Drift in MGEX base station coordinates?

For my work, I use the RTCM stream from my International GNSS Service as base input stream for my REACH rover to do RTK.

To get the base coordinates for input into ReachView, I run a separate RTKNAVI on my home desktop computer, with the base station as the input stream, set on Static PPP with and once the coordinates are resolved to <0.5 cm precision (usually through the night, set to ionosphere free LC since the base stations have L1 and L2 GPS, GLONASS, Beidou and Galileo), I input the coordinates in ReachView. In my experience, it has given me very accurate RTK float solution but I have to re-obtain the coordinates before every session of data collecting with REACH as I notice the PPP solution of the base station is drifting (> 10 cm within a few months).

Continental drift is at most 1-2 cm a year where I am at (Sunda Plate), anyone has any idea why the coordinate drift is so drastic?

You can see the ntrip caster table in mgex.igs-ip.net. I tap on the stream SIN17.

Comparing the PPP solution of the base station from my RTKNAVI output logs:
2016/03/18 14:59:30.000; 1.342979256; 103.679434330; 86.1680; 6; 16; 0.0043; 0.0058; 0.0128; -0.0009; -0.0008; -0.0009; 0.00; 0.0

2016/01/22 13:29:30.000; 1.342980936; 103.679438622; 86.6406; 6; 14; 0.0037; 0.0052; 0.0112; 0.0004; 0.0018; -0.0013; 0.00; 0.0

In the span of about 2 months, my latitude has drifted about 18 cm and my longitude has drifted about 47 cm. :scream:

Single frequency PPP is very challenging and accuracy is not 100% guaranteed as far as I understand it. The fact that your base is L1/L2 does not really provide any benefit.

1 Like

Thanks igor. Will try post processing the raw logs with the IGS clock products then.