# TDS localization

Discussion in 'Surveying & Geomatics' started by cboldman, May 5, 2012.

1. ### cboldman4-Year Member

Joined:
Mar 19, 2011
Messages:
41
0
Location:
Winterset,Ia
I posted a blog recently about trusting your data collector. Now I pose another topic of discussion "TDS localization". The terms "localization" and "calibration" are terms that have been thrown around for years by the survey equipment industry to explain "grid" to "ground" measurements. It seems most survey controllers have the means to perform the immense math to convert geodetic coords on the ellipse to "plane" grid coords that are scaled to ground. For me these grid coords are usually assumed coords based on a localized grid zone. Where i live in Iowa there are practically no GPS control points to calibrate on to set up a localized zone.

If i decide to use GNSS I usually just set the base, perform a "one point" calibration on the base point and go to work. Apparently where i live i can do this and still get within 2nd order accuracy (1;250,000?)within the range of RTK (about 2-3 miles). If i need to return to the same site then i can calibrate to my control points. In my software it "assumes" that ground and ellipse are basically the same and ground distances are scaled using the ellipse height and average earth center radius distance. My realm is around 41-1/2d N, 94d W. Here the height diff from ellipse is about 30m or 100' which should yield distortions in the range of 4.8 ppm according to the NGS.

Feel free to chime in and tell me that I am a bad Surveyor but be constructive in your comments.

Thanks

Joined:
Jul 1, 2010
Messages:
8,252
6,493
Location:
The Great State of Okie Homie
OK

Most of my work (35 N, 97 W) is extremely similar to yours. Although our GPS is providing some extremely tight locations..it is geodetic, not plane. The plane transformation occurs in the software.

Having started my survey career even before electronic distance measuring, I have always been skeptical of any measurement that was a calculation, not a true observation. Ground measurements of the relative locations provided by today's GPS provide very little, if any, distortions within most of our projects that are within a 2.5' mile radius, vertical and horizontal.

So we use it blindly on 'assumed' planes everyday. The only projects I have that positional error becomes noticable are long utility corridor projects. My most recent being 25 miles in a mostly east by northeast bearing. Operating on projected plane the coordinate positional error we noticed was in the neighborhood of about .25' horizontal in 30,000 feet. The vertical was slightly less. Although I don't believe it would have affected the overall integrity of the project, we contained the error in 2.5 mile control sections and worked from control within those sections. This was merely to minimize any discrepencies that construction might encounter using our established control for layout.

Whether or not we truly understand the capabilities of our equipment is actual secondary to the fact that they provide us good repeatable ground data. And for this old surveyor, that's all for which I need them.

3. ### LRDay7-Year Member

Joined:
Aug 26, 2010
Messages:
2,389
369
Location:
Sanpete County, Utah
UT
Most likely your localization is setting up a low distortion projection. Your base point that you calibrate to would be on the central Meridian and a scale factor calc'd to make your grid and ground distances equal at that elevation.

I don't know about TDS but in Trimble Survey Controller you'd be able to inspect the file and find out the projection parameters. Unless you do something like a OPUS solution on your base point or maybe use a state wide network to get the lat, long and height of the point you calibrate to there will be some error introduced.

You really should learn how to make your own projections and set them up, it's not that hard. On larger area projects its good to set up a coordinate system based upon a projection tuned to the area (especially the elevation). Then you are able to move the base point around and still have your grid directions the same. If you calibrate to a new base each time you are making grid north a bit different each time.

4. ### Loyal7-Year Member

Joined:
Jul 2, 2010
Messages:
2,540
1,170
Location:
Evanston, Wyoming
Different Data Collector Programs use DIFFERENT mathematical models (and projections) to generate and express these “calibrated and/or localized” coordinate systems.

Therefore, UNLESS there is quite a bit of metadata returned that fully documents what was done “behind the curtain,” it's pretty tough to “retrace” this work in a precise manner, or (more importantly maybe) to generate reasonably accurate geodetic coordinates from the returned data.

I think that this is probably a NON-issue for many Land Surveyors, and they are perfectly happy with relative “coordinates” that are unique to each site/project, and which are NOT [necessarily] related to the NSRS in any meaningful way.

I personally think that is a short-sighted position to take these days, but different strokes for different folks. I have found over the last 35+ years of referencing EVERY project to the NSRS, that it was worth every extra minute (or dollar) that it required. And back in the pre-GPS days, it was a LOT of extra minutes & dollars.

Like Leon says above, a formal LDP has all of the advantages of a "calibration," PLUS all of the advantages of SPC/UTM, while negating the disadvantages of NON-georeferenced FUGAREWE calibrations, minimizing the distortion of large scale projections expressed on or near the ellipsoidal surface.

Loyal

5. ### MightyMoe7-Year Member

Joined:
Sep 22, 2010
Messages:
4,890
1,162
You aren't doing anything wrong.
But, it's quite possible that you aren't getting the data that you want.
The distances you are inversing may be ellipsoid distances.
I don't know because I'm not seeing your numbers, but I do know other surveyors have had that problem when using similar methods.

