The PLS who sat in the chair before me, allowed the field crew I inherited to set a pair of 18? #5 rebar control points (CPs) with multiple GNSS occupations (NC's Network-RTK) separated in time appropriately. Field crew was then allowed to setup their S7 robotic total station on one CP, backsight the other, then proceed to locate boundary corners, buildings, valves, or anything else. Often times, the field crew was able to get everything needed from just this one setup getting high-fives from my predecessor.?ÿ They did not set additional stations and close.?ÿ They did not, and were not required to, setup on the second control point and tie to something located from the first.?ÿ Other than a direct and reverse shot from one CP to the other, no redundancy.?ÿ
North Carolina, like many states, has tiered accuracy requirements for Urban 1:10K, Suburban 1-7.5K, and Farm 1:5K. The requirement for an Urban survey, in its entirety, reads as follows:
?ÿ
Urban Land Surveys (Class A). Urban surveys include lands that normally lie within a town or city. For Class A boundary surveys in North Carolina, the angular error of closure shall not exceed 20 seconds times the square root of the number of angles turned. The ratio of precision shall not exceed an error of closure of one foot per 10,000 feet of perimeter of the parcel of land (1:10,000). When using positional accuracy standards for Class A control and boundary surveys, neither axis of the 95 percent confidence level error ellipse for any control point or property corner shall exceed 0.10 feet or 0.030 meters plus 50 ppm measured relative to the position(s) of the horizontal control points or property corners used and referenced on the survey.
?ÿ
My view is that my predecessor met the standards on the control pairs but not any boundary corners located using his method.?ÿ I'm now stuck trying to explain to the field crew how he's wrong and I'm right. Not an easy task as my predecessor was president of a local survey chapter and was quite proud of his procedure. I'm essentially asking the crew to setup and close in a day-and-age when field crews almost never have a misclosure.?ÿ Instead of being proud of tight closure, they view it as a waste of time, "Oh, what a surprise, EOC=0.008".?ÿ I don't blame them for their attitude but I do want to improve it.
I'm not a fan of telling adults to do something just because I say to do it, although this is essentially what I did when I told them that there is no scenario involving a total station where we don't close out the traverse (excepting route surveys with multiple GNSS pairs begging middle and end.?ÿ I'm new to Carlson SurvNet, but I thought of showing them that even though the two CPs met NC's 2 sigma accuracy standards, the boundary monuments did not. I'm trying to imagine how I would input it into Star*Net and I'm not sure that there's sufficient redundancy to run it.
Am I wrong to view this as unacceptable for boundary or engineering topo?
Any ideas for a simple approach to help them understand the positional uncertainty resulting from this method?
Thank you
Oh the private sector...how I do not miss thee.
Focus on educating them on redundancy, independent and dependent measurements. Explain to them that in order to meet the requirements set forth by the board one must measure more than once and it is even better if said measurements are from two independent sources.?ÿ
Baby step one would be to shoot everything from CP1, then locate all points again from CP2 and then just for $hits and Giggles make them set a 3rd CP to use at both set-ups as a common check point. As they start to collect the redundant shots they will start to see the errors. And remember - store another, do not merge at the controller.
to set a pair of 18? #5 rebar control points (CPs) with multiple GNSS occupations (NC's Network-RTK) separated in time appropriately. Field crew was then allowed to setup their S7 robotic total station on one CP, backsight the other,
You did not say how far apart these cp 5/8 rebar are.?ÿ
It would affect my practical mind to know...
Am I wrong to view this as unacceptable for boundary or engineering topo?
You're not wrong, you're just in the minority. WA has similar statutory language. It's not complicated, but there are plenty of licensees who are completely clueless about what error ellipses are, how to compute them, how to run and QC network adjustments, and how to analyze relative positional accuracy.
Simply put, if one doesn't have enough redundancy to compute error ellipses, standards cannot be met because one has no idea whether standards were met. I would be explaining to the field crew that I'm not about to gamble with my license, and I'm counting on them to do their due diligence so that I can prove I met the bare minimum of standards.
Also, I would remind them that a backsight check is not a closure, and a single setup is not a traverse, especially not according to statutory definitions.
And in the end, getting that redundancy and independent observations in this scenario is stupid simple. Just break setup at instrument and backsight, perform another station setup, turn another set of angles to the property corners, and done. That's the bare minimum to compute ellipses, and if a crew is too lazy to do that, then I'm not sure I trust them to set my cap in the right spot...
?ÿ
(As a side note, I absolutely love TBC's Allowable Relative Tolerance report. Once you have run your network adjustment, input the constant & scalar values, select which points you want to compare, and it spits out a report showing which ones meet/don't meet standards.)
Just some comments.
This is nearly un enforceable. Because probably the State BOR does not even know how to do this. And, if a vote were held, 90% of the surveyors would also not know if they had met standards. They would slip by, on a % compliance argument.
N
Well since statutes regulate licensed practice, we can take our pick of adjective.
