Exit From Eden
Six thousand years ago, the Sahara was a fertile savanna teeming with animals and people. How did it become a barren sea of sand - and when will it change back again?
DISCOVER Vol. 21 No. 01 | January 2000


Every year, twice a year if he's lucky, Hans-Joachim Pachur leaves his office in a pleasant villa opposite the Berlin botanical garden, leaves the exuberant plant life that shades the streets and pushes through fences and sidewalk cracks, and catches a plane for, say, Libya. In Tripoli he meets up with several colleagues and gets ready for the desert. A geographer at the Free University of Berlin, Pachur has been exploring the Sahara for the last quarter century, and he has been known to do it on camelback. These days, though, he mostly chooses Land Rovers. Supplies gathered and gear checked, the team drives south, away from the Mediterranean; they head out onto the sea of sand, where the waves crest at 600 feet.

To cross those dunes, Pachur and his group let most of the air out of their tires for better traction. When a dune is very steep - the slope can reach 36 degrees - they unstrap metal planks from the sides of the Rovers. They place a pair of these in front of the vehicle and drive onto them; then they pick up another pair from behind the vehicle and place them in front. This can go on for a quarter of a mile. "You have to have considerable experience to drive in this region and not come to grief," says Pachur. You also have to be willing to give up amenities such as beds, decent food, and showers - especially showers. Pachur and his group have to carry all the water they'll need for their several-week expedition, because they don't expect to find a drop.

But there once was water here - lots of it. Last March, in the Murzuq Sand Sea of southwestern Libya, Pachur found bones of crocodiles, hippopotamuses, elephants, and gazelles as well as windblown ridges of lake-bed chalk - evidence that the region had been dotted with bodies of freshwater. Years ago in the northern Sudan, Pachur found traces of a lake that may have been as large as Erie. In that same region he traced the course of a river that once flowed east into the upper Nile, crossing several hundred miles of what is now utter desert. He has found other rivers that flowed from the Tibesti Mountains 600 miles north to the Mediterranean Sea, through the core of the Sahara. That region now gets less than two-tenths of an inch of rain each year.

Along those lost rivers, between 6,000 and 9,000 years ago, giraffes munched on acacia trees, elephants sprayed water from their trunks, hippos wallowed in mud. And people lived there too. They were shepherds and cowherds, hunters and fishers, and they were starting to settle down in small villages and cultivate grains such as sorghum and millet. Pachur believes the Sahara then was an Eden. "People did not experience this region the way we do, as a hostile environment," he says. "For them, it provided enormous possibilities to blossom." On rock walls west of the Murzuq, or in the Gilf Kebir highlands of southwestern Egypt, the Saharans carved and painted scenes from their lives. They depicted themselves driving cattle, hunting, and swimming, or sometimes just sitting around drinking.

But then the climate began to change, and the desert came. It began sometime after 6,000 years ago. Within just a few centuries, a gentle, fertile region the size of the United States was transformed into one of the harshest, most barren places on Earth. The Saharans had to leave. Many must have migrated east into the valley of the Nile, their closest source of water. That exodus, some archaeologists think, may be the event that triggered the rise of the pharaohs in Egypt a little more than 5,000 years ago. Eden gave way to one of the planet's first great civilizations.

Miracle or catastrophe, the birth of the Sahara was not an act of God. At the Potsdam Institute for Climate Impact Research, just outside Berlin, a theoretical climatologist named Martin Claussen and his graduate student Claudia Kubatzki have recently been able to re-create the desert-building process on a computer. They say that ultimately what keeps rain from falling on the Sahara is a lack of plants - which is not as paradoxical as it sounds. It's even possible, they say, that in a century or two the desert might recapture its salad days.

Claussen, who uses Pachur's data to anchor his model to reality, has never been to the Sahara, never risked his neck trying to cross giant dunes in a four-wheel-drive vehicle. But then he has never been to Jupiter either, and in a way that's the starting point for any attempt to explain the Sahara. Jupiter's gravity and Venus's gravity pulls on Earth, causing our planet to tilt on its axis to varying degrees. The tilt is what gives Earth seasons: Summer happens in the northern hemisphere when it is tilted toward the sun. The sun's zenith during this period reaches as far north as the tropic of Cancer, which at present runs through the Gilf Kebir and the southern edge of the Murzuq. That latitude, 23.5 degrees, is the present tilt angle. But over the course of a 41,000-year cycle, the tilt and the tropic get as high as 24.5 and as low as 22.1 degrees.

While Jupiter and Venus are tilting the Earth's axis, the sun and moon are causing it to wobble like a top; that changes the time of perihelion, the point on Earth's elliptical orbit at which the planet comes closest to the sun. Together these two cycles - along with even slower perturbations in the shape of the orbit - determine how much sunlight falls on a given latitude at a given season. They are called Milankovitch cycles, after the Serbian mathematician Milutin Milankovitch, who proposed in the 1930s that such regularly occurring variations could explain why Earth's climate goes in and out of ice ages. According to this theory, the most recent ice sheets began retreating from Canada and Eurasia around 17,000 years ago because that's when the northern hemisphere started getting enough sun during the summer to melt them. About 9,000 years ago, the northern sunlight curve was reaching a peak; climatologists call this period the Holocene Optimum. The axial tilt was higher than it is now, around 24 degrees, and perihelion was in July. Both factors made for especially hot northern summers and - though it may seem surprising - for a greener Sahara.

Throughout the ice age, the Sahara was a desert. But as the summers got hotter, they also got wetter. During the summer, the African landmass heats up more than the adjacent Atlantic, and that temperature contrast drives the monsoon: as hot air rises up and away over the land, moist winds flow northeast from the Gulf of Guinea to replace it. Hotter summers mean a stronger monsoon that brings more rainfall farther north, into the Sahara.

