Trig Leveling Procedures

Howdy,

Two I'd recommend:

http://www.dot.ca.gov/hq/row/landsurveys/SurveysManual/leveling_specs.pdf - CA DOT specifications

http://gge.unb.ca/Pubs/TR142.pdf - an academic treatment

Enjoy,

DMM

Search online for Charlie Glover's and/or Jesse Kozlowski's work.

SS

>Specifically, how long can your site distances be?
>
>How many sets do you turn? Do you have to be very particular about balancing your BS and FS? Do >you factor in Refraction and Curvature if you don't balance your BS and FS? How important is it >to have identical rods on the BS and FS?

Always use both faces, the collimation error cancels out regardless of the distance. So there is no need to balance the sight distances as with a differential level.

The longest sight distance depends on the precision of your vertical circle and the steepness of the shot at hand.

Jesse and Charlie were using a Wild something with a 0.5" vertical circle and dual axis compensators. They chose to limit the sight distances to about 100 meters for what they were doing, which was to attempt to achieve first order results towards a standard for using trig levels for NGS geodetic work.

2 sets is usually enough.
Curvature & refraction just winds up lost in the noise because the sights are so short.
If the instrument applies it or the DC applies it, great, leave it turned on. If you are exporting the raw data to something else, you need to account for it in the "something else" -- for example, our Leica 1203 applies the C&R in its onboard collector & so the coordinates it calculates have those corrections. When we dump the raw data to Star*net, the angle & distance data do not have those corrections & so star*net applies them. Good to compare.

As far as identical rods, they don't have to be identical, just of a carefully measured height and not likely to get mixed up. Charlie & Jesse were using tall rods (2m+) to get the target up out of the hot air down near the asphalt. I tried with a pair of 2m GPS rods, each with a prism adapter and a Leica 0mm prism, and they wound up being 6.95' and 7.00', Middle of summer, it makes a difference, 50F and raining, not so much, so I mostly use the regular rod in its carefully calibrated "all the way down" position. Easier with one rod and the rodman walks a little more. Then you worry less about mixing up the rods. Or solo with robotics, one rod, also works.

If you have a reflectorless, you can put crow's feet everywhere and trig level between them with no rod.

> >Specifically, how long can your site distances be?
> >
> >How many sets do you turn? Do you have to be very particular about balancing your BS and FS? Do >you factor in Refraction and Curvature if you don't balance your BS and FS? How important is it >to have identical rods on the BS and FS?
>
> Always use both faces, the collimation error cancels out regardless of the distance. So there is no need to balance the sight distances as with a differential level.
>
> The longest sight distance depends on the precision of your vertical circle and the steepness of the shot at hand.
>
> Jesse and Charlie were using a Wild something with a 0.5" vertical circle and dual axis compensators. They chose to limit the sight distances to about 100 meters for what they were doing, which was to attempt to achieve first order results towards a standard for using trig levels for NGS geodetic work.
>
> 2 sets is usually enough.
> Curvature & refraction just winds up lost in the noise because the sights are so short.
> If the instrument applies it or the DC applies it, great, leave it turned on. If you are exporting the raw data to something else, you need to account for it in the "something else" -- for example, our Leica 1203 applies the C&R in its onboard collector & so the coordinates it calculates have those corrections. When we dump the raw data to Star*net, the angle & distance data do not have those corrections & so star*net applies them. Good to compare.
>
> As far as identical rods, they don't have to be identical, just of a carefully measured height and not likely to get mixed up. Charlie & Jesse were using tall rods (2m+) to get the target up out of the hot air down near the asphalt. I tried with a pair of 2m GPS rods, each with a prism adapter and a Leica 0mm prism, and they wound up being 6.95' and 7.00', Middle of summer, it makes a difference, 50F and raining, not so much, so I mostly use the regular rod in its carefully calibrated "all the way down" position. Easier with one rod and the rodman walks a little more. Then you worry less about mixing up the rods. Or solo with robotics, one rod, also works.
>
> If you have a reflectorless, you can put crow's feet everywhere and trig level between them with no rod.

