies might still be recognizable though. So MAYBE you could tell where you are in the Milky Way galaxy, and based on that, calculate how much the solar system has moved around it.

p.s. It may be better to measure the Moon’s orbital period. The Moon is moving away at a fairly steady speed. I think you can take a very accurate measurement with a simple transit.

You’d want this software

First of all, was that a Heinlein reference? My copy of Tunnel in the Sky is on loan right now.

Second, what resources do you have available? Extensive reference books? An appropriately-programmed computer? A telescope, and if so, what size? Can you get above the atmosphere, or take non-visible observations? How long are you willing to observe?

In the best case, start by getting above the atmosphere and measure the temperature of the cosmic microwave background. This will get you to within a precision of about a million years all by itself. No, that’s not very precise, but it’s an important start, because it’s not periodic at all: you can’t possibly be off by more than that approximately million years, from being in a different cycle than you think. It’ll also work at any point in time in the Universe’s history, until the CMB redshifts so far that it’s not detectable at all.

Now we need to start looking at periodic phenomena. Ideally, we’d find something with a period not much longer than the timespan we’ve just narrowed down to. Unfortunately, the best I can find are the period of the Sun’s orbit around the center of the Galaxy (225 million years), and the period of the Earth’s precession (26,000 years). So, OK, let’s work with that. To measure where the Sun is in its orbital cycle, you’d want to measure the angle between the center of the Andromeda galaxy (visible even to the naked eye) and SagA*, the center of our Galaxy (easily found using any radio telescope at all). You could probably measure this angle to within an arcsecond, if you’re above the atmosphere and have good instruments, but I don’t think that the Sun’s ephemeris is well-enough known to take advantage of that. Rough back-of-the-envelope, it’s probably going to be a bit worse than what we already have from the CMB. So it’ll let us confirm what we already have, but that’s not good enough.

So, on to the precession of the Earth. With a cycle length of 26,000 years, and a million-year window, we’re going to have about 40 possible solutions. But we’re not sunk: All we need is another cycle of comparable but different length from the Earth’s procession, and we can compare the phases of the two cycles. And fortunately, we have one: The perihelion precession of Mercury has a period of about 22,000 years (though it’ll take time to measure it). Note, by the way, that this is not the 43 arcseconds-per-century precession due to general relativity: The total is 5600 per century, with the rest accounted for by Newtonian sources. By comparing the two cycles, we can easily pin ourselves down to one precession-cycle, and to one portion of that cycle.

How small a portion? Well, I’m not sure offhand which one of the two we can measure more precisely, but if you’ve got the time to do the perihelion-of-Mercury measurement, you can certainly measure it to within one part in a few thousand (since the scientists of Einstein’s time were able to do so, to notice the 43 arcseconds-per-century discrepancy). So now we’re down to a window of mere tens of years.

From here, we’re sitting in butter. With a window of tens of years, the orbit of Jupiter, Saturn, Uranus, or Neptune would any one of them be enough to pin it down, certainly to the year, and even more certainly if you had all four of them. And once you have a one-year precision, the familiar cycles of the Earth itself will easily get you to a day, and if you know your longitude, to a minute or second.

Of course, that was all assuming the best-case scenario. Take away some of the advantages I used (above-the-atmosphere, microwave detection, and long observation times), and we’d have to see what else we can come up with.

EDIT: That post took a while to write-- There were no responses when I started.

There seems to be a lot of dispute about this (weird, since it would seem to have a simple answer one way or another), but from what I can tell with GoogleBing, the concensus seems to be that that wouldn’t matter: yes, we’ll have made about 2/3 of a galactic rotation, but everything in our night sky (barring faraway things like galaxies) is rotating with us, and you can kind of “cancel out” the galactic rotation. It might even help a little bit, since we can see a few other galaxies, which now give us a “millions of years” timescale to get started with.

