Geocentric Datum of Australia - FAQ

Jump to Background Questions

Jump to Technical Questions

[ Top ]

Background Questions - Answers

Q1. What is GDA?
The Geocentric Datum of Australia (GDA) is a new coordinate system for Australia which is compatible with modern positioning techniques such as the Global Positioning System (GPS).  GDA supersedes the existing Australian Geodetic Datum 1966 (AGD66) and Australian Geodetic Datum 1984 (AGD84) coordinate systems.  GDA is based on a global framework, the international Terrestrial Reference Frame (ITRF).

Q2. What is the advantage of GDA?
The main advantage is that GDA coordinates are immediately compatible with global coordinates obtained from GPS and with other coordinate systems adopted in many parts of the world.  GDA overcomes the artificial barriers placed by regional coordinate systems and provides improved opportunities to operate in an international community.  An additional benefit of GDA is that it will once again provide a single uniform coordinate system within Australia, allowing the efficient exchange of data and linking of products from information systems.

Q3. Why is Australia adopting the GDA?
Although Australia could continue to use the AGD, it is inefficient to work with AGD coordinates in an international environment where positioning, navigation and information systems relate to a global Earth model.  The longer it is left, the more inefficient it is to convert.

Q4. When did GDA occur?
A lot of work has occured leading to the adoption of GDA, which officially began on the 1st January 2000.  The progressive conversion to GDA is taking place and many authorities have already converted to GDA.  User demand plays a major role in the timing of conversion.

Q5. Do I have to change to GDA?
You don't have to change to GDA. As always, individual authorities will be able to work with project based coordinates (AGD or whatever).  However, as GDA is implemented, it will become inefficient to integrate these projects with other data sets.

Q6. What is MGA?
MGA is a grid coordinate system based on the Universal Transverse Mercator projection and the Geocentric Datum of Australia 1994.  The unit of measure is the metre. Because both AMG and MGA are Universal Transverse Mercator projections, the parameters for both grids are identical (zone numbers, zone width, central meridians, central scale factor, false easting & northing), but different elliposids are used.  AMG used the Australian National Spheroid (6378160, 298.25) but MGA uses the GRS80 ellipsoid (6378137, 298.257222101).

Q7. What is the difference between the existing AGD and the new GDA coordinates?
The Australian Geodetic Datum (AGD) was established before satellite techniques were available and was based on a model of the Earth, that is best suited the Australian region at the time.  GDA is based on an international mathematical model which 'best fits' the shape of the whole earth, with its centre coinciding with the earth's centre of mass.  Coordinates on the earth's surface change approximately 200 metres in a north easterly direction with the new datum (GDA).  The exact change and orientation of the change will vary slightly from area to area.

Q8. What is the difference between GDA and WGS84?
In January 1994, GDA94 and the International Terrestrial Reference Frame (ITRF) coincided.  At around the same time, the WGS84 reference frame was also aligned with ITRF.  In January 1997 WGS84 was re-aligned with the then current version of ITRF, within a few centimetres.  For most practical applications where an accuracy of only a metre or greater is required, GDA94 coordinates can be considered the same as WGS84.  However, for applications where an accuracy of better than a metre is required, the difference must be taken into account. Standard 7-parameter transformations from ITRF to GDA94 are regularly computed from the known GDA94 and continually updated ITRF positions of the Australian Regional GPS Network (ARGN). These parameters can be used to transform between ITRF and GDA94 positions, remembering that ITRF positions may be used as accurate WGS84 positions.

Q9. How will GDA affect maps?
The detail on maps will appear to shift, relative to the map grid. The magnitude of this shift depends on the map scale (divide 200 metres by the map scale to assess the effect).  At 1:1,000,000 the shift is barely significant at 0.2 mm. It is 0.8 mm on a 1:250,000 scale map and 2 mm at 1:100,000 scale.

Q10. How will GDA affect legal boundaries?
Legal boundaries are generally defined by physical features or marks on the ground. These boundaries will not change, though the coordinates of them may.

Some boundaries may have been defined by coordinates in terms of the AGD (or another system).  These boundaries too will not change, although the coordinates will be different when converted to GDA.  Some boundaries were defined by coordinates without reference to a coordinate system.  In these cases the physical location has always been uncertain and this predicament will continue until the boundaries are referenced to a coordinate system.

Q11. Why can't we continue using AGD?
Because of the convenience and efficiency, large organisations may ignore AGD and use global coordinates.  Many international organisations are already committed to a geocentric coordinate system, and the equivalent Australian organisations are following this user demand and international protocol (e.g. Civil Aviation Authority, Dept of Defence).  It is therefore sensible to provide the framework for this inevitable change to occur smoothly and economically and without fragmentation of Australia's coordinate system.

