> Would appreciate any suggestions on best ways to proceed.
Unless I'm missing something, the very first thing to do is to figure out why your client wants the design survey mapped in state plane coordinates. Is this a DOT requirement? Do they need it to be able to combine other imagery and/or data with your mapping?
Most of the time a LDP will be say a Transverse Mercator Projection or with a semi major axis and flattening numbers which will set up your system to be at sea level causing a adjustment to be made to "get it up" to the surface you want it to be at. The way computer programs I've worked with handle this is a scale factor. Probably there is a better way to do it, but for most this seems to be the simplest solution.
You would make Dr. Herb very happy;-)
Big Al
Again...yes and no.
A N/E coordinate value can be considered a “distance” (dN/dE) FROM the 0/0 coordinate. In most ALL cases, that is a DIFFERENT point, than the Projection ORIGIN (point).
This can (and does) get a little confusing, and there is more than one way to “scale” (modify) State Plane Coordinates (some worse than others).
I am of the opinion that anytime you have to “scale” your coordinates to satisfy your NEEDS, then you are using the WRONG coordinate projection to begin with.
But that's just ME!
Loyal
OK, Moe, so here's an example. I just want to make sure I understand the method that you're using for your DOT projects:
I've got a nail in the ground, on which I ran a short GNSS session on Sunday. Submitted to OPUS and I get the following results in State Plane:
Northing (Y) [meters] 869108.516
Easting (X) [meters] 47399.742
Convergence [degrees] -1.23848884
Point Scale 0.99996765
Combined Factor 0.99994115
So, you'd take the above grid coordinates (Y,X)
Y = 869,108.516 x (1/.99994115) = 869,159.666 meters
[times (3937/1200) = 2,851,568.004 feet]
X = 47399.742 x (1/.99994115) = 47,402.532 meters
[times (3937/1200) = 155,519.806 feet]
Effectively, a horizontal scaling. I could follow this same procedure for multiple coordinates at which I conducted GNSS sessions, and the resulting inversed distance between the two points would approximate the ground distance between the two points.
The disadvantage is that I wouldn't be working "in the State Plane", so if my client wanted to overlay the CAD work onto aerial imagery, for example, it wouldn't look right, especially in the Northing.
It does look correct, but look at the numbers. The scale factor is quite small.
I'd really consider just using state plane on this project. .06' per 1000' isn't much. .6' per 10,000' is all that you're dealing with. I never see numbers like that.
You should really think out the advantages of tweaking state plane for this situation. There may not be any.
> It does look correct, but look at the numbers. The scale factor is quite small.
>
> I'd really consider just using state plane on this project. .06' per 1000' isn't much. .6' per 10,000' is all that you're dealing with. I never see numbers like that.
>
> You should really think out the advantages of tweaking state plane for this situation. There may not be any.
Moe, that's a good point. So, if I wanted to work in State Plane, would it be correct then to apply a scale factor to my work in the field?
I would start with this:
http://www.oregon.gov/ODOT/HWY/GEOMETRONICS/docs/ocrs_handbook_user_guide.pdf
So, if I wanted to work in State Plane, would it be correct then to apply a scale factor to my work in the field?
No, don't do that!
The scale factor is informational on the OPUS printout. Keep in state plane and don't apply scale factors in the field. If you want a ground system then you are no longer in state plane.
A properly set up LDP will have an average scale factor of 1 for the project elevation. This allows the drawing to be drawn at ground and still be connected to the geodetic.
OK, then, so I think you're suggesting that where the project size is small and where the combined scale factor is negligible, that it might be appropriate to hold the grid coordinates of one point which is central to the project and run everything on ground from there. That, of course, means that only one point is "on the grid" but the differences are negligible. It would allow the engineer to use aerial photos, etc., and as long as the methods used were disclosed in my submission, would allow any user of the data to understand what was done....
> I get nervous about scaling CAD drawings. particularly 3D CAD drawings, because it will eff up the elevations by said scale factor (unless there is some 2D only scaling routine I'm unaware of).
Would/Could the AutoCAD command ALIGN perform a 2D only scaling?
Moe
I've never seen such a thing in these parts, but your mileage may vary.
A Transverse Mercator or Single Parallel Lambert can (and should) be designed such that the “developed surface” (GRID) is set to whatever “elevation” (actually ellipsoid height) that suits your purposes.
Screwing with semi-major axis values, and/or flattening numbers, “divorces” you from NAD83, and constitutes a “custom ellipsoid.” Mike Potterfield once commented that “using a custom ellipsoid to define an LDP, is like using an H-Bomb, to kill flies” (or something along those lines).
By using a “scale 'reduction' value” ('k') that is greater than 1 (Transverse Mercator or Single Parallel Lambert), you MOVE the developed surface (GRID) UP to whatever height suits your needs. Here in Evanston, the 'k' value would be something like 1.000 333 (elevation ~7000 ft.), and you are still "in" NAD83.
