Mapping and GIS Solutions Community

 View Only
Expand all | Collapse all

Finding the same point with a DA2/QField and Zephyr 2.0/ Geo7x

  • 1.  Finding the same point with a DA2/QField and Zephyr 2.0/ Geo7x

    Posted 03-04-2024 07:50

    Hi,

    The organisation I work for has historically used a Zephyr 2.0 receiver with a handheld Geo7x to both find and map points with 1cm accuracy. We specifically use these instruments for long-term agricultural field trials, some of which are over 100 years old, so accuracy is important and we need it to be within 1cm because this is also the tractor systems work to (also some of our trail plots are only 1m by 1m, so even 10cm can make a big difference).

    With a lot of help from our IT team around software compatibility, the Zephyr/Geo7x system has worked really well for us over the last 10 years, but as they are now switching off the 3G network in the UK, we need to upgrade and have chosen to pair a Trimble DA2 receiver with a tablet that runs QGIS/QField. 

    My problem is that I cannot get the two systems to find the same point location with 1cm accuracy. To test this we uploaded the same map (in WGS84) onto both systems and marked out several points using the old Zephyr/Geo7x. I then tried to navigate to these points using the DA2 and QField, but was consistently about 20cm out (see below photographs). I am almost certain there is a difference between the settings but I can't figure out what it is. Below is the device setup as I have figured it out on both systems so far. 

    DA2/ Trimble Mobile Manager/ QField:
    Trimble Mobile Manager settings: GNSS correction source = Auto (Trimble Correction Hub); GNSS output = Auto; Geoid = EGM96 (Global);
    QGIS settings: Layer CRS = WGS84, Project CRS = WGS84
    QField settings: Positioning device in use = Internal Device, Antenna height = 2m (height of the pole that came with the DA2), Measure (M) attached to vertices = Timestamp (milliseconds since epoch)

    Trimble Zephyr/ Geo7x:
    Geo7x: GNSS Receiver port = COM3: GNSS Serial port (internal GNSS receiver); DOP type = PDOP, Max PDOP = 99.0; Min SNR = 12.0; System: Latitude/ Longitude; Datum: WGS 1984; Altitude reference = Height Above Ellipsoid (HAE)


    Any insight anyone has would be hugely appreciated. 




    ------------------------------
    Aislinn Pearson
    ------------------------------


  • 2.  RE: Finding the same point with a DA2/QField and Zephyr 2.0/ Geo7x

    Posted 03-04-2024 08:49

    Hi Aislinn

    This looks like a datum shift error which should be easily fixed

    If your GNSS output in TMM = Auto in the UK, then this will be using the OSNetv2009 (2009.76) datum, BUT your data and QGIS If understand correctly are using WGS84

    This will add a shift in decimetres (depending on where you are in the UK) between your GNSS position and the data. 20cm sounds about right

    In TMM try this

    • Under the GNSS Configuration menu
    • Change your GNSS OUTPUT to 'Select from list' and choose WGS84 (G1762) current
    • This should at least align everything

    I'm guessing you were formerly a Terrasync user on the Geo7x? 
    If yes then have you looked at Trimble Terraflex (and the Trimble Terra Office ArcMap and ArcGIS Pro extension for ESRI users)?

    I understand QGIS is free but If you are working to 1cm level you really need a professional system that handles geodetics and precise navigation with the highest level of accuracy and simplicity.
    I'd recommend looking into Terraflex - there is a free 30 day trial available and migration tool to convert you existing Terrasync data dictionaries (or SSF) to a Terraflex project to speed up the process

    Take a look at the Terraflex Skillbuilder series for quick start information and the online help HERE for more details



    ------------------------------
    Tom Harrison
    ------------------------------



  • 3.  RE: Finding the same point with a DA2/QField and Zephyr 2.0/ Geo7x

    Posted 03-05-2024 02:36

    Hi Tom,

    Thank you so much Tom. I'll go out into the field later this week and try what you suggested. I'll update this post with an answer once I've done the testing. 

    We were using Terrasync, so thank you for the info on Terraflex.



