Bad (?) GPS accuracy

The biggest problem here is that the stated "accuracy" is actually a probability. And no manufacturer, including Garmin, will tell you what parameters THEY use. The actual meaning of "6-10 meters" is that "my GPS-reported position is within 6-10 meters of my real position with a probability of x," where is x is what they don't disclose. But it is typically a fairly high percentage (85 - 95 %?). (It's actually more complicated than that, but that's the short, understandable version.)

That number depends on a lot of factors, including atmospheric conditions, the geometry of the GPS constellation at the time, multi-path issues (reflections off of nearby terrain or buildings), and so forth. Without knowing the details of the manufacturer's computations, you never know how much to trust the absolute numerical result. The best way to think of that number is in relative terms. If the GPSr reports 3m, that's better than a situation where it reports 10m. 

Realistically, 5m (16 feet or so) horizontally is about what you can expect. It is safe to assume that the actual horizontal error is no more than 15m (50 feet). Vertical error is typically larger (by perhaps a factor of 2) because of geometry.

Hi,

Ok ok. Thank you for the answer.

Regards,

Were you in the canyon there?

If so, those canyon walls are about 300ft higher than the river below, and narrow, so the results you report are as good as it gets for those conditions.

You may want to try the new GPSMAP 66sr if you desire better accuracy in those conditions.

Yes and no. In the canyon the accuracy was bad but i think expected.

I was more surprise by the "bad" accuracy in the wood.

I didn't know the sr model. I will take a look but since my 66s is quite new, no sure i will change

Regards,