Probably everyone has seen this. We had a project to provide property lines, control and photo-control for two road reconstructions. An east-west road and the first mile of an intersecting road that heads south. They (county) didn’t want to spend money to run levels and “close” control for the project so we did state plane (modified to surface on the county system) and Geoid03 and set four control points with static. The road project looks like this:
Years go by and the contractor hires a surveyor to stake the road, we give him all the control with meta-data. Hand him the Lat, Long, ellipsoid height, northing, easting, and elevation for the four points. Also, the HARN points and bench marks we used to establish the control.
He starts staking the project and something goes very wrong. The road is higher than it was intended to be by 1’ to 1.5’.
Of course, It’s the control (our control) that’s “wrong”.
Since I didn’t work on the project, I was chosen to go out and check what happened. I told our other surveyor to not even give me anything but the HARN points and the bench marks. I went out and occupied all the points and did fast static on the four points. Downloaded everything and ran an adjustment (over a weekend of course). They all checked within a few hundreths of a foot.
So what went wrong? It appears that the contractor’s surveyor did a calibration or localization and somehow got “tilted” with respect to real elevations. Geoid heights change in this area sometimes as much as .7’ per mile so you need to be careful how you treat elevations using GPS around here. Why he tried to localize? I just don’t know-he should have been able to create a file, with the meta-data parameters, set on the control and go. Check in to the other control points and start staking.
Anyway, the problem went away; at least he found it before the thing got built.
I hope that they paid you for every minute you spent helping them! And then they paid for you and your family to go out for a nice dinner as a THANK YOU for saving their A$$. Do you think they will get any GPS training courses? No, probably not. Excuse me for being this way, but I remember a surveyor that messed up at least 50% of all the GPS work he did and he had been to GPS training four times!
Hey, I'm curious. If he localized on all your points, it should have been fine. Did he skip the southerly point? I had that happen once, localizing along a line. The error was dramatic!
No, I didn't charge for it. In some way I'm just glad to check stuff just in case there is a problem with what we did. I just didn't want to get into a he said, he said. The original static network had so much reduncency that I knew there wasn't a problem with control, but 5-6 years after; you just don't know what might have moved. The surveyor is a good guy and he may have not been all that involved, trusting his field guys. I think they figured out the problem pretty quickly, after I let them know everyting was good, and they probably spent a bunch of free time restaking.
Yeah, some training would be nice. The calibration stuff makes me nervous. I avoid it if I can, but, saying that it looks like I may be forced into calibrating some old control next week. I'll have to see how it works. Not looking forward to it. I hate trying to use old control, but sometimes you can't avoid it.
I think so, but I don't know. I'm just guessing. But in this area creating a calibrated plane is going to introduce error. The Geoid slope is so extreme in some places it looks more like a rollercoster than anything, so you need to really restrict a calibrated file to small areas. He may have just typed in a bad height or elevation, that would do it also.
> Hey, I'm curious. If he localized on all your points, it should have been fine. Did he skip the southerly point? I had that happen once, localizing along a line. The error was dramatic!
I would think the same thing. The project area is only a couple mile triangle, so localizing that shape should not be a problem for that project. If the guy only used the 3 along the E-W road, shame on him. He'd be better off just using the one in the middle, and checking into the other points. If not the whole thing seems odd.
Using GPS to check your GPS will just make you feel better. Do you still "really" know what's what? Maybe that guys localization was right and the control is goofed up? Probably not, but it still seems odd.
I once did a 5 point calibration along a line about 1/2 mile long, vert residuals were a couple hundreths. I went 500 ft away, perpendicular to that line. The inclined plane was 4 ft off, as determined conventionally. Very bad, but caught it and fixed it and never did that again.
Using GPS to check your GPS will just make you feel better
Maybe, except for this project all the topo, panel points and design were established using the GPS derived points. So, to build the road the control needed to be held. Which means Geoid03 and the given numbers. Otherwise, you won't match the topo. If you don't match the topo your quanities and slopes won't work, pipes won't fit into the existing flowlines and on and on. Apples to apples. As an aside; during the project we checked things conventionally which also checked.
