Hi FX, does the area have any multi-path, i.e. trees, buildings… I’ve experienced this before and have been meaning to verify with some local passive control marks. They also have varying clear sky areas. I use my RS2 using our state RTN VRS service.
From my experience, just comparing control marks on my home site, I’ve seen errors too. These control marks have been established by terrestrial traverse and also static GNSS observations. These areas have verifying clear skies and low to high multi-path.
I’m skeptical using the RS2 in any kind of multi-path areas. I’ve check a few passive control marks and got good results on these marks. They mostly have very little multi-path, however one point is about 25% multi-path. It took about 10 minutes to get a good comparable result. This depends on PDOP values from what I’ve seen… high values indicate bad or low accuracy results. I suggest to look at your PDOP values before accepting any results.
Just be careful in using any GNSS receivers in multi-path areas. We have JAVAD machines here. They are the only receivers that can go into places that the sun doesn’t shine and have a reliable-verifiable RTK RTN solution.
Indeed, one of my points might be affected by a multipath area, could explain the results.
In another hand, i have a really clear point, with 3cm difference in Z between the first and second survey, and 12cm between the second and third survey. This one is strange, i’ll try to investigate more, but i’m a little bit disapointed about this one.
Fun fact, i’m trying with an other method and it works perfectly …
I’m connecting my rover straight to an open base in my area, and i’m doing the same measurement and guess what ? Less than a centimer difference between my points.
Have to investigate if the virtual station is not the problem with Reach.
I also noticed that my XY position is really good bewteen each survey. Shouldn’t the multipath area impact my 2D posisition also ?
Yes. On clear area stations where I’ve compared with passive control marks using static and RTN VRS, I’ve seen 1-2 cm in horizontal/vertical error. That’s pretty damn good. The RS2 is a capable receiver in open areas but it doesn’t like multi-path.
I checked and my network provider works with all GNSS systems, and all are ON in my settings.
Looks like i have better results with other correction systems than virtual reference station (MAC and FKP).
Have to to more observation to be sure about this idea.
Reach receivers don’t have any special requirements for working with VRS networks. So, if the VRS provides standard RTCM3 messages, Reach should work fine with it. Please note that to work in RTK, Reach should receive at least RTCM3 messages with GPS observables and ARP station coordinates.
Let’s investigate the issue to determine why the rover calculates its precise position only with a local base. To collect the data for the investigation, please connect Reach to the VRS and the local base and record these logs for 10-15 minutes in each case:
Raw data
Base corrections
Position
Please place the rover in the same environmental conditions so that we can compare the results.
It’d be great if you collect several points during each test and attach the CSV files as well. It will allow me to examine how the points were averaged and check their precision.