It would interest me in this case considering the nature of the traverse conditions and the probable limits on redundancy.
It's not likely a 2' error. It's probably a 0.05' error in a short backsight propagated by a long traverse. The misclosure is meaningless until this data is simultaneously adjusted and analysed.?ÿ?ÿ
The worst errors are not zeroing on the BS and inputing wrong BS point numbers.
I always use tripod and tribrach and tilting prism sets for BS and FS and it really helps closure.
This is my goto traverse analysis proceedure
What bothers me is 0.05?? on distance. Even not being careful should get better than that. ?ÿI also agree short sights will cause problems. It??s in a cave what else can you do. Interesting job thanks for the pics.?ÿ
There is no good way to adjust a point to point traverse. If you had a point outside the cave, then entered the cave and came out at another point you should be able to have an azimuth in and an azimuth out. It is correct to hold both azimuths and adjust and a good adjustment can show you where the primary source of error is. To be reasonable you cannot hold for a maximum of 0.5"angular error as it will adjust all the good angles also.?ÿ I would suggest allowing 5".?ÿ
Are your traverse points on the floor or the ceiling? How precise were your plummet sights? Coming out flat on elevation and distance says to just turn the whole traverse, point to point and call it good until you find otherwise.
Paul in PA
Hold 5 seconds on an angle with a 20 ft leg? ?
Yes the 0.05 miss on a check back is concerning. Should be no more than 0.01, maybe 0.013 on the far outside.
Caves can have weird refraction too.
Bill93,
A minute allowable error might be more appropriate. But even at 5", if the adjustment throws the 5" into one specific angle you will know where he problem was. It is better to give the LS program some room to work in.
Paul in PA
It would interest me in this case considering the nature of the traverse conditions and the probable limits on redundancy.
It's not likely a 2' error. It's probably a 0.05' error in a short backsight propagated by a long traverse. The misclosure is meaningless until this data is simultaneously adjusted and analysed.?ÿ?ÿ
Yes but the misclosure has been the starting point for him to know he has a problem. I also question whether there is enough redundancy, the OP has not specified.
It also looks from the pictures that a 360 prism on a pole may have been used for the traverse. From reading other posts here it seems that traversing with a Trimble using Autolock and a 360 is a bad idea.
I suggest there may be a problem here with procedures that are unlikely to be fixed by a LS adjustment.
All-
Thanks for your input and ideas.?ÿ This work was all done with our daily driver along with a 360 prism pole.?ÿ The accuracy that I got out of it was good enough for what i needed it for. Looking back there are a few things that I should have done to get better results.?ÿ I knew what most of these things were going in but it was a test to see "how close" I could get without trying too hard.
Next up we are going to scan this cave with the new Leica 360 scanner.?ÿ Supposedly this scanner will tie to itself with auto registration.....Like doing a resection using 1000s of points. Then you adjust the cloud to the control.?ÿ Obviously I'll used the control outside the cave to hold and then see how it fits the adj Trav control.?ÿ The dealers have told me they did about 120 scans and closed on the original scan by 0.02'.?ÿ I'm super interested to see how this turns out and will report back to you all.?ÿ
I don't do scanning - yet - but I'm close to someone who does. I'm told that the autoregistration works best with well defined shapes. You may not get good results with cave walls alone.?ÿ
"It also looks from the pictures that a 360 prism on a pole may have been used for the traverse. From reading other posts here it seems that traversing with a Trimble using Autolock and a 360 is a bad idea."
Just for the record, I have about 5 years experience with a Trimble S6 and 360 prism. Our traverses always closed good. 1:20,000 to 1:40,000 were common. I don't think that was his issue. On short sights, though, sighting the center of a prism pole can be an issue. We kept our level bubbles well adjusted.
Precisely?ÿsighting the center of the prism pole on short sights can be difficult.
And just for the sake of covering all the bases:
You did convert your GPS control points from Grid coordinates to Ground coordinates. Right?
Ahhh, now that I understand exactly what you did, two feet doesn't sound out of the ordinary. I would suggest not using bipods or a 360 prisim under such adverse conditions. You say under normal conditions your closure is as low as 1:20,000. That may be accurate enough for what you need, but that's really low for modern equipment. When you put more stress on your procedures by the conditions in the cave that's going to go down significantly.
Traversing with tripods seems to be out of fashion, but it will greatly increase your precision.?ÿ?ÿWhen I traverse with my 1990's conventional gun my absolute bottom is 1:60,000, but if I switch to bipods I am down near what you are getting. I have traversed in similar environments as your cave. It took two tries, and painstaking effort, but we were able to get 1:30000. The problem was the inordinate amount of time it took. It sounds like it may not be worth it for what you are doing.
Ahhh, now that I understand exactly what you did, two feet doesn't sound out of the ordinary. I would suggest not using bipods or a 360 prisim under such adverse conditions. You say under normal conditions your closure is as low as 1:20,000. That may be accurate enough for what you need, but that's really low for modern equipment. When you put more stress on your procedures by the conditions in the cave that's going to go down significantly.
Traversing with tripods seems to be out of fashion, but it will greatly increase your precision.?ÿ?ÿWhen I traverse with my 1990's conventional gun my absolute bottom is 1:60,000, but if I switch to bipods I am down near what you are getting. I have traversed in similar environments as your cave. It took two tries, and painstaking effort, but we were able to get 1:30000. The problem was the inordinate amount of time it took. It sounds like it may not be worth it for what you are doing.
aliquot,
I think you have confused me for the original poster.
I use a 360 prism for control and boundary regularly. I acknowledge that there is some small loss of precision because of it. But other items, such as centering error, ppm corrections, and even tripod stability are much greater concerns.?ÿ ?ÿ?ÿ
My comments about sighting the center of the pole are irrelevant since he was using a robot.
I would bet money on his level bubble being out of adjustment.
Not very great test on my part, but I'm a new trimble robot user.?ÿ I tested our robot with tripods and tribrachs with?ÿregular prisms.?ÿ I carefully turned the angles 4 times with a 3 second Nikon gun.?ÿ Then let the robot turn the same set of angles 4 times all by itself.?ÿ Got within the same angle by 1 second.?ÿ Then with the 360 prism on the foresight. I got about 9 seconds different.?ÿ Not very exact test as might backsights and foresights were only about 150 feet.
Given what has been posted so far, my initial go with Starnet would be fix the known control coords ?ÿYXZ- ?ÿangular error of 180 seconds zenith error 20 seconds, inst. centering error 0.005
tgt centering error 0.05
height error 0.05
and see what the residuals were on my angles- then start looking from there.
cloud to cloud registration is going to be difficult in this environment. You need well defined planes in each x,y,z axis for C2C to work well, which is why it works really nicely for interior buildings or urban environments. You may be able to get some things to match up, but the 2' error you're seeing in your traverse will be nothing compared to how your scans don't match up. I would target the hell out of the cave for the registration. it would be a shame to waste all of that time scanning and be left with 100 scans that don't tie together and are useless.
?ÿ
you dont think an RTC 360 can handle it?
