The creation of a national cadastre in Australia was always going to be a problem. In 1787 - the year before the first European settlement - there was just one land parcel on the mainland, and it had no Western-style title. But as European settlement spread from the various coastal cities, the colonists solved their cadastral problems on the basis of their own, local, knowledge.
There was no effort to make the various cadastral systems the same, or even compatible; at the time there was no need. It was an era when business - including the recording of land ownership - was conducted on paper. Communications between businesses was measured in days or weeks and the focus of life was limited by the difficulty of travel.
But evolving technologies, and the requirements of nationally focused organisations, has led to a revolution in the way the nation does business. Part of that has been the idea of a national cadastre. The potential value of such a dataset has been recognised for some time. Even as state-level digital cadastres were being introduced, in 1994, Telstra was investigating the possibility of a national cadastre for its Cable Plant Record Project.
In the event, it has taken a further six years to get it off the ground. The first national cadastre in Australia is called Cadastral Lite. It was launched late in 2001 by PSMA Australia.
PSMA Australia was incorporated on 20 June 2001. It replaced the original Public Sector Mapping Agencies, a consortium formed in 1993 to create an integrated national digital basemap in support of the Australian Bureau of Statistics 1996 Census of Population and Housing.
PSMA Australia's objectives are to co-ordinate the assembly of fundamental national datasets from its member jurisdictions, and to facilitate access to such datasets.
Two key problems stand in the way of a national cadastre: consistency of data and consistency of access.
During the drafting of the cadastral data model, the Intergovernmental Committee on Surveying and Mapping's Cadastral Working Group found that there were very few nationally consistent cadastral data attributes. These inconsistencies would make the population of the full cadastral data model and the creation of a fully attributed national digital cadastral database an immense task. Members of the group were trying to find common elements over the eight jurisdictions, but in many cases there were none.
However, a much bigger barrier for prospective national users was negotiation with the custodians. There were at least eight separate licence terms and conditions. The administration associated with this task proved unmanageable, even for the most persistent user.
In the light of these findings a feasibility study was conducted by PSMA into a dataset that could be built from the ground up. It would offer a graphic representation of land parcel boundaries on a broad scale to enable businesses to integrate it with other data. Initially it would be made up of cadastral polygons, and it would incorporate a unique parcel identifier - a key to link back to the rich data held in jurisdictional DCDBs.
ERSIS Australia won the contract to put the dataset together in 1999. MapInfo bought ERSIS during the course of the creation of Cadastral Lite. The company put together a basic data model in four months, drawing on the cadastral data model developed by the working group. The first version of data went out for evaluation and comments by the industry in early 2000.
Early steps involved importing the mixed format data using some in-house translational software to back up the commercial systems. Peter Clark, now the national data manager of MapInfo Australia, but then the data manager at ERSIS, called the process a 'stocktake of cadastres'. When anomalies were found, the information was fed back to individual jurisdictions for corrections.
Clark said the process allowed the states to expose their products to the market, and enabled them to look at issues such as completeness, currency and accuracy based on external objective reports and from a national perspective.
Cadastral Lite contains all Australia's legal cadastre parcels, showing the boundaries that define land ownership. It's a geometric representation - that is, it is made of geometric shapes (polygons) that represent each parcel of land. It shows density of settlement, and uses nothing more than lines.
The aim in putting together the data was to attach to each polygon as many nationally consistent attributes as possible. At the moment, the parcel's state and local government area, its calculated land area, centroid and legal identifier, are all available. More attributes can be added as they became available.
Dan Paull, PSMA Australia's chief executive, says the legal identifier is critical, because it can be used to identify any individual parcel of land in Australia. Although the dataset does not yet, in itself, offer anything close to all the information about a parcel of land, the identifier is the key that can be used to access it within the richer jurisdictional databases.
'Although there are technical restrictions to Cadastral Lite, even in its current form it is still a very important resource. What it offers is a fundamental national dataset at a price which does not prohibit access,' he says. 'And the way it's being created means it's built on a strong foundation.'
There are plans to lift the level of content to include land use and tenure, and PSMA Australia is currently finalising a study into the feasibility of assembling and maintaining a definitive geo-coded national address file (G-NAF).
But Paull says the tenure question is a good example of the problems that have been faced, and will be faced in the future. 'At this moment, the definition of tenure varies from state to state. There needs to be consistency. There is limited commonality in land data across Australia, and this is an issue that needs to be dealt with now, if we wish to continue to build on the quality of our data. It takes time to implement change,' he says.
The importance of tenure and the need for national consistency have been acknowledged by ANZLIC. At a recent meeting of ICSM a working group was established to examine a tenure standard.
At this stage the plan is to update the national cadastre only once a year. State cadastres are typically updated on a daily or weekly basis. The states will submit new information each July, which will be incorporated into the dataset by September. Paull says at this stage there has been no call for it to be updated more regularly, but if feedback suggests a need, it will certainly be looked at.
Release of the final product, the framework dataset, occurred around the same time as the incorporation of PSMA Australia, in October last year. The actual product is 4-5 Gbyte. This file has all the basic elements; more will be added, but they will not increase the file size overly much.
Paull says the impact of Cadastral Lite is already starting to be felt. He says that as value adders begin selling the benefits of the dataset to business, some of the states, having had to push the boundaries this far, are now seeing the value of going out and looking at the market, and working in response to market trends.
There are currently two value-added resellers - MapInfo and Navigate. MapInfo's CadastralPlus incorporates a number of value-added features, such as enhanced searching and background information and references. It enables users to retrieve information by street name, locality and state.
Navigate has used its multi-vendor and spatial database experience to create the NAVData - Cadastre product from Cadastral Lite. NAVData - Cadastre, like the other Navigate datasets derived from PSMA data (NAVData - Reference and NAVData - Streets) has been migrated into an in-house Oracle Spatial model, at the same time undergoing quality checking procedures developed by Navigate. This Oracle Spatial model is then used as the source for the creation of data and models to support formats used by all the mainstream GIS and spatial database vendors.
The release of Cadastral Lite is a landmark in the context of national information infrastructure development. It is by no means the ultimate land information product, but ten million plus parcels built to a standard specification in a common format, and referenced to a single datum, is an admirable starting point - and a milestone for Australia.
