OPUS variation
I read a thread here from 2019 regarding variation in OPUS solutions where commenters discussed using other software in post to eliminate specific CORS stations (which I see OPUS Beta is offering) or burning 7+ hours among other things. 5 Years ago now, though. Saturday I did a topo for a small commercial site using my old R8-3's and an S6 to check stuff and get to some difficult shots. After 3 hours I replaced the battery in the base and at the same time rotated the head 180 degrees 'just for the halibut' and went on for nearly 3 more hours.
I converted both files to Rinex (v3.02 if that matters) and submitted as TRMR8_GNSS3.
I got a difference of nearly 7cm ortho height and, maybe worse, over 13 cm horizontally (nearly half a foot).
Is that normal or acceptable? Fortunately I shot the base and corners with th S6 but how would I justify an elevation (which isn't necessary for this but could be someday)? I had quite open sky, one large building on a neighboring property but getting good fix on 8-10 sv all day.
Could it be my oldish equipment? I did lose radio link several times with it coming back online on it's own within a couple minutes.
Thanks.
No, that's not normal, expect to see differences under 1cm. And from your description of what you did with the two sessions, that would mostly be rotation of the base receiver. I assume you meant you rotated the tribrach. It's always best to rotate the tribrach, the receiver, and "break" the legs and raise or lower the tripod. When I measure the receiver, I will use meters and feet. You can write down the foot measurement in the book, then take the meter measurement, input the number in the DC with a m behind and it will automatically convert it to feet so you can compare the two.
The difference you describe does relate to the distance I would see with a "bad fix" back in the early days of GPS. I had a Trimble rep tell me to expect one wavelength error with a "bad fix". I recall that one wavelength is 21cm. Someone here will probably correct that, it's been years since I cared to look it up.
So that does look like a bad fix, I can't imagine it persisting for 3 hours using R8-3s which aren't even close to the old 4400s of the 1990 where I would see them happen and stick for 20 minutes or so before failing. R8-3s aren't really old tech, they are great for a base receiver.
I think 8-10 satellites is pretty bad. I'm assuming that's an instrument or software limitation and not the result of having the base in a bad spot.
Also, I thought OPUS required a minimum of 4 hours of data in order to generate a solution. Did you upload your files separately or did you find a way to combine them first?
Lastly, I wouldn't spin the base like that.
I think 8-10 satellites is pretty bad. I’m assuming that’s an instrument or software limitation and not the result of having the base in a bad spot.
Also, I thought OPUS required a minimum of 4 hours of data in order to generate a solution. Did you upload your files separately or did you find a way to combine them first?
Lastly, I wouldn’t spin the base like that.
10 probably is low. They do only get GPS and GLONASS but still thought it should be better than that.
2 hours min
?? Not sure how spinning the base (the tribrach) could make any difference in ortho height. If it does, we've got serious industry issues. I did it to get a second collection under 'different' conditions, without changing the height.
<i style="background-color: transparent; font-family: inherit; font-size: inherit; color: var(--bb-body-text-color);">No, that’s not normal, expect to see differences under 1cm. And from your description of what you did with the two sessions, that would mostly be rotation of the base receiver. I assume you meant you rotated the tribrach. It’s always best to rotate the tribrach, the receiver, and “break” the legs and raise or lower the tripod. When I measure the receiver, I will use meters and feet. You can write down the foot measurement in the book, then take the meter measurement, input the number in the DC with a m behind and it will automatically convert it to feet so you can compare the two.
<i style="background-color: transparent; font-family: inherit; font-size: inherit; color: var(--bb-body-text-color);">The difference you describe does relate to the distance I would see with a “bad fix” back in the early days of GPS. I had a Trimble rep tell me to expect one wavelength error with a “bad fix”. I recall that one wavelength is 21cm. Someone here will probably correct that, it’s been years since I cared to look it up.
<i style="background-color: transparent; font-family: inherit; font-size: inherit; color: var(--bb-body-text-color);">So that does look like a bad fix, I can’t imagine it persisting for 3 hours using R8-3s which aren’t even close to the old 4400s of the 1990 where I would see them happen and stick for 20 minutes or so before failing. R8-3s aren’t really old tech, they are great for a base receiver.
Thanks. That's helpful info. I'll probably look into it with my service guy. The radio link drops made me uncomfortable.
10 probably is low. They do only get GPS and GLONASS but still thought it should be better than that.
Yeah, you should get around 8 just from GPS. OPUS only looks at GPS data still as far as I know so if you're showing 10 and half of those are GLONASS then OPUS is processing 5 which might account for the poor results?
?? Not sure how spinning the base (the tribrach) could make any difference in ortho height. If it does, we’ve got serious industry issues. I did it to get a second collection under ‘different’ conditions, without changing the height.
My mistake, I misread and thought you spun the base while it was collecting static.
I did lose radio link several times with it coming back online on it’s own within a couple minutes.
Lost radio at the rover? That shouldn't affect the static collection at all.
Good points. Thanks. I also just noticed that in the 2 OPUS reports, where they show the Base Stations used, they share 2 of 3 and the third base station listed is different between them, I'm assuming that should also never cause a discrepancy like this but I wonder why they changed between back to back readings.
In the 2nd static reading that differing base station is more than 30 times further away than the one in the first reading.
Are your opus results different or the RTK points.
If it’s OPUS results differing then when you state you rotated the head that could cause an issue. Opus now uses the absolute antenna calibration file instead of the relative. So you need to make sure your antenna receiver is facing north according to the NRP in NGS guidance on that particular model.
It is not unheard of for the "base stations" that OPUS uses to have bad information. Antennas have been changed without their positions being updated resulting in faulty reference data. You could select your own CORS (base) stations to be used for the processing. You might find by using other stations with your resubmitted data that you get consistent results by eliminating one of the stations that was initially used in your prior OPUS reports.
How far away are the CORS stations used to process the OPUS results?
I'm in a pretty sparse area as far as national CORS nearby, so OPUS pulls stations from 50 to 70+ miles away. In those areas, 4, or preferably more, hours really was needed to get good results from OPUS. Less than 4 hours often returned results that would be in line with what you are discussing.
If I didn't have time for 4 hours of data, I would download the static from the my home state's CORS stations (20 to 40 miles away) and process at my office. This would usually provide very good results.
