The Centre for Resource and Environmental Studies (CRES - now the Fenner School of Environment and Society) at the Australian National University extracted catchments with an average area of approx. 5000 km2 from an unpublished, interim version of a nested catchment framework (Stein, 2006). Main stem sub-catchment units less than 2500 km2 in area were combined with the upstream tributary catchment, and closed (internally draining) basins of combined area less than 10,000 km2 were iteratively aggregated with a lower neighbouring catchment via the lowest point on the drainage divide. The catchments were derived using a multi-flow extension of an early version of the GEODATA 9 second Flow Direction Grid (ANZLIC unique identifier: ANZCW0703012015) associated with the GEODATA 9 second Digital Elevation Model (DEM) Version 3 (Hutchinson et al. 2008; ANZLIC unique identifier: AW0703011541).

Geoscience Australia then used ESRI ArcGIS® software (v.9.1) (ESRI, 2007) and ArcHydro® Tools (v.1.2 beta) (Maidment and Djokic, 2000) to generate a higher resolution catchment coverage of Australia (threshold area 500 km2) from the 9 second DEM. This high-resolution coverage was used to guide the disaggregation of over-sized polygons in the CRES coarse-scale catchment coverage and bring them within an optimal area range of 4000-6000 km2. Catchment polygons with areas less than this range were, where appropriate, aggregated with neighbouring polygons to again achieve an optimal area. Decisions on which polygons to aggregate or disaggregate were guided by information obtained from these datasets:

- NATMAP Raster Premium 2005 (includes 1:250,000 scale topographic and cultural information for the whole of Australia), along with a Landsat 7 national mosaic (Geoscience Australia, 2007);

- Shuttle Radar Topography Mission (SRTM) 3 second (~90 m) v.2 digital elevation model (NASA, 2007).

Applying the method outlined above, an iterative series of edit, clean and build processes in ArcINFO® resulted in the final national catchment coverage. ArcHydro® Tools (v.1.2 beta) was then used to generate an outlet point for each catchment. These points were used as target sites for field sampling of catchment outlet sediments. Point location adjustments were made manually, if necessary, following consultation with State/Territory Geological Surveys and/or assessment against other environmental datasets. Adjustments were made to account for factors such as proximity to sources of contamination, inaccessibility, or false outlet point placement (e.g. points on the Gulf of Carpentaria which were auto-generated in tidal mangrove swamps).

Subsequent to field sampling, minor adjustments to a small number of catchment boundaries were made manually using ArcGIS® software. These adjustments related primarily to catchments where samples had been collected just outside their designated catchment bounds, and generally involved extending the boundary closest to the catchment outlet downstream, usually resulting in a net change to the affected catchment areas of <100 km2.

Notes:

References are listed under Additional Metadata.

The current nested catchment framework is distributed by the Bureau of Meteorology as the 'Geofabric': http://www.bom.gov.au/water/geofabric/index.shtml (last accessed 9 Nov 2011)

Further information on the methods used to generate the NGSA catchments and target sampling sites can be found in Lech, M.E., Caritat, P. de, McPherson, A.A. 2007. National Geochemical Survey of Australia: Field Manual. Geoscience Australia, Record 2007/09, 53 pp.