It would be bad?? Try to imagine all life as?ÿyou?ÿknow it stopping instantaneously and every molecule in your body exploding at the speed of light.
Oh wait... that only applies to crossing the streams.
I don't remember hearing the legs can't cross.?ÿ I must not have been paying attention that day.
(Pedant) - a person who is excessively concerned with minor details and rules or with displaying academic learning.
You get an adjustment when you do least squares, but that is not why you do it. You do it to get a report on the accuracy of your points.
To be even more pedantic; I would argue that you don't "change the measured values" (the angles and distances between the points) with least squares; you get estimated coordinates for the points, and inverses between these estimated values.
As opposed to (in a closed figure) actually distributing the differences between the measured figure and the theoretical error-free figure amongst the measurements and recomputing the traverse with adjusted values.
In a closed figure deflection angle traverse the sum of the deflections in one direction (usually clockwise) should equal 360?ø.
If the traverse crosses itself an even number of times the sum should still equal 360?ø when considering one direction positive and the other negative.
If the traverse crosses itself an odd number of times, the sum of the right deflections should equal the sum of the left deflections.
So you prefer to report measurements you know are wrong, to the most likely correct values based on your measurements?
I don't think you can infer that the measurements are wrong. My survey techniques and equipment get me unusually high precision. To adjust the coordinates, or distribute the error by some algorithm is not necessary.
As an example, the perimeter of my latest traverse got a precision of 1:290,000 and contained about 7 acres.
Analysis with the method of least squares assured me the precision of the work exceeds minimum standards. Therefore I did not adjust any coordinates. I do print the analysis for the file.
If the coordinates are adjusted, then I need to go back and see if the adjusted coordinates are wrong, not the other way around.
It is like working with negative values for northings and eastings, it happens and you work with it.
0.02
easier said than done
I see pi all day & everywhere!
no matter at what angle I look
"Traverse Adjustment: the art of taking a blunder and smearing it thru the entire project"
Bob Curtis, 1970 something.
I don't think you can infer that the measurements are wrong. My survey techniques and equipment get me unusually high precision. To adjust the coordinates, or distribute the error by some algorithm is not necessary.
As an example, the perimeter of my latest traverse got a precision of 1:290,000 and contained about 7 acres.
Analysis with the method of least squares assured me the precision of the work exceeds minimum standards. Therefore I did not adjust any coordinates. I do print the analysis for the file.
If the coordinates are adjusted, then I need to go back and see if the adjusted coordinates are wrong, not the other way around.
"Traverse Adjustment: the art of taking a blunder and smearing it thru the entire project"
Bob Curtis, 1970 something.
You certainly should make the effort to identify and rectify blunders.?ÿ
But there is always random error, too.?ÿ Is it better to distribute the random error or to effectively assume it was all in one arbitrarily chosen leg of the traverse??ÿ If you run the traverse data backward, you end up with a different solution, no better. So isn't it better to distribute the where the probability is for a closer match to the actual dimensions and positions?
I prefer to note the "total accumulated error" rather than a ratio.
as in IF I "accumulated" ... 1 foot over 290000 total ... who is to say that it is actually a 1.5 foot blunder (between two objects 10 feet apart) in the opposite direction?
there was a very large control net back home (published of a Record of Survey) in the 1970's with a GREAT ratio, but in reality there was a significant blunder in a short course, something on the order of 20 FEET!
hope that makes sense.
with enough redundancy in our measurements we will know those anomalies...
under Bob's guidance we ignored the "ratio" of error and looked to the direction and magnitude of total accumulated error.
then, we looked for a possible Distance blunder on courses similar to that direction
and then we looked for possible Angular blunders on kinda perpendicular courses
we actually found (and corrected) numbers of errors in odd places, ones that really mattered.
Bob Curtis, brilliant man. Thanks again Bob!
@dougie same here. If the closure resulted in meeting tolerances established for the project, why change each angle a few seconds or a distance a few hundreds? If I go back out and remeasure, I'll more likely be consistent with actual measurements rather than "pencil-whipped" ones. Its all about employing the proper methodologies and error checking during the measurements.
