Contractor I work for was awarded a state contract for widening an existing state road and adjoining side streets in a nearby college town. About 1-1/2 miles long running west to east, earthwork, dga, drainage, curb and gutter, asphalt. Plans have a control sheet with coordinates for control points and centerline geometry, like a hundred others I’ve worked on. Has this statement:
Coordinates for horizontal control were obtained from GPS methods and adjusted to the NAD83/1994 SPC System.
A project datum factor of 0.9999979631 was computed from the central most point to convert the SPC to Project Coordinates. Coordinates shown are on Project Datum, to obtain SPC, divide the Project Coordinates by the project datum factor.
The SPC coordinates and the project coordinates check mathematically using their project datum factor for CP1 and CP2, which are the only points showing both sets. These 2 points are at the very east end of the project within sight of each other. There are about 25 other control points shown on the same sheet, but they only have project coordinates, as does all the centerline geometry. So I converted all the other control points into SPC using their project datum factor, loaded them into the controller, and went looking for them. Using my TSC3 with Access and R8 with VRS, I found and checked CP1 and CP2 within a couple hundredths. I could not find a single other control point listed. Some have obviously been destroyed by other construction, but there were a few in open yards that should still be there, but nothing.
I called Jeff Clark with Precision Products because I knew he had done some work with Ky. DOT on how to get onto project coordinates with GPS using the project datum factor. He sent me a pdf with the details, I followed his instructions, and in another job file with only project coordinates, I again checked CP1 and CP2 within a couple hundredths. I could not find any existing point identified on the plans that I could check. One of the side streets has curb and gutter that new radii should tie into, so I marked those 2 radii and missed the existing by about 5-6 ft, shifted to the east. Checked another side street wshere the new centerline appeared to match the existing, and again was shifted 5-6 ft. to the east. Hmmm.
So on a hunch I changed the project datum factor to the reciprocal of that shown, checked the street centerline and radii, and was hitting where I should have been. Went back to CP1 and CP2 and missed both about 8 ft. Thought maybe now I would find the other control points, but still nothing.
So I emailed the highway dept. in charge of this one and laid out what I found. They responded that the District Surveyor could meet me there to “discuss ideas, etc.” Uh-oh. Asked if he knew what the problem was, and they said he never had much success either and had been using robotics but not GPS. Said he might know of some other control points. And that the project has been in the design stage for about 25 years and therein lies the problem. They had been staking R/W for utility relocation over the past couple years. I’m betting theres some stuff where it’s not supposed to be.
to obtain SPC, divide the Project Coordinates by the project datum factor.
I'm no expert, but that doesn't sound right to me. As I understand it there has to be an origin point for the scale, probably the same point they used to determine the average combined factor. It seems to me that the scale error would be zero at that origin point and would increase with distance.
I'll NEVER understand why they wouldn't provide SPCS for all of the control points. The whole point of using State Plane is that it can be easily re-established as long as a couple control points are still extant.
Sounds like you have a real mess, correcting it may be impossible, you will have to rely on the DOT to figure out just what control was used to establish R/W takings, mapping, ect. Is there any way to tie into non control such as property corners, section corners that kind of monument?
> Sounds like you have a real mess, correcting it may be impossible, you will have to rely on the DOT to figure out just what control was used to establish R/W takings, mapping, ect. Is there any way to tie into non control such as property corners, section corners that kind of monument?
Good luck with that.
This is a good example ...
of why you NEVER EVER multiply coordinates by a CSF!
This is a good example ...
:good: :good:
Central Most Point?
The central most point could be the calculated midpoint of the East and West control points or it could be a control point nearest the midpoint.
The given scale factor does not seem correct to go from SPC to local.
First check is to occupy the East and West control points with static GPS to verify the correct factor. It could be n incomplete correction or miss-applied correction. You will have no idea how many times the local coordinates were miss-corrected over the many years.
I would say it may be up to you to re-establish all new local control points.
Paul in PA
"...and they said he never had much success either"
But they issued the plans anyway?! What morons... Change order #1 coming right up!!
There are two conflicting statements: one that the project was scaled from a "central point" the other that ground coordinates could be converted to state coordinates by dividing them using the scale factor.
Dividing them implies that the coordinates are calculated from the origin point (0,0).
If so, and if the coordinates are around 1,500,000, then I can almost guarantee that there are some coordinates using us survey feet and another set using international feet.
