Activity Feed › Discussion Forums › Strictly Surveying › Static GNSS post processed Grid Coordinates to Ground Coordinates
Static GNSS post processed Grid Coordinates to Ground Coordinates
Posted by hellboy2703 on August 15, 2017 at 9:28 pmHi, everyone.
First let me describe in the situation I am working in.
I am from Bangladesh. We have very few CORS in our country, 6 to be exact, and also there is no geoid model for our country yet. I am aslo working at WGS84 UTM Zone 46N coordinate system.
I am using 2-Trimble R10 and 2-R8 Model 2, TSC3 controllers with Trimble access and Trimble Business Center (TBC) for post processing work.
Now about the problem-
I am tasked with establishing Control Points by Static GNSS Observation for a Rail Line development project that runs about 30Km long. My job is to setup control points at about 5km apart along the route and for every control point a auxiliary point for back checking with total station.I have completed the observation and post processing work and got the grid coordinates (X,Y and Ellipsoidal Z) for all the control points and their adjacent auxiliary points from TBC.
As the TBC provided the coordinates in grid system, typically back checking with total station shows discrepancy for the auxiliary points for the shots taken from Control Points, and vice versa.
Yes i know i should use the combined scale factor for this problem. And while using the combined scale factor the discrepancy goes away.
But the biggest problem is that— The surveyor of my client does not want to use the scale factors. That means I have to provide all coordinates in ground level.As i previously mentioned that for my country there is no geoid model, so i cannot transfer the grid coordinates to ground level via TBC or I do not know how – if it is possible to do so in TBC.
So if any of you guys have any idea how to grid coordinates to ground level, without using the established geoid model please share with me.
Thanks.
jhframe replied 4 months, 2 weeks ago 12 Members · 18 Replies- 18 Replies
hellboy2703, post: 442283, member: 12445 wrote: Hi, everyone.
First let me describe in the situation I am working in.
I am from Bangladesh. We have very few CORS in our country, 6 to be exact, and also there is no geoid model for our country yet. I am aslo working at WGS84 UTM Zone 46N coordinate system.
I am using 2-Trimble R10 and 2-R8 Model 2, TSC3 controllers with Trimble access and Trimble Business Center (TBC) for post processing work.
Now about the problem-
I am tasked with establishing Control Points by Static GNSS Observation for a Rail Line development project that runs about 30Km long. My job is to setup control points at about 5km apart along the route and for every control point a auxiliary point for back checking with total station.I have completed the observation and post processing work and got the grid coordinates (X,Y and Ellipsoidal Z) for all the control points and their adjacent auxiliary points from TBC.
As the TBC provided the coordinates in grid system, typically back checking with total station shows discrepancy for the auxiliary points for the shots taken from Control Points, and vice versa.
Yes i know i should use the combined scale factor for this problem. And while using the combined scale factor the discrepancy goes away.
But the biggest problem is that— The surveyor of my client does not want to use the scale factors. That means I have to provide all coordinates in ground level.As i previously mentioned that for my country there is no geoid model, so i cannot transfer the grid coordinates to ground level via TBC or I do not know how – if it is possible to do so in TBC.
So if any of you guys have any idea how to grid coordinates to ground level, without using the established geoid model please share with me.
Thanks.
You don’t need a Geoid model to get ground XY coordinates, you do need it to get elevations. Are you going to run some kind of levels along this corridor? If not how do you plan to get elevations? Without a Geoid model it will be difficult using only GPS. Are there good elevation points nearby?
Mighty is right on.
Orthometric Heights do NOT factor into “grid-ground” or Low Distortion Projection SCALE computations. It is strictly an function of Ellipsoid Height.
How much Ellipsoid Height variation is there in your Project?
Loyal
Assuming 7 stations 5Km apart for the 30 Km line and 7 matching auxiliary points. Your post processed vectors from each control point to auxiliary are at ground and should match the distance and change in elevation using a total station. Compare that with the distances between grid coordinates to get 7 scale factors. You may be able to live with a single combined scale factor, and hold the central control points, It may be necessary to mean the scale factors for the line to either side of the midpoint. Worst case is 6 different scales, one for each 5 Km segment. From your center point adjust your GPs to your ground system. You are in effect using ellipsoid height as geoid height. Vertical errors, end to end, should be within tolerable levels.
I doubt it, but you may have sufficient observation precision to establish separate grid and elevation scale factors.
Paul in PA
hellboy2703, post: 442283, member: 12445 wrote: The surveyor of my client does not want to use the scale factors. That means I have to provide all coordinates in ground level.
The surveyor needs to addressed as to professional execution of his duties.
It is the 21st century. The surveyor should have the technical skill to perform a rudimentary task.
MightyMoe, post: 442286, member: 700 wrote: You don’t need a Geoid model to get ground XY coordinates, you do need it to get elevations. Are you going to run some kind of levels along this corridor? If not how do you plan to get elevations? Without a Geoid model it will be difficult using only GPS. Are there good elevation points nearby?
I am only tasked for the X,Y coordinates for the control points, the client will do their own leveling work.
but they want to match the GNSS processed points to total station points.
here is the coordinates difference between GNSS processed points and total station points,I would never ever adjust conventional to GPS coordinates without including elevation differences.
How were azimuths from station to auxiliary established without a traverse?
Given a traverse how do the GPS to EDM station coordinates all match?
When all is said and done, I have no problem adjusting elevations to level runs, because I would use that as an opportunity to verify my traverse run precision.
