Previous Page Table of Contents Next Page


(Item 8 of the Agenda)

67. The document APCAS/98/7 “Geocoding of Agricultural Establishments in Australia” was presented to the Commission by the Australian Bureau of Statistics.

68. The Commission was advised that the ABS had made a decision in principle to geocode farm businesses on the ABS Business Register. Geocoding involved a process of identifying farm properties by spatially coding the farm entrance, the farm house, the centroid of the property, or property boundaries.

69. The main benefits of geocoded data to the ABS and users of agricultural data were the ability to: (i) classify units to non-standard classifications and provide users with data on user-defined geographical areas; (ii) provide a coverage check; and (iii) produce time series data for geographic areas.

70. The Commission was informed of the difficulties experienced by ABS in geocoding agricultural properties. These included location of the farm entrance in relation to the property, location of the centroid of an irregularly located property and multiple parcels for many farms. Ideally, the ABS would want to capture property boundaries and store them digitally in a Geographic Information System (GIS). The Commission was advised that, in Australia, the development and maintenance of land owner registers was a State Government responsibility and each State had developed its own Digital Cadastral Data Base (DCDB). However, as information about agricultural farms on the ABS’ Business Register was based on the land operator, and not on the owner of the land, very low matching rates had been achieved.

71. The Commission was advised that a number of geocoding trials had been conducted by the ABS with the most recent being done in 1997, using a MapInfo GIS containing physical data (roads, railway lines, rivers), Local Government Area boundaries, cadastral line work and locality names. The GIS plotted a map of the area based on this information and information from the Business Register. Respondents were asked to mark on these maps either the boundaries of their land holdings or a single point corresponding to their farm gate. The trial confirmed the feasibility of this methodology and enabled the ABS to determine the costs of geocoding farms.

72. The Commission was advised of ABS’ involvement in a joint project with the Bureau of Resource Sciences (BRS) to map land use throughout Australia. The proposal involved geocoding data from the 1996-97 Agricultural Census so that land use, commodities produced and land management practices available at the farm level, could be used along with a range of other data, including remote sensing data, to produce detailed land use maps.

73. It noted, however, that a critical issue for this activity would be the need to ensure that confidentiality requirements for use and release of information relating to individual farms were preserved.

74. The ABS advised that a recent proposal from BRS to produce both broad national land use maps and finer detailed maps for selected Key Implementation Areas (KIA’s) using this approach had been approved. This work would enable the activities on land use mapping to progress. However, the major obstacle to advancing this work on a national scale was the initial cost of geocoding all agricultural establishments.

Previous Page Top of Page Next Page