Leveling Adjustment

A level circuit should not use the same turning points in both directions. It should start and end at the same Bench Mark. Preferably using two Bench Marks with one at each end. If by "stations" you mean turning points, that is not the recommended procedure. The return trip should be along a different path.

Also I always do the two peg test before each level circuit. That's just me.

To answer your question, I think both accumulated and compensating errors are possible. The two peg test would answer the question definitively.

?ÿ

The addition or cancellation of errors depends on the instrument locations. If every setup exactly balances the backsight and foresight distances?ÿ the errors cancel at each setup.

If you have unequal sights, and use the same TPs and same setup positions, you will get good closure but have accumulated but undetected error that matches in each direction's run.

I see no problem with using the same TPs in each direction, and doing so makes blunder detection easier.

But if you do reuse them, balance the sights or randomize them by not picking the same setups.

Posted by: @bill93

I see no problem with using the same TPs in each direction, and doing so makes blunder detection easier.

I agree. I have always heard to not use the same turning points but that is for someone sitting in the office telling his crew to go out and do the work. It is harder to fake and easier to prove if your crew uses different turning points. Since I am the sole person that I have to satisfy, I do it my way.

James

If you were running a level circuit between two BM's 6 miles apart and establishing elevations on 39 points in between them then you had better use the same turning points (the new BM's, D'oh!) running forward and back.

?ÿ

Using the same fixed turning points has benefits, in blunder detection and correction. If something goes wrong, you may only need to run 1 leg 1 direction to remedy the problem. Depending on your adjustment- they can cause some heartburn. Using digital levels and a least squared adjustment program, using the same turns, becomes another good analysis tool.

Posted by: @rankin_file

Using the same fixed turning points has benefits, in blunder detection and correction. If something goes wrong, you may only need to run 1 leg 1 direction to remedy the problem. Depending on your adjustment- they can cause some heartburn. Using digital levels and a least squared adjustment program, using the same turns, becomes another good analysis tool.

Absolutely. Not to mention if you use the same TPs out and back it cuts down on the number of unknowns in your analysis and therefore tightens up your accuracy on the run.

I don't know where the idea comes from, but I have worked with multiple folks who insisted that a "level loop" had to look like a loop on the ground; that is, going out and back on the same TPs does not look like a closed loop when viewed from above and is therefore somehow "wrong" and "not a loop". I never heard exactly why this was so.

?ÿ

The idea of using different turning points in?ÿ each of the two directions of the leveling circuit comes from a definition published by the American Congress of Surveying and Mapping.

The reason being that using the same turning points might lead to cheating. The other benefit being differing atmospheric conditions as the circuit traverse along a separate path.

I was fond of using the same turning points in my work for redundancy until I read the precise description. If you are using a turtle I don't see how you could repeat on the same turning points anyway.

Posted by: @mikesurveying12

The bubble is is something that I sometimes look at, but being honest, only from time to time.

Danger!

Posted by: @not-my-real-name

The reason being that using the same turning points might lead to cheating. The other benefit being differing atmospheric conditions as the circuit traverse along a separate path.

I would hope all professionals and their crews were above that kind of cheating.?ÿ And if someone would rather sit in the truck and dry-lab the return trip they could still do it on a loop.?ÿ That just takes more mental (not physical) effort.

I don't buy different atmospheric conditions resulting from using a loop unless you are leveling around a whole county.?ÿ There would be change in conditions due to time in either case..

@bill93

If you go back through your same points then what you really have is a series of daisy-chained small loops. Fundamentally, that's fine. Maybe even better. But if you try to apply a simple overall misclosure adjustment factor you are going to end up with 2 different elevations on all those turning points.  Even that is rarely a big problem, but nobody wants to see it. 

In an area with a lot of relief, such as PDX, reusing old turning points isn't always convenient.

I do like to repeat a point every third or forth turn so that errors can be isolated.  

