[Proj] Google Earth Accuracy

Wed Dec 17 10:08:44 UTC 2008

Andrew,
I may have misunderstood you. And maybe the Starfire differential service 
really does give you WGS84 coordinates, rather than some ground-attached 
variant like ETRS89 or NAD83.  

> The variation between WGS84 and ITRF of the day varies because of 
> plate tectonics. That is WGS 84 framework was set at a time in history, 
> but ITRF if calculated daily. The variable is the plate movements between 
> them.

I still think you are wrong here. Let me quote:
	"WGS84 is maintained by the US. Like the ITRS it is a global system.
	 Initially it was coincident with ITRS to an accuracy of about 1.5 meters.
	 It is currently maintained within 10cm of the ITRS."
	 http://www.ogp.org.uk/pubs/373-10.pdf (section 2)
If the differences were caused by tectonic plate movements, then it
would be quite difficult to maintain the two systems to within 10cm.
You would need to connect all continents by very strong steel chains, to 
prevent them drifting apart more than 10cm from their original position ;-)

No, if the difference between WGS84 and ITRF varies from day to day, I 
think that such differences are caused by something like athmospheric 
conditions, not tectonic plate movements. (Radio waves slow down in the 
athmosphere, by an amount depending on air pressure and humidity, I think.)

> I'm not expert all I know is that we have to write some pretty special 
> code to account for this variation so that when we use this GPS for
> locating with WGS84 coordinates we can find them.  

All right. If your special code is something like a time-dependent 
7-parameter datum transform (like those given in section 6.5 of 
http://www.ordnancesurvey.co.uk/oswebsite/gps/information/coordinatesystemsinfo/guidecontents/guide6.html ),
then you are indeed correcting for tectonic plate movements. But I would 
say that if so, then it is not the difference between WGS84 and ITRS
that you are correcting for.  Rather, your original coordinates must 
be expressed in something like ETRS89 or NAD83, not WGS84 in its 
strict sense.
On the other hand, if your special code is something else, 
then I don't know what it does... 

Disclaimer: I am a GIS programmer, not a surveyor. 
Maybe some surveyor on this list can give a better answer.  

Best regards,
--
Mikael Rittri
Carmenta AB
Box 11354
SE-404 28 Göteborg
Visitors: Sankt Eriksgatan 5
SWEDEN
mikael.rittri at carmenta.com
www.carmenta.com

-----Original Message-----
From: proj-bounces at lists.maptools.org [mailto:proj-bounces at lists.maptools.org] On Behalf Of Andrew Williams
Sent: den 15 december 2008 23:17
To: PROJ.4 and general Projections Discussions
Subject: Re: [Proj] Google Earth Accuracy

Mikael,
Perhaps it's my terminology.

As I understand it the GPS we have used Starfire differential service to correct it's position. It is a dual frequency GPS. The output coordinate is ITRF of the day.

The variation between WGS84 and ITRF of the day varies because of plate tectonics. That is WGS 84 framework was set at a time in history, but ITRF if calculated daily. The variable is the plate movements between them.

So when I said varies I meant the difference between the two frames varies between them because of plate tectonics.

I'm not expert all I know is that we have to write some pretty special code to account for this variation so that when we use this GPS for locating with WGS84 coordinates we can find them.

-----Original Message-----
From: proj-bounces at lists.maptools.org [mailto:proj-bounces at lists.maptools.org] On Behalf Of Mikael Rittri
Sent: Friday, 12 December 2008 11:53 PM
To: PROJ.4 and general Projections Discussions
Subject: Re: [Proj] Google Earth Accuracy

Andrew Williams wrote:
> ITRF varies over time compared to WGS84 based on tectonics, correct?

