It has always been my understanding that the compass rule is for closed looped traverses where you distribute the error of your misclosure.
I'm being asked to do a compass rule adjustment on control I located along a highway corridor where I can't practically do a loop. I have a list of given coordinates for the control. How do I properly do a compass rule in this scenario? What am I closing against? If the closure is bad, how do I know what's a result of errors in my traverse vs the control I located in the field being disturbed/shifted?
Any help understanding this would be appreciated.
Thanks
Yes, you can do a compass rule in an "open" traverse. Simply adjust the line from the beginning point using given coordinates to the terminus of the traverse and hold those given coordinates. Distribute the error same as looping back to the beginning.
In my software, PCSurvey, each traverse point must be connected with an inverse in the order that the traverse was run. Then those inversed lines must be selected and an alignment created. Then select adjustment "Compass Rule" and then the option for "closed" or "open". Under "open" there is an option for selecting a coordinate point from the existing point list or type in a coordinate to close on. So like @mightymoe said, the open compass rule closes onto a coordinate point of your choice near the end of the traverse.
Reaching way back into my memory, the difference between the closing bearing and the calculated bearing is your angular error and must be distributed among the traverse points. Exceptions to strict distribution would be a suspect observation like sighting a target that is low to the ground and was affected by the thermocline, or a short backlight. Recalculate your traverse after correcting the angular error. The difference in distance between your closing coordinates and the control point coordinates is your linear error.
By the way, this is one of the topics that frequently falls victim to loose terminology. There are closed traverses and open traverses, and there are loop traverses and link traverses. Loop traverses are by definition closed: ending on the same point at which they began. Link traverses can be either closed or open. If they begin and end on points with known locations (e.g. passive control monuments or GNSS-observed points) then they are closed. Otherwise, they are open. An oriented, radial shot from a known location is the simplest open traverse.
Once I used the term "open" traverse in an examination with the Board and one guy got up, banged on the table with his fist, and yelled at me so much that his face was turning red. I tried to explain that I had started and ended on points with known coordinates, but it was too late.
All I could do was leave the room, hoping the guy didn't stroke out from the extreme amount of stress he got from what I dared to say.
A traverse that starts on a station of known coordinate and ends on another station of known coordinate is a closed traverse. A traverse that starts and ends on the same station is also a closed traverse.
A traverse that starts on a station of known or adoted position but does not end upon such a station is an open traverse or an open ended traverse.
Historic boundaries and conservation efforts.
Aka the Bowditch adjustment, where you are beginning your traverse off of a known coordinate and closing into another known coordinate. Quite common these days when traversing between points set with GPS where a traverse is run between them through an area not conductive to using GPS such as under dense canopy. The assumption is that the error is evenly distributed between each set up in both angles and distances and the error correction is applied proportionally to each station based on the distance from the starting point to the closing point. The closest station to the starting point receiving the smallest proportion of adjustment and the closing station, the farthest, receiving the full amount to bring it into harmony with coordinates that are being closed upon. A significant difference in the total distance inversed between the two sets of traversed and record coordinates, starting point and closing point and the point being closed into, indicating a bust, in which case this should not be used because somewhere your traverse likely got boned and adjusting out blunders is bad.
Just because I'm paranoid, doesn't mean they aren't out to get me.
@not-my-real-name Yeah that's kind of why I put open in quotation marks. I just kind of struggle understanding that even though I'm ending on a known coordinate- how do I know that the misclosure on that point is a result of the error in my traverse vs the physical point being off, I know it's both but how much-
@verplanck with the software that I use any adjustment will provide that solution as long as I specify the two known positions in the input file. Just make sure the last front sight observation data is given the same point number as the given closing position.
Historic boundaries and conservation efforts.
@not-my-real-name Yeah that's kind of why I put open in quotation marks. I just kind of struggle understanding that even though I'm ending on a known coordinate- how do I know that the misclosure on that point is a result of the error in my traverse vs the physical point being off, I know it's both but how much-
Thats the 64,000 dollar question. This is where someone provide "GPS" control to be held fixed for a traditional traverse and then complain the traverse "doenst look good". I always ask, what's the 95% station coordinate error ellipses of the control you want me to hold pure(fixed).
