A rover using GEODNET corrections can reach centimetre-level positioning in suitable conditions. The result depends on the receiver, satellites, baseline, correction age and local obstructions. Its estimated precision and its error against a known coordinate are different measurements.
Measure the error you care about
Choose an independently known point in the same reference frame and height system. Record the fixed result, reinitialise the receiver and repeat. Compare horizontal differences and height differences separately. A wrong pole height can produce a consistent vertical error even when successive fixes agree closely.
Report conditions with the number
Note the time, receiver, solution type, antenna height and control used. For a job near trees or buildings, repeat checks in representative conditions. A single open-sky result should not be published as a guarantee for every site.
State what you measured
On a known point, record the coordinate reference system, height reference, antenna height and solution state. Compare the observed easting, northing and height with control, then repeat after a fresh initialisation. Report the observed differences separately. A receiver’s displayed precision estimate alone cannot establish the accuracy of the delivered drawing. Check the export workflow as well.