Encountered an interesting (troubling) scenario today while reviewing a survey by another surveyor. Roughly 20 months ago I had reviewed a nearly identical survey by the same company. Looking carefully at the new plat I noticed that it said "Revised to correct location of west quarter corner." I pulled out the previously recorded plat to discover the "new" corner was about four feet further to the north and about three feet further to the west than previously documented. My preliminary assumption was that they had recently discovered two nearly identical monuments that distance apart and had used the wrong one the first time.
Due to a significant error on the new drawing which the surveyor would want to correct prior to filing, I called him. What had really happened was that the GPS data for the corner was wrong the first time. They could find nothing to indicate that it was in error. However, a different firm was working recently on a nearby survey and had tied into the same corners. They could not get the location of the west quarter corner to match up with the record location on the 20 month old plat on file. The surveyor went out and reshot everything and agreed with the other firm's location. Somehow, someway, an erroneous GPS determination was wrong by the approximate distances mentioned above.
Short of shooting everything a dozen different times on different days under different conditions would you catch this kind of error on any GPS shot. In this area, it could be many years before someone else was tying into the same set of corners and discovering such an error for a high percentage of our corners.
> Short of shooting everything a dozen different times on different days under different conditions would you catch this kind of error on any GPS shot.
There are well-documented ways of ensuring reliable GPS positions that don't involve an absurd number of occupations. Some of them don't even require more than one occupation, though they do involve substantial occupation time.
My guess is that the blunder is the result of employing RTK in an inappropriate manner. That and a dose of negligence and/or incompetence.
How do you spell 'oops'?.......R.T.K.
> My guess is that the blunder is the result of employing RTK in an inappropriate manner. That and a dose of negligence and/or incompetence.
Yes, that sounds like a classic RTK misadventure.
SHIFT happens.....
Yes, a classic operator (or processor error). Whenever I hear that "the GPS is wrong" by a metre or three, the first thing I think of is that an autonomous position is involved (relative to other observations) or a float solution accepted.
In the 'olden days' with static data, many users would simply process the baselines and export the coordinates for use without doing a least squares adjustment. Without the adjustment, or alernatively setting a seed point before processing, it was likely that the exported coordinates were be the autonomous receiver estimated positions. Hilarity would ensue. Not.
Lots of money spent on equipment, not so much on training...
N 4' W 3' Sounds Like They Used ITRF Lat and Lon
Or NAD 27 instead of NAD 83.
Paul in PA
N 4' W 3' Sounds Like They Used ITRF Lat and Lon
Someone got an ace chewing over this one, if not fired.
N 4' W 3' Sounds Like They Used ITRF Lat and Lon
Not a GPS user myself, but I have to say that I have to give up some measure of respect to the surveyor who got called out and then corrected his f-up.
Seems mistakes like these will be more and more common as inexperienced crew members use RTK GPS. My guess is a crew used the "here" function to set up the base, or in Carlson did not have the correct localization file. I have seen this many times (luckily not from us). Both above "errors" are usually around 5'.
N 4' W 3' Sounds Like They Used ITRF Lat and Lon
> Or NAD 27 instead of NAD 83.
>
> Paul in PA
In this part of the world the error would be more along the lines of approx 50' in the northing and approx 30,000 ft in the easting.
N 4' W 3' Sounds Like They Used ITRF Lat and Lon
> > Or NAD 27 instead of NAD 83.
> >
> > Paul in PA
>
> In this part of the world the error would be more along the lines of approx 50' in the northing and approx 30,000 ft in the easting.
In this part of PA the error is several thousand feet too (NAD 27 vs NAD 83)...mostly an east/west difference
N 4' W 3' Sounds Like They Used ITRF Lat and Lon
I greatly dislike working in two datums. Almost everything we do for the Corps has to be submitted in both datums. The accuracy of Corpscon is only 12-18 cm so that's usually not an option. It's a matter of tying to legacy (NAD 27) control in NAD 83. Just have to pay attention to the scale factors. Sometimes it's like doing the job twice.
N 4' W 3' Sounds Like They Used ITRF Lat and Lon
The accuracy of Corpscon is only 12-18 cm so that's usually not an option.
You mean you don't just get an 88 number and crank out a conversion to 27 with a program?
You mean you actually tie into the 27 monument system? You mean you actually do a survey? Horrors!
When I did that (for an airport project) I was told my numbers were wrong because the person reviewing my work couldn't make them match Corpscon. I kept trying to explain that the Corpscon conversion was not accurate, but they just wouldn't hear it. Even though Corpscon itself says so. They said my numbers had to match Corpscon because it was offical!
I have tied first order NAD27 monuments with good 88 coordinates and run them through the Corpscon program and been 5' "off" the 27 number. Usually I see about .6' to 1'.
I quit looking at it years ago because I could never make the Corpscon numbers match very well, so what's the point?
Mainly I get stuck in NAD 27 world when I'm working with Mine control. Then it's hoping there is some control left on-site and that I can warp into it with GPS. But Corpscon never works.
N 4' W 3' Sounds Like They Used ITRF Lat and Lon
Most of the Corps projects' mapping is still in NAD 27. We have worked with others in updating some of it but most remains in 27. We have 7 "boundaries" we are working on right now. I say "boundaries" because it's more of just a re-monumentation. The maps are in NAD 27 so we just use Corpscon to get us close enough to do recon. That is about the best use I have found for Corpscon.
GPS works and has been working for decades to CM accuracy if you follow correct procedures.
GPS works and has been working for decades to CM accuracy if you follow correct procedures.
AMEN!Jp