[talk-au] Australia "changing coordinates"

Warin 61sundowner at gmail.com
Tue Aug 2 10:43:15 UTC 2016

On 8/2/2016 3:37 PM, Andrew Davidson wrote:
>> WGS84 is a datum fixed; as in dated 1984 and the data does not change.
> I'm not sure that you've understood exactly what a semi-dynamic datum 
> is. Have a read of this:
> http://ceur-ws.org/Vol-1142/paper6.pdf
> particularly sections 3.4 and 3.5.

I'm reading
The document is dated in 1996 so more recent things are not in it. But I 
think the principles would hold.

It lists;
WGS 84  1 Jan 87 - 1 Jan 94
WGS 84 (G730) 2 Jan 94 - 28 Sept 96
WGS 84 (G873)  From 29 Sept 96

You will observe the refinements of WGS 84 (to reduce 
error/uncertainty). I would expect that suffix would be used in 
professional circles.
The (Gxxxx) simply gives a refinement to WGS84, does not change the 
basis, and the refinements are getting smaller.
This document also introduces the concept of  'uncertainties', a good move.

Page 16 of 175 states in part
"As of 2 Jan 94, a set of geodetic coordinates shall include a 
designation of the reference frame and epoch of the observations"

Note that this is for any datum. I expect this date requirement will 
become more common and a requirement with earth centric datums.

"In summary, the refinements which have been made to WGS 84 have reduced 
the uncertainty in the coordinates of the reference frame, the 
uncertainty of the gravitational model and the uncertainty of the geoid 
undulations.They have not changed WGS 84. "

  In section 3.5 the error is introduced in making the first transform 
from GDA94 to WGS84. Not clear if this transform was made at all, or if 
made was not to the correct date.
The error could have been avoided by the correct transform in the first 
datum conversion?

>> Or you have access to survey points/marks and their data  in the local
>> area.
> LOL. Reminds me of this quote:
> "You'll never find a survey mark on the ground with coordinates 
> referenced to WGS-84. If you do, it's wrong."

A global datum will have more compromises than a local datum. For that 
reason alone I would expect survey marks to be in the local datum..
e.g. NSW survey marks are GDA94 using MGA 
I also expect these will be translated to the new earth centric datum as 
a mathematical operation some time in 2017.
I would expect that any local datum can be transformed into WGS84 with a 
sufficient level of accuracy for the task, if required (i.e. if the end 
equipment requires WGS84, or its refinements)

More information about the Talk-au mailing list