Bulk composition of northern African dust and its source sediments — A compilation

This paper presents a review of bulk compositional data of northern African dust and its potential source sediments and includes elemental, isotope and mineralogical data. Northern African dust represents about one half of the total global atmospheric mineral dust burden, and its uplift, transport and deposition have strong impacts on climate and various terrestrial and marine ecosystems. The chemical data set shows, that an ‘average northern African dust’ exhibits comparable Si, Fe and Mn contents with respect to the average composition of the upper continental crust, is slightly depleted in the alkali metals K and Na, and enriched in Ti and P. However, the complete data set yields clear evidence that northern African dust and its source sediments are compositionally heterogeneous on a regional scale and that this heterogeneity can be used to differentiate between major potential source areas on the basis of so-called source markers. An evaluation of these compositional fingerprints shows that the following parameters and especially their combination are effective in the discrimination of the most active source areas in northern Africa: ratio of (Ca + Mg)/Fe [wt.%], calcite (or carbonate) content, palygorskite occurrence and abundance, illite/kaolinite ratio, ƐNd(0) value, and 87Sr/86Sr ratio. For example, the data set corroborates previous ideas, which assign carbonate-, illite- and palygorskite-rich mineral dusts to north(west)ern source regions. Because most of the above listed source markers do not change substantially during transport, even far-traveled dusts may be assigned to specific potential source areas in northern Africa. Some limitations of the presented data set are also discussed. Our compilation reveals some substantial gaps in the knowledge of the composition of source sediments and mineral dusts from important potential source areas that should be filled in the future. re compiled data set can be used as a reference frame, when incorporating the composition of source sediments (e.g., mineralogy) into global or regional dust transport models and can be compared with source analysis by remote sensing or back-trajectory analysis. However, source apportionment studies supported by our data set will not only be useful for actual dust samples, but will also be helpful for the understanding of paleo-wind directions and hence paleo-climatological conditions through the investigation of Quaternary eolian sediments deposited in and around northern Africa.