The Total Ozone Mapping Spectrometer (TOMS, a satellite-based instrument) shows that most atmospheric dust comes from a very few places (Fig. 1). By far the dustiest is the Bodélé depression (in the Djourab of northern Chad). At the Bodélé, strong winds, funnelled round the eastern slopes of the Tibesti Mountains, pick up fine diatomites and clays from the bed of a now-dry lake (Washington, et al., 2003, Fig. 2). Some 5000 years ago, when it was full, the lake was "Mega-Chad", then the size of the Caspian Sea (Grove and Warren 1968). Today, the Bodélé contributes well over half of the c. 400-700M tonnes of dust that leave West Africa each year. Much ends in the Atlantic, but some reaches South America (Swap et al. 1992; Todd, et al., 2003, Fig. 3). The process is year-long, but it is most intense in spring and early summer, when the Harmattan cloaks West Africa in dust.
Dust and Global Climate
Because dust interferes with radiation, it is crucial to global climate. Sediments and soils are only one source of dust (or atmospheric aerosol). Others are volcanic eruptions, pollution (largely sulphates), bush/grass fires, and the oceans. The climatic impact of dust from soil and sediment is much less well understood than that of pollution, even though the two may be comparable in magnitude (Ramaswamy, et al., 2001). In many places dust from soil is by far the bigger source (Tegen, et al., 1997). It has also been suggested that dust can further influence climate by suppressing rainfall (Rosenfeld and Lahav, 2000), possibly creating in a positive feedback cycle. Moreover, there is evidence that more of this kind of dust is being produced (Goudie and Middleton, 1992; Prospero, 1996). Dust has probably played and will probably play a pivotal role in climate change, but there is uncertainty about the magnitude and even the direction of its influence (Ramaswamy, et al., 2001).
Dust is also important to Earth's nutrient budget, especially in the supply of iron, potassium, phosphorus and calcium to the oceans. The fertilisation of the ocean with iron during the Quaternary may have been one of the key processes responsible for glacial-interglacial cycles (Ridgewell and Watson, 2002). Dust from the Bodélé is thought to be the main source of nutrients in Sahelian farming systems (Chappell, et al., 1998) and Ghanaian forests. It is also an important source for nutrients in parts of the Amazon (Swap, et al., 1992) (Fig. 3).
Dust from the Bodélé
The production of dust from the Bodélé, as from anywhere else, is influenced by two sets of factors: the erodibility of the surface; and the erosivity of the wind. As to erodibility in the Bodélé, recent high-resolution MODIS satellite data lead us to hypothesise that the primary sources of dust there are evaporite pans or sediments near the shoreline of Lake Mega-Chad (Fig. 4). This idea fits with a map of the shoreline platforms (Ghienne, et al., 2002). As to erosivity in the Bodélé, our own research indicates that dust production is related to the strength of a topographically constrained low-level wind jet (Goudie, et al., 2003).
Fundamental questions about these processes are unanswered. Long-term averages of different satellite aerosol data (TOMS and the Meteosat Infrared Different Dust Index) may agree that the Bodélé is the major dust-producing region, but their estimates of the variability in dust production at sub-monthly to interannual timescales differ markedly (Washington and Todd, 2004). Nor do we know the history of supply over the last 5000 years, or how size of the remaining reservoir of dust. Although more is being learnt about the recent climatic history of the area (Ghienne, et al., 2002), including the recent discovery of a seven million year old hominid skull (Sahelanthropus tchadensis) (Vignaud, et al., 2002), little of this has been related to the history of dust production.
Without field observations we can only speculate about how and from exactly where dust is released from the Bodélé. For a start, dust obscures the surface processes that create it on satellite imagery. If it is like other the dry lakes we know, its sediments are likely to vary according to distance from the former shoreline and other factors. In other dry lakebeds, there is a complex interplay between the production of sediment that can become dust and its removal (Chappell, et al., 2003). The patterns of each are very spatially variable. If the Bodélé is anything like Owens Dry Lake in California, which has recently been intensively studied (Gillette, et al., 1997), bombardment by saltating sand is strongly implicated in the liberation of dust. We hypothesise that the intense flows of sand in the trains of barchans, which traverse the Bodélé, are one of the chief liberating mechanisms (Mainguet 1968).
Because of the global dominance of the Bodélé as a source of dust, quantitative confirmation of these hypotheses (or discovery of unsuspected sediments and processes) would have impacts well beyond the region. Estimates of dust production in the past, present and future, and their climatic impact have so far depended only on models, which in turn depend on estimates of dust production and transport available from satellite data. There are important limitations to these techniques, notably their coarse spatial resolution (>1 degree in the case of TOMS) and poor vertical resolution (TOMS only produces estimates of dust above 1000 meters). There is precious little data on dust movement in the all-important lower levels (in the atmospheric boundary layer), let alone an understanding of the interplay of erosivity and erodibility that releases it from the surface. Of the 500 or so ground-based radiometers in the AERONET network currently providing the best validation data on dust in the atmosphere, none is located in the Sahara, and none in the dustiest of all places, the Bodélé. This calls for a multi-disciplinary research effort.