I am struggling to understand how Trimble’s Pathfinder Office transforms and exports GNSS data and I’m hoping someone here can explain. I've seen posts like this and this, and documents like this, but there must be something I'm fundamentally missing here, because after reading information provided in the links, I'm still not getting the result I'm expecting.
We collect our GNSS positions in autonomous mode using a Trimble 7X Centimeter Edition rover unit and then use Trimble Pathfinder Office to correct these positions against a CORS base station. Our unit collects in the WGS84 datum (we confirmed this by checking the rover settings). Our correction process has always used the reference position provided from the downloaded base files, which in this case is in the NAD 1983 (2011) datum. We do not choose to use the reference position provided from Trimble’s base station list, which provided the reference position in terms of ITRF00 (1997) (which I believe is synonymous with WGS 1984 (G730)).
After differential correction, we provide a coordinate system and datum to export features. We’ve always exported in latitude/longitude in the WGS 1984 datum. However, for a recent project we’ve changed our procedures and exported data in latitude/longitude, but in the “ITRF to NAD 1983 (2011)” datum provided by PFO.
We’ve noticed that these positions have shifted about one meter when compared to our original procedures of exporting in WGS84. This generally makes sense to me because I know that WGS84 and NAD83 have about a 1m shift, but I get into the weeds when I process/export our data in other ways.
We collected a position over a NHDOT benchmark. We’re less than one mile from the CORS base station. Our unit delivers +/- 1.0cm horizontal accuracy after post-processing. I’ve exported our data four different ways: (1) using the downloaded reference position and exporting in WGS84 datum, (2) using the downloaded reference position and exporting in ITRF to NAD 1983 (2011), (3) using Trimble’s reference position and exporting in WGS84 datum, (4) using Trimble’s reference position and exporting in ITRF to NAD 1983 (2011) datum. I also downloaded from NH GRANIT a list of benchmarks that includes the one we’ve used here, so I can compare our outputs.
Position #3 from above, that used the Trimble-provided reference position (in ITRF00 1997) and was exported in the WGS84 datum, aligns with the benchmark shapefile (there’s a 1cm shift, which is within the margin of error).
The positions that were corrected according to #1 and #4 above align with each other (+/- 3cm), but are about 1m south of the benchmark. The position that was corrected using #2 above is about 2m south of the benchmark.
Can someone please explain to me what’s going on behind the scenes with PFO to cause the points to change position, and what is the “correct” workflow?
