Testing the Sediment Fingerprinting Technique Using the SIAR Model with Artificial Sediment Mixtures


Yanchong Huangfu, Michael E. Essington, Shawn A. Hawkins, Forbes Walker, John Schwartz, Alice C. Layton


Purpose The accurate identification of primary sediment sources in a watershed is necessary to implement targeted management practices that will reduce erosion and restore water quality. Sediment fingerprinting is a commonly used tool to accomplish this task. However, the accuracy and precision of different procedures to select tracers for un-mixing sediment sources are still a largely uninvestigated area in relation to sediment fingerprinting. The goal of this research was to validate a sediment fingerprinting methodology by applying it to the Oostanaula Creek watershed in southeast Tennessee, USA. Materials and methods We assessed three method protocols (soil digestion procedure, objective source grouping, and tracer selection) that are utilized for assessing the performance of fingerprinting in terms of apportionment outputs. The major and trace elemental composition of sediment source and suspended sediment were determined by total dissolution and nitric acid extraction followed by analysis with inductively coupled plasma-optical emission spectrometry (ICP-OES). The Kruskal-Wallis (KW) test as well as stepwise discriminant function analysis (DFA) was utilized during tracer selection. The source un-mixing model utilized was a Bayesian mathematical model within Stable Isotope Analysis in R (SIAR). Sediment fingerprinting in the Oostanaula watershed proved to be difficult due to the chemical and mineralogical similarities of the potentially erodible source material. Results and discussion Upon analysis, it was found that the sediment tracers identified as those with low misclassification during cluster analysis would not guarantee a high degree of accuracy during source apportionment. However, there are certain outputs with low errors as compared with the real proportional contributions in artificial mixtures, for example, findings showed that bank erosion is a primary source of suspended sediment in the Oostanaula Creek. Conclusions Source apportionment from sediment fingerprinting was sensitive to the digestion procedure, objective source groupings, and the tracer selection. Our research provides a quantitative approach for assessing the validity of the sediment fingerprinting technique.

Access Full Publication


Huangfu YC, Essington ME, Hawkins SA, Walker FR, Schwartz JS, Layton AC. 2020. Testing the sediment fingerprinting technique using the SIAR model with artificial sediment mixtures. Journal of Soils and Sediments 10.1007/s11368-019-02545-7.