We've just done a survey with two different GPS brand RTK systems - one was a Trimble Field Level System using the survey plug-in and one was, I think it's probably Novariant or AutoFarm System, an Australian guidance system.
The operator drove around the boundary with both systems very closely to check the difference and the MB is here. The GPS Ag System saved the coordinate system in Eastings and Northings, see these numbers here, and using the GPS datum. Then we shifted the multiplane so that they matched up at the Master Benchmark and we want to see the distance at the extremities.
As you can see, there is a significant difference at the furthest point away from the MB, which is about 1400m, there's a 21m difference in horizontal. Those lines should overlap perfectly if the coordinate systems matched. The difference is Trimble Field Level System uses the local plane coordinate system which is slightly different to the other coordinate system - the UTM Easting and Northing. So you can see a 21m difference there and at the southern edge it's 17m difference.
The problem that produces when we try and survey with one coordinate system, do a design, and then export to back out to load into another GPS system to do the earthmoving, is that these difference can end up producing significant differences in what the design should be. On this field it was quite flat, so this 20m horizontal error we tested and found the following.
The Proposed Topography Map and contour intervals is at 30mm or about 1inch. Where the contours change, the closest together that's where we will end up with the largest error design because of that horizontal error. Looking at these contours, the worst vertical error could be 80mm in a couple of locations and 40mm at a few locations. The rest of the field will be less error than that.
We could add a feature in OptiSurface to translate these systems from one to the other but there's always the risk that we don't know the exact coordinate system that each system uses. So just be aware of this issue and the easiest solution is to survey with the same type of system that you wish to do the earthworks with.
One other comment is the further you get from the MB the greater the error. If you had a small field in here or around the MB, the error is quite small and you may get away with it. A field that's maybe 400m x 400m would not be too bad at all, but with large fields it is more significant.
And the other thing is depending on how large your slopes are. If your slopes range from say 10% say at this point it's 20m out and sloping 10% then you are going to be 2m out in the vertical, you can see that's going to be a massive amount of error. This problem is least significant when you have smaller fields and flatter topographies. It is most significant when you got larger fields or larger slopes (anything above 0.10% or 0.20%).
I hope that's useful and thanks for watching.
Also see this article for additional information: GPS Coordinate System Check