Sciencemadness Discussion Board

star tracker positioning

Polverone - 17-7-2003 at 16:40

I know that some satellites/spacecraft obtain positioning information by looking at the stars and deriving position from their appearance. I am in no way an astronomer, so let me ask on the off chance that someone here will know: would a similar scheme be possible for terrestial navigation? That is, given a portable computer with a CCD camera for taking pictures of a (clear) night sky, an accurate internal timer, and a database of stars, would it be possible to derive position on the surface of the earth?

GPS is nice, and more convenient, but it seems that something that doesn't rely on satellites might have uses during time of war or major disaster. I don't know if such a thing is possible. If nobody's built one yet, I bet this would be an interesting master's thesis project...

EDIT: Another interesting application came to mind. Cruise missiles or drone aircraft flying at night, above cloud cover, could likely use this for navigation as well (if the concept is sound), and it wouldn't be dependent on US-controlled positioning information.

[Edited on 7-18-2003 by Polverone]

I am a fish - 18-7-2003 at 00:38

This is perfectly feasible. In fact, you don't need to do it at night.

If you constantly measure the altitude (the angle from the horizon) of the sun, you can determine local time by noting when this reaches a maximum (local noon). By comparing local time with GMT (from a clock), you can calculate your longitude. By comparing the sun's altitude at local noon to its declination (the latitude of the point directly below the sun, which can be calculated with a computer) you can calculate your latitude

Remember however that you would need to take very accurate measurements to obtain a reasonable estimate of position. One minute of arc (about 1/30 of the sun's apparent diameter) equates to 1 nautical mile on the earth's surface (by definition).

raistlin - 18-7-2003 at 03:55

Yes Polverone, it is possible, though would be a very difficult undertaking. For starters, it would be similar to a planetarium in the sense that you would need to be in the country away from the glare of lights and the heat waves as they will distort the stars, causing your calculations as to where you are to be very messed up.. a few degrees difference in the atomospheric temperature would be enough to distort the appearance of the stars to the point where you could end up being several hundred miles off from where you really are...

Also, you would need to create a internal timing mechansim that would change as you passed through different time zones and back.. and if you could create one that changed automatically, then there really wouldnt be a need to use the stars...

VERY good idea though:D

[Edited on 18-7-2003 by raistlin]

Lugh - 18-7-2003 at 04:49

Aint that the kind of thing (without the tecnology) that has been used for thousands of years by the sailing fraternity?
But a fully automated and portable device would be dandy.
Getting away from GPS would be a good move. There's too much reliance on a technology run by the US military. Does anybody else find this a little bit worrying.
Perhaps a clock that sets its time from standard timers (cant remember the name - picks up radio broadcast from international agencies), would be nice and compact way of getting accurate times.

I am a fish - 18-7-2003 at 08:24

Originally posted by Lugh
Aint that the kind of thing (without the tecnology) that has been used for thousands of years by the sailing fraternity?

Yes. However, until the 18th century, determining your longitude was extremely difficult as you needed to know the local time at some fixed reference point. An error of one minute in your clock could put your position out by as much as 15 nautical miles. Today, accurate clocks are readily available and so the ancient methods can easily be improved upon.

Lugh - 18-7-2003 at 13:31

Very true Fish.
What I was trying to say is that for a very long time satalites weren't around and we got on fine. So it's not a very new idea.
I supose that 15 miles aint much if your just trying to hit a large land mass, but then trying to find a small atol in the pacific would be a bit tuff.:)

Polverone - 18-7-2003 at 13:59

I mentioned the potential weapons applications just because they leapt to mind. I really have no interest in creating weapons. I find the idea of such a gadget interesting simply because it seems more reliable than an interconnected system like GPS and because the stars aren't controlled by any one nation, so you can get the same positioning resolution no matter who you are.

It was thinking of the old "navigate by the stars" sailing technique that brought this to mind in the first place. But as I haven't extensively studied such lore, I didn't know if there were any theoretical or practical limitations on the accuracy of such methods.

What would be needed to get decent accuracy from imaging the sun or stars? Accurate internal timer, easy. Some sort of internal "plumbline" (for lack of a better term, due to my own engineering/electronic ignorance) giving a vector toward the center of the earth. Imaging camera with long exposures (for night) and/or dark filters (for day).

I am a fish, it seems from your description of sun-based navigation that position can be determined only once per day, since local noon needs to be known and that changes as you move. Is that right?

Iv4 - 18-7-2003 at 21:07

I seriosly doubt the practicalety of this but it can be done.The camera would have to be of extremly high resolution and you probably would'nt be very acurate but it could be done.Basically you take a few reference photos of the stars and judge from there on.Sort of like a map of the world with your position super imposed on it.

Lugh - 19-7-2003 at 03:36

I suppose the best but not the most practical solution would be a telescope with a ccd attached.
There already is software out there that will automatically track a telescope to a star if you tell it where on the planet you are. Would it be possible to do the reverse? could you tell the program where say 20 different stars are and then let the program work out your positioning?
DOes any one have one of these programs that they could try it on?

