Community Forums


Traverse Adjustment  

Page 1 / 8

Posts: 33
20+ posts
Joined: 1 year ago

Need a little help and advise guys.  Its been years since I have done a traverse adjustment.(more like decades) most of our stuff is done with GPS or after 3-4 good turns checking in with no problems.  I have a site that has about 40 (extremely unbalanced) turns.  There are some with 20' back sites and 150' fore sites.   This was through a cave system so we had to stretch the limits of our hardware.  The traverse was ran with a Trimble s8 0.5 sec gun.  After I ran through the cave I come out on a known point and I miss it about 2' horizontally but flat vertical.  All of my distances where double checked through the traverse with none of them being off more than 0.05'.  I'm sure I'm dealing with an angular issue but my question to you guys is what is the best way to adjust this traverse?  Any advise?

46 Replies
Posts: 152
100+ posts
Joined: 8 years ago

Least Squares definitely and starnet if you can.  A first pass through a least squares adjustment will see if the 2' you are seeing will still be a problem overall.  What kind of accuracy are you trying to get to at the outer reaches of your control network?

1 Reply
Joined: 1 year ago

20+ posts
Posts: 33

Thanks DAN...thats what I was thinking but wanted to get some input before I proceeded.  Regarding the accuracy....its subject to whatever I feel is necessary.  I would like to get it within 0.25' spatially.  We are going to come through and scan the cave after the control and then run aerial LiDAR on top.  Should make for quite the map.

Posts: 5567
5,000+ posts
Joined: 9 years ago

I second the least squares recommendation, with reasonable standard error estimates for your conditions.  My gut tells me the gun was overkill for the job and the most critical issue is centering error.

If LS isn't available do an angle adjustment with the error distributed proportional to 1/bs + 1/fs at each point.

A centering error of maybe 0.01 ft on a 20 ft leg is a minute or two angle error and at a few points would easily accumulate 2 ft closure in a traverse of 3000 or 4000 ft.

After whatever adjustment your points are likely to be closer to the truth, but still probably significantly off.  LS will tell you what is likely.

Posts: 633
500+ posts
Joined: 6 years ago

I don't think this level of error is acceptable to adjust. It needs to be re-run. A point nodal prism for the shorter legs may help and also make sure autolock and or 360 prism is not used on the Trimble.

Norman Oklahoma
Posts: 4501
2,500+ posts
Joined: 4 years ago

I also use StarNet - I'm a big fan and proponent. I run virtually everything I do through it. That is more an effort to trap blunders rather than to adjust minor errors, but of course that also happens.  It is also a convenient place to fix point numbering issues, descriptors & measure ups, and so on.  I doubt that you have saved any time by not analyzing your work in this way. You have just shifted the work to other parts of your process.      

Page 1 / 8