The Bodele depression at the southern edge of the Sahara is a fearsome, forsaken place. Winds howl through the nearby Tebesti Mountains and Ennedi Plateau, picking up speed as they funnel into a parched wasteland nearly the size of California. Once there was a massive freshwater lake here. Now the lake is a shrunken puddle of its former self. Across most of the landscape, there is nothing.

Or so it would seem. But as the winds sweep the ancient lake bed, which has not been inundated in much of this area for several thousand years, they carry trillions of tiny particles skyward in vast, swirling white clouds. The dust then starts a mysterious journey—or a series of mysterious journeys—that scientists are trying to better understand.

Only a few decades ago researchers did not pay much attention to dust. Like the rest of us, they cleaned under their furniture and occasionally took note of drifting flurries of house motes—concoctions of particles that generally include bits of dead insects, plant
fibers and kitchen crumbs. Scientists studying the earth’s atmosphere were far more interested in man-made particulate matter—pollution. Few bothered to recognize that millions of metric tons of soil or mineral dust were circulating around the globe at any given time, affecting the climate, fertilizing the oceans and contributing vital nutrients to the Amazon rain forest, among other places.

Joseph M. Prospero was one of the pioneers. A professor emeritus in marine and atmospheric chemistry at the University of Miami, he has been called the grandfather of dust studies in the U.S. Yet he also recalls that when he published papers in the 1960s and early 1970s suggesting a massive transport of African dust across the Atlantic to the Americas, some of his colleagues were skeptical that this was a subject of significant scientific interest. “People used to find the topic of dust funny,” he says.

His was a lonely enterprise, monitoring dust stations in Barbados and other pristine locations, analyzing and measuring what he could catch in his air filters. Eventually interest grew, however, in part because satellite photographs showed in ever greater clarity what Prospero and a few others were describing: giant plumes of particles, hundreds of kilometers wide, being swept off the African continent like sea spray from a massive storm and falling on the other side of the Atlantic. At the same time, interest in climate change grew, and it became clear that dust played a key role in modulating the earth’s temperature.

“Now there are so many scientific papers coming out on dust, it’s impossible to read them all,” Prospero says. By one count, publications on Saharan dust doubled every four years from the early 1970s to 2001. Thomas E. Gill, an associate professor of geological sciences at the University of Texas at El Paso who helps to keep a database on dust, says he has a hard time keeping up. “You think it’s an esoteric topic, but every week I see somewhere between 50 and 100 publications on dust in some shape or form.”

What are all those studies telling us? The story of dust is actually about the challenges of trying to figure out what is happening to the planet we inhabit. It shows how an influence on one area of the earth’s ecosystem can have outsize effects on other areas. “The more our scientific tools encourage us to get to one answer, the more they lead us to three more questions,” says Robert J. Swap, a environmental studies professor at the University of Virginia. Swap, who co-wrote a seminal 1992 paper on African dust in the Amazon, says the study of dust leads to one conclusion: “We need to honor the complexity of nature.”

A way to understand that complexity is to follow a hypothetical handful of particles from the Sahara across the Atlantic. Along the way, and once our dust particles arrive at their next destinations (there are no final destinations), we can examine how they interact with the world around them.

We start in the Bodele because it is widely recognized as the dustiest place on earth. The broader Sahara and the nearby Sahel region also make their contributions: African dust is carried over much of the southern and eastern U.S. every summer and is responsible for 75 to 80 percent of the dust that falls over Florida. When it rains in Miami, and local residents clean a residue of reddish particles from their vehicles, they are wiping away a long-distance delivery from Africa. Walk across the islands of the Bahamas or the Florida Keys, and you will be hiking on African soil.

The earth emits an estimated two billion metric tons of dust a year, and more than half of it comes from African deserts and drylands. China emits dust that travels to Hawaii and western North America; Patagonia sends dust to Antarctica. Most of the dust that settles on Greenland comes from Asia, but when drought produced the American Dust Bowl of the 1930s, that dust also seems to have made its way to Greenland’s glaciers.

Much of Africa’s airborne dust takes a 6,400-kilometer ride on the westward trade winds across the Atlantic. By one estimate, roughly 40 million metric tons of dust, loaded with life-sustaining minerals, including iron and phosphorus, carpet the Amazon rain forest every year, and half of that amount may originate in the Bodele.

