-
Traverse network – leveling instrument after change of station
Dear all,
I have a question regarding leveling (making horizontal) of the Total Station instrument in traverse network.
Is it necessary to level instrument after changing the station – means when I transfer SNLL121 to the new station and instead of this I put TS16? Is it good practice to leave bubble compensation on or it should be off? I think I read somewhere I shouldn’t level instrument and compensator should be off – in this way we preserve the same conditions and minimizing errors?
We are using the following equipment:
- Leica TS16 1″
- Traverse kit Leica SNLL121 - http://leicatotalstation.org/tag/snll121/
Thank you in advance
Best regards
-
33 Replies
-
Compensator should be on and the instrument should be level before making measurements. There is a school of thought to not recenter the instrument over the point if it has had small apparent departure from the point, this approach is called “forced centering” but does extend to levelling of the instrument.
The only time to turn off the compensator is when you are on an unstable surface such as the deck of a boat, heavily swaying structure, top of a hay pile etc.
-
Compensator should be on and the instrument should be level before making measurements. There is a school of thought to not recenter the instrument over the point if it has had small apparent departure from the point, this approach is called “forced centering” but does extend to levelling of the instrument.
The only time to turn off the compensator is when you are on an unstable surface such as the deck of a boat, heavily swaying structure, top of a hay pile etc.
TNX for a quick answer. If I understood correctly – even if I level SNLL tribrach and prism, when I move instrument there I should correct level (always there is a difference more than I had when I level SNLL kit – approx 20 / 30 ” after instrument transfer)
TNX once more.
Best regards
Bojan
-
I will level the instrument, compensator on, but never move or recenter. You are traversing through the center of the tribrach some called forced centering as opposed to the center of the point on the ground.
If you look through the optical plummet and discover that you are considerably off center from the point on the ground (you should not be) I know of some folks who have carefully adjusted the point to be coincident with the plummet.
-
Compensator should be on and the instrument should be level before making measurements. There is a school of thought to not recenter the instrument over the point if it has had small apparent departure from the point, this approach is called “forced centering” but does extend to levelling of the instrument.
The only time to turn off the compensator is when you are on an unstable surface such as the deck of a boat, heavily swaying structure, top of a hay pile etc.
TNX for a quick answer. If I understood correctly – even if I level SNLL tribrach and prism, when I move instrument there I should correct level (always there is a difference more than I had when I level SNLL kit – approx 20 / 30 ” after instrument transfer)
TNX once more.
Best regards
Bojan
I have the same opinion but know some Surveyors telling me no need because if we measure in two faces it will be compensated…
TNX for reply
BR
Bojan
-
A tripod is not a fixed pedestal, it is a multi faceted mechanical device with adjustable parts. In fixed centering you are assuming the best condition and that you had not made a mistake, which may not be true. The typical variation is that the clamps are not tight and/or that the leg tips are settling. The best you can assume is that the movement was slow and steady and that at the time of observing that location of the tribrach was more likely closer to where it should have been and moved mostly in the interim. Therefor you must relevel and recenter so that you are getting the best observations possible at the position you are at from the position you were to the position where you are going. Not doing so is a case of hoping that gross errors cancel out. Please note I said relevel and recenter which are done separately. After releveling the instrument you observe your ground point and note and record any visual differences then take a backsight distance and record it. Compare it to your previous foresight distance. Now you recenter the instrument, take a backsight distance, record and evaluate any differences. You now have three probable distance for that leg.
When the traverse is complete you re-evaluate the differences in the bad leg before any other adjustment. First what is your raw error? Can it be reduced by using a different or meaned traverse distance? Once any obvious distance errors are compensated one can check for angular errors to those questionable traverse points. Having done both the you can then expect your adjustment to get you closer to the Truth. In general an averaged leg distance getting you halfway from perfect is the better solution.
Paul in PA
Bojan, measuring in two faces compensates for instrument errors not setup errors.
-
Experiment with it yourself. Turn a set of angles without re-levelling. Then fine level and do it again (just relevel, not re-center) without . I expect that you will find that, as long as you are within the instruments compensation range, the angles will be the same.
The question is how well centered over the point you are. After re-levelling you may appear to no long be over the point. If not, perhaps the thing to be doing is adjusting your tribrach. Then you won’t have to deal with this problem.
