Hello guys. I have noticed many people use resection in construction setting out. I understand you must use at least 3 resection points. My question is on the angular distribution of these points from your setup point so that you can get an ideal resection solution. Anyone to share best resection practices.
This Content Originally Published by a land surveyor to Land Surveyors United Network
I think you should ask yourself,Why am I using a resection.
Resections are used on many different surveys not just construction.
It was not the intent to use them as to start your survey every day routine.
Two points can work as well as three, the degree of accuracy is what matters.
But always check into other points to make sure your in reason of a closure.
No matter what the closure is on the resection, tie other controls or property points.
Never assume you are correct until you make these tie's.
Resections are use to aid a surveyor when points are hard to occupy or not accessible to occupy. There are lot of total Station,calculator or data collector to computer programs
available today. Triangle solutions ,mainly SAS,SSS can be used to check or calculate your
resection data or to determine best fit. Even Distance ,Distance and Bearing,Bearing
Intersections can also be used to check,the best fit solutions in a resection program. These resection programs auto adjust your coordinates of your occupy point,this is determine
by the accuracy of the points and your work and your equipment. I only have used them when there was no other choice, This is my opinion.I am sure there are others who will disagree and some that will agree with me.But like I said, asks yourself ,Why am I using a resection,is it to save time.That was not the intent, or the reason to use them. But they do have there place.
While in a perfect world they would be 120 degrees apart (or spread ) to cover a full circle, in reality we must live with less. First, try not to have less than 180 degrees or a half-circle spread, I try for more than 200 degrees. Numerous reasons abound, not the last of which is the old school "strength of figure" theory for any geometric figure.
Second, your distances should be simular in magnitute - ranges from 100 feet to 200 or 300 feet is okay, but avoid having a 30 foot distance to "A" and then 300 feet to "B" and "C", this is not good.
Third, if you are attempting to reslove an elevation in this resection you MUST use 4 points minumum.
Regarding resection: I only used this method for earth work. I double tied all control points in the process. I also leveled through control to insure that elevations hit. Not sure if you're using GPS or conventional methods....though my comment is based on conventional methods. Of course data processing proving you hit your control gnat's ass will more than likely prove ground work is good...especially if tie backs after topo work proves also your instrument didn't wander. I personally would never use resection for boundary work or lay out/stake out.
I am a cadastral surveyor and use resection in 99% of my work as the residuals are minimised to zero.
A others have said you cant get away from the basic good geometry of your point selection.
Set up in the 'middle' of the point spread and for accurate work a minimum of 3 points is required and always check other control points for a 'sanity' check.
Another point to be aware of is the "Danger" circle where all control points are equidistant from the setup. very remote chance I know but a point to be aware of when in the field.
Cheers, Phil (Australia)
Thank you guys for your contributions. I have picked up some principles regarding when to use a resection and good practices regarding the spread of the resection points.
reading through the replies most is said I work in a wide fields of Survey
and Engineering projects.
Lot of times when using Total Stations I choose resection all modern TS have
good programs to calculate your setup using least square method some have options like Helmet.
I never experienced much difference between them. Most imported is always what accuracy you
need to achieve position and elevation wise and with all setups never work outside your longest backside.
As mentioned 2 points work to calculate "a" position but bad practice min 3 in good arrangement not 3 points in one line 20 meters apart and with all Survey tasks "Check Check Check".
In theory, the closer to 90 degrees, the better your result. The resection uses the angle and the distances measured in your set up to resolve the position of the instrument. However, due to the set up and observation errors, mathematically, the lines of sight cannot be infinitely thin which is where the error comes from. Imagine (or sketch) two thick lines crossing at 90 degrees to each other...... The area where they overlap will be a square - mathematically, the smallest possible area. Imagine two lines that cross each other at an angle of about 175 degrees. Notice how the area for the possible solution is much bigger now? You can think about that as the area of the error. The closer to 180 degrees or 0 degrees, the worse your result. If fact if you set up too close to 0 degrees or 180 degrees between your backsights, you will get a message that the point cannot be resolved. This is because two parallel lines will never meet!
The above explains why mathematically you only need 2 points for a distance resection to work. However, you should always build in redundancy and take more points. The more points you take, the more accurate your set up. If it appears that your error gets bigger when you take extra points, who is to say that your original two points were good quality and your subsequent ones were poorer quality?
As other people mentioned, once you have set up you should always check to other control points an as-built features before continuing.
Never use a resection when you are establishing other control points. Only use it for tertiary surveying or setting out i.e. measuring survey points or setting final markers. Avoid using it for work of high accuracy e.g. structural monitoring.
I hope this helps.
Thanks Grant, quite helpful.
Resections work great - depending on the precision of the resection points, and the precision required for the project.
NEVER proceed w/o shooting that 3rd (or 4th or 5th) point for confirmation.
After all - ANY fool can get TWO points to line up!