I have five points that have GPS on them.
I have three with 3 hours and 2 with 12 hours of data that were sent in to OPUS.
I have a network with HARNS tied to this survey in which a network was adjusted.
These results match within 0.10' feet of the OPUS in vertical and 0.02'feet in the horizontal. Typical for OPUS verticals using only the geoid.
The delta's in vertical match almost exactly, so it's pretty much a geoid error.
A level line was run through all the points and looped back onto itself. It closed nearly perfect as most level lines do.
It has elevations some 0.5' different than the OPUS or HARN derived elevations on several of the points.
Crew set up on one and then did RTK to all the points and matched the HARN adjustment at each point in the 0.01-0.03 foot range.
I'm confident that I have three independent checks using GPS (OPUS, RTK and Network) but the level line does not match.
Suggestions?
Deral Paulk, PLS OK
And no, this is someone that e-mailed me his results and not one of my projects.
He said he borrowed the level from another surveyor, but did not peg it before using it.
Modified By Deral Paulk on 7/12/2005 at 6:15 PM
Without being able to review the level notes and understand how the run was made, I would trust GPS within it's tolerances.
Was it several loops? One continuous loop? Did he hit every point at least twice? Some people fudge level loops, because they cannot avoid blunders, figure out where they were did occur and properly correct them.
Paul in Pa
TM...
If he fairly carefully balanced the distances on this back sights and fore sights, the level would be fairly close, pegged or not...
Randy Green
any idea as to the rod used in the conventional leveling? I remembering using a fiberglass rod (only extended to the 12 foot position) in a loop several years ago that the hole where the spring pin worked had been somewhat "worn". The loop results were very deceiving as you might expect.
I think I would want to see the level, and verify proper working of the compensator as well. I have a Topcon ATF2 from 1984 that was "top shelf" most of the time. Had the compensator worked on by one of the "big supply bouses" the "monkeyed" with it to the point it sticks most of the time... at least that was the case several years ago when we just absolutely had to quit depending on it. Also had a Topcon TL6DE theodolite "monkeyed with" by the same "big supplier" that came back with a compensator that insisted on sticking, after being repaired for a faulty compensator....another story.
If it were me, I would have to question the tools of the other guy before I threw out that much GPS redundancy.
jd
They just borrowed a level from another person.
I've had ones that were bad that stick also. You can't peg them at all.
The compensator is broke or damaged.
They are going to redo the level line and I'm sure that they will find the error in the levels.
I've got to much information (redundant) to think otherwise.
Usually levels are about the easiest thing to run with great precision. This is the first time that I've questioned a closed loop using only GPS as my evidence.
I'll post when he redo's the levels and reports the results.
tm
"Monkey'd with".....too bad they aren't "so simple a cave man could fix 'em" so we could at least throw Darwin into the fray....
Dangerous Dave
I did NOT say the name of the "big supplier"...so don't go making any assumptions
:-)
I can imagine places where the geoid model might fall apart but it seems odd that OPUS and the local network match within 0.1 foot while the level loop differs considerably from that.
There are two projects that come to mind where our GPS work helped the level crew find and fix blunders.
Has anybody used OPUS on a first or second order bench mark in the neighborhood of this project?
I don't know about trusting GPS OVER conventional, but, I use my GPS for 2 things usually.
To get ON THE GRID and to have an INDEPENDENT CHECK of my conventional work.
There are times when I use the GPS to actually go places where it's very hard to survey and it's great for those situations, but I use the GPS, evidently, a little differently than most people.
Maybe that's why, my BAD setups aren't so bad . . . I verify them anyway.
In my conventional verifications, I see some intersting things happen to my data when I "play" with it, such as rinexing the data, etc..
I would have looked at mark to mark portions of the loop to see changes in vertical between stations instead of his loop closure. If there were consistent differences or two compensating blunders..Did I understand right that the loop went through all 5 points?
Couldn't you check half foot differences ubing vertical angles.
Has anybody used OPUS on a first or second order bench mark in the neighborhood of this project?
