Identifying eolian dust in the geological record

Recognition of terrestrial dust in geological records is essential for reconstructing paleoenvironments and quantifying dust fluxes in the past. However, in contrast to eolian sands, silt-sized dust is difficult to recognize in pre-Quaternary records due to a lack of macroscopic features indicating eolian transport and mixing with alluvial sediments. Windblown dust deposits are commonly identified by comparing their sedimentological and petrological features with Quaternary examples of dust known as loess. Here, we review the characteristics of terrestrial dust deposits and conclude that most of these features are not exclusively windblown and may be formed by alluvial deposits as well. We therefore synthesize a set of criteria which enable a reliable identification and quantification of dust while acknowledging potential contributions of alluvial components. These methods include quartz-grain surface morphology analysis to distinguish eolian and alluvial transport modes, provenance studies to identify local and extrabasinal sources, grain-size-shape end-member modelling to quantify the various sedimentary contributions to the record, and a basin-scale stratigraphic approach to derive regional patterns and avoid interpretation of local phenomena. We reassess the Eocene to Pliocene records of the Chinese Loess Plateau and conclude that these strata represent both alluvial and eolian sediments deposited in extensive mudflat systems. Quaternary loess, by contrast, is almost exclusively composed of windblown dust. The early Pleistocene shift from mudflat to loess deposits is associated with a significant increase in accumulation rates, likely due to increased dust production upwind, overwhelming and blanketing the local mudflat systems in central China.