Hello,

Can someone please provide some clarification on the below situation?

If I fly the M300 RTK using the default settings for geographic datum (I assume this is WGS84) and then use an NTRIP connection to a base station with broadcast coordinates in a different datum (in this case GDA2020) to correct the captured images, do the images now have GDA2020 coordinates or corrected WGS84 coordinates?

Cheers

Hi,

The RTCM standard sends messages with coordinates in ECEF (Earth Centered Earth Fixed) format referring to the theoretical center of the earth.  

Then the receiver converts them to LLH (Latitude Longitude Height).  And finally the conversion to GDA2020 is done in post process.  

Theoretically all coordinates that are provided by the M300 are WGS 84 in LLH format.  

It's not impossible that there could be NTRIP casters that broadcast something else than ECEF, but I doubt it.  

Where did you get this information that your broadcast caster in GDA2020? I imagine you use AUSCORS.  On the presentation web page of the stream it is written that they broadcast in GDA2020 RTCM3.  Which doesn't make sense.  Then on the base station page the coordinates are WGS84.

  I invite you to ask them directly.  In my opinion the broadcast in GDA is misleading information.  Maybe I'm wrong, or maybe they just refer to the Z reference which is also GRS1980

LV_Forestry Posted at 2-8 22:18
Hi,

The RTCM standard sends messages with coordinates in ECEF (Earth Centered Earth Fixed) format referring to the theoretical center of the earth.  

Hi LV_Forestry,

Thanks for your reply. Yes the AUSCORS website indicates that Broadcast Coordinates are in reference to GDA2020 as you mention (https://gnss.ga.gov.au/stream).

Is it possible that the base station is still broadcasting ECEF coordinates but in reference to the GDA2020 datum? I will reach out to AUSCORS directly and see what they say.

Also, when you say conversion to GDA2020 would be done in post-processing, do you mean as a separate process? I.e. DJI do not perform any geographic transformations as part of RTK correction?

Thanks for your help.

awarfield Posted at 2-8 22:44
Hi LV_Forestry,

Thanks for your reply. Yes the AUSCORS website indicates that Broadcast Coordinates are in reference to GDA2020 as you mention (https://gnss.ga.gov.au/stream).

ECEF is the basis of measurement systems. It is rather GDA2020 which takes refence on ECEF, rather than ECEF on GDA.

Australia is such a large territory that only a datum in degrees as GDA2020 can go over the whole territory. If you try to do it in meters (UTM) there will be a big difference over long distances. Such a datum exists, GDA94, do not be surprised to find deviations of several meters over long distances. Certainly there must be local referral systems.

Here it is well summarized:
Australian Datums (and how to move between them) | Intergovernmental Committee on Surveying and Mapping (icsm.gov.au)

In general, it is your client who will ask you in what format he wants the data. The drone will provide you with WGS84 ellipsoidal coordinates. It's up to you to convert them into the right format with software :

Global Mapper :


Agisoft Metashape :


DJI Terra :


There you have a good explanation of what the vertical datum is:
Australian Height Datum | Intergovernmental Committee on Surveying and Mapping (icsm.gov.au)

Just keep in mind that the data coming out of the drone is in WGS 84 Ellipsoidal format. The "RTK" correction has no influence on the format of these coordinates. Simply on their precision. By compensating for the defects in the observations of the signals from the satellites received by the GNSS receiver due to environmental disturbances and defects in the trajectory of these satellites.

In Australia specifically there is another problem, due to the size of the territory, there is no NTRIP caster available everywhere. No GSM network or simply no base station. In this case you can use your own mobile base station.

awarfield Posted at 2-8 22:44
Hi LV_Forestry,

Thanks for your reply. Yes the AUSCORS website indicates that Broadcast Coordinates are in reference to GDA2020 as you mention (https://gnss.ga.gov.au/stream).

If you need the geoid to correct the altitudes it can be downloaded here:

AHD
https://s3-eu-west-1.amazonaws.c ... id2020_20170908.tif

AVWS
https://s3-ap-southeast-2.amazon ... G/AGQG_20201120.tif

LV_Forestry Posted at 2-8 23:28
ECEF is the basis of measurement systems. It is rather GDA2020 which takes refence on ECEF, rather than ECEF on GDA.

Australia is such a large territory that only a datum in degrees as GDA2020 can go over the whole territory. If you try to do it in meters (UTM) there will be a big difference over long distances. Such a datum exists, GDA94, do not be surprised to find deviations of several meters over long distances. Certainly there must be local referral systems.

Hi,

Thank you, yes I am familiar with transforming between those datums. However, I have little experience working with ECEF coordinates, so I appreciate your explanation.

Your comment regarding the RTK correction having no bearing on the coordinate format in the M300 was what I after. Sounds like I am safe to assume the coordinates in the EXIF are indeed WGS84 regardless of the datum of the NTRIP caster.

Unfortunately I did not collect ground control, so I am unable to check for absolute errors. The nearest base station was <10km, so hopefully error is minimal.
  
Cheers