Below are the OPUS results for three 5-hour sessions on G 1200 (PID JS0755), one of the constraining benchmarks for a height mod project I'm working on:
SOFTWARE: page5 0411.19 master13.pl START: 2005/07/06 14:58:00
EPHEMERIS: igr13303.eph [rapid] STOP: 2005/07/06 20:01:00
NAV FILE: brdc1870.05n OBS USED: 10861 / 10927 : 99%
ANT NAME: TRM33429.00+GP # FIXED AMB: 39 / 42 : 93%
ARP HEIGHT: 2.000 OVERALL RMS: 0.013(m)
EL HGT: 47.399(m) 0.081(m)
ORTHO HGT: 77.280(m) 0.085(m) [Geoid03 NAVD88]
SOFTWARE: page5 0411.19 master10.pl START: 2005/07/07 18:55:00
EPHEMERIS: igr13304.eph [rapid] STOP: 2005/07/07 23:59:30
NAV FILE: brdc1880.05n OBS USED: 8075 / 8220 : 98%
ANT NAME: TRM33429.00+GP FIXED AMB: 35 / 37 : 95%
ARP HEIGHT: 2.000 OVERALL RMS: 0.014(m)
EL HGT: 47.368(m) 0.101(m)
ORTHO HGT: 77.249(m) 0.104(m) [Geoid03 NAVD88]
SOFTWARE: page5 0411.19 master19.pl START: 2005/07/11 15:11:00
EPHEMERIS: igr13311.eph [rapid] STOP: 2005/07/11 20:04:30
NAV FILE: brdc1920.05n OBS USED: 10752 / 11005 : 98%
ANT NAME: TRM33429.00+GP # FIXED AMB: 47 / 48 : 98%
ARP HEIGHT: 2.000 OVERALL RMS: 0.014(m)
EL HGT: 47.395(m) 0.079(m)
ORTHO HGT: 77.276(m) 0.083(m) [Geoid03 NAVD88]
JS0755 ***********************************************************************
JS0755 HT_MOD - This is a Height Modernization Survey Station.
JS0755 FBN - This is a Federal Base Network Control Station.
JS0755 DESIGNATION - G 1200
JS0755 PID - JS0755
JS0755 STATE/COUNTY- CA/PLACER
JS0755 USGS QUAD - ROCKLIN (1981)
JS0755
JS0755 *CURRENT SURVEY CONTROL
JS0755 ___________________________________________________________________
JS0755* NAD 83(1998)- 38 47 09.87346(N) 121 14 32.09509(W) ADJUSTED
JS0755* NAVD 88 - 77.38 (meters) 253.9 (feet) GPS OBS
JS0755 ___________________________________________________________________
JS0755 EPOCH DATE - 2002.53
JS0755 ELLIP HEIGHT- 47.44 (meters) (02/03/03) GPS OBS
JS0755 GEOID HEIGHT- -29.88 (meters) GEOID03
JS0755
JS0755 HORZ ORDER - B
JS0755 VERT ORDER - FIRST CLASS I (See Below)
JS0755 ELLP ORDER - FOURTH CLASS I
JS0755
JS0755.The orthometric height was determined by GPS observations and a
JS0755.high-resolution geoid model using precise GPS observation and
JS0755.processing techniques. It supersedes the leveled height previously
JS0755.determined for this station.
JS0755.The vertical order pertains to the first NAVD 88 superseded value.
JS0755
JS0755 SUPERSEDED SURVEY CONTROL
JS0755
JS0755 ELLIP H (05/14/98) 47.50 (m) GP(1997.30) 3 1
JS0755 NAVD 88 (06/15/91) 77.350 (m) 253.77 (f) ADJUSTED 1 1
I noticed that the ellip height and the geoid height doesn't equal ortho height.
The 0.1m between the observed and published does not surprise me.
How close were the lat and long?
Paul in PA
Modified By Lawrence Paul Lopresti on 7/14/2005 at 9:07 AM
I thought it was interesting that the published the GPS derived orthometric height but referred to it as a first order bench mark based on superseded differential levels.
But the point is that this makes it easy to see how well you can match a NAVD88 elevation using OPUS and the Geoid03 model at that point.
That is my idea of how to answer the question, "How good is it, really?"
I am never surprised that the NAVD88 published elevation does not match the elevation computed from the ellipsoid height and the geoid height. I can only imagine that it illustrates the same thing that many of us do when we are working on our own projects. It gets constrained once for the ellipsoid heights then the adjustment process begins again when we start fitting the project to our bench marks.
I left out an interesting bit of information: The 1991 leveled height of G 1200 was used to constrain GEOID 03 and (I think) the 2002 GPS project that was used to calculate the current ortho height.
I'm still trying to figure out what happened...
How close were the lat and long?
Well within noise levels; around 1cm, as I recall.
They reran the level lines with another level and found the error to be in the first level line even though it closed flat. The new run matched all the GPS points within 0.03'(feet)..
Weird stuff. I've always thought that levels were about as simple as they come and were almost impossible to screw up if you read the rod twice (normal and inverted) on each turn and if your level was pegged.
This is the first time that I have used GPS to disprove a standard level line.
I was a bit nervous to suggest he re-run the entire level loop at first just based on my checks using his GPS results. Nervous because of the extra day required for field work.
He's contacted the person he borrowed the level from (notified him)and it must have a sticking pendelum or such. The near perfect closure the first time must have been a fluke of compensating errors which balances out over the 6 mile run.
Murphy's law at work.
Deral Paulk, PLS OK.
Modified By Deral Paulk on 7/18/2005 at 2:51 PM