Is It Time for a New National Height Datum?

Personal Recommendations

The following recommendations are the views of the author, and I have arrived at these conclusions after the extensive and rigorous research undertaken as part of this thesis. There are many similar research papers by other authors which outline their views on the need for a new national height datum for Australia, and these have been consulted.

After completing the research as part of this thesis submission, it appears that there is a very clear need for an improvement in the Australian Height Datum, and it is not a matter of if it will be updated, but rather how and when. It is well understood that national geodetic datums have a finite life-span, and should be upgraded or replaced as required (Featherstone & Filmer, 2012, p. 21). An example of this is in the Geocentric Datum of Australia 1994 (GDA94). Whilst not height-related, GDA94 was considered cutting-edge when it was adopted; however it has since been shown to have limited internal accuracy and indirect linkages to the current global reference frame, the International Terrestrial Reference Frame 2008 (ITRF2008)  (Haasdyk & Watson, 2013). The Australian Height Datum has been a very good reference datum for Australia for over 40 years, however the use of heights in Australia has changed rapidly over this time, and as a result AHD71 no longer meets all the requirements of a modern vertical datum (Featherstone, 2006). It is for this reason that vertical datum improvements is of high importance for the future of Australian surveys.

Several different options for upgraded height datums have been presented in this thesis, namely from Featherstone & Filmer (2012) and Vanicek (1991). In my view, if an upgrade of the Australian Height Datum is to be done in the near future, the best approach for Australia is the second option outlined by Featherstone & Filmer (2012) – a combined vertical datum. A combined vertical datum is a datum that combines Sea Surface Topography (SSTop) corrected Mean Sea Level at multiple tide-gauges, and h – N at appropriately chosen points across the continent to define the zero reference surface, but also relies on the levelling network to propagate the heights throughout the network that connects bench marks (Featherstone & Filmer, 2012). Care must be taken as to which tide gauges are chosen to define the zero reference. Australia has a very large levelling network, and so unlike most other national height datums, it is not acceptable to fix a levelling-only datum to MSL at only one tide gauge. If only one tide gauge was used to fix the level datum it is almost definite that systematic and gross levelling errors will accumulate over large distances, causing it to be a very inaccurate datum – “where errors in levelling networks that connect tide-gauges become larger than MSL and SSTop errors at those tide gauges, fixing the levelling network at only one tide-gauge becomes a less attractive option” (Featherstone & Filmer, 2012, p. 24). The MSL in the current AHD was adopted through the local MSL at several tide-gauges around Australia’s coastline. As the vertical offset between tide-gauge-observed MSL and the quasi/geoid (i.e. mean local SSTop) is different at each tide-gauge, changing principally in the north-south direction in Australian coastal regions, the problem with the north-south slope is likely to be present in any new AVD that adopts uncorrected MSL as its zero reference – it is therefore not considered appropriate for future AVD (Featherstone & Filmer, 2012). Filmer (2010) stated that oceanographic SSTop models are now of sufficient accuracy that using multiple tide-gauges with an SSTop model is now a realistic option if a levelling-only vertical datum constrained at multiple tide gauges becomes the preferred choice for future AVD.

The practitioners interviewed as part of this thesis all believed that a new national height datum is not necessarily required. However all provided opinions on current restrictions and issues in AHD71 which need to be addressed if its use is to be continued. The issues which are important to be addressed in the current height datum are explained by Filmer (2010), and include the Australian National Levelling Network, tide-gauge records, terrestrial and satellite/airborne gravity, Sea Surface Topography Models and GNSS measurements. The practitioners interviewed all mentioned a shortcoming in the AHD71 in the form of a lack of quality benchmarks. While relating more to the network itself than to the datum,  it is obvious that increasing the number of high quality benchmarks across Australia would be of high importance for any future development of the levelling network, whether that be in the course  of defining a new national height datum, or simply providing improvements to to existing AHD71.

The author’s view is that a new national height datum is required in the future; however it is very unlikely and not entirely feasible in the near-term future, mostly due to the costs which such an upgrade is likely to incur. Ideally this upgrade would be as soon as possible, however there is not enough evidence to say that it is absolutely necessary immediately. Whilst it has been stated by the Intergovernmental Committee on Surveying and Mapping (ICSM, 2013) that the Australian Height Datum will remain Australia’s official vertical datum for the short to medium term, the AHD contains deficiencies that make it unsuitable in the longer term (Featherstone & Filmer, 2012). As more and more height uses are found and developed, and the need for higher quality height data becomes even more apparent, the need to upgrade the AHD may overwhelm the cost of doing so. In my opinion, therefore, of more current importance and the short-term need is the improvement of the current Australian Height Datum and the associated ANLN, with a view to eventually upgrading it.  Any improvements in the current height datum of Australia will likely help in future work for an upgrade to the AHD, and so any improvements are still positive steps towards the final goal of a new national height datum for Australia. A dynamic datum, whilst ideal for accurate height usage, is not deemed feasible due to the very high costs which would be involved in the determination and realisation of such a datum.

Whilst in the author’s opinion the most important short-term goal for Australia is the improvement of the current datum, it is vital that, if a new national height datum is to be produced for Australia, it is developed in such a way that a relationship could be established with a potential new world-wide height datum. In its current form, the Australian Height Datum is not compatible with any global vertical datum. The development of a new Australian Height Datum is necessary if Australia is to benefit from any future Global Vertical Datum (Filmer, 2010, p. 256). Future research and development to improve height datums in Australia should therefore involve connection to a Global Vertical Datum, when realised. This can be implemented on a global scale using a global geoid model and GNSS heights, just as has been done at a regional level, or by computing geopotential differences between a global geopotential value and the local vertical datum (ibid, p. 256).


Thomas Pollard z3323750 | School of Civil and Environmental Engineering | University of New South Wales