What are you using to calculate all this stuff?
compass rule? TBC? star*net? LGO ? some kinda data collector?
That's exactly what I was going to say, almost word for word. The results actually seem very good for the methods used and as ps8182 says, if the traversed distance from the initial setup to the check-in point in the middle is reasonably close to the value derived from the RTK co-ordinates, a little rotation holding the middle point and initial point will be like having a 2.5-mile backsight and I'll bet the initial backsight point 700' +/- from the initial setup comes in pretty darn close to the RTK co-ordinate.
I'm not saying this applies to the work situation of the original poster, I don't know enough to comment on that, but in general, with modern methods that have emerged in the last 20 or so years, I'm sure a lot of surveying firms don't have a person like Loyal who knows this stuff forward and backward. When I was starting my surveying career, the hardware was a transit and steel tape and the software was a pad of paper. The old-timers had used this basic technology for decades and could usually tell quite readily how to resolve problems because they had made every mistake known to man and found a way out of it. Just an observation from an old curmudgeon.
Ralph
Assuming the control points were at the only gps conducive areas, then sunshots to help control the bearings along the way is a great practice still in use today. For pipelines or other route surveys, there really handy.
We did them every mile to adjust bearings by. No flame took bro. You and me been straight a long time. 🙂
Your bro in arms 🙂
Static sessions can eliminate minor multipath.
I agree as well...It actually doesn't sound that unusual (around here) when starting from RTK pairs to be off a few tenths (over that distance). If you have the scale factor situation figured out, I'll bet the distances compare favorably between GPS and total station.
In AutoCAD, you can simply rotate your traverse graphically using the "far apart" GPS points. For me this method works great for long "corridor type" traverses (pipelines, roads, etc).
Naturally you'll want to go back and re-check, but I bet this method will tighten things up substantially.
As far as the vertical error...that could cause some concern...maybe a bad HI or something
Ralph
No worries Kris, you're still my Bro.:0)
Ralph
Just a little redneck math here...
You went east 13,200 feet or so traversing from the first pair on the west. You missed northing on the middle pair by 0.45'. That works out to about 7" bearing in 2.5 miles. Back at your west pair which are about 700' apart that works out to 0.02' due to the bearing error which is what you can expect from RTK.
Am I missing something?
Do I understand what you are saying?
Are you saying that you set;
1 RTK points at the begining of your job
2 RTK points in the middle
3 RTK points at the end of the job
You then traverse through the rtk points.
You did not hit you RTK values for the middle points
But,you did close well on the RTK points at the end of the job.
Is this what you are saying happened?
Simple things....
What version of Survey Pro are you using?
It is most likely just a coordinate system error...
SPC with the GPS and Scale Factor 1.000 in total station.
I agree with the scaling opinion.
Your DC can be set to either scale from State Plane (your RTK answer) or to State Plane (your Total Station answer).
You just decide which you want to work in, and set the scale factor to apply to the observations of that instrument.
In Carlson, you can choose a number of RTK observations, and have it calculate the combination factor that you would like to have applied, it is pretty slick.
Kind of hijacking this thread...RTK Standard Proceedures
> Oky doky:
>
>
> that I see between what you did (or I assume that you did), is that they use 5-15 (or 30) minute “RTK Infill” shots (a Trimble term I believe), AND they perform multiple observations (run South to North, and then repeat).
There is some interesting research out there that indicates observations over about 3 minutes produce very little increase in precision or accuracy. What gives you an increase in accuracy is to repeat the observation after "breaking lock" and allowing your receiver to resolve the ambiguity again, and then taking additional shots, and averaging/analyzing them. The BEST is to make those additional shots after a time break which allows the constellation to change its geometry.
His field practices appear to have accomplished these accepted practices, but...
> Now the “traditional rule of thumb” on RTK shots, is about +/- 0.010 meters (a centimeter), and in my experience that is pretty reasonable, even at 2 sigma (ASSUMING all is well). With a 700 ft. (~215 meter) backsight, that COULD mean an angular error of 15-20 arc-seconds from the get-go. Lets face it....you could have .010 meters one way on the Trav-station, and .010 meters the OTHER way on your backsight 700 feet away (it happens). That could throw you 1.23 feet in your 2.5 miles right there!
