# Is there a correlation between the absolute errors of nearby GPS receivers?

Suppose I have two different GPS receivers located x meters apart and their exact location is know. Then, would there be a correlation between the absolute errors of nearby GPS receivers? By absolute error I mean difference in distance and angle between the the GPS reading and the exact location.

If so, how to calculate this correlation? how much does it decrease with x?

Background: some one has told me that since the two receivers are close to each other, they would receive same signal with same error, thus they would most likely drift from the exact location at the same direction and error distance (assuming two identical receivers).

• Sorry, but humbly asking, does anyone have a method to coorelate GPS errors (for receivers x meters apart) as asked in the question? Commented Mar 6, 2017 at 14:11
• This does not really answer the question. If you have a different question, you can ask it by clicking Ask Question. You can also add a bounty to draw more attention to this question once you have enough reputation. - From Review Commented Mar 6, 2017 at 14:34

Yes this is how Differential GPS (DGPS) works. Reference stations for which the location is precisely and accurately known are used to apply corrections to the GPS 'rover'. This can be done in post processing or in real time with a data transmission from the (network of) reference stations. You need a special GPS receiver.

Here is a quick explanation from Trimble.

I test GPS receivers for a living, so I have experience with this. If we take two identical GPS receivers (same model, same firmware version) with a shared antenna, they usually give the same solution (within a fraction of an inch) but sometimes the solutions wander apart, maybe from clock drift in the receivers. Once you use separate antennas and start moving them apart, the positions they give correspond "less often". There's an element of randomness in this.

Simplistically, differential corrections could use a position offset but in practice corrections (RTCM 2.x or RTCM3.x or CMR etc) send errors per satellite.

• Are we talking about RTK (phase measurements) or DGPS (code measurements)? Commented Mar 7, 2017 at 9:13

This is generally correct assuming all receiver settings and equipment calibration are the same.

The general error GPS properties are very strongly spatially autocorrelated.

Your friend's statement is more true in an kerb area with a clear 360 view of the sky than in built up areas.

But yes out is a reasonable assumption.