I was thinking of using 4' section range poles with topo boots and mini-prisms attached with 5/8" thread adapter. Duct Tape on rod bubbles and there you go, 2 identical rods (hopefully). If I want them higher I could add another 4' section of range pole.

I did a test run on the repeatability of the elevation difference with my robotic total station. I turned the angles manually, because I don't trust the vertical tracking of my robot. I tested at a sight distance of a little over 500' and I was able to resolve the elevation difference to under a hundredth, but I don't think I would want to push it much farther than that.

The Bow Tie Surveyor

Third order trig levels are not rocket science.

It's basically 3-D traverse done to reasonable standards of work. A decade or two ago vertical angles were only taken seriously by one manufacturer, which I used but I am not going to name. It came down, in the end, to the accuracy of the vertical axis compensators, which to some total station manufacturers meant "good enough to reduce horizontal distances" but not good enough for trig levels. Today, all of the manufacturers probably meet the expectations.

You don't need scientific papers about achieving first or second order results about trig surveying in order for you to achieve third order results in 3-D work. And you need to understand the difference between accuracy and precision.

If you're in Florida, I'm going to take a wild guess that you won't be trig levelling across valleys from ridge top to ridge top or that any of the work will even be along lines that are other than basically flat.

That being the case, assuming that you have a total station that can measure zenith angles with a standard error of 3 seconds or better as the mean of both faces and can take care of the basics like determining the target heights, the critical part of the work will be the target and the quality of seeing along the line. Ideally, you'll want to use a target that you point the telescope at with much less uncertainty than the standard error of the zenith angle measurement. As for the seeing, that will vary by time of day and other factors such as cloud cover and the nature of the surface.

Ideally, you'll want to keep sight lengths so short that there is no discernible heat shimmer when you're looking at the target through the telescope. If there is shimmer, it means that the refraction effects along the line won't fit the standard refraction model since you'll be dealing with convex-downward curvature of the line of sight instead of convex-upward from the change in the index of refraction from just the adiabatic lapse rate. This is the result of the super-heated air near the ground that causes the shimmer in the first place being less dense than the air above it instead of denser as is the case under adiabatic conditions.

If you want to get fancy, you can make a card with a bar chart on it to measure the target resolution through the telescope. Ideally, you'd want it to be the same on backsight and foresight. An experienced observer should be able to get adequate results just by looking at the image. No shimmer, no distortion, and no blurring are what you want. If those refraction effects are present, the cures are to (a) raise the target or (b) shorten the sight length.

For notes I have always kept the left page the same as for regular levels, which for me is as follows:

+ HI - SS ELEV. DESC.

(I let the total station calc. the + and -, personally I have never used a trig level software routine but if I did my notes would be much more streamlined and I would only book point numbers and rod heights.)

For page right I normally just have "Remarks" but for trig levels I book:

Vert Ang SD ROD HT. REMARKS:

I do not set points to gun up over as I go, and I follow the same basic procedures as running traditional levels, still break the set up at the end and run it back if closing to the pob for example.

Running levels this way with calibrated quality gear has always got me comparable closures to differential leveling although I find that differential leveling is more therapeutic, something about working without batteries is good for the soul.

> Third order trig levels are not rocket science.
>
> It's basically 3-D traverse done to reasonable standards of work. A decade or two ago vertical angles were only taken seriously by one manufacturer, which I used but I am not going to name. It came down, in the end, to the accuracy of the vertical axis compensators, which to some total station manufacturers meant "good enough to reduce horizontal distances" but not good enough for trig levels. Today, all of the manufacturers probably meet the expectations.
>
> You don't need scientific papers about achieving first or second order results about trig surveying in order for you to achieve third order results in 3-D work. And you need to understand the difference between accuracy and precision.

:good: :good:

Trig levelling is not the same as trig traverse.

The moment you use a pocket tape for a measure up you lose all your precision, in comparison to trig levelling.

