High frequency un-mixing of soil samples using a submerged spectrophotometer in a laboratory setting—implications for sediment fingerprinting

Summary

This laboratory study evaluates the feasibility of using submersible spectrophotometers as a novel real-time method for tracing suspended sediment sources. The authors demonstrate that soil-specific absorbance patterns at different wavelengths can discriminate between sediment sources and that absorbance is linearly additive, enabling Bayesian mixing models to predict source proportions. The technique shows promise for high-frequency in situ monitoring, though accuracy declines with increasing numbers of source soils.

UK applicability

The methodology may be applicable to UK sediment tracing studies in fluvial environments and soil erosion monitoring. However, the paper is a controlled laboratory proof-of-concept; field validation in UK soil and water conditions would be necessary before operational deployment in UK catchment management or regulatory contexts.

Key measures

Absorbance at UV–VIS wavelengths (200–730 nm); soil sample concentrations; particle size; mean absolute errors for source apportionment (15.4% for two-source, 12.9% for three-source, 17.0% for four-source mixtures)

Outcomes reported

The study evaluated whether a submersible spectrophotometer can distinguish and quantify suspended sediment sources using absorbance patterns across UV–VIS wavelengths (200–730 nm). The research demonstrated that absorbance measurements are linearly additive and can be used with Bayesian mixing models to predict soil sample proportions in artificial mixtures.

Topic tags