The “no way to tell which is which” is almost certainly incorrect: Stars have near-fingerprint like spectra and magnitudes that vary in (mostly) well understood ways. Simple optical telscopes might have trouble but add some spectroscopic equipment and you’re golden.

Based on these responses, the correct answer seems to be: Take the membership of the SDMB and some extra equipment with you (it’s bigger on the inside), and you won’t have any problems.

I wouldn’t count on recognizing any individual star by spectral fingerprints. We understand some of the ways they vary, but not enough to say anything with any confidence over timespans of millions of years. Many of the brightest stars don’t even have lifespans that long.

SagA* and M31, though, you should be able to recognize at basically any time, especially if you have any sort of radio telescope (and likely even without).

And I hadn’t considered the recession of the Moon. I can find figures for the length of the Moon’s period out to a precision of a millionth of a day, and that period changes by about 23 microseconds per year. I don’t know how hard it was to measure the period to that precision, but assuming that we can match it, that’s enough to pin down our time to a noncyclic 4000 years or so. Not enough all by itself, but that’s still a huge help.

It’s really impossible, if we’re talking millions of years ago.

If we’re talking about within the last ±13,000 years you might be able to get a rough estimate by precession of the equinoxes. The position of the North Pole moves by 50.2 arc seconds per year, so with some good initial data you could measure the position of the circumpolar stars and get a good guess as to what position in the 26,000 year cycle you’re in. So see this article:, and check out this graphic:

But since it takes the Sun about 250 million years to make an orbit of the galaxy, you can see that when you’re millions of years in the past or future the Sun will be in a wholly different region of the galaxy. While it’s true that close stars are traveling more or less on a similar orbital path, they all have different proper motions. To give an example, here’s an article about how one star probably passed within a light year of the Sun only 70,000 years ago. However, since this star was a red dwarf it probably wouldn’t have been visible to the naked eye, even that close. But the point is, our familiar bright stars are going to be scattered all over the place a million years in the past or future.

If you had a whole astronomy department with you they might be able to compare the visible stars in your mystery time period and try to identify them with our current star catalog, and then work out from their current magnitude and their known proper motion how far in the past/future you are. But it wouldn’t be easy. It’s one thing if you’re fairly close to the current time and can identify positively identify Sirius and Alpha Centauri and so on, and do some math. But 100 million years ago Sirius and Alpha Centauri and Procyon were not close stars. So picking them out of the galactic background would be impossible for one guy with a telescope.

You’re going to be much better off trying to establish the time by looking at the plants and animals. And if you can take your TARDIS up to orbit, getting a look at the positions of the continents would really help.

Most stars in a given neighborhood follow similar orbits around the galactic center, but they move relative to each other at speeds of a few tens of kilometers per second (i.e. on the order of a light year per ten thousand years). That’s enough to thoroughly shuffle the stellar neighborhood over multi-million year periods. Most stars currently in the neighborhood would be visible only to a good telescope (and, as you noted, impossible to identify individually unless you’d been maintaining ongoing observations at least every few millenia); the few that are bright enough that moving a few hundred light years wouldn’t affect their visibility very much wouldn’t last over hundred-million-year intervals.

Lemur866, that’s why you don’t measure the precession relative to stars; you measure it relative to the center of the Milky Way.

Do we know how stable the 26,000 year precession cycle has been? Go back 20 million years and would you expect to see the same 26,000 year period?

But it’s awesome.

There are fluctuations (periodic changes and long-term decrease) in the rate of precession, but 20 million years is apparently not long enough to make a big difference. Disclaimer: I have not read the paper, just glanced at their result, but the point is that they have useful results for a 250-million-year period.

Fitting that Dopers with the handles of Chronos and Timewinder would chime in on this thread…

Plus the more luminous ones will probably be wayyy off the main sequence by then, either as supernova–>neutron stars or black holes, or just white dwarfs. Of the bright ones in our current sky, some are close but not very luminous, while some are far away but rather luminous. If we are talking 100 million years, the former may be thousands of light years distant, while the latter no longer exist (as main sequence members).