Hi all,.
When doing a GPS network adjustment (either free constraint/inner or full constraint), what value you looking at on the result to determine the quality of the observation and the network?
Cheers,
Tyas
I use StarNet to adjust vectors, and I'll be looking at N, E, & up residuals. Also, lower covariance numbers in the raw data generally indicate better quality data.
I like to make a plot of horizontal and vertical residuals versus baseline length. Making a plot of ECEF X, Y, Z residuals is much less meaningful. Yes, if they are large it indicates a problem, but it is difficult to tell if it is an HI bust, which the H and V plot will show. I also check repeat baselines before the adjustment.
Getting a reference factor (whatever the particular program might call it) near one is not that important to me. But, if it is way high then that is red flag.
John Hamilton, post: 343722, member: 640 wrote: Getting a reference factor (whatever the particular program might call it) near one is not that important to me. But, if it is way high then that is red flag.
I've seen good results with a reference factor (standard error of unit weight) as high as 7, though I typically expect to see it at 3 or less. GPS processors are notoriously optimistic regarding standard errors, so I don't often see the reference factor near 1 on the first go-round.