I used NRTK this morning to check elevations on 3 BMs. I explained to my
supervisor that he can expect a precision of only 0.06' <= 0.07' because that
was the VRMS for one of the BMs. The standard deviation during that 180 epoch session was
0.005', if I'm not mistaken. Does VRMS represent precision and the standard
deviation represent accuracy? Does VRMS represent the deviation from the
rover's true vertical position?
My experience with NRTK (both CRTN and a Cansel Trimble network) is that the reported vertical accuracies are quite optimistic, even for 300-epoch observations. 0.2' differences between observations 3 hours apart weren't uncommon on long vectors, and some were larger. Granted, these were all GPS/GLO only; I don't have any experience with more constellations.
Agree that results are optimistic when seeking true height. The precision of 180 observations do not indicate a truly accurate (correct) height. We performed days upon days of one second interval network observations at various distances from reference stations. While the RMS of all the observations were generally less than 0.06' of true height the difference between the highest and lowest observations were quite often 0.3' to 0.4' meaning there are times during the day when the quite precise observed height could be as much as 0.2' from truth.