When using a reasonably accurate Stochastic model (estimations of how well you can center the instrument, level, point and read the instrument) for Least Squares Adjustment and modern equipment the geometry of the angles being observed will still have a greater impact on the uncertainty of the measurements.
Unfavorable (weak) geometry will always have larger error ellipses than favorable (stronger) geometry. You can model this in your LSA software to see the results for yourself.
A better quality instrument could help compensate for poor network geometry used in tandem with increased redundancy. These procedures will never fully compensate for unfavorable geometry, but it will give you more confidence in your results.
No question that todays instruments are capable of measuring both distance and angle well.
However the constraints inherent in geometry of the points aren't changed by the type of instrument used to observe them.
Consider the following:
Triangle 1
angles roughly measure 45d, 45d, and 90d
Triangle 2
angles roughly measure 89d30', 89d30' and 1'
If you use the same stochastic model and instrument to observe these angles the uncertainty in the angles of Triangle 2 will be greater than the uncertainty of the angles in in Triangle 1. This is a mathematical fact and what I think Jeff's boss is getting at.
*edit... not uncertainty in the angles... I should have said results in a larger error ellipses at the traverse points*
The geometry has a significant impact. Sometimes we can't avoid it and we need to increase redundancy even when using say a 1" gun.
I would put a very redundant traverse run with a 20" gun up a against a minimally redundant traverse run with a 1" gun any day. Newer equipment will never outweigh good procedures.
Weak angles are a concept that has to do with intersecting directions such as happens in triangulation. The size of an angle is irrelevant when measuring the angle itself. All other things being equal, you can measure a 15 degree angle as accurately as a 90 degree angle.
Peter Lazio
Excellent point.
The uncertainty in the angular measurement is the same, but the resulting positional accuracy is very different depending on network geometry.
When traversing we are really concerned with the error ellipse at each traverse point. Again, smaller angles will result in a larger error ellipse.
Basically in any surveying situation redundancy is always our friend!
weak schmeek, tell your boss he can stick his weak angles up his arse--you did the best you could does he want to pay you a couple more hrs of overtime to send you back to do 3 or 4 100 ft sights--if he says yes, he's a bigger moron-- keep doing your thing you can work for me any day---tdd
I like Peter's answer.
If you measure a small angle or a large angle at some distance, then the sideways distance uncertainty is the same.
If you are intersecting two angles to find the distances to a point, and the lines form a small angle, then that sideways uncertainty translates into a large error in the length of the legs.
If you are measuring distance with an EDM then you are not relying on the angles for location on that axis and the concept of weak angles does not apply.
LSA is very valuable for properly combining all the angles and distances. Adjusting angles independent of distances is not as good. Do cross ties. Redundancy is your friend.
TDD answer is the best
answer of them all.
We as surveyors get so hung up on extracting the error the size of a gnats arse that we don't make any money.
Bottom line is he did the best job he could with the field conditions presented.
Closed the traverse and moved on.
With todays modern instruments, some of the old school ways of doing things are just not as important anymore.
Do we need proper procedure? Yes.
But we need to think about what we're doing and why we are doing the job!
> I like Peter's answer.
>
> If you measure a small angle or a large angle at some distance, then the sideways distance uncertainty is the same.
>
> If you are intersecting two angles to find the distances to a point, and the lines form a small angle, then that sideways uncertainty translates into a large error in the length of the legs.
>
> If you are measuring distance with an EDM then you are not relying on the angles for location on that axis and the concept of weak angles does not apply.
>
> LSA is very valuable for properly combining all the angles and distances. Adjusting angles independent of distances is not as good. Do cross ties. Redundancy is your friend.
Great answer Bill.
Farther up: I agree with TeDDy to a degree (get it?), but I don't agree with dismissing the understanding a discussion of the philosophy of good-quality and strength-of-figures. The math and geometry of the matter is still important.
TDD answer is NOT the best
> Do we need proper procedure? Yes.
> But we need to think about what we're doing and why we are doing the job!
That is just it, Joe; Jeff does not know the "why" of it. And if he does not know the why of it, he cannot begin to understand if it is close enough or if he needs to look for a better situation. Sometimes it is not about "does it close good?" But more about "how did you close it?"
