I was asked by a DOT person how they could deal with the following situation. They have old survey data that is on an assumed system, at ground. The project is getting resurrected, and they want to do new mapping. Because all of the old comps are at ground, they want the new data at ground. But they want it in NAD83 "SPC" so the machine control people can use it. What I don't like doing is scaling SPC coordinates by a combined factor because it changes the coordinates by many feet, sometimes hundreds depending on where in the zone and how high up.
So, here is the idea I came up with. They have already GPS surveyed a bunch of the old points, and also GPS surveyed photo control. It is a linear project about 2 or 3 miles long
1) Pick two points (P1 and P2) one at each end.
2) From a point in the middle (P3), compute state plane grid inverses to P1 and P2.
3) Use the combined factor (grid and elevation) at P3 to scale the inversed distances (the distances will get larger). These will now be ground distances P3?P1 and P3?P2
4) Compute new coordinates at P1 and P2 from P3 using the ground distances from step 3
5) In TBC do a site calibration to P1 and P2, using the coordinates from step 4 and the observed GPS data
This will create a scaled coordinate system centered on P3, so the coordinates won’t change much. They can then map using the slightly scaled control.
They want to rotate and translate the alignment data, but not scale it to grid.
Any comments on this idea? Or maybe a better way? LDP is not an option, they want it to be as close to SPC as possible.
This isn't my project, I was just asked for advice.
As I have said many times, it is unthinkable to me to have scaled coordinates look like true SPC coordinates, due to the immense likelihood of confusion. The closer they are to matching, the greater the confusion.
Subtract off the millions and hundred thousands from the scaled version so it is trivial to convert but you will always know which kind they are.
If you compute a transformation set the old coordinates will be fine. Share the new system parameters with the machine control guys and everyone is up and running pdq...
"...they want the new data at ground. But they want it in NAD83 "SPC""
I don't believe this is possible... they are two different things.
my guess is that he means "they want the all the maps, sketches, alignments, and conveyances to be at ground but the actual usuable field data to be at SPC."
have done this quite a bunch in order to satiate the GIS requirements of various local entities.
the proposed plan sounds fine to me.
i've done it in the past (on typically more rectangular sites than i'm guessing is the topic here) by establishing a "square" set of control outside the site and establishing a midpoint based upon the intersection of inverses between opposite points. the results have been pretty manageable- so long as you keep your databases and CAD files well organized.
I think your solution is the best. It may not be mathematically elegant, but it's practical. It's probably close enough for machine control grading. And any "plane" old surveyor can come in with a total station and not make a mess of it. And it fits into the GIS.
You're not building a piano.
Sounds like that will work - but I would add some redundancy - at least another couple of points at either end of the job. And a couple midway for checking.
One question - how do you intend to control elevation? That's often the most important factor for roading, and the machine control guys will need at least one site benchmark for checking against.
Also document everything really well. There is some huge liability here when it goes wrong.
Compute a DAF, scale all the coordinates around 0,0, don't truncate, it's how all the DOT's I work with do it, it works fine, just keep records.
Most software packages handle it best that way, you can't truncate cause you lose the connection to the State Coordinate system.
This way you have State Coordinate bearings and "surface" distances which is what they want, believe me, they do this every day.
Call it working plane coordinates, surface coordinates, it's all good;-)
I have a spread sheet that lists all the numbers and how they work.
And it's waaaaaay simpler than a conversion point anywhere else when you start using it.
What you describe sounds to me like translating and rotating the ground scaled data so that some chosen point in the middle has the same coordinate as it's SP coordinate. Sure. It will work as long as the extents of the project are not great. I've seen that done plenty of times. One of the local offices used to do it routinely.
But why not create a custom projection with an origin that corresponds to the 5000,5000 based control?
> Compute a DAF, scale all the coordinates around 0,0, don't truncate, it's how all the DOT's I work with do it, it works fine, just keep records.
>
> Most software packages handle it best that way, you can't truncate cause you lose the connection to the State Coordinate system.
>
> This way you have State Coordinate bearings and "surface" distances which is what they want, believe me, they do this every day.
>
> Call it working plane coordinates, surface coordinates, it's all good;-)
>
> I have a spread sheet that lists all the numbers and how they work.
>
> And it's waaaaaay simpler than a conversion point anywhere else when you start using it.
We use this exact scenario on all projects, private and State Transportation departments. Like you said...it's how they do it. Works great IMO.
Hence the reason I put "SPC" in quotes.
I am always against scaling from the origin, even if you chop off 1,000,000 or something like that. But, sometimes clients insist.
In this case I am just giving advice because they asked me, I am not working on the project.
Like I said, not my project. They want it to "look like" SPC.
I was involved in two very large projects in the 80's where they did that sometime in the past, but that fact got lost. Then new control was brought in, and didn't fit by 100 feet or so. As soon as they brought me in, I knew what the problem was, but they had spent a lot of time before that trying to figure it out.
It even happened recently when I was asked to look at a highway project where two sections adjoined, and they had different scale factors. The surveyor working on it was saying there was an 8' bust in stationing, but once the proper scale factors were applied, it all worked. It was all on the plan sheets, but not understood.
So the point is that these facts get lost or are not understood.
I like what you suggest, but for ease of use I would just use the site calibration to continue on the legacy coordinate system. Then, when the project is finished just run the data backwards through the calibration to archive all of the cords as true SPC in the current realization.
Sounds like blasphemy to me. But I'm becoming a curmudgeon about these things and a purist. They'll probably still be able to build the thing (whatever it is) with this method, I simply prefer doing things the "right way".
Now wait just a minute!
I claimed the title of "curmudgeon" some time ago. And now you're trying to claim the title. I guess we have to meet at high noon some where. Squirt guns at 2 paces (about 12' apart)?
> ... They want it to "look like" SPC.
Do they also want 7 perpendicular red lines, 2 with red ink, 2 with green, and the rest with transparent?
Squirt Guns?
Why not REAL REFLECTORLESS GUNS?
100 meters apart; Guns in the cases on the ground. At the mark, set 'em up, level 'em up and whoever gets a lock on the other's gun first, wins.
There Is No Such Thing As SPC Ground
I am very glad I do not have to explain that to them.
Paul in PA
Squirt Guns?
That sounds like a plan.