:good:
Figured all that. ...just sayin
I often wish the world was still flat, but that is not the case. Unfortunately. Thus these types of scenarios.
If you want to work from an inclined plane, you calibrate vertically to all points. If you do not, you should calibrate vertically to a single point and only check the other points.:-P
> No, I didn't charge for it. In some way I'm just glad to check stuff just in case there is a problem with what we did. I just didn't want to get into a he said, he said. The original static network had so much reduncency that I knew there wasn't a problem with control, but 5-6 years after; you just don't know what might have moved. The surveyor is a good guy and he may have not been all that involved, trusting his field guys. I think they figured out the problem pretty quickly, after I let them know everyting was good, and they probably spent a bunch of free time restaking.
>
> Yeah, some training would be nice. The calibration stuff makes me nervous. I avoid it if I can, but, saying that it looks like I may be forced into calibrating some old control next week. I'll have to see how it works. Not looking forward to it. I hate trying to use old control, but sometimes you can't avoid it.
:good:
Yes, that's my thinking, too. Localize on a strong network of 4 or 5 or more; otherwise just use ONE and the geoid.
The inclined plane just doesn't work so well in some areas I work in. There is just too much change in the Geoid slope. So you can have a hump or hole in the middle. Trimble had a routine where you could calibrate and apply the geoid but it never worked all that well for me. I prefer to use Geoid models on a project and a projection. No localization or calibration. Geoid03 was a big improvement over older ones and Geoid09 has refined the accuracy. I'm really looking forward to the new Geoid model. Now, if I can get surveyors who stake projects that I've set control on to stop using newer geoid models and OPUS; that would really help!
>... a routine where you could calibrate and apply the geoid but it never worked all that well for me...
I may get beat up here, so blast away. I was always under the impression that if you are going to calibrate/localize, you NEVER apply the geoid.
You are creating a local hor & vert datum - so keep it that way.
that if you are going to calibrate/localize, you NEVER apply the geoid.
Back in the day, I went to training and the way it worked was to calibrate bench marks around the site AND apply the Geiod model to try and warp the results to model the surface better. For the projects I tried it on I just didn't get great results. Before the newer models calibrating to good bench marks was about all that would get the verticals close, but now the newer Geoids work well and I can't see trying to calibrate vertically except for small sites like a building staking. But then the Geoid usually handles them well anyway.
Thanks Moe. Like I said, using GPS to check your GPS is still a mind bender for me. And I did like the world better when it was flat.
Had lots of training, and I know what I know about that round world and always keep within my limits. But I'm getting on in life so new tricks are tough for this old dog. As they (NGS'ers) keep moving the EC/EF point and reinventing Geoids like some kind of hybrid potato, but forgetting about the old stuff that makes them good - like ketchup. 😉
Since I have been using "GPS" for over 20 years, the problem is fairly common with surveyors using the tool "GPS" without understanding it and the consequences. Various manufacturers software that utilize calibration methods vs projection vs localization etc. ....it goes on and on. When I first discovered "localization" which was a local calibration with local control with autonomous positions, I found the gps/software manufacturers said it was to sell the GPS product using old TDS software methods that surveyors were accustomed to using with their edm's...otherwise surveyors would never buy it!
Pablo
> ... the gps/software manufacturers said it was to sell the GPS product using old TDS software methods that surveyors were accustomed to using with their edm's...otherwise surveyors would never buy it!
>
That's the way I heard it too, Pablo. The manufacturers invented techniques which allowed them to teach new methods to old-school surveyors by using old-school terminology and thought processes. They figured it was easier than teaching them geodesy. All it's succeeded in doing is creating more opportunity for failure.
The sooner you figure out how to avoid localization, the sooner you'll become productive and confident in the use of GPS.
JBS