Trimble Business Center

I hope this is the right pace to ask a question like this.

I have been given a "local" coordinate system (see screen shot "Coord System") showing the parameters used to go from UTM (Zone 11 N, Canadian Geoid HT2 etc) to this local system.

I have done some data prep on this project and I want to shift the local back to the UTM so that I can use Google Earth to underlay the alignment etc.

I cannot seem to duplicate the transformation. If I use Transform Survey Points I can only get a transformation if I do multiple point pairs. That moves the points exactly but the other factors such as X,Y,Z rotation, scale factor, and shift do not match what I was given. I tried 5 parameter, and 7 parameter but (I am assuming ) since all fields are "0" for rotations etc it doesn't work.

If I use a simple Move/Rotate/Scale it seems to work but the scale factor is rounded down to 6 places so the final result is a bit 'off'.

Am I missing something, or just going crazy from too much data prep this last week, lol?

I have attached the data as two CSV's and a screenshot of what I was given.

Marshall

Marshall (I dont have your contact details - could you email them to me at alan_sharp@trimble.com as I had a few questions here that I wanted to ask you off line)

Anyhow - this is a very similar request to one I had last year which had a similar look and feel to it

In the screen shot below, Points 1,2,3 are the first 3 UTM Coords in the list you sent me. This is how I got to this point in the first way of doing things

1. I opened the Project in Metric
2. I elected Coordinate System - World Wide UTM then selected 11 North as the Zone and  then selected NAD 1983 Canada as the Datum Transformation and then selected HT2 (Canada) as the Geoid Model
3. I then entered the 3 monument coordinates in the UTM projection (not local coordinates)
4. Having entered the 3 coordinates and the coordinate system, I ensure that I have a Connect connection set up in Support - Optons, External Services (my Username and Password) and test that it logs me in etc.
5. Now I Toggle Background Map and the 3 points show up in the ? correct location on the Trans Canada Highway as shown below

Are these in the right place?

In this project I back calculated a Site Calibration based on 2 or more point pairs - I know the Local NEZ values and I now have Lat Long Height Global Coordinates for those same points (from above process) so I can use the BC-HCE Site Calibration to map the Locals to the Globals (and keep everything Local Coordinates for the project. Note that this is a "Force Fit and is always a little risky with 2 points only - the more the better (so if you enter all the points in in the First Approach and get all the LLH Coords out - then enter them in as Point Pairs here you would have a decent number of control points as a check on the Calibration Results.

So I start a new Metric Project and do not set the Coordinate System

I create CAD Points and enter the local coordinates in N,E,Z

I create CAD Points and select Global for the Point Types and then I can enter the LLH values. Note when you enter the first point - it brings the Project to Global by matching the Local and Global point Pair (I called the points 1L (Local Coord) and 1G (Global Coord in LLH)

I enter the second Global Coordinate and when you complete that you will see it is a little off (This is primarily the Scale Difference provided)

Then I run the Site Calibration command and set the settings as shown

I check the Update Projection Definition check box above and then select Point List tab. It allows me to now select Point Pairs (a GNSS Point (The Global Points) and a Local or Grid Point. So I select 1G / 1L for the first pair and 2G/2L for the second pair etc. and then hit Compute - that computes the parameters to get the pairs to match up and you will see that 2G and 2L now are on top of each other whereas before they were separated.

Now that you have a Site Calibration in the project, you can import the CAD data into that project and should not have to Shift or scale it to make it work in Google Earth as the Calibration will take care of that.

Now you can see the Background Map again and it should fit, and in this project everything is still in Local Coordinates.

Hope that this now helps you going forwards on this project

Sorry for the disclaimer below, when you start computing projects in this way, people do get things wrong and use data in inappropriate ways and because I do not know / see your entire workflow I cannot forsee all potential downstream problems. There are scale factors in play here because of the methodology used, and I cannot guarantee that those are correct or usable only that they match up and put the points on the highway. Typically we would recommend that you have several points provided by the Surveyor in LLH and Local Coordinates and then do a full site calibration to match them using GNSS. The computations here should match with what you would see in the field. 

Disclaimer

Trimble does not take any responsibility for the project setup in the enclosed files. As a user of these files you take full responsibility to verify that the methodology used and results provided are accurate and fit for purpose.

Hope that this helps, let me know if you need more assistance

Alan

Alan: I stepped through the process and voila... Thanks for the tutorial. Much appreciated.

Any thoughts about why, if I do a point for point (13 points), using Transform Survey Points I obviously get a 'perfect' move/rotate/scale/ but the parameters in the Report don't match what I was given at all?

Thanks again. have a great day.

Marshall

Depends on the origin selected by our calibration as that will affect the shifts and rotations - provided the points match you are in good shape

Sent from my iPhone

I will take a look at it this week and try to give you a better understanding by using the data specifically you used - please send me your 13 point transform and I can use that to give you an explanation

The main issue will be the origin - if that is different then the shifts and rotations will of course be different - if the tilted plane has a different origin then the constant offset / shift will be different - so the origin is the key - I am not sure how we pick the origin for the Site calibration vs what someone else may use - for example you could pick the centroid of the point data provided or the Base Station location or the lowest E/N coordinate etc or the origin of a projection if you apply that etc - so the Origin selection is kind of random - but once selected the 7 parameter transform is computed from there - and because it can be different it will always give you different numbers for the 3 shifts 3 rotations and scale to make the data fit.

Alan

Sent from my iPhone

Alan: I appreciate the offer to look further at this but if you are pressed please don't take time away from more important things.

I have attached the Site Cal report done following your above instructions. I used all 13 points. These CSV's, initial information etc are also attached.

I tried a coordinate transformation using the origin at 0,0, the scale factor provided, and the 5,000,000 shift as specified and attached that report as well as the VCL of the points after that transformation. Not a good result.

Following along from what you said above about the origin, I could not see a way to use Transform Survey Points by forcing the origin to 0,0. But doing a point for point transformation this way does yield a "perfect" transformation but the resulting parameters in the report definitely do not match what we were given (ie origin at 0,0 no rotation, and shift of -5,000,000).

Again, I truly appreciate the insights but don't spend too much time. Ultimately we'll just have to go out there, calibrate to what they gave us, and get on with it.

Thanks Alan,
Marshall