Loyal
No I don't mean that. What I do mean is to stay on state plane for everything.
If the "error" is small enough then it really shouldn't matter for roadways and sewer construction. For large bridges and structures that .06' per 1000' may be too much and you would need to consider a ground system.
I may have misunderstood your prior question. Are you saying that your data collector and instruments don't automatically convert each measurment to SPC? If that's the case then you would have to srink each distance to properly calculate state plane coordinates. That would be a slow and unproductive way to work.
Physical fact - meridians are not parallel.
Basic concept - the difference between grid north and true north is determined by selection of the central meridian. To keep LDP projection grid north identical with SPC grid north one would need to choose the same principle meridian - thus obviating one of the reasons for choosing to use a LDP. Suggestion - use true north through the selected P.O.B. on the GSDM.
I scanned through all the answers and may be missing something....but it sounds like the designer uses state plane coordinates and designs his project as though it were on the ground....so that something meant to be 300' long he designs @ 300' long, and doesn't reduce it to a distance of (say) 299.98. If so, I might suggest that you use just the state plane coordinates as though they were on the ground surface. If you already have coordinates on the "ground" use a point that is at the center of the project as its own state plane coordinate, and convert all the other points radially around that by your scale factor.
Did that make sense? I would be very clear in my notes how you compensated for the confusion.
Interesting topic. Why would they want to design and build in State Plane Grid Coordinates? The local DOTs around here have required that also but we always create our end products such as topo and right of way data 'on ground'. Is the existing topo survey completed and prepared on grid? Or does the Engineer even know? Or are you setting up to work on the Topo?
Your 50 foot dimension at 1500 feet would not change enough at grid or ground to be noticeable to anyone not using a micrometer.
Regarding your CAD question: Do all your CAD work on Ground or on Grid, pick one and stay with it. Scaling of any type in Civil3d is sure to be a difficult and costly maneuver. You can adjust your point data elsewhere if you need to switch between grid and ground values. CAD has the ability to switch between Grid and Ground using the Transfomation settings but it has proved mostly futile in actual application. Your either on ground or on grid...depending on the coordinate type you are starting with. We always scale our point data to ground prior to any work being done in Cad or Microstation.
If your scale factors are small and the project is small stick with State Plane Coordinates (if your already too far along that you can't establish a ground system)
Just thinking out loud.
Even if you have a long project going in the most "distorted" direction.
(ex. N-S in a Lambert system)
Wouldn't you have a peppering of "precise" State Plane control points along the job for radial staking? (maybe every 500'-1000')
Even using uncorrected EDM distance would only affect the points located from that particular control point. These errors are minimal as you would never ray out a point more than say 500'.
You are not traversing from begin of job to end of job using ground distances accumulating error along the way. I don't think the "error" between grid/ground over several miles is relevant.
Even in a worst case scenario ray out of the theoretical 50' bridge would still produce a bridge that was really close to being 50.00'.
It's the SF applied against the difference that's the issue.
Check my math
Fairly severe scale factor of 0.999900 (i.e. one tenth per 1000')
Worst case is ray-out along axis of bridge. Gun is 1000' away.
1000.00 * SF = 999.900
950.00 * SF = 949.905
1000.000 - 950.000 = 50.000
999.900 - 949.905 = 49.995 (i.e. one half hundredth)
Even if each end of bridge had been staked from a different control point I doubt the error would be significant but I'd have to do the calcs to be sure.
Absolutely, a great question. One that derives from a requirement of the DOT here, and the need/want of GIS folks to be able to incorporate projects into a larger framework. Engineer might have some benefit of being able to use aerial maps, and other GIS data, etc. during the design effort.
> No I don't mean that. What I do mean is to stay on state plane for everything.
>
> If the "error" is small enough then it really shouldn't matter for roadways and sewer construction. For large bridges and structures that .06' per 1000' may be too much and you would need to consider a ground system.
>
> I may have misunderstood your prior question. Are you saying that your data collector and instruments don't automatically convert each measurment to SPC? If that's the case then you would have to srink each distance to properly calculate state plane coordinates. That would be a slow and unproductive way to work.
OK, now I'm confused. When I'm doing GNSS work, the data collector will automatically convert to SPC. However, when I'm doing total station work, unless I specify a scale factor which is to be applied to each shot, then the shots would be "on ground" (scale factor = 1).
So, I thought you meant that I could hold the state plane coordinates of one point on the project and run on the ground radially through the project, without applying a scale factor. Do you mean, rather, that applying a scale factor would be appropriate?
As I see it, to provide a product which is truly "on the State Plane", I would need to apply a scale factor when doing the field work (or, I suppose, working on ground, and scaling everything down just before sending to the engineer).
Thanks for sticking with me on this one....
