Activity Feed › Discussion Forums › Strictly Surveying › Long Backsight or long resection
Long Backsight or long resection
Posted by nettronic on October 2, 2021 at 4:54 amChoices are, a long backsight with a traverse to a flat angle or.. setup and shoot in two known points (the long backsight)
My boss is much smarter than me and confuses me all the time. Which is better, a long baseline with a flat angle. Or setup in the middle and double angle to the ends?
bill93 replied 2 years, 6 months ago 16 Members · 25 Replies- 25 Replies
Both is better
What’s a flat angle?
Nearly 0
I’d go the long BS. If possible, a POL, where I don’t touch the H tangent screw, so I wouldn’t double the angle either.
But I wouldn’t have an issue setting a point at near 0.
When you setup on the new point, after you BS, turn to the original BS and shoot that in as a check
- Posted by: @nettronic
Nearly 0
When you measure both angle and distance with comparable accuracy there is no bad angle to use in a traverse.
You can’t accurately resect from points without reasonably angular spread.
. You can not reliably resect off of two points with any level of confidence. Don’t get me wrong, Resections have their place it is just not the first tool to bring out of the box. Besides, I don’t think modern digital total stations have the problems associated with flat angles their glass plate theodolite predecessors had.
No matter what you do, ALWAYS check what you do.
And read everything you can get your hands on; written by Dr. Ben Buckner.
I hope everyone has a great day; I know I will!- Posted by: @john-putnam
I don’t think modern digital total stations have the problems associated with flat angles their glass plate theodolite predecessors had.
Educate me. Why would any angle, including a “flat” one, be measured less precisely by a T2 compared to a total station?
Are we talking about a “wiggle-in” here? (or something similar)
- Posted by: @bill93
You can’t accurately resect from points without reasonably angular spread.
This is a misconception that I held for a long time until someone challenged me on it and I ran a Star*Net preanalysis to check. As long as you’re including distances in the resection, you’re actually better off being close to the line defined by the two control stations (flat or shallow angle) than you are being offline (large angle).
In the following diagram, points 1 and 2 are fixed (no error). Points 3, 4 5 and 6 are resected points.
Using the default Star*Net standard errors and the preanalysis function, here are the resulting error ellipse values:
Station Coordinate Error Ellipses (FeetUS) Confidence Region = 95% Station Semi-Major Semi-Minor Azimuth of Elev Axis Axis Major Axis 1 0.000000 0.000000 0-00 0.000000 2 0.000000 0.000000 0-00 0.000000 3 0.015811 0.010297 0-00 0.010435 4 0.014625 0.010273 0-00 0.010260 5 0.034726 0.013410 0-00 0.013984 6 0.014869 0.012963 0-00 0.014585
The smallest errors lie with the points closest to the control line.
Resection is angles only?
Including distances is sometimes called Freestation.
Personally, I’ve never understood the hesitance turn flat angles.
As for the precision of angle measurements between a modern MS60 and a T2, while they may both be 1″ instruments they have differences in sources of non-random reading error bas on how they obtain their angle measurements. The T2 used an angular gradient imprinted on a plate. The plate was a possible source of non random error. To minimize the error, you would turn multiple sets of angles, re-zero the T2 in different quadrants of the plate between sets (DBS-DFS-RFS-RBS, new quadrant and repeat). Modern total station utilize digital incoder for reading the angles and does not have a plate per say, minimizing that possible source of error.
The first concern is always eliminating blunders. It is the blunders that get you in trouble. Lack of precision can be a bit embarrassing but that is about as far as that will go in most cases.
Resectioning is very powerful, and, done correctly, aught to work just fine. But it is also very easy to screw up. The statistics that your dc returns for a resection are somewhat more complicated than for a simple backsight check. Not rocket science, but somewhat more complicated. So it is probably a phobia about screwed up resections, and not a serious concern about precision, that has your boss resistant to them. Personally I use resections routinely but I have a rule to always use 3 control points. An error in a 3 point resection is likely to cause it to outright refuse to resolve, thus reducing the chance of a blunder being perpetrated. I will bend my rule and use 2 points in a low leverage situation when I’m doing it, personally. But people that work for me must use 3 points until I’m 100% confident in their ability to understand the statistics, and their willingness to review them in a timely manner.
Finally, the “flat angle” thing is a throwback to the days of open vernier transits which read angles to the half-minute. It’s not really a thing to concern yourself with when using a half-way modern total station.
@jim-frame Bingo. Math doesn’t lie…
- Posted by: @john-putnam
As for the precision of angle measurements between a modern MS60 and a T2, while they may both be 1″ instruments they have differences in sources of non-random reading error bas on how they obtain their angle measurements. The T2 used an angular gradient imprinted on a plate. The plate was a possible source of non random error. To minimize the error, you would turn multiple sets of angles, re-zero the T2 in different quadrants of the plate between sets (DBS-DFS-RFS-RBS, new quadrant and repeat). Modern total station utilize digital incoder for reading the angles and does not have a plate per say, minimizing that possible source of error.
I understand the mechanics. But if I’m not out of date total stations do have a plate and measure angles by means of electro-optical scanning of precise digital bar-codes etched on their rotating glass cylinders or discs, the same medium as the glass plate on a T2. The key is the TS “spins” the plate for each pointing and reads the entire plate digitally, integrating any plate anomalies so a single pointing is essentially free from plate errors. The T2 mimics such integration by observing multiple pointings around the plate as you say; with 8 sets I suspect plate error drops below other non random instrument errors.
Don’t get me wrong, modern TSs have many features that also reduce instrument errors, automatic mislevelling corrections, trunnion axis error compensation, etc. I’m saddened when folks poop on pop and say the T2 is less precise than modern TSs; may be (barely) true but the T2 is rugged, lightweight, uses no electricity, more waterproof and the optics are divine. The real reason TSs including robots took over is they are much faster and can be operated by less skilled operators.
@mark-mayer There were two issues with ‘flat’ angles. Even after open vernier transits, most instruments would not produce accurate near zero angle measurements. That is pretty much solved. The other issue is strength of interpolated functions. Carrying around tables isn’t completely a thing of the past, but we aren’t slaves to them any longer.
A good exercise is to perform a strength of figure calculation using stanard formulas from old geodesy text books. It will open your eyes to the other effects of geometry. Small errors can cause big problems on near zero angle measurement. If it’s too close I simply mark the intersect on line and pull over to set…
Nah… “Freestaion, also called resection…”
Ghilani/Wolf Elementary Surveying text book describes method of resection as using angles and distances.
- Posted by: @thebionicman
Carrying around tables isn’t completely a thing of the past, but we aren’t slaves to them any longer.
It’s certainly a thing of the past with me. (Okay, technically I carry them — they’re in the back pages of my field books — but I haven’t referred to them in decades.)
@jim-frame I have a half dozen versions I use for cadastral work. The approximations match precise compution well and help isolate probable sources of error.
Log in to reply.