-
I always relevel and center back over the point. A couple of reasons, the tripod may have moved slightly, and the level bubble on the gun is often more accurate then the one on the tribrach.
-
A tripod is not a fixed pedestal, it is a multi faceted mechanical device with adjustable parts. In fixed centering you are assuming the best condition and that you had not made a mistake, which may not be true. The typical variation is that the clamps are not tight and/or that the leg tips are settling. The best you can assume is that the movement was slow and steady and that at the time of observing that location of the tribrach was more likely closer to where it should have been and moved mostly in the interim. Therefor you must relevel and recenter so that you are getting the best observations possible at the position you are at from the position you were to the position where you are going. Not doing so is a case of hoping that gross errors cancel out. Please note I said relevel and recenter which are done separately. After releveling the instrument you observe your ground point and note and record any visual differences then take a backsight distance and record it. Compare it to your previous foresight distance. Now you recenter the instrument, take a backsight distance, record and evaluate any differences. You now have three probable distance for that leg.
When the traverse is complete you re-evaluate the differences in the bad leg before any other adjustment. First what is your raw error? Can it be reduced by using a different or meaned traverse distance? Once any obvious distance errors are compensated one can check for angular errors to those questionable traverse points. Having done both the you can then expect your adjustment to get you closer to the Truth. In general an averaged leg distance getting you halfway from perfect is the better solution.
Paul in PA
Bojan, measuring in two faces compensates for instrument errors not setup errors.
Thank you for the detailed description… really appreciate
For distances I don’t have a big concern – we tried traversing and distances are never more than 1 mm difference (always under) so in my traverse, I will have a probably small adjustment for the distance of around 3 – 4 km traverse (distance between points approx 100 m). For angles, I have bit concern because we tried and I found our ability to measure an angle is around 20 – 30 seconds on 100 m – means up to 1.5 to 2 cm error in angle and if I add let’s say on each station this error in 3 – 4 km traverse I can get up to 50 – 70 cm at the end. Is my thinking right…?
I’m not sure is it because we didn’t level instrument after changing the station or because it’s not possible to orient better on that distance (100m)…?
One more thing, if we can’t make better angle measurements on that distance – what if I use Trimble SX10 that has I think have more zoom comparing to TS16…?
-
Experiment with it yourself. Turn a set of angles without re-levelling. Then fine level and do it again (just relevel, not re-center) without . I expect that you will find that, as long as you are within the instruments compensation range, the angles will be the same.
The question is how well centered over the point you are. After re-levelling you may appear to no long be over the point. If not, perhaps the thing to be doing is adjusting your tribrach. Then you won’t have to deal with this problem.
Hi and thanks for the reply.
For sure we will test and I’m here to gain more knowledge because I want to master this subject where I’m really week, to be honest… If this is a case – with or without leveling angles will be the same then probably we getting errors because we can’t better measure the angle at 100 m distance (distance between points in traverse network)… for sure I must check more and test…
Thanks once again
-
I always relevel and center back over the point. A couple of reasons, the tripod may have moved slightly, and the level bubble on the gun is often more accurate then the one on the tribrach.
I have a similar opinion but still, I wish to check with experienced Surveyors…
TNX For comment
-
Guys,
any good practical literature on this subject and traverse adjustment…
TNX
-
In developing networks underground we traverse through pedestals or plates with fixed bolts that the tribrach is threaded onto, after wrapping a set of angles to “break” the set up the instrument is removed from the tribrach and the bottom is rotated 120 degrees before putting it back on the tribrach to wrap another set.
Even using top of the line instrumentation that is well cared, on a fixed plate the level will be different when the gun is put back on the tribrach, so we level it up and wrap another, now if the level were bad at the end of a set that would be an alarm to investigate the set up. If you never fine leveled it in the first place then you would not know.
It all depends on what you are doing anyway, for a lot of work you could very well get by without fine leveling at all once you are in the compensated range, but you are already looking at it and it is hardly an effort to turn a couple foot screws.
-
I liked Plane and Geodetic Surveying, second edition, by Aylmer Johnson.
The book is easy to follow and only the size of a novel, which is important if I am the one reading it.
