Last Glacial Maximum dune activity in the Kalahari Desert of southern Africa: observations and simulations

It has long been understood that as ephemeral landscape features sand dunes are highly sensitive to environmental change, and thus their distribution and the timing of their development may provide clues to past climate dynamics. The relationship between climate and dune activity, however, is neither simple nor straightforward, with a range of controls affecting the balance between erodibility (the availability of sediment for deflation) and erosivity (the potential for sediment transport). To explore such complex systems over large spatial and temporal scales, a number of dune activity indices (DAI) have been created that incorporate wind speed and moisture balances to calculate the potential for, and degree of dune mobilisation. Using modern weather station data, these indices have generally been shown to provide reasonable indications of dune activity potential. Until recently, however, the detailed quantitative data required to inform these equations has not been available for past climate scenarios, and attempts to determine the relative importance of the various controls of dune activity have relied on rough estimations of climatic parameters. This paper combines data from monthly general circulation model (GCM) outputs from the coupled Ocean-Atmosphere GCMs for 21 ka with the most detailed DAI equation presently available to calculate the potential for dune reactivation in southern Africa during the Last Glacial Maximum (LGM, 18–24 ka). Based on these data and calculations it is indicated that there was significantly less potential for dune activity across southern Africa at 21 ka. When compared to the aeolian sediment records from the region, this study poses serious and fundamental questions about: 1) the reliability of the model outputs, 2) the degree to which DAIs are able to account for the complexity and dynamics of aeolian systems, and/or 3) the interpretation of dune records as palaeoclimatic proxies at millennial time scales.