Either way, with respect to accuracy standards, the word "shall" means "mandatory".
statutory
1: of or relating to statutes
2: enacted, created, or regulated by statute
regulatory
?ÿ
Respectfully, I disagree with your interpretation of this rule. If you derive two control points by GNSS, and locate points terrestrially from that pair of points using a total station, you don't have a traverse. I believe the rule regarding 20 seconds per turn and 1:10,000 closure ratio is specific to surveyors who are traversing and do not wish to perform least squares analysis on their survey. As long as their closed traverse is within those specifications, it meets the standard. It sounds like a grandfather clause for the "old" way of surveying a boundary.
?ÿ
The rule also gives an option for positional tolerance at (2 sigma) of 0.10' + 50ppm. This can be determined without repeat observations using proper error estimates for random error sources (occupation, sighting, measuring, etc.) in a least squares adjustment.
?ÿ
?ÿ
Other than a direct and reverse shot from one CP to the other, no redundancy.?ÿ
set a pair of 18? #5 rebar control points (CPs) with multiple GNSS occupations (NC's Network-RTK) separated in time appropriately.
How does having multiple GNSS occupations = no redundancy??ÿ But who cares where the second control point is anyway, as long as you know the bearing between the two, if you're never going to occupy it??ÿ It's just like sighting a distant tower that you have a coordinate on.
The field procedure described is very similar to my practice, except that for boundary corners I turn a D/R set.?ÿ Example:
1.?ÿ Get GNSS positions on 2 control points.
2.?ÿ Set total station on CP1, shoot D/R to CP2, zero the gun.
3.?ÿ Shoot D/R to boundary point 1 (BP1).
4.?ÿ Backsight BP1 (D/R), shoot D/R to BP2.
5.?ÿ Backsight BP2, (D/R) shoot D/R to BP3.
6.?ÿ Repeat for all boundary points.
7.?ÿ Shoot other features (e.g. building corners) with a single D sideshot.?ÿ
8.?ÿ Dump all data into Star*Net and adjust.
P.S.?ÿ All of the above assumes that the boundary and other features are to be located relative to some local orientation, like one of the monumented lines.?ÿ If the survey has to be oriented to a published geodetic framework (e.g. SPC), then at least 1 other GNSS-tied control point is needed.
This is nearly un enforceable. Because probably the State BOR does not even know how to do this. And, if a vote were held, 90% of the surveyors would also not know if they had met standards. They would slip by, on a % compliance argument.
Well, considering that the BOR likely had a hand in writing and/or approving the language, I'd say it's unlikely the BOR "does not know how to do this". I can pull up network adjustment and positional accuracy reports for my boundary work.
How is applying basic statistical testing to prove a minimum standard difficult? Statistics is a 200 level class, Spatial Data Adjustments are 200-300 level courses. Hardly graduate-level material. If a 20-year-old can learn to apply these concepts, surely licensed professionals can do the same.
I'm unclear on what it is that you would want them to do. Set up on another point and tie the boundary points a 2nd time??ÿ Tie them a 2nd time with a different backsight?
I'm a proponent of setting a minimum of 3 control points.?ÿ Mostly to have a point to 3rd point check to, partly to have a backup plan in case one of them gets destroyed.?ÿ Also, I use resection a lot and I require that each resected position reference 3 points.?ÿ So always a minimum 3 controls.?ÿ
Also, I always tie all control and boundary points with doubled angles, usually 2 sets.?ÿ?ÿ
The field procedure described is very similar to my practice
I should note that it's a rare job in which I can tie all boundary control from a single setup.?ÿ It happens, but not often.?ÿ Most jobs require ties from multiple locations, and I always get at least one or two redundant ties from each setup in the process.
Every surveyor has developed their own methods for reassuring themselves that they continue to meet regulatory standards. Meeting the standards is what is required. The method of doing that is left to each individual surveyor.
The rule also gives an option for positional tolerance at (2 sigma) of 0.10' + 50ppm. This can be determined without repeat observations using proper error estimates for random error sources (occupation, sighting, measuring, etc.) in a least squares adjustment.
It's important to note that, in this particular case (no redundancy), there is no adjustment being performed (number of unknowns = number of measurements), and error propagation is totally dependent on a priori estimates, with no post-adjustment statistics to review and analyze for outlier/blunder trapping.
It doesn't mean that it's necessarily wrong. I think the letter of the law can be satisfied. But the operator had better have excellent a priori error estimates. The short amount of time it takes to perform at least one or two repeat observations is worth it for me in order to check my work, and more defensible if that work should ever be challenged.
The rule also gives an option for positional tolerance at (2 sigma) of 0.10' + 50ppm. This can be determined without repeat observations using proper error estimates for random error sources (occupation, sighting, measuring, etc.) in a least squares adjustment.
It can be predicted that in the absence of a blunder, bubble that has been bumped, EDM reading a reflection, etc that the measurement will be within the tolerance, but it can't be determined that no blunder or unexpected error source has influenced this measurement without some redundancy.