Q12. Will GDA affect heights?
The GDA will not affect heights.  The Australian Height Datum (AHD) is based on mean sea level and will remain so.

[ Top ]

Technical Questions

Q1. Are Australian NTv2 data files compatible with North American files and software?
The files of coordinate shifts conform to the Canadian format known as National Transformation version 2 (NTv2).  The Australian NTv2 transformation files are available as binary files from the GDA Technical Manual.  Free software provided by ICSM jurisdictions can readily convert them to ASCII format as well as carry out the transformation process.  An in-depth explanation of the format can be found in Appendix C of the Victorian GDAit (Lands Vic) or GDAit (Melbourne University) GDAit Users Manual.

In the NTv2 format, south latitudes are stored as negative values, but for ease of operation in North America, western longitudes are stored as positive values.  To maintain compatibility with NTv2, Australian distortion grid files store the(eastern)longitudes as negative values.

Q2. How do I transform from AGD66 and AGD84 to GDA94?
There are 4 methods to transform from AGD to GDA94, they are:

1. National Grid Transformation: The National Transformation grid is the preferred transformation method.  It is the simplest and most accurate and for consistency should be used in preference to less accurate methods. This national grid, which is in the Canadian NTv2 format, has an accuracy of a few cm in dense urban areas.  In October 2001 the national grid to transform from AGD66 to GDA94 and the national grid file from AGD84 to GDA94 were released to replace the previous State and Territory grid files.  These transformation grid files, and free software to use them with, are available from Chapter 7 of the GDA Technical Manual.
   
   
2. 3-Dimensional Similarity Transformation: Also known as 7 parameter transformation.  This transformation option is commonly supported by GIS and Mapping packages and is suitable for "thematic" type data (i.e. vegetation, geology, soil types etc).  National parameters have been developed for AGD84/GDA94, but due to inconsistencies in the AGD66 network it is not possible to have accurate national AGD66/GDA94 parameters, though some states/territories have published their individual state/territory AGD66 parameters.  The official parameters are published in the ICSM GDA Technical Manual.  This method of transformation is less accurate than the National Grid Transformation Option, having an uncertainty of the order of a 1 metre.
   
   
3. Molodensky Transformation: Also known as 3 parameter transformation.  This option is sometimes used in mapping and in hand-held GPS.  Molodensky transformation parameters are available from either AGD66 or AGD84 to GDA94.  The official 3-Dimensional Similarity Transformation Parameters are published in the ICSM GDA Technical Manual. They have an accuracy of about 5 metres and supersede previous parameters published in the United States Department of Defense's, WGS84 report.
   
   
4. Block Shift: The results of this transformation process from the ICSM GDA Technical Manual have produced an accuracy of approximately 10m.  However, if similar grids were developed over a small area, the block shift may have better accuracy.  This method is only recommended where it is not practical to use one of the other transformation options.

Q3. What strategies are available for the conversion of existing digital data?
To convert to GDA94 you must firstly know what coordinates system has been used for your existing data.  You may then immediately convert all your data to GDA94, or convert on demand. Conversion of all data produces a consistent data set, but conversion on demand may minimise the amount of data to be converted.  Whenever practicable, new data should be collected in terms of GDA94. Whatever strategy is adopted, data must be labelled with the coordinate reference, and if necessary the transformation model, used to derive it.

Q4. What will the effect of transformation between AGD and GDA be?
Transforming coordinates generally will not affect their relative accuracy.  The relative accuracy of your positions will be retained because the transformation is similar for all points in a region.  However, the absolute accuracy of the transformed data set will depend on the accuracy of the transformation method used.

Q5a. What is NTv2?
NTv2 (National Transformation version 2) is a grid of accurate datum shifts in a specific format originally devised by the Geodetic Survey of Canada.

Using Australian distortion grids (in NTv2 format), the change in latitude and longitude from AGD to GDA94 can be determined for any position by simple interpolation and is then added to the AGD position to give the GDA94 position.  The grid is generally regular, but where necessary may contain "sub-grids' of greater density to cater for particular areas.  A description of the NTv2 algorithm can be found in the manuals for the Victorian GDAit software GDAit (Vic Lands) or GDAit (Melbourne University).

Q5b. Where do I get the national transformation grids?
These are available from the GDA Technical Manual

Q6. Can I use Molodensky or 7 parameter transformation instead of NTv2?
The NTv2 transformation method is the simplest and most accurate and for consistency should be used in preference to less accurate methods.  Other methods, such as the Molodensky or 7 parameter transformation, can be used if it is not possible use NTv2 (e.g. transformation in a hand-held GPS receiver) provided the accuracy limitations are understood and where appropriate they are included in the meta-data.