Oklahoma's DOT insists that you show the levelling results as 2 different one way runs, and then average the results of the 2. To that end the crews would run the levels in one direction, then return to the beginning and run them again. Which involved a lot of walking, or a lot of shuffling of trucks, usually both. It took me some time to convince them that they could run out for an hour or so, then back to the start, and thus eliminate the truck shuffling. I could arrange the data in a spreadsheet to the DOT's satisfaction. I never did convince all of them. And I'm sure they went right back to their old ways before I crossed the Kansas line.       

First Order leveling requires the forward run to be done, say, in the morning and the backward run to be in the afternoon, or vice versa.

To accomplish this is easy if you drop off a vehicle four or five miles from the starting bench, run levels to it and then go pick up the first vehicle at the start.?ÿ Then, tomorrow, reverse the procedure.?ÿ The hardest thing to do is know where to drop off the end-of-run vehicle.?ÿ Four or five miles a day is not being overworked.

@norman-oklahoma

In PDX...use a trig loop. I know this is sacrilege, but it is also a much better way to transfer elevations across areas with a lot of relief. Just because it isn't how the ancient Egyptians did it doesn't make it wrong.

Set up gun, hit bench (+) to a fixed rod (F&R), hit TP (-) with fixed rod (F&R). Leap frog. Just like your automatic level, but better. Use SurvCE, and it will even run an adjustment. 

@norman-oklahoma

DOT

"I don't know where the idea comes from..."

Well, I guess we know where ideas come and die, and I think most DOT's and government agencies have the same issue. Interia. But, it isn't all bad, it provides consistency. But it doesn't help innovation.

@dmyhill

I've done work for the Oregon DOT and thought that the institutional inertia was strong. But it is nothing compared to that in OK. A correspondent to this board from Norman, OK, who shall not be named at this time, has written of it, some of it on the board but much of it in personal communication. The OK DOT survey standards are well tailored to the technology of the 1960's.

I have had occasion to use the TS to transfer elevations as you describe. It works well.?ÿ?ÿ

The point of a longer level run is normally to establish bench marks along it. I've always run back through them. The turning points may be a turtle or just random features that are stable so I don't use the same turning points but every 1500' or so a bench is occupied both directions.?ÿ

If leveling is an important part of your operations I would HIGHLY recommend trying to attend one of the free leveling training classes put on by NGS at their Corbin, VA facility.?ÿ Next one is planned for Oct (hopefully) -- https://www.ngs.noaa.gov/corbin/class_description/leveling_0420.shtml .

@norman-oklahoma

Just curious, do the DOT standards/specifications in OR and OK differ from those of the FGCS here: https://www.ngs.noaa.gov/FGCS/tech_pub/Fgcsvert.v41.specs.pdf

I certainly find the OP’s reference to only occasionally referring to the level bubble to be an innovation.

I did a great deal of leveling in my career. Never liked it. Doing it right is hard. With the attrition of the network of monuments obtaining the required checks has been increasingly difficult. 

BTW, in addition to the specifications linked above, there are a number of presentations on the NGS describing how to perform geodetic leveling. I used them to prepare this 2007 lecture: http://geodesyattamucc.pbworks.com/f/Lecture24_2K7PreciseLeveling.ppt

I looked at the NGS Presentation Library to find them but did not find them after a cursory look. See: https://geodesy.noaa.gov/web/science_edu/presentations_library/

Cheers,

DMM

Posted by: @mightymoe

The point of a longer level run is normally to establish bench marks along it. I've always run back through them. The turning points may be a turtle or just random features that are stable so I don't use the same turning points but every 1500' or so a bench is occupied both directions.?ÿ

Yep, that's how it should be done. Establish TBMs every mile or so but don't reoccupy turning points because it introduces bias.

Spirit level surveying for elevation is the gold standard in limited locals; GNSS is a poor second and allegedly trig level surveying is equivalent but I've seen bad results due to unaccounted for diurnal atmpospherics

@geeoddmike

No, nothing that technical .... This is project level benchmarking.?ÿ

It has been 6 years since I was last in OK and I see that there has been some slight movement since I was last there. But they still have a long way to go.