Inaccurate GPS leading to unreliable data

Robert_A_R

Hello all,

I've been having many activities that the GPS is off and after doing an interval training i couldn't believe the average pace it gave, so i looked at the route it tracked me. In the image below you can see that not once, the path that i ran was followed (the square road), making the entire run significantly shorter and subsequently giving a lower pace.

Now this wouldn't be a problem if this happened once or twice, but it often happens on various tracks. Am i the only one with this problem? I.e. GPS being significantly of track, leading to unreliable data?


Regards,

Robert

elbo

What's the scale of the map? How long is the square? Are there trees? Was it a cloudy day?

GPS accuracy is not within the meter. It could be off for a couple of meters to even ten meter. If one side of your square is 100 meter the GPS could measure that as anything between 90 and 110 meter.

Also GPS accuracy could be influenced by clouds, trees or high buildings. I also noticed mij runner 2 is struggling with GPS reception when it is under my clothes (something I have never experienced with my Suunto Ambit 3). Try to keep it somewhere with a clear view to the sky.

Robert_A_R

There are trees (though not fully covered) and the square is 500meters. I indeed know/expect to see some deviance between actual distance and gps measured distance, however this was in some instances rather big. Seems to be in some laps of 1000m a difference between 50-100m, that's rather significant. What's your experience?

I also noticed that the GPS reception is quite sensitive indeed, hope that they find a solution for this.

elbo

Last sunday I ran the Rotterdam Urban Trail and on Strava I noticed people reported length between 12.1 and 13.7km. So a 10% deviation could be expected.

gl00

Elbo is pretty right here. Also note that your example is definitely not the best scenario if you want good accuracy: signal is deteriorated by the foliage, sharp changes of directions are difficult to track for GPS and the errors are repeated each lap you do.

If you have to run on a track, better find a proper track at the stadium with big curves instead of sharp turns and no tree coverage: the watch will still measure a bit short, but that will already be closer to reality. Not much can be done appart from that: the issue is with the technology used. note that in such conditions, my Tomtom watches usually give better results than the other brands I own (probably better GPS chip and/or error correction algorithms used)

best case for accurate tracking would be to run in straight lines, far from buildings and trees (I run some of my repeats doing back and forth along a canal for example)