The implementation of GDA2020 is drawing near! In an effort to be ready before time I have downloaded the NTv2 Grid Shift Binary (GSB) transformation grid from ISCM (available here) who are the official custodians of all things GDA2020 related. I am using the GDA94_GDA2020_conformal_and_distortion.gsb as well as AUSGeoid2020 (available here).

ArcGIS 10.5 does have GDA2020/MGA2020 in the list of available coordinate systems.. a good start. ArcGIS 10.2 does not recognize GDA2020 calling the CRS "custom" but locating correctly.

When transforming from GDA94 to GDA2020 ArcGIS has a default transformation GDA_1994_To_GDA2020_1 which doesn't match a GSB file in the australia folder but may be a parameter only transformation. I am uncertain as to the accuracy of this transformation and so followed the instructions in this Esri support post about installing a new NTv2 file attempting to install the downloaded GSB file as the assumed to be more correct transformation between these datums.

Has anyone verified that the default Esri transformation is close enough (4th decimal place in MGA2020) to the ISCM GSB for all of Australia so that I don't need to jump through all these hoops on every workstation with ArcGIS installed?

Note: I understand that this isn't much of a question; I have posted this mostly as a resource for users who will be implementing GDA2020 after frustrating hours encountering numerous broken links and reading through verbose documents in a quest for a correct transformation and I'm hoping that this vague question saves others a considerable amount of time and frustration by posting the working links... at least until the data moves again and these links become broken too.

  • Is it an option for you to upgrade to 10.6? Esri Australia seems to be saying that it "just works" there: esriaustraliatechblog.wordpress.com/2018/01/04/gda2020-arcgis – PolyGeo Nov 2 '18 at 3:27
  • I'd like to know if the pre-packaged transformation is a parameterized or NTv2. Working with datum transformation isn't something that I have done before so I'm unsure how to prove/disprove the existing transformation. "Just works" makes me cringe, I'm concerned about processing months worth of data and then having rejected by a client because the shift isn't correct.. considering the development of GDA2020 has been in progress for a significant amount of time if Esri adopted an earlier model released by ISCM there could be problems. Perhaps it's something I should contact Esri support on. – Michael Stimson Nov 2 '18 at 3:40
  • 1
    I would definitely contact them. Also, if mkennedy looks in on this and answers/comments I would take that as gospel. – PolyGeo Nov 2 '18 at 4:14
  • 1
    The link PolyGeo posted states 10.6 / 2.1 only supports the parameter transform out of the box, unless you manually install the grids (includes instructions). – user2856 Nov 2 '18 at 10:07

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.