Q7. Why is the ITRF92 used for GDA, instead of the WGS84?
The International Terrestrial Reference Frame (ITRF) has been adopted in favour of WGS84 because it is more recent and is supported by the International Union of Geodesy and Geophysics (IUGG), whereas WGS84 was developed by the US Department of Defence years ago.  This decision was affirmed in early 1994, when WGS84 was modified to align it more closely with ITRF.

Q8.Why is the GRS80 ellipsoid used for GDA, instead of the WGS84 ellipsoid?
The two ellipsoids are effectively identical.  When the WGS84 system was developed it was based on the GRS80 ellipsoid, but computational techniques resulted in a small difference in the flattening.  When used to express earth-centred Cartesian positions (X, Y, Z) as latitude, longitude and ellipsoidal height, these two ellipsoids result in a difference of less than 1 millimetre.

[ Top ]

Q9. Why don't the MGA grid coordinates use a different false origin to that used by the existing AMG grid coordinates, so that the difference will be obvious?
he values for the false easting and false northing are defined as part of the Universal Transverse Mercator projection.  While different false origin values could be used, the resulting grid coordinates would no longer be Universal Transverse Mercator coordinates, and would not be globally compatible/icsm defeating one of the principal aims of the GDA.

Q10. Can the AMG neatlines be retained while using the MGA grid?
Yes, but these neatlines (map edges) would be re-labelled with the new GDA values, which would be non integer numbers (e.g. 30°30'05.6218")

If the neatlines were given the same GDA values as they had in AGD (e.g. 30°30'), then the detail covered by the map will appear to shift (by about 200 metres).

Q11. How will 'continental drift' affect GDA?
The Australian plate is known to be moving in a north easterly direction at a rate of about 6 cm per year, but there is little known distortion within the plate.  This means that continental drift will have no real effect on relative GDA positions, but over time the absolute position of all points will change by a similar amount.

Those seeking accurate positions from GPS generally work with differential positions and will not detect any change in absolute position.  With a single hand-held GPS receiver the best absolute position that can be obtained by routine means is about 10 metres.  This means that it would take about 150 years before the change could be detected.

With more sophisticated GPS equipment (dual frequency surveying recievers) and long observing periods (at least several hours) it is possible to obtain positions with an accuracy of few centimetres.  With this accuracy, the shift due to "continental drift" since GDA94 was introduced (1994) can be measured.  However, GeoScience Australia's on-line GPS processing system provides accurate positions in the global system (the International Terrestrial Reference Frame) and also transforms them to GDA94, thus eliminating any possible confusion due to 'continental drift'.

Q12. Will zero degrees longitude still pass through Greenwich?
No - the diffence is well defined and is important scientifically - but for most day to day purposes it has no real consequence.  Today, it can be confusing as there are four Meridians all passing through the Old Royal Observatory.

The earliest is Flamsteed's, named after the first Astronomer Royal, which was established in 1675.  In 1725, Edmund Halley, the second Astronomer Royal established a second Meridian.

The third was defined by another Astronomer Royal, James Bradley, in the mid-18th century, and is still used as the basis for map-making in Britain today.

The fourth Meridian was established in 1851 by yet another Astronomer Royal, Sir George Airy and is the basis for the monumented zero meridian at Greenwich Observatory.

To further confuse the matter zero degrees longitude as determined by space techniques gives a different result.

The difference in longitude between the two systems, that available in 1936 (Airy) and WGS84, at the Airy Transit is 102.478 metres.  Therefore the International Reference Meridian (IRM) is 102.5 metres east of the Airy Transit at Greenwich.  As the IRM is tied to the definition of time the real reference meridian to be used for the Millennium is the IRM which, as shown above, is about 100 metres east of the Greenwich Meridian.
http://www.rog.nmm.ac.uk/leaflets/longitude/longitude.html

Q13. Will the equator be in the same place?
No, but the equator as defined by the AGD coordinate system was incorrect.  The GDA equator is the best estimate.

Q14. Why is the size of the AGD-GDA vector calculated from the geodetic coordinates, different from that calculated from the grid coordinates?
When a position is transformed from AGD to GDA94 the magnitude of the apparent difference in position can be calculated from the difference in the original and transformed latitude and longitude.  This is done by converting the difference in latitude and longitude to metres and using trigonometry to calculate the distance.

If the AGD latitude & longitude is converted to AMG and the transformed position is converted from GDA94 to MGA94; the AMG and MGA coordinates are differenced and used to calculate a vector, it may be significantly different to the vector calculated from the latitude & longitude.  This is because the two grid systems use different ellipsoids and this directly affects the meridian distance used in the calculation of the grid coordinates.

This difference is most significant near the zone boundary and increases as the location moves away from the equator, being about 15 metres at 40° latitude.  The grid differences shown in the spreadsheet of Chapter 7 in the GDA Technical Manual are still valid as a simple, low accuracy method to convert between AMG and MGA, provided the values are only used in the zone for which they were calculated.

[ Top ]