The term calibrate really means tying into some existing xyz control and letting the dc controller program preform a kind of rubber-sheet adjustment.

It really doesn't take a lot of time to place any survey with GPS on a projection (assuming there was static data collected during the survey)-state Plane or one of your own. In fact, before even going to the field the data collector can be loaded with a projection that "fits" the area you are surveying in.

Then, even with a "here" coordinate on a new control point you can go. Later, the set-up coordinates can be further refined with an OPUS submission or you can run an adjustment yourself. This isn't gold standard control because it doesn't have redundant measurements, but it works well.

Before you send the numbers to a CAD program they are already on a projection and have a connection to a repeatable system. This can help save money and time.

I would say it's more time consuming to calibrate. And often more frustrating.

Joined:
Apr 15, 2011
Messages:
1,199
0
Location:
Mohave County, AZ
The only issue I see that could impact future work is the one point calibration/localization for horizontal. Me, I often start out a project in such a manner just to get started. Especially a one point vertical calibration.

But I'll use another point for horizontal calibration. That gives me a bearing basis and rotation. Assume a bearing, pick one off the plat of record, use the one your GPS god gave you, pick one from twixt thyne buttocks - it doesn't matter. Just cogo in the point based on your chosen bearing basis, and measured distance and calibrate your GPS shot in point to that calc point. It basically is assumed and your closure will be zero.

But without that second point you will have no way to reproduce your bearing basis and could wind up in a world of doo doo. Especially if some super button pusher lowballer tries to duplicate your datum, and you are stuck answering questions.

If you venture into the 3, 4, 5 plus calibration points you then enter all the adjustment factors that ruin many peoples days.

Good luck

7. ### Luke CO PLS6-Year Member

Joined:
Feb 10, 2011
Messages:
216
17
Location:
CO
But without that second point you will have no way to reproduce your bearing basis and could wind up in a world of doo doo. Especially if some super button pusher lowballer tries to duplicate your datum, and you are stuck answering questions.

Or better yet, some knowledgeable surveyor tries to follow in your footsteps trying to duplicate your datum and you are not able to answer the questions!

8. ### JBStahl7-Year Member

Joined:
Jul 21, 2010
Messages:
1,292
218
Location:
Salt Lake City, Utah
MT, UT
> If i decide to use GNSS I usually just set the base, perform a "one point" calibration on the base point and go to work. Apparently where i live i can do this and still get within 2nd order accuracy (1;250,000?)within the range of RTK (about 2-3 miles).
Pressing the FUGARWE button in TDS will set you up similar to the FUGARWE button in Trim*#&, but you aren't controlling anything from a projection standpoint. It's creating an LDP on its own using an approximate (w/in 10ft H, 20ft V most of the time) of reality. Like you say, that's OK for obtaining fixed solutions and relative points, but not good for a large or long project.

I'd suggest that you at least turn on the data file so you can download it after the fact to determine a better position through OPUS. You can always adjust your system later to use a 'real' position. For the time being, you can proceed as you've been doing on the first setup. Set control, gather points etc. Just make certain that you have dual observations on your control so you've got come confirmation.

>If i need to return to the same site then i can calibrate to my control points.
This is the common mistake made by most. You've already got control points which have local xyz and llh positions computed based upon the first LDP the data collector has chosen for you. There is no reason to calibrate on the second visit. All the calibration will do on the second visit is take any observation error on the second shot and distribute it throughout your project at random. It will also, as Leon pointed out, change your grid rotation slightly each time. Yes, it "should be" negligible, but it doesn't have to exist at all.

To eliminate re-calibration errors, simply make sure that you've set your control points during the first visit in convenient locations that are good for GPS and will make good radio broadcasts. Then, on the return visits, always set your base over a known control point. Tell the DC to use the known point number and select the same job you had before. It will use the same parameters as determined from the first FUGARWE visit, but they are the same parameters from the same LDP.

All you have to do after your setup over the control point (with measured HI) is to go to a couple of control points (or any other observed point) and, rather than doing a "calibrate" function, just do a "check control point" function. Check, check, check, and you're ready to go. Same LDP, same grid parameters, no projection error at all.

The only thing you don't know is how the "software engineer" really is determining your points. I'd much prefer to set my own LDP parameters before my first visit and avoid all the unknowns. I get to choose which projection to use, which average project elevation to use, and which "north" to use.

JBS

Joined:
Apr 15, 2011
Messages:
1,199
0
Location:
Mohave County, AZ
> Or better yet, some knowledgeable surveyor tries to follow in your footsteps trying to duplicate your datum and you are not able to answer the questions!

Agree. Case in point. Small site calibration has it's place, but is not the cure all people think it is. It is still a radial survey, no different than setting up a total station on a point, shooting a BS, assume North.... there you have your horizontal orientation.

10. ### Shelby H. Griggs PLS7-Year Member

Joined:
Jul 8, 2010
Messages:
1,358
256
Location:
Prineville, OR
:good:

11. ### Shelby H. Griggs PLS7-Year Member

Joined:
Jul 8, 2010
Messages:
1,358
256
Location:
Prineville, OR
:good:

Joined:
Apr 15, 2011
Messages:
601