Before liftoff, the Bodele dust has been in a geologic waiting room. As each layer gets skimmed off, a new layer becomes exposed. The wind speed necessary to dislodge dust particles on the soil surface and to start them bouncing varies depending on surface and climatic conditions, but generally speaking, the threshold is in the range of four to 12 meters per second. As the particles start to jostle, they loosen others. The smallest ones float upward. Once airborne, the dust begins to mingle—first with other swirling particles from the Bodele, then with dust and pollution from elsewhere in Africa. Eventually it becomes part of a huge dust front moving across the Atlantic.

When I met with Prospero at his office at the University of Miami, he pulled up satellite photographs on his computer to show me this phenomenon and shook his head. “It’s sort of a mess,” he said, pointing at plumes of various colors and origins over Africa and the Atlantic. “It’s difficult to point your finger, in a quantitative sense, to what’s going on there. It all just gets mixed up. The whole of North Africa is blowing away all the time.”

Once in the air, dust that may have done nothing for millennia suddenly starts to modulate the earth’s climate. It absorbs radiation from the sun, including some that is reflected off the earth, warming the atmosphere. And it reflects other radiation back into space, which has a cooling effect. What proportion of radiation gets absorbed or reflected depends, in turn, on the chemical composition, mineralogy and size of the dust, as well as on the wavelength of the light. For the most part, dust has a propensity to reflect short-wave radiation from space and to absorb long-wave radiation coming off the earth’s surface. If the particles have mingled with soot, they will absorb even more heat.

Other factors also come into play. dust traveling over darker areas, like the oceans, cools the planet because it reflects some light that would otherwise be absorbed on the surface. Yet dust traveling over light-colored areas like ice and sand tends to have a warming effect because it usually absorbs more light than the surface. If dust falls on snow or ice, it leads to more warming. “Any aerosol, any dust, any dirt will darken snow,” says Charlie Zender, professor of earth system science at the University of California, Irvine. “If you walk through a snow field in the morning and put a little dirt on top of a small patch of snow, leave it there and come back in the afternoon, that part of snow will have sunken in.” Several scientists I spoke with believe the overall impact of atmospheric dust is probably a cooling of the planet, but not nearly enough to compensate now for the warming effect of greenhouse gases.

Airborne dust influences climate in indirect ways, too. It has a vital part, for instance, in cloud formation. Moisture in the air does not form into droplets on its own. It needs to attach to particles. Scientists disagree on the extent to which dust acts as “condensation nuclei.” Natalie Mahowald, a Cornell University professor who develops atmospheric models, firmly believes that both water and ice condense on dust. Paul Ginoux, who produces climate models at the Geophysical Fluid Dynamics Laboratory at noaa, agrees that dust acts as a condenser for ice but believes water will condense only on dust that has been mixed with sulfates, mainly from pollution.

On at least one point, Mahowald and Ginoux concur: there are tremendous gaps in our knowledge about cloud formation. When large numbers of tiny particles are suspended in the atmosphere, they can help form big concentrations of water droplets, but because those droplets are so small, they are less likely to fall as rain. Clouds of small droplets, moreover, are brighter than clouds of large droplets—so they scatter more radiation back into space. If dust particles absorb heat, however, the moisture they attract will evaporate faster. The clouds will not last as long. “Dust can make precipitation more likely or less likely, depending on what the rest of the atmosphere is doing,” Mahowald says. “It’s even more complicated than you might think.” Ginoux points out that even the best computer simulations do not give us a full picture: “We know the physical processes, but it’s difficult to evaluate what’s happening with any precision.”

It is hard to overstate the importance of clouds to the earth’s climate—and not just because they produce rain or snow. Roughly 60 percent of the earth’s surface is covered with clouds at any given time. Small changes to the formation and properties of clouds could dramatically alter the role they play in reflecting light and heat back into space. By one estimate, a 5 percent increase in “short-wave cloud forc­ing” would cool the earth enough to compensate for all the increases in greenhouse gases that occurred between the years 1750 and 2000.

