Are source groups always appropriate when sediment fingerprinting? The direct comparison of source and sediment samples as a methodological step

The classification of sediment source groups is often the least thoroughly considered part of a
sediment fingerprinting methodology; however, the use of inappropriate source groups can be
the cause of significant uncertainty. In many catchments, source groups based on land use or
geology are a poor fit for their geomorphological processes and the nature of the tracers used.
Against this context, this study directly compared the average percentage difference in the
standardised concentrations of all tracers between a sediment sample and each individual source
sample, to map the similarity between the properties of sources and sediment in 3 study catchments.
The environmental significance of individual tracers and their similarity between individual
samples were also examined in order to identify functionally important source groups. In the
River Nene, UK, the mean percentage differences between source and sediment tracer concentrations
were primarily controlled by the presence of distinctive ironstone and urban sources,
which had very dissimilar properties to the target sediment. However, a generally consistent
trend of certain source samples having more similar properties to multiple target sediment
samples than others was also found; a finding that could not be identified when using conventional
source groups. In the Sywell reservoir catchment, UK, sediment originated from throughout
its catchment, apart from in the case of damaged road verges, and there was little indication of
any major change in sediment sources through recent time. In the Vuvu catchment, South Africa,
there was a larger contribution from distal igneous sources during high‐flow events. The trialled
method, however, provided little advantage over the standard fingerprinting approach in this
case, due to the existing good fit between catchment geomorphology, the tracers used, and the
geological source groups. The method trialled herein can provide distinct advantages over the
conventional fingerprinting approach and, although it should not replace it, provides a useful
supplement by permitting an assessment of whether potential source groupings make best
environmental sense and providing increased resolution of sediment provenance.