Yesterday I did a 6-hour OPUS session on a disk that has poor sky visibility, but is on the NGS priority list for GPSonBM because there is nothing else found with decent sky for some distance. I wanted to find out how bad the results would be, and indeed they are bad.?ÿ
The rapid-orbit result using Geoid12B is 0.288 meter (0.945 ft) higher than the leveled elevation.
Only 83% ambiguities fixed, EL HGT pk-pk = 75 mm showing that it isn??t a very stable measurement. I had a similar ultra-rapid solution and one (expecting rapid to be available) that was even higher.?ÿ I??ll re-run the solution again when precise orbits are available, but don??t see it as likely that will make much improvement in matching up.
I have done a session 30 km (20 miles) away and came out 45 mm lower than the leveled elevation, not unusual for the fit of Geoid12B, so I don??t think the geoid could be so far off at this new session as to explain most of a foot discrepancy.
?ÿI broke it into 3 files of 2 hours each and got these differences from leveled: +305, +226, +252 mm (0.74 to 1.00 ft, giving a different average). Not great repeatability.
This mark is a RESET done by a local surveyor in 1980 and is rather far for a reset, almost a quarter mile from the original that was on a school building that was torn down. If I hypothesize a 1-foot blunder in his readings or calculations, I would be 17 mm lower than leveled (2-hour sessions 0 mm, 79 mm, and 53 mm lower than leveled).?ÿ Those values would seem reasonable for a poor site.?ÿ I still wouldn't want NGS to use my measurement to constrain their new geoid.
So the questions are:
- Does this environment explain such a huge difference, or did the guy doing the reset make a 1-ft blunder?
- Should I go ahead and share my solution when precise orbits are available, or does this environment invalidate any possible GPS solution?
?ÿ
?ÿ
Looks like a STINKER to me.
A good candidate for an OFFSET observation if I ever saw one.
Loyal
I would suggest if you want to find out out bad the results would be, then you need a control for the experiment.?ÿ For example a second receiver close with good sky observations, and check heights between the two with a level, or other suitable device.
A good candidate for an OFFSET observation if I ever saw one.
If only NGS would approve an offset procedure for GPSonBM that didn't involve an official reset and new PID.?ÿ
Since we are hoping for a cm at best I think they should allow one level turn to a temporary mark, measured from two setups.?ÿ
That would open up several sites I know of that would benefit the program, but are on/too close to a building, under trees, or even some too close to a RR.
A good candidate for an OFFSET observation if I ever saw one.
If only NGS would approve an offset procedure for GPSonBM that didn't involve an official reset and new PID.?ÿ
Since we are hoping for a cm at best I think they should allow one level turn to a temporary mark, measured from two setups.?ÿ
That would open up several sites I know of that would benefit the program, but are on/too close to a building, under trees, or even some too close to a RR.
I agree Bill, and feel your pain.
Keep up the good work though, and hope for better candidates in the future.
Loyal
if you haven't already done so, you may want o check the times series on the CORS stations OPUS is selecting., and possible pick a "better" site and see if that helps.
If you have CORS close, do a 10-15 minute static in clear sky nearby and level to the mark, then you would know if there is a 1' bust.
I have 2 suggestions, one add some extensions to get your antenna above the roof, and set a new mark in the open near by. ?ÿDo the best that you can leveling to it then observe on both marks at the same time. I would suspect that there was a bust in the reset but there is only one way to prove it.?ÿ
The mark is so close to the building you??re going to need the antena up above the roof line to get a direct observation. But I like the idea of setting a mark in the open and leveling between the two as a check on your original observation versus the published value.?ÿ
I'm not going to work hard enough to get an antenna stabilized 20 ft in the air.
I only have one receiver (this activity is a retirement hobby, more interesting to me than golf).?ÿ I don't see that simultaneous measurements are essential, as the level makes the comparison and any session over the mark seems imprecise.
I will plan to do the offset measurement with one receiver when I can get back there.?ÿ One turn would get me into a fairly open area.
Dave Doyle might have been able to find the reset data when he was at NGS, but I doubt anyone there now can or would take time to do it.?ÿ
I'm also considering a trip to the County Engineer's office to see if they have the reset data (field book) from 1980.?ÿ I think most or all Iowa counties just have an engr office now, no separate surveyor.?ÿ I'm afraid they might not even understand what I'm up to.
You are overthinking the need to stabilize the antenna. A little horizontal wavering is not going to affect the elevation, so get it above the roof eave. As long as this is no hard steady wind that keeps the pole leaning the horizontal position also gets meaned out. In looking at the CORS map you have a chance at getting 6 CORS OPUS-RS solutions so give it a shot. Even a 3 CORS OPUS-RS can give a better elevation than OPUS because of the software and the increase in observations used. OPUS-RS uses the C1/P1 and C2/P2 ranging observations (measured distances from satellite and antenna).
Send me you 6 hour RINEX file, I would like to look at your data.
The setup is on the North side of the municipal building, so you have thrown away at least 4-6 good satellites.
Also that mark is close enough to the building that it is possible that frost could readily affect a less then stabile mark.
After a closer look I see that that building has a metal roof, so I would prefer an antenna with a large ground plane to block multipath from the roof which may be overwhelming.
Paul in PA
Dropbox File (Private or Invalid)Even a 3 CORS OPUS-RS can give a better elevation than OPUS because of the software and the increase in observations used.
NGS hasn't come to that conclusion.?ÿ They still require a 4+ hour OPUS-S session for the GPSonBM program.
I created 4 1.5 hour files.
My OPUS solution used 92% of observations fixing 83% of ambiguities.
My first OPUS-RS aborted loading a CORS position, which is unusual because the next 3 OPUS-RS solutions used that CORS. I resubmitted excluding that CORS but it still aborted for the same reason. The next 2 OPUS-RS solutions each used 9 CORS?ÿ?ÿand were rather precise, 68% and 72% of observations used. The last OPUS-RS used only 6 CORS and was rather sloppy. For the 2 good solutions ortho elevations were 261.289m and 261.290m, so compare that.
My suggestion is to break it into 6 1 hour files, starting at 15:00 naming it F86R136o.18o, through 20:00 naming it F86R136u.18o and submit to OPUS-RS. You should at least get 4 rather good solutions in the middle. I use WordPad for editing RINEX files.
If you check your OPUS solutions you will see that NGS relables your input per the 24 hour RINEX naming convention with the 8th?ÿcharcter ranging from A to X for 00:00 to 23:00, the time of each first observation.
Hope this makes your OPUS work more fun and more useful.
Paul in PA?ÿ
OPUS-RS?ÿ 261.289m and 261.290m
That is 230 mm or 0.75 ft above the data sheet NAVD88 at 261.06.?ÿ I think your results show my 6-hour session (288 mm high) isn't meaningless, just sloppy, and there is a serious problem with this disk.
If there was a 1-ft bust, then the OPUS-RS would be 74 mm below the corrected data sheet.?ÿ That's a bit large compared to the mark 20 miles away, but certainly in the ballpark.
I still think I need an offset session as the "cleaner" method to demonstrate the problem to NGS rather than cherry-picking data out of the first run.
?ÿ
Any RESET mark should not be in the GPS on BM project IMO
that depends- we have resets from the 50's that were leveled through in the late 70's early 80's, but I would agree that many newer?ÿresets need to be viewed skeptically.