Bioregional_Assessment_Programme_Land Use of Australia, Version 3 - 2001/2002

Abstract

This dataset and its metadata statement were supplied to the Bioregional Assessment Programme by a third party and are presented here as originally supplied.

The 2001/02 Land Use of Australia, Version 3, is part of a series of land use maps of Australia for the years 1992/93, 1993/94, 1996/97, 1998/99, 2000/01 and 2001/02. The non-agricultural land uses are based on existing digital maps covering four themes: protected areas, topographic features, tenure and forest. Time series data at relatively high temporal resolution were available for the protected areas and forest themes. The agricultural land uses are based on the Australian Bureau of Statistics' agricultural censuses and surveys for the years mapped. The spatial distribution of agricultural land uses is interpretive and has been determined using Advanced Very High Resolution Radiometer (AVHRR) satellite imagery with ground control data. The maps are supplied as a set of ARC/INFO (Trademark) grids with geographical coordinates referred to GDA94 and 0.01 degree cell size. For each of the years mapped there is a set of probability maps, one for each agricultural land use and a single summary map showing the non-agricultural land uses and a likely arrangement of the agricultural land uses. The arrangement of agricultural land uses in the summary map was determined from the probability maps using some simple rules to make an approximation to a maximum likelihood land use map. As supplied, the probability maps are floating point grids with cell value between 0 and 1 and no value attribute table while the summary map is an integer grid with a value attribute table with attributes defining the agricultural commodity group, irrigation status and land use according to the Australian Land Use and Management Classification (ALUMC), Version 5 (http://www.daff.gov.au). Prospective users of the data should note the caveats and additional metadata, which are included in the document entitled 'User Guide and Caveats: 1992/93, 1993/94, 1996/97, 1998/99, 2000/01 and 2001/02 Land Use of Australia, Version 3' (BRS, 2006c). The caveats are also available as a separate document entitled 'Caveats: 1992/93, 1993/94, 1996/97, 1998/99, 2000/01 and 2001/02 Land Use of Australia, Version3' (BRS, 2006a).

Dataset History

I. For each year mapped, four thematic layers were constructed in raster form with 0.01 degree pixel size and overlain to determine the non-agricultural land uses and the distribution of agricultural land. The layers were a topographic features layer, a protected areas layer, a tenure layer and a forest type layer. They were based on a 1999 update of TOPO-250K (Series 1) and a 2005 update of TOPO-250K (Series 2), 1:250,000 scale vector topographic data sets published by Geoscience Australia (GA); the Collaborative Australian Protected Areas Database data sets for 1997, 2000 and 2002, 1:250,000 scale vector protected areas data sets published by the Department of Environment and Heritage (DEH); Australian Tenure, a 250m raster tenure data set compiled by BRS in 1997; agricultural land use status information for aboriginal freehold and leasehold land from state and territory agencies; forest extent data compiled by the Department of Environment and Heritage for greenhouse accounting purposes for 1992, 1995, 1998, 2000 and 2002 - 25m raster data; crown cover data from Vegetation: Present (1988) and Vegetation: Pre-European Settlement (1788) published by GA; land use data from the collaborative 'Land Use Mapping at Catchment Scale' project managed by BRS (BRS, 2002) and from the collaborative 'Land Use Data Integration Case Study - Lower Murray NAP Region' project managed by BRS and from the Agricultural Land Cover Change: 1995 Land Cover data set compiled by BRS; and plantation forest data from BRS's Plantations 2001 data set.

II. The spatial distribution of specific agricultural land uses for each of the six years was determined using SPREAD II, developed by Simon Barry of BRS. SPREAD II, like the SPREAD algorithm of Walker and Mallawaarachchi (1998), uses time series NDVI data with control sites (ground control data comprising records of the agricultural land uses that existed at specific points in specific years) to spatially disaggregate agricultural census or survey data. The SPREAD II methodology is statistically based, using a Bayesian technique - a Markov Chain Monte Carlo (MCMC) algorithm. It has been implemented in R. NDVI images were obtained from AVHRR data processed to correct for cloud cover by DEH. Control site data were collected by State and Territory agencies for the Audit (project BRR5) and relate to the years 1996, 1997 and 1998. The irrigation status of most control sites is known and the method was used to determine the distribution, not only of commodity groups, but also of their irrigation status. Agricultural census and survey data reported on Statistical Local Areas (SLAs) were obtained from the Australian Bureau of Statistics. Modifications made to the agricultural census and survey data are the same as carried out during the construction of the 1996/97 Land Use of Australia, Version 2 (Stewart et al, 2001). The irrigation boundaries data set published by the Audit, the Australian Irrigation Areas, Version 1a, with some additional polygons incorporated for irrigation districts in Victoria, served as an irrigation constraint to refine the prior probabilities used in the MCMC algorithm. A horticulture mask constructed using some of the data sets listed in section I served as a horticulture constraint to refine the prior probabilities used in the MCMC algorithm. For each of the six years, SPREAD II generated outputs comprising the 42 probability maps described in the abstract and a summary agricultural land use map (the agricultural component of the summary map described in the abstract).

III. Land uses were assigned to pixels in the summary grids with the aid of a macro, which assigns land use categories from the Australian Land Use Management Classification Version 5 (search the website http://www.daff.gov.au/ for the term 'ALUMC') according to the attributes of the four layers overlaid in step I and of the summary agricultural land use map made in step II.

Dataset Citation

Australian Bureau of Agricultural and Resource Economics - Bureau of Rural Sciences (2013) Bioregional_Assessment_Programme_Land Use of Australia, Version 3 - 2001/2002. Bioregional Assessment Source Dataset. Viewed 21 December 2017, http://data.bioregionalassessments.gov.au/dataset/98cea3eb-61a2-4e77-ad90-c280ec645c3a.