Paul in PA
Loyal, post: 442287, member: 228 wrote: Mighty is right on.
Orthometric Heights do NOT factor into “grid-ground” or Low Distortion Projection SCALE computations. It is strictly an function of Ellipsoid Height.
How much Ellipsoid Height variation is there in your Project?
Loyal
Here are some Ellipsoidal Heights for 4 points at approx 10 km apart.
Point ID Ellipsoidal Height
PGM 01 -46.473
PGM 03 -46.241
PGM 05 -44.980
PGM 07 -45.247Paul in PA, post: 442288, member: 236 wrote: Assuming 7 stations 5Km apart for the 30 Km line and 7 matching auxiliary points. Your post processed vectors from each control point to auxiliary are at ground and should match the distance and change in elevation using a total station. Compare that with the distances between grid coordinates to get 7 scale factors. You may be able to live with a single combined scale factor, and hold the central control points, It may be necessary to mean the scale factors for the line to either side of the midpoint. Worst case is 6 different scales, one for each 5 Km segment. From your center point adjust your GPs to your ground system. You are in effect using ellipsoid height as geoid height. Vertical errors, end to end, should be within tolerable levels.
I doubt it, but you may have sufficient observation precision to establish separate grid and elevation scale factors.
Paul in PA
You are almost correct, there are total of 8 control points about 4 – 5km apart, and 8 matching auxiliary points.
4 of the control points are statically observed for about 8hr-10hrs, then they are post processed. And control points between and matching auxiliary points are then observed for about 1-2 hrs. and post processed. See the screenshot bellow.Vertical errors are not my concern as i am not doing the leveling work, but difference between GNSS observed points and Total station points in field of the auxiliary points are the main concern. please take a look at previous reply.
I observe you are below the ellipsoid, which means your elevation scale would be opposite what we consider normal. I worked below the ellipsoid along the shore in New Jersey years ago and I always had to sketch out and think carefully about my scale factors.
I would say you can go with a single elevation scale correction. Your grid to ground reduction end to end is in the range of -0.2m.
Paul in PA
You want to get a snakegrid for the project.
This project is very flat, just a quick look I see it running NE-SW, I would think one Datum Adjustment Factor would be good for the entire 30km. I would figure one out that made the ground distances and grid distances match as close to possible. Then I would sit down with the project people and run through the options to apply a correction. The easy way in TBC is to apply a ground factor in the project settings tab under Coordinate System/Local Site.
You will see a slide come up with three options Project Location,,,,,Ground Coordinates,,,,,,,,,Coordinate Display.
Under Ground Coordinates you can apply a Ground scale factor, this will make your ground and grid be the same if properly applied. Remember this is NOT, NOT, NOT a combined scale factor, if you want your ground to be larger than grid the number has to be larger than one (the reverse is done in auto-cad).
Project Location will allow you to pick where the coordinates are scaled from (you can pick 0,0 or some point on the project, I will usually pick an even lat, long if I do that, say lat. 45degrees, long.-100degrees).
Coordinate Display will allow you to apply a N, E offset if you wish.
Keep records of what you do and be sure that other users can use your data as intended.
If you want to know what ground vs grid in your project is don’t look at total station numbers, use TBC and inverse points using the measure distance function, it will show you the ground and the grid distance anywhere in the project, use those values to figure out your Datum Adjustment Factor. It will show you the grid distance, ellipsoid distance and ground distance, those are the numbers to use to get a factor.
Also keep your scale factor as short as possible. 6 decimal places is 1ppm, after that it’s just noise in the system but the longer you make it the more it can get misused.
And if you find a scale at 5 places works use that. There is no difference between 1.000110 and 1.00011.Multiply the inverse of the average combined factor by the grid (projection) scale factor. Create a new projection holding all other parameters the same.
This gives you a simple system approximating ground. You can switch between grid and ground with simple arithmetic.
Some folks condemn the practice as the grid and ground coordinates are too similar. This isnt alien. Rogue coordinates do not come bursting out of the chest of crew members or scurry about the rig on their own. Manage lists with meaningful names and you won’t have any problems.Friends don’t let friends scale to ground without truncating.
I’d like to hear more about how control densification between these 5km apart control monuments is going to happen before offering further advice. But this may well be a really good candidate for a Low Distortion Projection. Although it does seem that if the team cannot deal with scale factors, custom projections are out of the question.
Local, project grid coordinates.
Truncate, subtract a large constant. Make the ground coordinates visibly distinct from grid coordinates. Absolutely.
It’s really pretty straight forward.
very interesting issue! how did it go at the end? This is the first time I see someone in BD talking about elevation factor issue! people @DLRS don’t know about this and facing lots of trouble while importing old cadastral mouza maps in GIS system (recorded surveyed parcel area does not match with calculated area from GIS).
If not using a Low Distortion Projection or snake grid. I would use the 1/6 rule to get a good project factor and choose a point at some location on the project define its lat long and the average from 1/6 rule ellipsis height for computing the factor. Plug it into TBC and do some checks inverses all around to make sure grid and ground are within your tolerances. But it’s fairly flat from the numbers. Just define all this so whoever follows you knows how you did it. So they can reinvent everything. I am away now out on farm but the 1/6 th rule works fairly good for this scenerio
“I observe you are below the ellipsoid, which means your elevation scale would be opposite what we consider normal. “
Who’s this “we,” sucka? I’ve been working 30m below the ellipsoid for most of my career.
Log in to reply.