ASSESSING THE PARTICLE SIZE AND ORGANIC MATTER SELECTIVITY OF SOIL EROSION USING A RAINFALL SIMULATOR AND ITS IMPLICATIONS FOR SEDIMENT SOURCE FINGERPRINTING

At present, one of the main limitations of the sediment source fingerprinting technique is the ability to link sediment back to their sources due to the non-conservative nature of many of the sediment properties. Ideally, sediment properties do not change as it moves through the watershed allowing for a direct comparison between sources and sediment. However, eroded sediments are often found to have a finer grain-size distribution and a higher organic matter content compared to the source material as the smaller and less dense particles are preferentially mobilised and transported. Accounting for changes in both particle size distribution and organic matter content are important as many tracers are correlated with both properties. Understanding how best to deal with these changes in sediment properties has been receiving more attention recently as it is unclear what is the best approach. Sieving (e.g., <63 µm, <10 µm) is a common procedure used to limit the differences in the particle size distribution but there are limitations in terms of the representativeness of transported sediment. Even with sieving, small differences in the particle size distribution can have a large affect on tracer concentrations so correction factors are often applied. Many correction factors assume that the relation between particle size and tracer concentrations are linear and that it is similar for all tracers. However, such correction factors are controversial because little is known about the potential errors associated with its use. There is currently a need to understand the processes that drive particle size and organic matter selectivity and its implications for sediment fingerprinting. A rainfall simulator experiment was setup to investigate the particle selectivity of water erosion in an agricultural setting. Both the source of sediment and the generated runoff were characterized in terms of their particle size distribution, organic matter content and geochemistry. This research can be used to guide sampling protocols, fingerprint selection, data correction factors and data interpretation.