The unit of output normal

[FFFFFG-IGG]

Hello developers,
I've tried to compute ZTD from GNSS PPP, aming to get the post variance through inversion of the Q(which is the normal equations):

But I found the values of Q is so large that the inversion of it will be very small, which can reach 10-E6, so I wonder the unit of them. The unit of computed ZTD is 'm', so does unit of Q is same as ZTD and reveal the precision of ZTD?
regards,
Yongzhao Fan

[tmayerguerr]

Hi Yongzhao Fan,

Yes, the units should be [m].
In our experience, the estimated accuracies are an order of magnitude too optimistic (i.e. the variance is too small by a factor of 10^2). We suspect that this is because we neglect the correlations in the observations.

regards
Torsten