TDD answer is NOT the best
sorry...your right I'm just cranky..this weather is getting to me.
TDD answer is NOT the best
Back at ya. I may be a tad cranky too. Had a migraine all day.
I think there is a time for saying good enough, just think you need to know what you are getting with it. And I always say that the guy signing decides if it is good enough. If the guy with the ticket says he does not like weak angles, then you try to find a way around it. Jeff said his lot had tall fences, I wonder if the neighbor had tall fences too. If not, maybe run around the outside, fire two control points into the back from the front, tie the back of the house and improvements from both positions... There are sometimes ways around it. But unfortunately we do not know if these were options here. For that matter, we really do not know enough to say that the closure was really good or cancelling error. But I am willing to give him the benefit of the doubt on that one.
How are things in CT?
I can pretty much guarantee that TDD's angles are always strong, I mean look at those guns! 😉
Agree adamsurveying. One should have the simple habit of setting up a traverse considering strength of figure, it should be ingrained and of little hardship. Even when using GPS it needs to be considered, especially if the data is going to be used for other things, leave a strong figure for the following work using other tools to use. Hand drafting is a good place to learn how and why. I have always thought that a good hand draftsman would be a natural in setting up any survey. Don't know if mechanical drafting is taught in school any more, that class provided the base for many things to be built on later.
jud
TDD answer is NOT the best
This winter has been tough weather wise. really hard to get things done. I actually got work, but I can't make any money because of the weather.
Could be worse, as I get older I don't enjoy the cold much anymore.
I got surveys basically in a holding pattern...
Working on networking and trying to build relationships while I got the time.
Joe
Hows things in Georgia...sure is pretty down there..
I HAVE ALWAYS TAUGHT GOOD GEOMETERY WHEN TRAVERSING- HOWEVER, ONE HAS TO WEIGH THE EFFORT AND THE DESIRED RESULT, WHY SHOOT TO TOP 1:90,000 WHEN 1:10,000 IS BALLS ON ON MOST ANY SMALL LOT THE POINT IS ON THE CAP, WHY DO BETTER, WHAT DO YOU PROVE--THAT YOUR AN EXPERT MEASURER THAT EATS BUDGETS ?? TDD
I worked with a couple of guys that were kind of screwy......One had never worked in the field before. The other one took him out to locate some property pins. He trained the first one. The first one came to me and said that he got 1/1,000,000 closure. I had him show me what they had done.
These two guys went out and set up the instrument and shot in the property pins with sideshots from one setup. They ran the coordinates, and then inversed between each property pin. They then took those inversed bearings and distances, and ran them through a traverse program.....and voila....they got 1/1,000,000 closure.
I just shook my head. I don't know if side-shots are good enough for what they were doing; but obviously they had not concept of error propagation and misclosure.
let alone the instrument error, parallax, prizm error, pole error, every point can easily be out 0.04, then add procedural error---they just don't have a clue--tdd
A weak traverse angle is one that needs to get Charles Atlas's program, OH HAR HAR HAR.
Did the big burly angle kick sand in it's face?
I did not read in your post anything on traverse line lengths. But you stated that you had obstructing fencing.
Your senior surveyor may be commenting on a weak angle where at least one traverse line is extremely short and you turned the angle from say a 30 ft line to a 300ft line.
Perhaps he also was commenting that on a short leg to a proposed station that you should have picked a distant foresite B and lined in the traverse station C so that when you occupied that station C you then turned the angle from the distant foresite B. That way you had better length angle legs.
RADU
Thanks for all of the replies.
I'm still a little confused. Most of the posters on this board agree that "weak angles" should be avoided in a traverse, if possible. I always look for good strength of figure and try to avoid weak angles because the collective survey wisdom says to (as does the guy who signs my checks).
I still don't get the math/logic behind it though. It seems to me, like other posters have noted, that a modern instrument (calibrated and used correctly, of course) can measure an angle and distance to a new point just as well regardless of if the angle is 15deg or 90deg. It doesn't seem to be analogous with swing ties.
What I am really looking for is not whether using strong angles in a traverse is good or not. I want to know what makes it good. If you think that it is important, tell me what makes it so. Or, if you think that it doesn't matter with a modern total station, tell me WHY it doesn't matter.
Thanks Again!
Jeff