An assessment of catchment condition in Australia

Abstract

Abstract The condition of Australia's water catchments was assessed in a 2 million km2 intensive land-use zone (ILZ) as part of a National Land and Water Resources Audit (the Audit). The assessment used existing biophysical data obtained from satellite imagery, digital elevation models, computed or derived values, surrogates and existing spatial data from state and national databases. The data were used to derive indicators and an index of relative catchment condition for 3718 sub-catchments and 197 basins. The spatial patterns for catchment condition were then mapped across the ILZ. The term indicator is used consistently to mean a single attribute (even if it is derived from several variables), whereas a sub-index or an index is an aggregation of scaled indicator values. An inventory was made of all potentially relevant biophysical measures of the water, land and biota in catchments that existed in archives. Criteria to assess data quality including a national coverage, ease of interpretation, scale, relevance to policy development and catchment management were applied to some 110 measures from which 21 indicators were ultimately selected and used for tabulation, mapping and interpretation. The data were assembled at 250 m cell resolution and aggregated to 5 km cells as the base unit for analysis. To map the indicators or indices, a five-point classification (quintiles) was used, ranging from relatively poorer to better condition. The class boundaries were based on equal intervals or equal areas under the frequency curve of each indicator or index. Indicator values for each 5 km cell ranged from 1 to 5 (1 being poorer, 5 being better). Index values were calculated by adding the scores for a set of indicators, for example, all the 21 identified and taking a mean for all the 5 km cells within each catchment. Data manipulations and analysis used a multi-component decision support system (DSS), geographic information system (GIS) called catchment condition (CatCon). Numerous maps were generated for each indicator and for indices comprising different combinations of indicators, ranging from the complete set (21), a core set (14) to relatively few (6) and for different regionalisations. Cross-comparisons between catchment condition classes and some consumptive outputs from catchments were also carried out. Only maps for the index based on 14 indicators and a single cross-comparison are presented here as illustrations. The web site www.affa.gov.au/catcon/ has more examples. The main points to emerge were: clear interpretable patterns of better to poorer condition were defined for the intensive land-use zone; visualisations of catchment condition were markedly different for classes defined by equal area or equal interval; the broad patterns of relative condition could be captured with relatively few aggregated indicators; the cross-comparisons suggested areas where land-use changes are needed to develop sustainable land-use systems. The study illustrates that relatively low quality but extensive data can be used to provide a rational basis to identify priority areas for management action and for natural resource management policy development. We suggest that the development of a higher quality national data set in Australia to assess the condition of the Nation's catchments may be best achieved through the co-ordination and collection of core data sets (indicators) at regional scales.