PPRuNe Forums - View Single Post - GPS Based Navigation Accuracy Question
View Single Post
Old 28th Jul 2008, 18:06
  #11 (permalink)  
TeachMe
 
Join Date: Nov 2007
Location: Seoul
Posts: 105
Likes: 0
Received 0 Likes on 0 Posts
I believe the position error of say 5 meters X/Z and 9 meters Y stated above is a statistical model in which the position is calculated from the various solutions.

One way of looking at it in a stationary unit is like a shooting target. With a stationary shooter and one shot on the target, you can not estimate the centre, with 10 you can make a very rough guess, with a thousand you can be quite close, and with an infinite set of shots you can nail down the location of the centre of the target to the point where Mr Heisenberg gets involved.

Now, with GPS add to this that the satilites are orbiting, and the plane is flying and you get an idea of how difficult things can be and how your accuracy is always limited.

This gets at one reason that GPS units can measure the rise of a volcanic dome in milimeters, get only give it in meters on an airplane.

Now for the original question, GPS works based on the time stamps of the signal, based on an atomic clock on the satilite, sent by the satilite. If that signal for some reason took more or less time to reach the reciever then the accuracy would be altered.

If that was a constant situation then the actual result would be constantly wrong, if it was transient, then the situation would be transient.

As c is different in different mediums, I would not be surprised to learn that results varied based on the atmospheric density, but I would expect that to be in mm over a day and not more. Ionospheric refraction, as noted above, would cause the signal to take a longer route to the reciever and thus result in a differnet solution.

Hope that give some background. And to those who know more, sorry if I am a bit hazy about some parts of it, I sold them about 10 years ago for Garmin, but things have cahnged and I have forgotten much.
TeachMe is offline