Quick Dumb Question

[jmfargo]

If you were to go far back in time, GPS systems would be useless because there are no satellites, correct?

Is there a system, in use today, that can tell you your relative position on the face of the Earth without use of satellites, and not requiring any input from the user?

Comments

[elffin.livejournal.com]

HiYouDon'tKnowMeIReadYouonFlemco'sFList.

In short, no.

In long: That depends. What do you mean by "relative position"? How accurate do you want that to be?
What do you mean by "input from the user"?
There are systems in use today that could be combined together with software to find a very imprecise location - Know what date and time it is, take a snapshot of the visible sky, correlate from the height and declination of visible stars on the horizon. That would give you a figure probably somewhere on the accuracy of within forty square miles. We used to practice using lookup tables and star charts for Orienteering in Boy Scouts. I don't think anyone's actually implemented such a thing in hardware because A: image processing not terribly advanced until recently and B: GPS satellites.

There's also the possibility of using which craters are visible on the surface of the moon, along with the elevation, and knowing the date and time (the moon currently shows the same features to a given longitude on every orbit - it "rolls" around the earth), but again no extant portable off-the-shelf hardware modules and the moon is technically a satellite.

so, not really, no. And the computational methods for working it out don't work if you don't have a way of keeping and knowing time.

[jfargo.livejournal.com]

HiYouDon'tKnowMeIReadYouonFlemco'sFList.

Actually, I recognized you from the comments on Flemco's blog.

As for relative position, I'd be happy to have something within 100 miles of position, the closer the better of course.

Input from the user? Well, it seems my dream machine doesn't exist, but if it did, the most input a user would have to do is turn it on, and maybe point it at a few things (like the stars). I was trying to figure out if there was something as simple as GPS, but using a different system that would work without external mechanics (like radio frequency, or what-have-you).

The moon would be fine as a point of reference for me; I should have said man-made satellites.

[elffin.livejournal.com]

It occurs to me that if you had a 360 degree camera, some image-processing software to pick out reference stars (Current research telescopes/facilities probably have such software), and a set of pre-computed hash tables for dates/times/stars/locations, then your proposed device would probably be workable. The heavy lifting of computing your astronomical rainbow table would be performed by number-crunching in the current time. The database would probably be very large and you'd have plenty of collisions on individual star/reference point's hash tables to dates and locales, but probably very few duplicate collisions if you implement every reference point as it's own hash table and then take the union of the results, or the union of the results widened by a margin of error that is relative to how distant you are from known (current) space/time.

This presumes that the models used to compute the hash tables are highly accurate in their predictions.