I didn't make it clear enough in my last post: don't worry about special rods, contrive a way to get a good known height on your regular TS rod, maybe in something repeatable like the "all the way down" setting, or maybe now it makes sense why you might hand someone a 2m GPS rod with a prism atop.

In its simplest form:
If you use the SAME rod at each end, you don't even need to measure the rod height.
Start on the BM, shoot the slope distance, use the zenith angle to get the horizontal & thus the difference in elevation. Shoot in the other face, mean them. Move the rod to the foresight. Shoot in both faces again, calc the DE to get the turning point elevation. Move ahead. Loop it just as you would a regular level loop.

It gets much more interesting if you put it in least squares. You can observe the same BMs and turning points from different setups, and 3 or more at a time. This gives a redundancy that allows you to essentially "loop" the trig level run as a one-way trip. Simultaneous equations allow you to be everywhere at once, so the order you create the redundancy in does not matter.

With trig levels, you are 2/3rds of the way (zeniths + EDM distances) to 3D free stationing (just add horizontals). This was where I found the 2m GPS rod to be less practical, because it didn't seem to add any precision in the vertical but got more wiggly in the horizontal. Somewhat less thermal effect at 50F and overcast. So here we are back to the regular TS rod.

With the trig level technique, you can run a level loop solo using a robot.

Nice note format, something that allows a double-check on the DE vs the slope/zenith if needed later.

"The moment you use a pocket tape for a measure up you lose all your precision, in comparison to trig levelling. "

Why do I get that scared feeling that I can't accurately read a steel tape...yeesh

Because if you do the redneck science experiment where you try it both ways, and if you gather enough data, you will see that on average you cannot read the steel tape as accurately as you think you can. It will evolve from a "feeling" to a confidence in the error of the method. Maybe even 95% confidence.

Most people won't do the experiment, though.

For precise trig levelling (third order), my site distances are typically 600 feet or less, although I have had much longer legs; Typically I turn 2 sets, but longer legs may turn 4 sets for check if there is any variation in d/r DIE readings; I use the same rod, either a prism pole locked tight or a fixed height GPS pole. The rod must be the EXACTLY same height for your entire traverse, and the prism should preferably be a tilting prism. I use a full round prism versus a mini-prism to increase range and improve ATR pointing. Run the dual-axis compensator calib routine at each setup.

My fieldnotes are simple BS (+/-), FS (+/-) and D.I.E. with space to average the various sets. I have had excellent results running up and down hills, around city blocks, levelling across water bodies, etc. using precise trig levels as a robotic surveyor. Jesse's article is a simple guide.

Regards,
Mike Moran

"The moment you use a pocket tape for a measure up you lose all your precision, in comparison to trig levelling. "

Ahh..I see, "Measure-Up" = H.I. I mistakenly thought you meant i couldn't reliably measure the rod height to the 0.001'. Not that it matters with a static height.

Carry on 🙂

Redneck science experiment, indeed. 🙂

Here's Jesse Kozlowski's 1998 article from Professional Surveyor magazine.

Until now I have not even thought that trig levelling is different from trig traverse.
I just measure each traverse point using crosshair to point at the center of the prism.
Usually I get closures of less than 5 cm in elevations for a loop of several hundred meters.

Usually I have not intention of running a level check then I place the prism as close to the ground as possible. This way errors in vertical of the rod is minimized since the vertical distance will be computed by the gun from the slope distance.

Actually reading scales is a major source of fluff dating back to the days when you had transits and theodolites

> Until now I have not even thought that trig levelling is different from trig traverse.
> I just measure each traverse point using crosshair to point at the center of the prism.
> Usually I get closures of less than 5 cm in elevations for a loop of several hundred meters.
>
> Usually I have not intention of running a level check then I place the prism as close to the ground as possible. This way errors in vertical of the rod is minimized since the vertical distance will be computed by the gun from the slope distance.

There are a few differences in measuring rod heights, but if properly done, both can be done during the same traverse.

I prefer 10+ mile, 1000' elevation change, trig leveling with a T3. 12-16 sets minimum.

And oddly the best time is midday, not before 10:00 AM, or after 3:00 pm.

Go big or go home.