Atmospheric mineral aerosols influence climate and biogeochemistry, and thus understanding the impact of humans on mineral aerosols is important. Our longest continuous record of in situ atmospheric desert dust measurements comes from Barbados, which shows fluctuations of a factor of 4 in surface mass concentrations between the 1960s and the 1980s <Prospero and Nees, 1986>. Understanding fluctuations this large should help us understand how natural and anthropogenic factors change mineral aerosol sources, transport, distributions, and deposition, although we are limited in our ability to interpret the results as there is a quantitative record only at one location. We test the hypothesis that dry topographic lows (and not disturbed sources such as cultivated areas or new desert regions) are the sources of desert dust, using a hierarchy of models as well meteorological data sets to look at decadal scale changes in the North Atlantic desert dust. We find that the inclusion of a disturbed source improves our simulations in many (but not all) comparisons. Unfortunately, we are severely limited by the accuracy of the available data sets and models in making definitive statements about the role of disturbed sources or anthropogenic activity in changing the atmospheric desert dust cycle. Processes that might change the size or intensity of desert dust sources in North Africa (such as new sources due to desertification or land use) may be difficult to distinguish from topographic low sources in models due to their similar geographical locations and impact on atmospheric aerosol distributions.