Since there is currently a lot of heat about this, let me fan the flames:
Actual factual example:
One day we ran a 19 leg traverse through the controlling subdivision monuments, and through our plat, including closing the loop. Because of a ravine, to the west and east, the figure was basically a "U", with each leg of the U being about 1000' feet.
The traverse was balanced in Starnet by an experienced and expert user.
Later that week, I went to the site and used this control traverse to create a localization for some RTK work (not setting of plat monuments). I returned to tell the PLS that the traverse had a rotational error that resulted in the two north points of the "U" being off by about 0.15'. How did I know this???
So, who cares might be the answer, since 0.15' will never be missed in a ravine now dedicated to open space, but the decision was made to traverse across the "U" and strengthen the figure...and guess what...it was a 0.15' miss.
Now...if one cared about the accuracy of each plat monument in relationship to the controlling subdivision monuments, I submit that you would use RTK and a localization that included these monuments.
If one cared about having the monuments in the road be all off by the same amount, (with LESS accuracy BUT...they would match record distances/angles between each other very closely) you would use a traverse and total station approach.
The old guys looked and me, looked at my "black box"...and ran the traverse across the ravine, and used a total station to set the monuments. (Which were installed by a fly by night company hired by the contractor, and are all spinning freely now, so @#@$##%% on us for caring at all.)
My conclusion:
There are some times when using RTK to set centerline monuments and final corners in a plat is the most accurate and reasonable way to monument a plat.
Sooo...let the flames begin!!!
> Later that week, I went to the site and used this control traverse to create a localization for some RTK work (not setting of plat monuments). I returned to tell the PLS that the traverse had a rotational error that resulted in the two north points of the "U" being off by about 0.15'. How did I know this???
>
So, basically, you took any "error" in the conventional traverse and ADDED it to any error in the RTK work by using a localization, and are wondering about a few hundreths?
You lost me at localization.
Not something I will do.
But otherwise; mixing GPS and instruments, that I do all the time.
> You lost me at localization.
>
> Not something I will do.
>
> But otherwise; mixing GPS and instruments, that I do all the time.
Yeah, the obvious way to have done that was to add GPS vectors to the traverse network to tighten it all up. We do that all the time on just about every project, i.e. mixing conventional measurements and GPS vectors. The method described sounds like the worst of both worlds.
In 1990, I proved another crew's static GPS control wrong using a T2000 via a resection. At the time I had been working there less than 6 months and knew very little about GPS (having turned on my fist receiver just in June of that year), I did know when the conventional work didn't fit the GPS control points done earlier that year that something was wrong.
They are all just tools, poor survey procedure is poor survey procedure rather it be traverse, static GPS, or RTK.
I agree with Kent on this one, if I am going to prove something wrong, I want to use the best method possible AND static GPS vectors are the best way to tighten up sloppy conventional work.
SHG
My conclusion: an experienced surveyor knows the value of cross ties in a long "U" shaped traverse.
> > Later that week, I went to the site and used this control traverse to create a localization for some RTK work (not setting of plat monuments). I returned to tell the PLS that the traverse had a rotational error that resulted in the two north points of the "U" being off by about 0.15'. How did I know this???
> >
>
> So, basically, you took any "error" in the conventional traverse and ADDED it to any error in the RTK work by using a localization, and are wondering about a few hundreths?
No, I did not. What are you talking about? My analysis of the control via RTK revealed the error, and error is never in quotes. When you calibrate or localize, the underlying "tightness" of your control becomes evident very quickly.
> My conclusion: an experienced surveyor knows the value of cross ties in a long "U" shaped traverse.
And he knows that a 19 leg traverse is much less accurate than RTK.
> In 1990
Dude, so what??? Comparing your work with GPS in 1990 with today is like...I cant even come up with anything...I cannot believe you are serious.
You must be making a joke, because it is like saying that you listened to music on your 8-track, and it kind of sucked so anyone that uses an ipod is stupid.
Or better yet, you had a brick phone in 1990, and man it sucked, it dropped calls all the time! So, anyone that uses a smartphone and trusts what it tells them is just silly.
Any others that anyone can think of?
> Yeah, the obvious way to have done that was to add GPS vectors to the traverse network to tighten it all up. We do that all the time on just about every project, i.e. mixing conventional measurements and GPS vectors. The method described sounds like the worst of both worlds.
The method I described was to localize to the subdivision, and set the points. How is that the worst of any world? The accuracy would exceed your example in most real world situations.
> You lost me at localization.
>
> Not something I will do.
>
Never?
Do you only work on State Plane Grid?
My point is, state of the art in 1990 can be wrong and so can state of the art in 2012 or any other year. You stated you proved a traverse wrong with RTK, I stated I proved static GPS wrong with a total station, the year really isn't relevant (and since 1990 totally threw you, I apologize, just read what I posted without the year, does it change the facts?)
Is RTK so good in 2012 that it is the best tool 100% of the time, no, of course not, neither is a total station, nor any other tool.
Point being, no one tool is right for 100% of surveying tasks, all measurements have error and using the right tools or combination of tools appropriate for the task at hand is the correct thing to do.
SHG
> Or better yet, you had a brick phone in 1990, and man it sucked, it dropped calls all the time! So, anyone that uses a smartphone and trusts what it tells them is just silly.
Well actually as far as a phone goes, you know, one of those devices for talking on instead of a million other tasks, the brick phones of 1990 for the most part were actually much better than a current smart phone for voice calls. And no, I am not still using a brick phone, have since upgraded to a smart phone, now it is dated too, but it gets the job done 🙂
SHG
yea, you should get some red/blue tights with RTK sewn on the chest.
If we are talking about what I understand localization to mean; I've been told Lieca calls it that, while Trimble calls it calibration.
Set up the base on a point, locate another point, or a number of points, then have the data collector "calibrate" the coordinate file to "fit" the GPS data to the control points-then no I never do it.
However, you may mean something else, I've had people tell me they localizated a project and what they did was a one point location with no projection.
I used to calibrate or localize (back in the day) pre 21st century-but not any more.
And no, I work at too high an elevation to use state plane grid very often. But state plane grid isn't the only projection available.
I've even heard of some crews doing a "calibration" each day-one after another. That kind of makes my skin crawl. I can't beleive that is a practice, but I've heard it's done.
Calibration is a very handy tool. One party chief told me if you miss your control point just include it in the calibration and all is well. 🙂
One party chief told me if you miss your control point just include it in the calibration and all is well.
I don't care who you are; that there is funny!
I went out with another company to decide together about a Section corner. I had my base running about 1-1/2 miles from the corner on a system that I had been working in for years, I checked into a point and went to meet the other crew. The party chief walked up a little rise next to the road and took about a 1 second shot.
I asked if he was checking in and he said no "orientating".
I didn't ask-I didn't want to know.