No. Neither ITRF nor WGS84 is attached to a tectonic plate.
The difference between them is very small (I've heard that it is at most something like 10 cm, but I don't remember where I heard it), and it does not increase over time. The reason for the difference is just that coordinates in WGS84 cannot be measured as exactly as in ITRF.
(Source: http://www.ordnancesurvey.co.uk/oswebsite/gps/information/coordinatesystemsinfo/guidecontents/guide4.html )

For WGS84-based mapping in Europe, we use ETRS89, which is attached to the Eurasian tectonic plate.  ETRS89 coincided with
WGS84 on January 1, 1989, as if the WGS84 graticule had been painted on the ground that day.  Since then, the ETRS89 (and Europe) has drifted away toward northeast by 2-3 cm/year.  So it is the difference between ETRS89 and WGS84 that increases due to continental drift. (Similarly for NAD83 in North America.)

However, I am writing about WGS84 in a strict sense above.
Many people who use GPS are familiar with WGS84, but they have never heard of ETRS89.  So these people will say "WGS84"
when they really mean "ETRS89".  I believe differential GPS in Europe will really give ETRS89 coordinates, although most users would call them WGS84 coordinates.

Duncan Agnew wrote:
> Rather weirdly, Australia has chosen not to define a local frame, even 
> though it has a plate all to itself.

I am not a plate tectonician, but I believe this is wrong.
The Geocentric Datum of Australia 1994 (GDA94) is attached to the Australian plate.
(Source: http://www.icsm.gov.au/icsm/gda/wgs84fact.pdf )

Best regards,
--
Mikael Rittri
Carmenta AB
Box 11354
SE-404 28 Göteborg
Visitors: Sankt Eriksgatan 5
SWEDEN
mikael.rittri at carmenta.com
www.carmenta.com

-----Original Message-----
From: proj-bounces at lists.maptools.org [mailto:proj-bounces at lists.maptools.org] On Behalf Of Andrew Williams
Sent: den 11 december 2008 05:36
To: PROJ.4 and general Projections Discussions
Subject: Re: [Proj] Google Earth Accuracy

Thanks Duncan. That "docking" scenario was how the system was sold to me by the developer of the hardware (Navcom).

Guess the brochure sounds better than the technically correct description;)

So just to make sure ITRF varies over time compared to WGS84 based on tectonics, correct?

-----Original Message-----
From: proj-bounces at lists.maptools.org [mailto:proj-bounces at lists.maptools.org] On Behalf Of Duncan Agnew
Sent: Thursday, 4 December 2008 5:24 AM
To: PROJ.4 and general Projections Discussions
Subject: Re: [Proj] Google Earth Accuracy

Since I am in the crustal-motion business I want to correct the previous posting about ITRF. This is "space-based" only in the sense that it uses data from outside the Earth (GPS and other satellites, along with distant radio sources).  It was not designed with space use in mind. Since ITRF and
WGS84 are both space-based in this sense, they align closely: after all, the center of the Earth and the rotation pole are both non-arbitrary.

ITRF was designed to be an ultraprecise reference frame for measuring motions of the Earth as a whole, and of different parts relative to each other.
Since
the frame is designed to be fixed relative to the velocities of all the tectonic plates, averaged over the whole Earth, any particular point will have time-varying coordinates in this frame--as is is in fact true for any frame not tied to a particular plate (WGS-84 included), though many softwares do not take account of this. The people who develop and update ITRF are working at the millimeter level.

Two "plate-fixed" frames are NAD (at least east of the Rockies), and EUREF (for Northern Europe). Rather weirdly, Australia has chosen not to define a local frame, even though it has a plate all to itself.

For more on ITRF, see http://itrf.ensg.ign.fr/

Duncan Agnew

_______________________________________________
Proj mailing list
Proj at lists.maptools.org
http://lists.maptools.org/mailman/listinfo/proj
_______________________________________________
Proj mailing list
Proj at lists.maptools.org
http://lists.maptools.org/mailman/listinfo/proj
_______________________________________________
Proj mailing list
Proj at lists.maptools.org
http://lists.maptools.org/mailman/listinfo/proj
_______________________________________________
Proj mailing list
Proj at lists.maptools.org
http://lists.maptools.org/mailman/listinfo/proj