I am a fish - 20-7-2003 at 12:35

Originally posted by Lugh
Very true Fish.
What I was trying to say is that for a very long time satalites weren't around and we got on fine. So it's not a very new idea.
I supose that 15 miles aint much if your just trying to hit a large land mass, but then trying to find a small atol in the pacific would be a bit tuff.:)

That's assuming the clock was only a minute out. If it was an hour out (which could easily happen after weeks at sea in a violently moving boat), the error would be 900 miles. Before John Harrison invented the first clock capable of accurate timekeeping at sea, it was a common occurance for ships to be wrecked as a result of navigational errors.

I am a fish - 20-7-2003 at 13:02

Originally posted by Polverone
I am a fish, it seems from your description of sun-based navigation that position can be determined only once per day, since local noon needs to be known and that changes as you move. Is that right?


You could find your position at any time by measuring the sun's altitude and azimuth (angle from due north). However, to do this, you would need to know which way was north, which is not easily done to the degree of accuracy required (using a compass would throw you hundreds of miles out).

At local noon, you know the sun is pointing due south (or north) in the northern (or southern) hemisphere and so this problem is solved.

Organikum - 21-7-2003 at 11:15

As a matter of fact nuclear warheads of intercontinentals work this way - determining position and accurate reentry by the position of stars.
The developers of these machines have concluded very wisely that if those are fired there probably is no way left to determine the warheads position in a earthbound way anymore.

I see it as astonishing that it never came in mind of those that if it is impossible to determine position in a eartbound way, it might be time to disarm those which are still on the way. No - manhood will be done already, but they don´t even have mercy with the environment.
Reminds me on the good old tradition not only to overcome an enemy but to kill every living being in his capitol and carry the plow over the ground. Only more largescale is understood.

Organikum - 21-7-2003 at 11:27

much cheaper as any device and always working is to carry a piece of paper with one wheron the right answer is written.
The always right answer is:
"You are nowhere"
"nowhere is located exactly in between or a little right from this"

further questioning may be seen as heresy and a violation of the patriot act (in the US) or as bad manners (rest of world) ;)

Polverone - 21-7-2003 at 13:25

I took some 12-second exposures of the clear sky last night with a digital camera. Unfortunately for me, there is no way to tell the camera to accumulate a longer exposure. That still yielded a fairly nice image of stars. Worse, though, is that I can capture only a relatively narrow section of the sky at a time. I suppose I need a fisheye lens or similar. But that would cost money, which I don't have in abundance right now.

Orgy - I've done my share of reading about nuclear weapons, but I'd never heard about ICBM reentry being star-guided. Any websites, articles, or books you'd like to recommend on this?

blip - 21-7-2003 at 14:32

Interesting topic! :) I think this is doable, but I was also thinking that obstructions like trees would make it have to compensate.


Organikum - 21-7-2003 at 14:53

Sorry - I plain forgot where I read this as it is far from my interests. I can say it was a halfway relieable source (I just don´t read trash - exept on bordersciences) and that my memory serves me usually very well. I remember it was a article on the latest generation of warheads which are active in direction changes and aiming against the old generation resemblimg more or less ballistic units.

I believe cruise missiles have at least been tested which such systems as additional feature.

You might first look under astronomy for usable stars and ensembles of stars for where you live, fixed and nonfixed ones and you will either go by triangulation wherby a good telescope is helpful and a stand with the possibility to read out the angles + very good math abilities, or you go by completeness of visible stars - what you plan by now I guess.
In this case you wil have to reduce the part depicted by a mask and you have to make the camera look straight up. Pendulum principle - easy. A stick and a weight fixed to the camera in line with the optics - a improvised cardanic fixation - ready. This segment of the sky can be analysed and/or by picturerecognitionsoftware compared to stored ones. Of course the stars visible at the border of the mask tell your location best.
And I have a right as needed software handy for ya which enhances picture quality of stand-pics by night!
from SVXb-Labs: nightvision

Attachment: SVBxLabs_Nightvision.rar (17kB)
This file has been downloaded 779 times

raistlin - 21-7-2003 at 20:33

Organikum is right. I read it somewhere, I think it may have been Popular Science, that a new technology the government was working on was a guidance system based on the maping of the stars.

The only problem I really see with it is the fact that there are astronmicall anomalies such as shooting stars, that could alter the readings of such. Also, the effect of distortion from heat, lights, weather, etc, could also be a problem causing you to get readings miles from where you actually were.

Iv4 - 2-8-2003 at 10:57

I think that if this is a last reort it should have several tracking mehods.For the Sun thing you might try puting a box with a circle painted at one point ad judge the position from its shadow.I'm guessing that there are already programs out there that can detect constelations so you could just ran those through a databse with the locations(I fergot what tjose are called).

vulture - 3-8-2003 at 10:54

The European Union is building it's own commercial sattelite navigation system (called galileo IIRC) which should be more accurate than GPS and should have no accuracy restrictions for private users.

Iv4 - 3-8-2003 at 22:26

Is it more accurate because of more satelites,better algorithm,or better clocks?