FingerPro: an R Package for Tracking the Provenance of Sediment

Abstract

Soil loss by erosion processes is one of the largest challenges for food production and reservoir siltation around the world. Information on sediment, nutrients and pollutants is required for designing effective control strategies. The estimation of sediment sources is difficult to get using conventional techniques, but sediment fingerprinting is a potentially valuable tool. This procedure intends to develop methods that enable to identify the apportionment of sediment sources from sediment mixtures. We developed a new tool to quantify the provenance of sediments in an agroforest catchment. For the first time, the procedure for the selection of the best combination of tracers was included in the tool package. An unmixing model algorithm is applied to the sediment samples to estimate the contribution of each possible source. The operations are compiled in an R package named FingerPro, which unmixes sediment samples after selecting the optimum set of tracers. An example from a well-studied Mediterranean catchment is included in the package to test the model. The sediment source apportionments are compared with previous results of soil redistributions where 137Cs derived rates validate the unmixing results, highlighting the potential of sediment fingerprinting for quantifying the main sediment provenance. Fingerprinting techniques will allow us to better comprehend sediment transport to water ecosystems and reservoirs and its detrimental effect on the quality of the water and aquatic habitats. The FingerPro package provides further understanding of the unmixing procedure through the use of graphical and statistical tools, offering a broader and easier application of the technique.