|Documentation :||Hardware | Software | Tutorials||Wardrobe :||Jackets | Applications|
|Papers :||Notes | Proposals | Research||Administration :||Stock | Minutes | Progress|
One of the most common questions we are asked is "how accurate is GPS?". This is not a straightforward question to answer - there are many measures in use. The surroundings of the user are a factor too - tall buildings and trees can obscure satellite signals and hence degrade the accuracy. This page details some of the results of comprehensive tests which were carried out in January 2002. These give an indication of accuracies to be expected from the receivers we currently have in use.
Each test was carried out over a 24hr period with readings taken once per minute. The receivers were placed on a south facing roof (with a 30deg slope), with an otherwise clear view of the sky. A reference position of 51deg 28.8046min N, 2deg 36.4323min E. was determined from the average of the readings from a dGPS receiver over 24hrs.
The results are shown as the 95% and 50% accuracy, a line plot showing the actual error in metres, and a scatter plot showing the distribution of readings with latitude on the y axis and longitude on the x axis. The scales are in metres.
For mobile applications where size is critical.
95% accuracy: 36.41m ; 50% accuracy: 17.45m.
n.b.1 Satellite fix was lost for two periods of aprox 30mins.
n.b.2 Similar results were obtained with a second iGPS receiver.
Garmin's economic GPS receiver.
95% accuracy: 8.20m ; 50% accuracy: 3.57m.
n.b. quantisation errors due to .001min resolution.
Garmin's top of the eTrex range receiver.
95% accuracy: 7.80m ; 50% accuracy: 3.19m.
Requires (wireless) connection to a dGPS server using the internet .
95% accuracy: 6.23m ; 50% accuracy: 2.29m.
n.b. these readings used to obtain reference position.
Only suitable for static applications with a local dGPS receiver.
95% accuracy: 5.10m ; 50% accuracy: 1.95m.