Curious pattern of GPS location variance

[ST Gui]

I blame space lasers.

 

[Gerhard]

I blame space lasers.

Good point Marjorie

 

[CYYJ]

...the curious part to me is the heavy clustering of locations in parallel lines running north and south. The lines are 10 to 15 feet apart.

Hello Dave:

Sorry for the late reply, I just saw this post today.

There could be any number of reasons for the "north-south" issue that you mention. My guess is that the most probable reason has nothing at all to do with the performance of the GPS constellation, instead, it has to do with how your service provider (Spotwalla) truncates or rounds off position reports when they create the map you are viewing.

The service provider doesn't want to be sending out (and recording, and storing) data with extraordinarily high precision (to the foot) because that is going to create a great increase in the size of the data they store in your account, and the size of the data they transmit back and forth. I suspect they figure that for their clients, 15 feet of accuracy is sufficient, and that is why you are seeing the data points so evenly and regularly spaced apart on both the north-south and the east-west axes. As for all of the highly scattered points (way off your location), that's probably just the result of bad satellite geometry at the time the location was recorded, or perhaps a tree in your yard blocking view of a satellite, etc.

I've seen very similar results using a Spot tracker when I have forgot to turn it off and left it on the glareshield of the aircraft overnight. The aircraft is chocked and it ain't moving (it weighs 17,500 pounds fully fueled), but the tracking record looks like the plane spent the night dancing in a discotheque.

Michael

 

[dduelin]

Hello Dave:

Sorry for the late reply, I just saw this post today.

There could be any number of reasons for the "north-south" issue that you mention. My guess is that the most probable reason has nothing at all to do with the performance of the GPS constellation, instead, it has to do with how your service provider (Spotwalla) truncates or rounds off position reports when they create the map you are viewing.

The service provider doesn't want to be sending out (and recording, and storing) data with extraordinarily high precision (to the foot) because that is going to create a great increase in the size of the data they store in your account, and the size of the data they transmit back and forth. I suspect they figure that for their clients, 15 feet of accuracy is sufficient, and that is why you are seeing the data points so evenly and regularly spaced apart on both the north-south and the east-west axes. As for all of the highly scattered points (way off your location), that's probably just the result of bad satellite geometry at the time the location was recorded, or perhaps a tree in your yard blocking view of a satellite, etc.

I've seen very similar results using a Spot tracker when I have forgot to turn it off and left it on the glareshield of the aircraft overnight. The aircraft is chocked and it ain't moving (it weighs 17,500 pounds fully fueled), but the tracking record looks like the plane spent the night dancing in a discotheque.

Michael

Thanks Michael. Rounding of coordinates was pretty much the conclusion the hive had come to. Location errors have always been a part of GPS navigation as you know. When I got my first GPS Selective Availability dithering was in effect and the “dancing” was over a much wider wide area and Cessna 150’s were recording max speeds approaching 500 knots. In the attached picture locations painted over an area approximately 75’ x 75’ with outliers beyond that.

 

[CYYJ]

....When I got my first GPS Selective Availability dithering was in effect and the “dancing” was over a much wider wide area ...

I remember flying Twin Otters for the Red Cross in the Liberian civil war in the late 1980s - early 1990s. This was just before GPS became available for civil aviation use - we had a long range navigation system called Omega in our aircraft that gave us accuracy of about ±2 miles at the best of times, and ±10 miles at the worst of times. One of the 8 worldwide antenna stations for Omega was located in Liberia, not too far from the airport. It was the tallest antenna in Africa at the time, and my biggest worry was that I would hit the darn antenna due to the inaccuracy of the system.

We've come a long way from those days... now we can talk about position errors of ±75 feet as if they are problems!

Michael