As others have said, scaling coordinates by simply multiplying them by some factor scales them about the origin 0,0. Think about it 0 x Scale factor = 0, so that is the point that does not change. That is not the correct way to scale a job like this. That number can probably be applied to distances inversed between two sets of SPCs to get ground distances.
Why don't you plot the two that with known SPCs in CAD? Then plot all the other coordinates in the local system. Scale the local coordinates by the scale factor, and fit them on top of the SPC control. That will probably relate the local system close enough to SPC find any physical points that may exist on the ground. Then you can try to identify further problems, which will inevitably exist.
If you are hitting within 6 ft of some points, couldn't you find more of the missing control by going to where your GPS says and swinging the magnetic locator?
The more control points you find, the better your chance of figuring out where the scaling went wrong and being the hero who straightened it out.
It sounds like they derived the factor at a central point and applied it to whole ordinates. The error described is nominally equal to the shift between US and international feet in my area. Not a rare mistake given the default settings of many programs.
The other issue I see is the formula for conversion. SPC / CAF = nominal ground. While not perfect it is fine for a project of this size if relief is not severe.
As for not scaling coordinates I simply don't agree. If I provide correct data and metadata there shouldn't be a problem. Shifting between grid and nominal ground is 6th grade math (even here in idaho). With DOS long in the rearview we can assign meaningful file names with ease. I've used this method for most of my career with no problem.
My .02, Tom
Yeah, me too, I've been using them for over 30 years now and no problems, not sure why it's an issue. As long as the DOT gives you the metadata then everything will work.
However, if international feet is used it can cause problems, the first time I saw that was staking a short tie-in street between two highways for an in town project that tied to a new DOT highway. Engineer No.1 did the highway with a Scale Factor, Engineer No.2 later did the short tie in project using the same Scale Factor (around 0,0) but he used international feet and declared that Engineer No.1 was "off" by 5-6 feet.:-(
I got in the middle trying to figure out what was up. And it didn't matter if you looked at state coordinates or surface coordinates they were both different for the same point.
In my state US survey is what is to be used, I think other states are different.
We have Oregon next door at I feet, the rest I work in are USFT. As an aside, are you by chance from Missouri?
25 yrs of data huh? I wonder how much of it is out of date?
Time to submit a proposal, to RESURVEY it, on one datum!
N
Already did that and found nata.
As for the scale factor issue, Ky uses 0,0 as the origin for scaling. The scale factor for each project is based on what it would be in the middle of the project. Then any project will pretty much tie into any other project. You just figure out the Lat-long of 0,0, put the scale factor into the controller, say they are ground coordinates with keyed in scale factor, and assuming they gave you correct data, voila, you are on project coordinates. As evidenced by the fact I hit the 2 control points using both SPC and scaled. So the scale factor seems to be correct. The location of the project from the standard parallel means that SPC should be slightly larger than project coordinates.
I'm guessing at some point somebody shifted everything in Autocad, or whoever set the targets for the aerial mapping did a "here" instead of using SPC. Most of those extra control points look like they may have been target locations, and maybe they never set anything there, just located the targets.
Gonna meet Monday to discuss.
Not gonna happen. Work order already received. Boxes and pipe ordered and being delivered. Traffic control up. This is not unusual that there would be some screwup with the control vs mapping, but usually not too hard to figure out a solution. This one may take a day or 2 to work out.
That's why we get the big bucks.
"This is not unusual that there would be some screwup with the control vs mapping..."
Well it certainly ought to be unusual!! If they do not know what they are doing someone needs to call them on it...
Your original post still leaves a lingering question. The CAF you posted is less than 1 and you stated you divide ground by the CAF to get Grid. Later I see you are in a 'ground is smaller than grid' location. Are you using 1/CAF? That would make sense when its all put together, otherwise I'm not following you...
Is there a difference in the meanings of "project factor" and "scale factor"? I seem to remember discussing with MightyMoe a similar situation where the software used the reciprocal of the "scale factor" as the "project factor."
Maybe I misread, but I think you said that you hit two radius points by using the reciprocal of the "project factor." If this is so, then two possibilities present themselves:
1. Project coordinates for points other than the two control points with complete data were calculated backwards.
2. Coordinates for points other than the points with complete data are actual State Plane coordinates.
With project coordinates, there's always the question, "Do I multiply, or do I divide?"