No problems to complain about, but have a question about point collection.
I’ve been experimenting with my two Reach RS units for a couple of weeks now and have a few questions. I’m new to this so they’re probably very basic but here goes.
Here are my settings in case you notice I could improve something:
I’ve got the base set up like so:
Positioning Mode: Kinematic
GPS AR Mode: Fix and Hold
Glonass AR Mode: Off
Elevation Mask: 15 deg
SNR Mask: 35
Max Accel: 1m/s
GNSS: GLONASS/SBAS/QZSS
Update 5hz
Base Coords: Average Single for 10minutes
The rover is being used for point collection within 1km of the base in an open field:
RTK Settings are the same as base.
Survey settings on the rover unit:
Autosave rules:
FIX
Time 1 minute
Precision 0.003
DOP 2
My questions:
-
Each of the points I’ve collected has a σE, σN, σU value that looks very nice (like 0.003 for the σE and σN and 0.004 - 0.008 for the σU) does this imply the error of the position relative to the base unit? Or does it imply only the standard deviation of the measurements taken in the one minute observation period I’ve specified?
-
Relates to the operation of the Reach unit: When I’m using it in survey mode, is the RTKLIB correction is happening in real time between the units? Meaning that at least relative to my base unit, my position is truly as accurate as the σ values suggest? (like 0.002 to 0.008 at worst).? Does this also imply that if I post processed my raw data on my laptop I would get the same sort of result?
-
In this project I left the base running in one spot for 7 hours. I marked the spot with a piece of metal that I can come back to. I submitted the raw data to our Canadian correction service online Reach Tests with NRCan Corrections like another user on here. This yielded a pretty big correction.
Peter’s results (a 24h observation)
95% Error Estimates:
Lat : .161m
Long: .107m
Ell. Z: .299m
Mine (a 7 hour observation)
95% Error Estimates:
Lat : .180m
Long: .147m
Ell. Z: .357m
Looks great to me.
The correction service returned a pretty big correction.
Latitudinal difference from Apriori: -1.921 meters
Longitudinal difference from Apriori: 1.605 meters
Ellipsoid Height -10.197 meters
Am I correct in assuming that the ellipsoid used by the Candian correction service is different that what the Reach uses (they use NAD83(CSRS 2017)? Is this why the vertical is so different?
Finally the big question… If I take my collected points put the into Arc or QGIS with the correct NAD83 Zone 17N coordinate system, then shift the points I collected by -1.921 meters north, 1.605 meters east and -10.197 meters down… would I then have points that are absolutely positioned to the accuracy of the 95% confidence interval listed above (having an absolute error of about 18cm lat, 15cm lon and 35cm vertical but a relative error in the 1-2 cm range)?
Thanks for reading, I’m new to ‘proper’ GPSing 
I cannot attach the corrections PDF as a new user unfortunately.