For example from StarNet:
Station Coordinate Error Ellipses (FeetUS)
Confidence Region = 95%
Station Semi-Major Semi-Minor Azimuth of Elev
Axis Axis Major Axis
HLS2 0.000000 0.000000 0-00 0.000000
HLS1 0.012088 0.000002 0-00 0.009971
HLS3A 0.027135 0.014356 10-45 0.012410
HLS1A 0.065914 0.023242 86-34 0.021457
HLS4 0.025872 0.016293 165-52 0.014776
Now I know how "good" the control is to evaluate my work against the control.
Most of the time I hear well we used GPS its good just hold it fixed.
My two cents.
Gary
how do I know that the misclosure on that point is a result of the error in my traverse vs the physical point being off
Does it matter? If the misclosure is small enough that it will not impact the project, then it doesn't matter who's work was "off", the traverse or the original given points. The adjustment can be made and no one will ever know or care about the misclosure.
If on the other hand the difference is actually material, then for the sake of the project the traverse and the original control has to be redone to determine where the correction needs to be made so that it will not impact the project.
This uncertainty of whether the error is in the starting point or the ending point is why IMHO this is always an open traverse. The only way to not have this uncertainty is to traverse back to the beginning making a closed traverse.
@lurker I get that and I agree, I'm just trying to see if I'm fundamentally not understanding something here. I had already traversed through these points, did a best fit least squares rotation and everything checked pretty well. This is how I usually do it. Now I'm being told that's not good enough and to re-do the traverse and do a compass adjustment- I'm just not sure what that's accomplishing.
@lurker I get that and I agree, I'm just trying to see if I'm fundamentally not understanding something here. I had already traversed through these points, did a best fit least squares rotation and everything checked pretty well. This is how I usually do it. Now I'm being told that's not good enough and to re-do the traverse and do a compass adjustment- I'm just not sure what that's accomplishing.
Why would you have to redo the traverse ? ! Get a demo of Starnet with your data.
@verplanck I agree. I'm not sure that you are actually accomplishing anything other than satisfying someone who doesn't really understand what you have already done.
@landbutcher464mhz So if I come off of a given starting point and traverse through another given point and miss it by 1.00' and then traverse back to my starting point and close balls on what does this prove? Without compensating errors it proves my traverse was good but I still have no idea whether the starting point has bad coords or the intermediate point has bad coords. All I have accomplished by closing back to the starting point is redundant traverses. This gives weight to the traverse being correct but doesn't answer which of the 2 given points are bad. Again assuming no compensating errors.
If the closure is bad, how do I know what's a result of errors in my traverse vs the control I located in the field being disturbed/shifted?
If you have 2 measurements that disagree (by more than can be attributed to random error) you may have to do a third to determine which of the first two contains the blunder.
@Lurker wrote "So if I come off of a given starting point and traverse through another given point and miss it by 1.00' and then traverse back to my starting point and close balls on what does this prove? Without compensating errors it proves my traverse was good but I still have no idea whether the starting point has bad coords or the intermediate point has bad coords. All I have accomplished by closing back to the starting point is redundant traverses. This gives weight to the traverse being correct but doesn't answer which of the 2 given points are bad. Again assuming no compensating errors."
It is possible that the office wants to hold that starting point and that ending point for some reason and they want your traverse adjusted to fit using the Compass Rule. Traversing back to the beginning to close your work will tell the office that the traverse does not have any blunders and then they will have confidence that doing a Compass Adjustment will give them a control traverse to do other work from.
If the closure is bad, how do I know what's a result of errors in my traverse vs the control I located in the field being disturbed/shifted?
Good point. Reading through this thread reinforces my belief that one should know the purpose of a job in order to formulate a plan to carry it out. I'm sometimes faced with a lack of communication, and I don't like that.
I feel your pain on being told to redo a Least Squares adjustment with a Compass Rule adjustment. It feels like taking a step backward just to satisfy someone else's "check-box" requirement, especially since Least Squares is generally more rigorous for identifying where the error actually lies.
If you just need to get the math done quickly to move on, I’ve used this little web tool before- https://thetoolapp.com/compass/ - it’s handy for a quick "link" or "open" traverse calculation without having to fire up StarNet or fight with your main software's settings for one small task.
Ultimately, though, if the misclosure is huge, no adjustment method is going to tell you if it's your traverse or the original control that's shifted without going back out and hitting a third point.