I am using a GPS receiver to perform a speedtest over a few meters. I’ve read that the GPS is accurate up to 4 meters. Is the deviation the same in an open area and also over a short period of time? Or can this vary everywhere and also over a (short) period of time?
GPS accuracy varies based on several parameters, most important the number of satellites you can get signals from, and their geometric distribution on the sky.
Some factors affecting this are:
- Line-of-sight. You need line-of-sight to the satellites to get any useable signals. Different locations will have different horizons (i.e. a hill, mountain or building may cover parts of the horizon in one place, but not the other).
- Time of day. There are many GNSS satellites in orbit, and as they move along their distribution may vary depending on time of day. There are tools to check satellite coverage in a location at a set time based on orbit data (ephemerides).
In addition there are other variables that may affect accuracy:
- Atmospheric noise. Changes in solar winds and other atmospheric variations.
- Local noise. Signal noise from e.g. reflections from a building.
- Quality of receiver.
- Measuring procedures. The accuracy improves with data over time, i.e. logging over time gived better data than a snap measurement on the move.
In your case (measurements within a limited open area within a short time) I'd say most of the above points should not affect the results. The main question would be how much the accuracy is affected by the movement involved.
I've seen GNSS used for very accurate measurements in e.g. alpine skiing. The accuracy here is obtained by using a local base station and differential GPS. Is this an option here?