Of course, dust has been swirling around the globe for all of its existence. So why should it have any greater or lesser effect now than it has had before? Mahowald argues that, over much of the planet, more dust is in motion now than at other time in recent history. “It looks like we had about a doubling of dust over much of the planet in the 20th century,” Mahowald says. “We don’t know exactly what caused the 20th-century increase, but human activity could be fueling the change.”

Joseph R. McConnell of the Desert Research Institute–Reno in Nevada has been working on precisely that question of cause and effect. To get answers, he analyzes the dust embedded in the ice of Greenland and Antarctica. He begins by taking ice cores, anywhere from 20 meters to three kilometers long, depending on how far back in time he wants to probe. Then he flies them to his lab. He has two $400,000 machines—high-resolution mass spectrometers—to measure the concentrations of elements found in the ice. These elements include aluminum and rare-earth elements such as cerium found in dust but not in sea salt, industrial pollution, or emissions from volcanoes and forest fires.

The machines work like this: glacial water from the ice cores is injected into a plasma that is as hot as the sun’s surface—about 6,000 kelvins. “This vaporizes almost everything, and we count the ionized atoms of each leftover element based on their atomic mass and electrical charge,” McConnell says. “It’s extremely sensitive. Some elemental concentrations are as low as parts per quadrillion. We’ve applied it to shallow ice cores covering the recent centuries and just now are applying it to deep ice cores spanning the last ice age.”

What McConnell is trying to measure is dust levels over time so that he can figure out what might have caused them to rise and fall. From his results it would seem that desertification and changes in land use in Patagonia (including the expansion of sheep farming in the early 20th century) correspond with a doubling in dust levels in Antarctica during that period. It might be tempting to argue for a simple process of cause and effect: overuse of land leads to desertification, which produces more dust, which then fuels climate change. McConnell warns, however, that “there are a lot of drivers of dust.”

Climate itself is one of those drivers, but its role is not entirely clear. Rising temperatures, by reducing soil moisture and fueling desertification, might contribute to increasing levels of dust, which could be just a short-term phenomenon. Over the long term, dusty periods correlate with cooling. McConnell sees evidence that Antarctica was less dusty, for instance, between the 10th and 13th centuries—an era of moderate warming and higher precipitation in the North Atlantic region—and more dusty between the 13th and 19th centuries, a period of modest cooling and lower precipitation. His study of central Greenland ice records showed an increasing trend in dust levels over three centuries until the 1930s, followed by a mysterious decline.

But our hypothetical particles tumbling and swirling out of Africa—part of the largest and most persistent migration of dust anywhere on the planet—do not just play a vital role in the atmosphere. They also act like an enormous spray of fertilizer over both the oceans and the land.

As they ride westward, many dust particles fall into the Atlantic. Here they perform a climate-regulating function that is different from what they do in the atmosphere but that also has a cooling effect: they provide iron, spurring the growth of phytoplankton, which consume carbon dioxide, die and take that carbon down to the deep, dark ocean depths. There the carbon remains isolated from the atmosphere for centuries.

The ocean contains nearly 85 percent of the carbon on the earth that is not held in rocks, and ocean phytoplankton are “responsible for ... a majority of all carbon sequestered over geologic time,” says a 2011 paper in Aeolian Research. Yet whereas large areas of the ocean have high concentrations of the nutrients nitrogen and phosphorus, they also have shortages of iron, limiting the amount of plankton that can bloom. That is where wind-borne dust comes in. African dust is high in iron content.

A few years ago there was so much excitement about the discovery of the important role of iron in the carbon cycle—and the indirect role of dust—that some scientists began to dream about ambitious geoengineering projects. The thinking went like this: in the large areas of the southern oceans and the northwestern Pacific called high-nutrient, low-chlorophyll zones, where plankton blooms are much reduced, we humans could just dump big loads of iron. Then plankton would bloom like crazy, consume carbon dioxide, die and sink to the ocean bottom. Good-bye, greenhouse gas problems.

It did not take long, though, to see the dangers in this approach. “There are many possible unintended consequences,” Prospero says. These include a drastic change in the current species distribution of microorganisms in the water column. That is not necessarily a bad thing, but the impact is unpredictable; new ecosystems often are not as diverse and productive as those they displace. Also, if iron is dumped in zones that are deficient in iron but rich in other nutrients, the new plankton plumes will draw down to the depths not only carbon dioxide but also phosphorus and nitrogen. Those nutrients will not then be available elsewhere in the oceans where they are needed.

