Spatial uncertainty of the137Cs reference inventory for Australian soil

[1] Soil erosion threatens the productivity of agricultural land, reduces the soil's ability to store water, promotes losses of nutrients, organic matter and soil biota and increases the amount of dust carried by wind. It is difficult to measure directly, particularly over time periods with sufficient duration to represent its variability. The caesium-137 (137Cs) technique estimates net soil redistribution by the combined effect of wind, water and tillage erosion. Used with a single sampling campaign, the 137Cs technique provides retrospective estimates for a given location on medium-term (30–40 years) erosion or deposition rates. In 1990 a national reconnaissance survey of soil erosion in Australia was performed using 137Cs. The aim here is to use the results of that national inventory to establish with spatial uncertainty a preliminary baseline map of the 137Cs reference inventory for Australia to subsequently estimate net (mid 1950s to early 1990s) soil redistribution for Australia. We used an indicator co-simulation between 141 samples of the 137Cs reference inventory and gridded mean annual (1954–1990) rainfall and compared the results with a deterministic method using rainfall and latitude. The former produced a baseline map which reconstructed the spatial data with less bias and smaller uncertainty than the latter. The Australian 137Cs reference inventory map can be improved over time with additional samples and the improved map used for subsequent soil redistribution assessments. Where sufficient data exist the geostatistical method may provide similarly reliable 137Cs reference inventory estimates at unsampled locations around the world.