Chemistry and grain-size distribution of terrigenous sediments deposited at Site GeoB9501

The Sahara Desert is the largest source of mineral dust in the world. Emissions of African dust increased sharply in the early 1970s, a change that has been attributed mainly to drought in the Sahara/Sahel region caused by changes in the global distribution of sea surface temperature. The human contribution to land degradation and dust mobilization in this region remains poorly understood, owing to the paucity of data that would allow the identification of long-term trends in desertification. Direct measurements of airborne African dust concentrations only became available in the mid-1960s from a station on Barbados and subsequently from satellite imagery since the late 1970s: they do not cover the onset of commercial agriculture in the Sahel region ~170 years ago. Here we construct a 3,200-year record of dust deposition off northwest Africa by investigating the chemistry and grain-size distribution of terrigenous sediments deposited at a marine site located directly under the West African dust plume. With the help of our dust record and a proxy record for West African precipitation we find that, on the century scale, dust deposition is related to precipitation in tropical West Africa until the seventeenth century. At the beginning of the nineteenth century, a sharp increase in dust deposition parallels the advent of commercial agriculture in the Sahel region. Our findings suggest that human-induced dust emissions from the Sahel region have contributed to the atmospheric dust load for about 200 years.

(Table 1) Stable isotope ratios of Globigerinoides ruber and pore water, and Cl measurement at ODP Hole 133-817C

The oxygen isotopic composition of pore waters squeezed from sediments in Hole 817C co-varies with the oxygen isotopic composition of Globigerinoides ruber below 8 mbsf. The magnitude of the variation in the pore water d18O is approximately 30% of the variation in the foraminifers. Overall, the d18O of the pore waters increases down the core, a trend that is also present in the Cl- concentrations. The variations in the d18O of pore waters may be the result of either of two phenomena. First, these may reflect original variations in the waters, the magnitude of which has been subsequently reduced by process of diffusion. Second, these may reflect recrystallization of the precursor sediment and isotopic exchange between the fluids and the recrystallized sediment. At the moment data are not available to ascertain which process is responsible although the correlation between the Cl- and the d18O data suggests that these values reflect the original composition modified by diffusion.

Reconstructions of eolian dust accumulation in northwest African margin sediments

Reconstructions of eolian dust accumulation in northwest African margin sediments provide important continuous records of past changes in atmospheric circulation and aridity in the region. Existing records indicate dramatic changes in North African dust emissions over the last 20 ka, but the limited spatial extent of these records and the lack of high-resolution flux data do not allow us to determine whether changes in dust deposition occurred with similar timing, magnitude and abruptness throughout northwest Africa. Here we present new records from a meridional transect of cores stretching from 31°N to 19°N along the northwest African margin. By combining grain size endmember modeling with 230Th-normalized fluxes for the first time, we are able to document spatial and temporal changes in dust deposition under the North African dust plume throughout the last 20 ka. Our results provide quantitative estimates of the magnitude of dust flux changes associated with Heinrich Stadial 1, the Younger Dryas, and the African Humid Period (AHP; ~11.7-5 ka), offering robust targets for model-based estimates of the climatic and biogeochemical impacts of past changes in North African dust emissions. Our data suggest that dust fluxes between 8 and 6 ka were a factor of ~5 lower than average fluxes during the last 2 ka. Using a simple model to estimate the effects of bioturbation on dust input signals, we find that our data are consistent with abrupt, synchronous changes in dust fluxes in all cores at the beginning and end of the AHP. The mean ages of these transitions are 11.8±0.2 ka (1Sigma) and 4.9±0.2 ka, respectively.

Grain size distribution and bulk chemistry of eolian dust

Atmospheric dust samples collected along a transect off the West African coast have been investigated for their physical (grain-size distribution), mineralogical, and chemical (major elements) composition. On the basis of these data the samples were grouped into sets of samples that most likely originated from the same source area. In addition, shipboard-collected atmospheric meteorological data, modeled 4-day back trajectories for each sampling day and location, and Total Ozone Mapping Spectrometer aerosol index data for the time period of dust collection (February-March 1998) were combined and used to reconstruct the sources of the groups of dust samples. On the basis of these data we were able to determine the provenance of the various dust samples. It appears that the bulk of the wind-blown sediments that are deposited in the proximal equatorial Atlantic Ocean are transported in the lower level (>~900 hPa) NE trade wind layer, which is a very dominant feature north of the Intertropical Convergence Zone (ITCZ). However, south of the surface expression of the ITCZ, down to 5°S, where surface winds are southwesterly, we still collected sediments that originated from the north and east, carried there by the NE trade wind layer, as well as by easterly winds from higher altitudes. The fact that the size of the wind-blown dust depends not only on the wind strength of the transporting agent but also on the distance to the source hampers a direct comparison of the dust's size distributions and measured wind strengths. However, a comparison between eolian dust and terrigenous sediments collected in three submarine sediment traps off the west coast of NW Africa shows that knowledge of the composition of eolian dust is a prerequisite for the interpretation of paleorecords obtained from sediment cores in the equatorial Atlantic.