>
This is probably more likely the case. The better question to ask this guy, is not how much did you miss, but HOW did you miss? Angle? Is your inverse distance the same?
> Now all of this ASSUMES that you are making the proper environmental corrections (temperature & barometric pressure) at the Total Station, have NOT done some lame-o “calibration” to the GPS data, HAVE used the same coordinate system/projection in both data sets, AND made SURE that your tribrachs, rod bubbles, etc. are in CORRECT adjustment.
>
> All in all...700 foot traverse legs are kinda short,
Spoken like a guy who lives and works in the open country!
>
> Given 4 hours, and two receivers, I would have opt'ed for a modified leap-frog static (or fast-static) approach to the problem. That would have given you a LOT more data to play with. Of course the REALLY HOT SETUP, would be to have 6 receivers and then you would be walking in tall cotton!
>
And thousands in the budget to post-process it 5 ways from Sunday! This is actually my favorite, the results give me great pleasure in whatever part of the brain likes really tight control work. But the pocket-book says, "Fire up the network RTK and get me a coordinate RIGHT NOW!"
That's an easy one... Loyal
I would say coordinates for GPS geodetic and coordinates for TS plane.
The difference due to scale factor not being applied.
RADU
You're not missing a damn thing. Pretty funny reading all of the early replys here.
Until he rotates TS traverse to the GPS points and reports the distance matches he and we don't know anything. He is most likely perfectly fine.
Stephen
> You're not missing a damn thing. Pretty funny reading all of the early replys here.
>
> Until he rotates TS traverse to the GPS points and reports the distance matches he and we don't know anything. He is most likely perfectly fine.
>
> Stephen
You're probably right, this guy's probably the field operator and posted this in a panic.
Ralph
That's an easy one...
> Are you comparing these coordinate values as SPC or to a local datum? Those departure values seem excessive for just a scale factor bust...
I would say it's a combo of scale/rotation, I believe it's probably inexperience.
Ralph
For the original question: If you have to ask it is apparent you need more experience working under a qualified professional to understand the many procedural and mathmatical relationships in play with an apparent attempt at acheiving high order results. To achieve high order results with consistency, you first have to understand completely the methods, procedures, equipment and mathmatics involved.
Traverse accuracy and precision is the base upon which a surveyor is built. To not understand that is a sure sign the basics have not yet been learned.
After you set your GPS points did you compute a combined scale factor to apply to your total station traverse . In my area of NC in 13,000 feet I would be looking at .06' per 1000' so if I did not apply scale factor I would see around .78' in horiz error in 13000'.
For your elevation error I would set more checks along the route . Trig leveling can be a problem on long traverse loops . You need to keep the traverse legs under 700 feet for elevations to be decent. Your GPS has better elevations over a longer distance than your conventional traverse . Go Back and do static GPS session on the control points . You will have a great elevation check .
Double all angles on your traverse . run all through least squares .
Are you able to RTK you traverse points from the beginning to the mid-point? If your data consistently gets worse as you progress, it's probably a scale factor issue.
Combining EDM traverse data with gps data over a long distance is tricky. I'm assuming you set up you GPS autonomously, calibrated to known control and then took off with you traverse. Set up a new job for EDM work and set the scale factor to one. You may be able to manipulate the data, but if it were me, I'd eat it and run the traverse again.
....sunshots
Have we learned yet what the traverse distance from the initial setup to the midpoint is compared to the RTK value? Why would you want to re-run the traverse if all you had to do was rotate the traverse about 7"? Then keep on running the traverse the second 2.5 miles and see how that works out. Or just run the whole darn thing unadjusted and deal with it all at once. I don't see a big problem here, but I guess it depends on the project specs and such. The vertical sounds like more of a factor to deal with than the horizontal but I've found that running careful trig levels results in better vertical data than horizontal most of the time, when your equipment is adjusted and set properly.
That is a good point. Being such a long traverse, he prolly doesn't want this screwed up. I guess I'd rotate the first half, run the second half traverse and see what happens. Rerunning the whole thing would give me a little piece of mind
What is this project anyway? Is trig leveling really what you want to do