Related Experiment Video
Updated: Apr 25, 2026
11:19
Measuring Carbon-based Contaminant Mineralization Using Combined CO2 Flux and Radiocarbon Analyses
Published on: October 21, 2016
12.5K
Accuracy of mixing models in predicting sediment source contributions
Arman Haddadchi1, Jon Olley1, Patrick Laceby2
1Australian Rivers Institute, Griffith University, Nathan, QLD 4111, Australia.
The Science of the Total Environment
|August 17, 2014
Summary
Selecting the right sediment mixing model is crucial for accurate source attribution in catchment management. The Distribution and Modified Hughes models proved most accurate for determining sediment sources using geochemical properties.
Area of Science:
- Environmental Science
- Geochemistry
- Hydrology
Background:
- Sediment source attribution is vital for effective catchment management.
- Geochemical properties are commonly used to trace sediment origins.
- Model selection significantly impacts the accuracy of sediment source apportionment.
Purpose of the Study:
- To evaluate the accuracy and robustness of four sediment mixing models.
- To compare the performance of Modified Hughes, Modified Collins, Landwehr, and Distribution models.
- To assess model dependence on geochemical fingerprint properties for sediment source determination.
Main Methods:
- Artificial sediment mixtures were created from three distinct geologic sources.
- Four sediment mixing models were applied to analyze known source contributions.
- Model accuracy was assessed using Mean Absolute Error (MAE) and standard error (SE).
Main Results:
- The Distribution model showed the highest accuracy for individual mixtures (MAE=10.8%).
- The Modified Hughes model was the most robust for grouped samples (5.4% error).
- The Modified Collins model demonstrated significantly weaker performance compared to others.
Conclusions:
- Sediment source attribution is highly dependent on the chosen mixing model.
- Testing models with known samples is essential before field application.
- Distribution and Modified Hughes models offer the most reliable geochemical-based source attribution.