    ------------------------------
    Aislinn Pearson
    ------------------------------



  • 4.  RE: Finding the same point with a DA2/QField and Zephyr 2.0/ Geo7x

    Posted 03-06-2024 05:48

    Hi Aislinn,

    As a general comment I would avoid using WGS84 and particularly if you are looking for cm accuracy. A one-of test between two devices at the same time might be ok as it's only a relative comparison, but as you have seen even that can still open you to other issues.

    For your data it's definitely not good, WGS84 coordinates are space based and not fixed to the ground. And that ground that you and your points are on is surfing through space, in the UK around 2.5cm every year. So anything you record in WGS84 will never be at the same coordinates on the ground when you go back, and the amount of error will depend on how long it's been. For example, after the hundred years you mention those points would be 2.5 metres away.

    I would standardize to your national datum for everything, including testing as that way you will know everything is set up right and in sync.

    Cheers



    ------------------------------
    John Winkler
    ------------------------------



  • 5.  RE: Finding the same point with a DA2/QField and Zephyr 2.0/ Geo7x

    Posted 03-06-2024 08:09

    Hi Aislinn,

    John, and Tom comments are right on in terms of putting data in the "right place".   Coming from years of use of the Geo7x/Zephyr/TerraSync/Pathfinder Office mapping system, your test of Precision above, (precision being the relationship of returning to a point on the ground over and over) is just that.  A precision test.  Your assumption that the Geo7x is TRUTH, would be, in this case a test of accuracy, but only if you assume the XYZ coordinates from multiple visits to that same point on the gound were at a higher level of precision than the DA2, would it only be called a test of accuracy.  But your navigating, and navigation using whatever GPS is incumbent on the AUGMENTATION being borne from space or post-processed coordinates.  Tom asked if you were using Pathfinder Office.  I would also ask to describe in more detail, and as "accurately" as possible what your "TRUTH" coordinate is derived from.  I, like the above would ask if you Post Processed, or did you use, as its inferred from your post, that you used some sort of real-time internet connection to some source to derive your Geo7x coordinates.  I'm assuming the latter - and in that case, you have to know what the reference frame (datum) of the Geographic Coordinate system of the realtime source was in.   I'm not from UK, so assume, as Tom mentions you were indeed "in" OSNetv2009 (2009.76) datum.   Then for a precision test to be valid, you must set your current realtime augmentation to store, with no shift into that same GCS.

    I do get though, being a long time user of Trimble software (and ESRI software), the inappropriate language that you were in WGS84.  Thats just a placeholder, and a poor one at that, because, like John said, you can't be in "WGS84". Plain Vanilla WGS84 is not possible since WGS84 has gone thru multiple reference frames since that datum tag was first rolled out by the US military.    Software uses that term loosely, and its caused decades of confusion.  Fast forward to today, standing out in that field, you must know what you were "in" from history, and what you are "in" today.   I have no idea how your current software is handling such transformation (or not), so you will chase your tail with "accuracy" until you set a control (as a surveyor to do this). on that point and collect a trusted coordinate.  Then you can assess "accuracy". Until then, your tests now are precision.    Unless you used a Geo7x Centimeter edition of TerraSyncy, there is no way your 1cm precise.  At best, your 2 - 5cm precise in that "error" budget.  If you could, you would average all your occupations of the Geo7x, and determine a precision error budget amongst all your observations, applying a 2-5 cm error budget (being conservative here), and then you may find the mean and standard deviation of all your observations would be +/- X with a SD of X.  Then, your precision test could proceed, and I think you'll find your not 20cm away for maybe within 5-10cm away (given no blunders in all your tests).  It does sound like your tests are in the open, thats good, but if you want to evaluate Accuracy, true accuracy, you'll need to bring in a "surveyor" like approach to this question and understand exactly how your Geo7x handled the coordinate and this new system you have.     Winkler is right on! 

    Apologize for this long post, but there are many datum hoops you have here, everyone can move that coordinate from 20cm to 100+cm.



    ------------------------------
    Joel Cusick
    ------------------------------