Other new knowledge further undermined the iron solution. “There’s been a complete change in the way we see ocean biochemistry,” Cornell’s Mahowald says. “What we thought was going on 10 years ago is completely different from the way we see it now.” One of the bigger revelations is that “not all dust is equal in terms of the iron it makes available.” It turns out that acids in the atmosphere—from biomass burning and other pollution—interact with dust to make iron more soluble. So when we burn fuel and waste, we contribute to the production of available iron in the atmosphere and the oceans. “The amount of iron being deposited in the oceans may have already about doubled because of humans,” Mahowald says. “At the same time, sedimentary iron in the ocean is a much larger amount than previously thought. There is much more iron coming off the ocean shelves. So atmospheric iron is less important than we thought it was.”

For those particles that make it all the way across the Atlantic, the journey can take a week or more. It is common to see an African dust haze over Miami in the summer or to find a film of such particles on your vehicle after a rainstorm in the Amazon. That is how Swap of the University of Virginia got interested in the topic of dust transport back in the late 1980s. He was working in Brazil as a graduate student when he and others noticed that after days of rain, dust would continue collecting on their white Volkswagens. “We were 1,000 miles inland, where it would rain like hell, three to five inches a day,” Swap recalls. “We’d look at our cars after a rain and find red dust. And we’d think, ‘What’s going on here?’”

That question was linked to another that had long festered about the Amazon. The basin consists of old soils continually battered by rains that probably should have drained out many of the key nutrients long ago. So how was the Amazon getting replenished? How did it remain so fertile? Some think it may replenish itself as plant matter decomposes. Others think that is unlikely and wonder how it became so fertile to begin with. “It’s a very viable hypothesis that a lot of the fertility of the Amazon can be explained by the transport of African dust,” says Daniel Muhs, a scientist at the U.S. Geological Survey. “How else does the Amazon support that unbelievable diversity of plants and animal species on such a hot, humid and old landscape, where the soils are highly leached?”  

New studies have confirmed similar intercontinental dust deposits in other areas. Muhs took “geochemical fingerprints” of the soil on several islands in the Caribbean. “In some places, African dust is the sole source of the soils; in others, it’s a partial source,” he says. Some islands are made of limestone, coral reefs and sand, yet their topsoil is rich in unrelated clay and aluminum silicates. There are two possible sources, Muhs says: ash from a volcanically active part of the Caribbean or dust from Africa. In some places, including Barbados, the soils are composed of both. In others, like the Bahamas and the Florida Keys, it is almost all from Africa. “Our work on Barbados, with fossil reefs of different ages, indicates that the process [of African dust transport] has been going on for hundreds of thousands of years,” Muhs says.

How long will the process continue? Here is the last thing you need to know about our traveling dust particles: not only do they have a profound effect on the earth’s climate, but the earth’s climate can also have a profound effect on them. “Dust is different from other aerosols because dust in the atmosphere—unlike man-made pollution—is dependent on climate itself,” Prospero says. “If climate change affects wind velocity and rainfall, it can have an immense impact. Dust is extremely sensitive to small changes in wind and rain. It’s the ultimate feedback loop.”

Evidence of such relationships can be seen in ice core and other records. Glacial periods were much dustier than interglacial times. “But we’re still trying to figure out the chicken and the egg of that,” Muhs says. “Did glacial periods lead to more dust or more dust to glacial periods? There are a variety of feedbacks. It gets very complicated very quickly.” That is what makes scientific solutions to climate change—dreams of a simple, elegant feat of bioengineering like the iron solution—so troublesome. “With all the feedbacks within feedbacks within feedbacks, what unexpected feedbacks might we have?” Muhs says. “We might solve one problem while creating another.”

Prospero has already noted some unexpected weirdness going on. During the 1970s and 1980s dust concentrations at Barbados and Miami were highly correlated with drought and rainfall in North Africa: more drought, more dust. But all of that changed starting in the 1990s. “Now there is no correlation at all, and we don’t know what’s going on,” Prospero says. “I am concerned and confused.” He worries that dust might be yet another indicator that our complex earth systems could be getting out of whack, making predictions impossible and the future increasingly uncertain.