[talk-au] Mass realignment of roads? (was Mapping things by importance)

[Elizabeth Dodd]

On Fri, 12 Jun 2009, Delta Foxtrot wrote:
> --- On Fri, 12/6/09, Liz <edodd at billiau.net> wrote:
> > I'll ask the offspring. He is 7/8 of a surveyor
> > with one semester to go!
>
> Can you also ask the bearing the drift is going, all I've been able to find
> out is approx north or north north east, but haven't been able to find
> anything on the net that clarifies it.


adrian advises
Making me drag out the Geodesy notes....
Ok, here we go. This may be incredibly boring, but it was boring in
class. When GDA94 was defined one of the driving factors was GPS and
the need because of it for an Earth centred datum that matched WGS84
closely. GDA94 takes the same origin as WGS84, the centre of mass of
the Earth and the same semi-major ellipsiodal axis, but a tiny
difference in the semi-minor to match the geoid over Australia better.
This results in the original position difference between the 2. This
positional difference is fairly irrelevant to any but Geodetic
surveyors, as most people do not have equipment to measure that
accurately and the general Cadastral surveyors use predetermined marks
to reference their surveys to.
        For those who are not following at the moment, an ellipsiod is
a lot like a circle, but has a varying radius. It is defined by an
origin, the largest radius (semi-major axis) and the smallest radius
(semi-minor). All Earth models assume it is an ellpsiod of revolution,
around the semi-minor axis to create a flattened sphere. It is a purely
mathematical surface. The geoid is a surface defined by being
perpendicular to the line of gravity at every point. It also is a
surface where gravity has the same force at any point on the surface. If
you level something with a spirit level, it is level with respect to the
geoid.
        Back to the datums, I don't have any current misalignment data
between GDA94 and WGS84. GDA94 is fixed to the position of reference
stations on the Australian continent as of 1st January 1994. At that
point the difference between the 2 was regarded as irrelevant at the
DGPS level, so it should have been within 0.3m. I have a document from
late 2002 saying that the difference was negligible, the continental
drift was approx 7cm per year across the continent, and at that point
the 2 datums had diverged by a further half-metre. Assuming the same
continental drift, it has diverged by another half-metre. So,
currently, GDA94 and WGS84 are within 1.3m of each other. Who would
know this? In Australia, the Intergovernmental Council on Surveying and
Mapping (ICSM) is supposed to deal with datums. Someone in there should
be keeping track on the shifts. The shift at a given point can be
calculated quite accurately with a survey grade receiver and a few
hours worth of recordings.
        On to continental drift. While I mentioned above that it is
approx 7cm per year, I can do better then that. The International Earth
Rotation Service is responsible for keeping track of various receivers
that measure the differences between their locations by several
different methods, each to at least centimetre position. The data from
these stations is freely available, but a pain to find. By finding the
released position data used as the basis for the establishment of
consecutive International Terrestrial Reference Framework (ITRF) datums
the continental drift can be calculated for any participating station.
From the horrible to navigate SINEX files provided by the IERS, in 1997
the continent was doing about 6cms/y at Canberra and Hobart, 7cms/y at
Pearth, and 13cms/y at Darwin. In 2000, the values are almost
exactly the same. As for what direction it is going, the
IERS releases files in 3D X,Y,Z Earth centred co-ordinates. These are a
pain to work with, and I can't tell you from them what the direction is
in a nice easy co-ordinate system. I always thought that it was in a
North to North Westerly direction, an azimuth of around 330 degrees. 

References:
IERS: http://www.iers.org/MainDisp.csl?pid=163-253
ICSM: http://www.icsm.gov.au/
Going Geocentric, Version 1, October 2002. Queensland Government
Department of Natural Resources and Mines for the ICSM.
(I should see if I can put this cd and it's information somewhere that
you can all read it. It is quite a good overveiw of GDA94, and also
Datums in general.)
Geodesy Notes, Geoff Sandford, 2008 and 2009.
        

-- 
BOFH excuse #288:

Hard drive sleeping. Let it wake up on it's own...

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-au/attachments/20090613/515430a3/attachment.html>