-
The question is are you locating the tribrach on the tripod, or the monument under the tripod. Your closure will be better if you leap-frog from tribrach to tribrach, your traverse on the monuments will be better if you change tribrachs, rotate them, relevel them for each measurement set. Introducing redundancy to the measurements decreases the accuracy and increases the validity of the traverse. Much like a GPS antenna “over” a monument, you can locate it with one session sitting on a tribrach or a pole to many CORS station and have each baseline check, but there is no proof you were over the point, only that an antenna location in space was good. Doing it twice with different tribrachs, different HI’s zooms up the validity of that monument’s coordinates.
-
The question is are you locating the tribrach on the tripod, or the monument under the tripod. Your closure will be better if you leap-frog from tribrach to tribrach, your traverse on the monuments will be better if you change tribrachs, rotate them, relevel them for each measurement set. Introducing redundancy to the measurements decreases the accuracy and increases the validity of the traverse. Much like a GPS antenna “over” a monument, you can locate it with one session sitting on a tribrach or a pole to many CORS station and have each baseline check, but there is no proof you were over the point, only that an antenna location in space was good. Doing it twice with different tribrachs, different HI’s zooms up the validity of that monument’s coordinates.
In an ideal world and with perfectly calibrated and maintained equipment the tribrach and monument underneath should be coincident with each other.
However it is not a perfect world and upon snapping your instrument into the next tribrach you will typically have to relevel the instrument and by doing so you will see the optical plummet may not be centered on the point below anymore. If your equipment is in good shape it is usually negligible and for most projects not to be concerned about.
To answer though, I traverse through the center of my tribrach and unless I have a brain fart and booger up the setup I do not move the tribrach at all.
-
In my opinion, it depends on what your goal is for the traverse. If you relevel and recenter each setup, you are introducing accumulating error into the traverse, but each ground point is slightly more accurate than if you didn’t.
I am currently doing a high accuracy traverse once per week where I am trying to get 5 mm accuracy or better for a bunch of sideshot stations. It is along a stretch of interstate highway. I have two stable endpoints (monuments) outside of the area being monitored, and two monuments inside the area, and then about 20 targets that are used to control mobile lidar. We have to run the traverse each time they do a mobile run (once per week). Except for the four monuments (which also have long session static GNSS), all of the other traverse stations are not marked on the ground. This eliminates any HI/HT errors and any centering errors. We keep the sight lengths around 300 m between traverse stations in order to get vertical accuracies equivalent to leveling, We did one run so far, next one is tomorrow. Prior to the first run we ran levels with invar rods. All of the targets were within 2mm vertical between leveling and traverse when I fixed the two endpoints 3D. All of the targets (on the outside shoulders) are shot as sideshots to 0.10 m HI stakeout prisms. Tomorrow we will see how the horizontal compares between the two runs. Problem is that now they have started longwall mining beneath the highway (interstate). They are expecting the road to drop 3 to 5 feet, but are saying it may take up to two months for that to happen. Meanwhile they are down to one lane each direction and 45 mph speed limit. Someone is on site 24/7 checking for any damage to the road.
Total length between endpoints is 2.6 km. Using a high accuracy S6, being very careful about T and P, etc. First adjustment indicated error ellipses all under 5 mm at 95%.
We had previously planned on doing dual occupation RTK and differential leveling each time until they decided they needed 5 mm accuracy horizontal, which cannot be done with GPS. Turns out the traverse is faster than running levels, and just as accurate. And no RTK.
-
A lot depends on your goal.
If you want an astronomically high closure go with forced centering. If you want to find and remove errors, break your setups, relevel and center every time. The exception would be when setting on fixed pillars as opposed to tripods…
-
This is a great explanation, I have definitely learned from it, thanks !
-
It is standard practice among every person I know to level the instrument when moving forward to the foresite. That said, traversing along flat ground, (and assuming the tribrach is well calibrated) a modern instrument with a calibrated compensator that is working within specs will probably provide basically the same results. If you are outside the limits of the compensator, it should not let you continue (at least for the instruments that I have used).
-All thoughts my own, except my typos and when I am wrong. -
@bpetkovic: 20 seconds of arc in 100 meters?!?!? I can point better than that just looking over the top of the scope. You may want to look into your sighting methods.
20 seconds per set-up in 3.5 km figures to be 9?ñ cm. The variance of the sum is the sum of the variances.
Log in to reply.