Question about field setup: base coordinate transformation

What you do when you arrive on site and are given a base coordinate or site control coordinate that is not WGS84?

  • In what coordinate system are the coordinates you are given?
  • Do you translate it to WGS84 or do you fudge it because it is ‘close enough’?
  • What tool do you use for the coordinate transformation?
  • Is it ‘online’ or in a smartphone app or part of your data collector or fieldwork software?
  • Or do you always get the coordinate beforehand or take it with you and transform it back at the office?

It would be nice to get a variety of answers so that we can see the different ways this is handled in the field.

1 Like

no.

  • In what coordinate system are the coordinates you are given?

ETRS89/UTM33

  • Do you translate it to WGS84 or do you fudge it because it is ‘close enough’?

Since I want to work again within that coordinate system I convert from UTM to degree and neglect the ellipsoids. The UTM to degree conversion produces the same results as if I order it converted from the official administration. When I work like that I did not see height differences to the official 1m dem.

  • What tool do you use for the coordinate transformation?

Python

  • Is it ‘online’ or in a smartphone app or part of your data collector or fieldwork software?

no

  • Or do you always get the coordinate beforehand or take it with you and transform it back at the office?

I always to PPK back in the office

1 Like

Could you just average in the point and use their elevation?

Thanks for your input guys.

Anyone else? And what about RTK?

I do not understand the scenario, do you want to use your known point as point for your base or will you receive your corrections from another offsite source?

If I need RTK I prepare the coordinates in the office before I go into the field. Up to now that is only for revisiting points or flight paths I have done before.

I think the main factor that he mentioned is that he’s being given a point for his base from another source. Whether it be in NAD83 or something else. The deal may be to recognize that some projections are grid and some are surface. This is where we rely on localization to assign a WGS84 coordinate to a grid point. If you’d don’t know the point beforehand then it could be inconvenient to try to do it in the field, but I think it should be fairly easily accomplishable in QGIS. All you really have to do is start a new project on the coordinate system from the given point, create a new shapefile layer, input the point and then export it on WGS84. Here’s where the problem arises. If it is not a true one to one direct transformation then because reachview does not supply localization any variance between that direct transformation and the actual wgs84 coordinate that you can collect then none of the other points that may be given will match. Especially if those points came from a localized scenario.

Ah ok, thank you for the clarification!

I’m personally very unsure whether the transformations offered in QGis and ArcView are good enough to provide cm accuracy!? At least in Germany I had the impression that the points which we tried to translate in both programs are at least 30 cm off. Maybe that is a special problem of Europe/Germany.

There are specialised programs to perform such transformations which also considers differences in different WGS84 realizations, e.g. TRANSDATpro.

BTW.:(I’m still not sure which WGS84 realization RTKlib uses and whether the ZED-F9P uses the same.

Thanks again for your comments.

I didn’t want to frame a specific scenario so that anyone could describe their own scenario and how they deal with it, or where they are having problems/difficulties. More like a request for people to share their workflow as it relates to the coordinate system given vs. what you use, vs. what you provide ( or what the customer wants ) in your finished product.

But maybe it will help for me to suggest a scenario. Feel free to explain your own or modify to suit you own answer:

Say you show up on a job site with several contractors (and you are one). You are given coordinates of a control point and a PDF site plan by the general contractor. He forwards them to you by email right there in the jobsite office. The coordinates are not WGS84 - let’s say they are UTM grid coordinates; or let’s also say they are in a specific geoid for the local region (based on NAD83 for North America); or they could be stated as 0m,0m,0m for that point.

Now you’ve set your base up on the control point (what would you do for base coordinate here? average single? quick QGIS work back in the van using the laptop computer? survey app on phone/data controller?)

What do you do with ReachView? Do you set up bluetooth output/mock location and run a 3rd-party app? Is it free / or paid one-time / or subscription? Or do you just use ReachView on-site and QGIS back at the office (or a non-free GIS/Survey/CAD application?)

Now you are starting to work (maybe you are laying out GCPs for a drone flight). Now you get interrupted by someone who needs you to take your rover over to a spot and they need the coordinate for that ‘right now’ for whatever reason. Let’s assume they might want it in the given reference system, or let’s assume they want measurements in meters from the two nearest easy reference points.

Can you accomodate them with an answer immediately? Can you accomodate them with an answer after you get back to the office? Or do you tell them you can’t do it?

Interesting and vague question. What do you mean by “use” ?
Realization is just a adjusted version of the previous one, hence adjusted and maintained by DoD and not by the reciver it self which uses the orginal WGS (WGS84 original in this case). Therefor the RTKlib should work similar.
That is how i understand it.
WGS84 (latest realization) is just a adjusted version for where the mass of earth center is, all versions of WGS84 use the same refrence frame formula.

That’s pretty much the scenario that I was going after. I can’t speak for the UK, but in the United States pretty much every construction project uses XYZ coordinates and a localization when using GPS. Having just the base station coordinate will do you no good and Reachview does not allow for localizations so all you could really do is locate your GCP’s and then convert the processed data to whatever coordinate system the site is using, but you still won’t be guaranteed that you are 100% in the same place.

That being said you will need a third-party software like Microsurvey Fieldgenius that allows for localizations. There are several threads in this form mentioning it and other softwares that allow for connection to the Reach receivers and also provide localization. I’ve been on here for over a you’re stating the need for that feature and little has been said officially so you will definitely want to investigate third-party software.

More than likely the GC will be using Topcon, Trimble or Leica and the localization files for those are all proprietary so you would have to localize yourself anyhow. It would be worth requesting the WGS84 coordinates for the control points. They have the information, but may have never been asked for it before so they may not even know how to get it out of their data collectors.

2 Likes

Based on your comments @michaelL, is this a fair summarization:

For construction site work in the US, you can use Emlid Reach with FieldGenius (or equivalent) and localize with two or more site control points. Then you can do RTK work and have useable coordinates available in real time.

The provided control points will likely be in state-plane XYZ grid format.

update below

Correct, but the best result is from having 4 control points that “box-in” the site. Two of which should be good vertical and preferably on opposite sides of the site. Three will work, but 4 is better.

That makes sense. So how are you able to handle a control point with ‘bad vertical’ differently than one with ‘good vertical’?

In the localization you see the values for horizontal and vertical individually and can select/deselect them as you like. I might hit 20 points, but I really only want 5 or so depending upon the shape of the project. I can go through the control point list and start knocking off the points with the worst values. I can stop taking them out (if I want) once my tolerances get down below 0.10ft. Preferably 0.05ft.

1 Like

Let me update my summarization then:

For construction site work you can use Emlid Reach with FieldGenius (or equivalent) and localize with ~4 of the best site control points. Then you can do RTK work and have useable coordinates available in real time.

Thanks for your input @michaelL.

@tobias-dahms already mentioned ETRS/UTM33 but he is using free software and a PPK workflow. Does anyone else have a workflow for doing RTK work in a coordinate system other than WGS84?

The concern is not about the information you produce with your processed data, but more about how you or how your software deals with coordinate transformations, either before or after the RTK session.

Any input is welcome, even more info about PPK workflows. Additionally, does anyone know of any free software that does least squares fitting? Can QGIS accomplish the localization that @michaelL is talking about for PPK work?

3 Likes

This topic was automatically closed 100 days after the last reply. New replies are no longer allowed.