Influence of sampling approaches on physical and geochemical analysis of aeolian dust in source regions

The characterisation of mineral dust at emission sources is essential for quantifying the wider scale environmental impacts dust has, as well as improving its incorporation in modelling. Methods of sampling sediments at these source areas for the purposes of dust characterisation are varied and can produce different representations of emitted dust. This study systematically compared dust characterisations from three established approaches for estimating dust emission potential, namely: Big Spring Number Eight (BSNE) traps, a Portable In-situ Wind Erosion Laboratory (PI-SWERL) wind tunnel, and bulk surface sampling, at a known Namib Desert dust source. Individual particle analysis by auto-SEM (QEMSCAN) allowed comparison of size, shape, mineralogy and elemental composition at micrometre-scale for samples from the three approaches. BSNE samples consisted of a lower proportion of fine sediment (<20 µm particle diameter) than PI-SWERL-derived samples, with this performance possibly influenced by atmospheric humidity. In comparison, PI-SWERL characterised the dust with relatively fewer particles between 63-100 µm, a fraction that was more evident in both the BSNE-derived and surface sediments. The reduced representation of this coarser fraction resulted in appreciable differences in particle shape and mineralogic characteristics compared to BSNE and surface-derived samples. The different representations by the three methods return variable dust characteristics at source across fundamental properties of particle size, shape and mineralogy. Awareness of the different representations of dust caused by sampling technique remains essential for the appropriate physical and geochemical characterisation of aeolian dust and highlights how standardised techniques are important for meaningful comparisons, while methods to achieve accurate characterisation remain a priority for the discipline.