Ground Projection and Curvature of the Earth
> > I can see a time, maybe in the near future, as computing and modelling continue to develop that projections will become redundant and we'll be surveying, modelling, designing and staking out on a virtual earth, where all distances are 'ground'.
>
> Actually, I believe that we've been doing that since the dawn of time. The problem we have is overcoming our "flat earth" mentality.
Working in an X/{Y or N/E) numerical coordinate system, is a flat-earth mentality. You have parallel norths and a grid whether you want to call it a grid or not. Spherical coordinates are the only way you can truely get out of the 'flat earth' mentality.
Most every survey job I do, can be handled with x/y coordinates and inversing "flet-earth" bearings. But having said that, I agree that we need to keep in mind that when we go very far, we are "off the grid" so to speak, and need to figure things out when it's important.
The biggest problem I see is the gps junkie or someone who doesn't know how to confirm if their numbers "look right" or not. Many huge mistakes are made that are so far off, that a transit and a chain could have been more accurate.
Ground Coordinates
I hadn't heard of an LDP. Looks very interesting! I found this online:
http://www.oregon.gov/ODOT/HWY/GEOMETRONICS/docs/OCRS_Handbook_User_Guide.pdf?ga=t
LDPs
The term “Low Distortion Projection” (LDP) is reasonably new (I believe the term may have been coined by Micheal Dennis), but the idea and implementation thereof, has been around for MUCH longer. These types of projections have been discussed in great detail on the old RPLS/POB board for over a decade, but usually referred to as “custom projections.”
I started working with LDP(s) back in the late 1970s, and never looked back. Software support was non-existent (for me), until I wrote some of my own software in the early-mid 1980s. Sometime shortly thereafter, I discovered a program called PROJCTR, which was written by Norm Olson (which I still use quite often). [REDACTED] also has excellent support for LDPs.
Most EVERY RTK Data Collection Software now supports LDPs, although you sometimes have to be rather clever to set them up properly.
There were many years that it “appeared” (to clients) that I was using State Plane Coordinates (that's what they wanted as a deliverable), but in fact, ALL of the real work was done in an LDP and transformed before they ever saw it.
There are several LARGE Public Works Projects (Freeways and Railroads) in Utah that are mapped, designed, staked and BUILT in either NAD27 or NAD83 LDPs, and I haven't heard much in the way of complaints.
You can also UPDATE a well designed and realized LDP from NAD83(1986) to NAD83(HARN), to NAD83(CORS96) to NAD83(2011) without losing any significant accuracy or precision, and ALL of the Northings/Eastings remain essentially the same (+/- a millimeter or two).
LDPs are a powerful tool, and VERY versatile, but are still limited by terrain and your distortion budget. On some route projects, you can go for 100+ miles, on others, only a few miles, before the distortion becomes unacceptable. For retracement purposes, they are GREAT, because you can keep the convergency values at trivial levels, and a FOOT = a FOOT for most all practical purposes.
Loyal
LDPs
Thanks, Loyal! I'm excited to learn how to do this. I think I'll still have the same problem of grid vs. ground on this site, though. 20 sq. miles with 2000 feet of relief. But I can't figure out any other way besides manually calc'ing every point using dist/dist intersects and the ground distances reported by TTools. PITA. I tried to find a way to use the CGF for each point to modify each coordinate, then truncate the values. But that gave me insane answers. I'll try making an LDP and compare the distances to my "single-mean-CGF-modified, truncated state plane coords". The worst thing is to give a contractor coords on control monuments, have them get a different distance with their EDM, then get a call from the client asking why the project is on hold while the "error" is corrected. Not fun.
LDPs
On a project with ~2000 feet of relief, you are going to see ~48ppm (.048 ft. per 1000 or 0.25 ft. per mile) unless you can chop it into TWO (or more) Zones.
It all really depends on what your distortion budget is.
I have come to the conclusion over the years, that MOST folks (and projects) can live with 20-30 ppm, but much more than that can be an issue on some projects.
Of course there are still folks running around with Total Stations that DON'T own a thermometer or barometer, so what makes them think they can “see” a few 10s of ppm in scale?
Loyal
Ground Projection and Curvature of the Earth
> The problem we have is overcoming our "flat earth" mentality.
[sarcasm]We?
You got a mouse in your pocket?
[/sarcasm]
😉
LDPs
Dang, looks like Topcon Tools can't do this. Probably just as well considering how flakey TTools is. If I set the coord system to state plane and inverse 2 points 7000' apart, it gives me grid and ground distances. If I toggle on "grid to ground" and inverse those same points, it gives me the same grid distance but a ground distance 0.2' longer. They can't both be right. Back to square one...
Ground Projection and Curvature of the Earth
:good: :good: 😉
LDPs
> ....Of course there are still folks running around with Total Stations that DON'T own a thermometer or barometer, so what makes them think they can “see” a few 10s of ppm in scale?
>
> Loyal
:good:
LDPs
I'm not a fan of the term "modified State Plane" in any sense. I tend to feel you're either in State Plane, or you're not. I prefer to call it the "Project Coordinate System", which can then be tied to State Plane in a manner specified in the metadata (at least, HOPEFULLY the metadata is specified clearly on the plans). I've seen some horror stories about people who see "modified State Plane", and if not in the "truncated" form, they think they're using State Plane coordinates. Or even worse, they get multiple surveys from different firms, all using a different version of a "modified State Plane", and then make all kinds of major mistakes.
I'm planning on re-submitting a version of the "Grid to Ground" seminar I did at AU 2010 for AU 2012. This time, though, I'll probably split it into two separate one-hour classes... The first on the general theory, and the second on how to deal with it in Civil 3D. I got several comments last time from people who already understood the general theory, but just didn't know how to deal with it in C3D, and didn't need to sit through the first part to get to the second part.