Timeline for My time machine needs to know when it is
Current License: CC BY-SA 4.0
15 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jul 2, 2018 at 5:31 | comment | added | aroth | And a potential 5th issue - If you go back far enough, your reference points won't have been born yet. But then probably the solar system and the Earth would be in the same boat, which would likely be the bigger problem. | |
| Jul 2, 2018 at 5:26 | comment | added | aroth | Source of error 4 - If one (or more) of your reference stars happened to pass close by to a black hole (or neutron star, or other suitably massive but generally invisible object) and had its trajectory altered millions (or even thousands) of years ago, it would throw off your linear regressions past that point. If this happened to a large enough number of your reference points, it could prevent the algorithm from producing a result at all. Astronomically unlikely to occur, but not physically impossible. | |
| Jun 30, 2018 at 1:47 | comment | added | Mark Olson | Actually, that turns out to be the case only if you do a bad job of choosing the stars you use. The idea is to use a large amount of computation to match 100,000 star positions and magnitudes recorded by the telescope with the known "trajectories" through the sky of the stars from our day. Most of the stars in the then-day sky won't match, but for the right date, a big bunch will. And that ties it down. This is not a technique that can be done "by hand". | |
| Jun 29, 2018 at 22:00 | comment | added | Keith Morrison | This will only work over a surprising limited period of time in the past. Consider Arcturus, 5th brightest star in the sky, magnitude -0.05 at 11.25 parsecs. Given its current motion, in a million years it will be around 127 parsecs away at magnitude 5.33. Now think about how all the other stars are going to move around, so how do you identify which magnitude 5+ star is Arcturus in order to calculate your time? | |
| Jun 29, 2018 at 10:17 | comment | added | ths | "wide angle" is a relative term. the LSST has a 3.5° FOV. that said, it should be possible to get some estimates from simple high-res wide-angle photos with an ordinary camera. fans of the xkcd comic "time" determined location and period of the comic from a night-sky animation. | |
| Jun 29, 2018 at 8:21 | comment | added | mpasko256 | I had very similar idea but instead of posting another answer, I decided to put a comment instead: You don't have to focus oneself on visible light spectrum only. You can perform astronomical observations using ultraviolet, radio frequencies, gamma rays or even elementary particles. Note that, if you choose proper medium, you can decrease detector size and increase measurement quality. | |
| Jun 27, 2018 at 19:28 | comment | added | Mark Olson | @DJClayworth Not at all. For a given detector size (array area) you can trade field of view and magnification pretty much as much as you want. To get higher magnification at the same field of view (or a wider FoV at the same magnification), you need a larger detector -- more pixels. The LSST (en.wikipedia.org/wiki/Large_Synoptic_Survey_Telescope) is a great example. (Probably bigger than you want to include in you time machine, though.) | |
| Jun 27, 2018 at 18:52 | comment | added | DJClayworth | I'm sure "wide angle telescope" is a contradiction in terms. | |
| Jun 27, 2018 at 12:45 | comment | added | Andrew Brēza | Thank you! This is the best answer for a few reasons: 1. I don't have to leave anything behind and find it later. 2. I don't need any more dangerous materials. My wife doesn't like that she already has to wear a radiation badge when she's in the house. | |
| Jun 27, 2018 at 12:43 | vote | accept | Andrew Brēza | ||
| Jun 26, 2018 at 23:39 | comment | added | Makyen | Note that I'm not saying that what you have here is incorrect. It just assumes that the machine is dumped in the middle of a random date/time without further information. That may have been what the OP wanted, but I view such a calibration as a multi-data-point process, which is easier to determine by building upon the data that you already have (fewer unknowns), rather than jumping to some distant point without first hitting some intervening points. | |
| Jun 26, 2018 at 23:32 | comment | added | Makyen | The question says "So far I've used a radio clock hooked up to the onboard computer". This implies that there's, at a minimum, of "short" and "long" settings. For me, I read this as meaning that they, at least, have a general idea of one date/time setting being shorter or longer than another, rather than just random. This would imply that they can progressively determine what the date/times are, while updating their methodology along with the additional data they've acquired. | |
| Jun 26, 2018 at 23:28 | comment | added | Mark Olson | The question says that no one knows how far the machine travels until the measurements are done. The whole point is to calibrate it. | |
| Jun 26, 2018 at 23:20 | comment | added | Makyen | You have a time machine. You don't need to be able to predict 1M+ years into the past from only the data we have today. You can acquire whatever additional data points you need/desire starting from today and working backwards/forward. Doing so allows you to refine your models to account for whatever variation actually happened, rather than what we currently believe happened. | |
| Jun 26, 2018 at 21:45 | history | answered | Mark Olson | CC BY-SA 4.0 |