Until now, that has been the whole explanation for why the desert once bloomed - but it just doesn't work, says Claussen. "When you try to reproduce that process in a computer-model atmosphere, the results are disappointing," he says. "You end up with a Sahara desert more or less where it is today." A bit more rain does fall on the desert, but not nearly enough to support abundant vegetation or to create lakes. A few years ago, Claussen had a hunch: Maybe the Sahara wasn't turning green in the models simply because the models, unlike reality, weren't letting it.

Every gardener knows, he says, that climate zones dictate what plants can grow where. But what about the reverse process - might vegetation have an influence on climate? At the small scale the answer is obviously yes: A forest is a cooler, damper place than open rangeland. But what about at the global scale? Climate models have tended to ignore this possibility. The ones that tried to calculate how much rain would have fallen on the Sahara during the hotter summers of the Holocene never actually allowed the model desert to turn green. They left it empty sand. "If you sow a desert, then naturally you'll reap a desert," says Claussen.

There are at least two ways in which vegetation itself might procure the rainfall it needs. First, ground covered by plants is darker than a desert; if you look at satellite images, the Sahara is the brightest feature on Earth after the polar ice caps. It reflects away as much as 40 percent of the sunlight it receives and sends most of the rest right back into space, through cloudless skies, as infrared radiation. Human bodies get fried in the Sahara, but as far as the atmosphere is concerned, the desert is a net cold source - and so most of the time air flows into the region at high altitude and sinks, forming a dry region of high pressure. If there were enough plants there, the situation would be very different: Darkened ground would absorb more sunlight, the atmosphere above the plants would be heated, and the warm air would rise - which is a prerequisite for clouds and rain.

The other prerequisite, of course, is water. Plant-covered soil is invariably wetter than desert sand. When moist soil absorbs sunlight, the water evaporates. In the atmosphere it condenses again, releasing energy that fuels updrafts, and thereby forms clouds. The clouds rain, returning water to the soil. The updrafts strengthen the monsoon, pulling in more moisture. "It's a positive, self-reinforcing feedback loop," says Claussen. "As the vegetation gets thicker, the ground gets darker and wetter, which leads to more precipitation, which causes the vegetation to get thicker again."

With his colleagues at Potsdam, Claussen developed a model that incorporated this positive feedback. They turned its clock back to 9,000 years ago, and started it running with the sunlight distribution prescribed for that date by Milankovitch, and lo! they got a Sahara savanna. There was enough rain to fill lakes and rivers, enough grass and trees to feed Pachur's elephants and giraffes. Claussen let the model keep running.The positive feedback works as well in reverse. By 6,000 years ago, Earth's axis was tilting less than it had been, heading toward its current angle, and perihelion was heading out of northern summer and toward winter. Northern summers were getting cooler, the African monsoon a bit weaker, and the vegetation in the savanna sparser. All this was happening very gradually; the orbital cycles are subtle. But at some point - in Claussen's model it was around 5,500 years ago - the system crossed a threshold: the feedback kicked in. Enough vegetation had been lost from the Sahara, enough bare ground exposed, that precipitation declined sharply and barrenness began to spread like wildfire. "Suddenly things went downhill very rapidly," says Claussen. Within a few centuries the cool moist soil had become sand.

Over the years some archaeologists have suggested that the people who lived there might have helped create the Sahara by cutting down the trees. Claussen is skeptical. "We don't really need people," he says. "We've shown that this can be described as a natural phenomenon." The people, more likely, were just caught in a vast and weird clockwork, one that connected Jupiter to their acacia trees, with Earth's tilted axis, the summer monsoon, and the brightness of desert sand all acting as intermediate cogs. The wheels turned and took the Saharans' landscape away from them, and it all happened very fast, in a time span well within the limits of cultural memory.

In the last couple of decades archaeologists have done a fair amount of digging in the Sahara, especially in Egypt. At Nabta, 60 miles west of the Nile near the Sudanese border, Fred Wendorf of Southern Methodist University, Romuald Schild of the Polish Academy of Sciences, and Angela E. Close of the University of Washington in Seattle have found the remains of a small settlement established along a seasonal lake between 5,000 and 6,000 years ago. The people there engaged in some odd activities: ceremonial burials of cows, for instance. They also hauled multiton slabs of sandstone, some of them 10 feet tall, from hundreds of yards away and stood them on end in the lake mud. The standing stones may have served to point out north or to "acknowledge the zenith Sun near the onset of the rainy season," as Wendorf and Schild have recently proposed. In any case, the backbreaking labor required to erect them suggests a social hierarchy. "Nobody's going to do that to fill up a quiet Sunday afternoon," says Close. "Somebody has to be sitting there at the top saying, 'You. . .will. . .do. . .this!'"

Farther north and around 120 miles west of the Nile, near Farafra Oasis, a team led by Barbara Barich of the University of Rome has been excavating another lakeshore settlement from the Sahara's greener days. The people there kept sheep, goats, probably cattle, and perhaps ostriches - Barich has found an inordinate number of eggshells. They built houses with stone foundations and hearths, and they were starting to cultivate the sorghum and millet that grew wild along the lake. These rudiments of agriculture put them well ahead of people then living in the Nile Valley. "There they were still chasing gazelles up and down," says Close. And perhaps fishing too: During the Holocene Optimum, when even the Sahara was wet, the Nile may have been too wild and too prone to flooding to permit much else in the valley.

Close and Barich both believe that Saharans fleeing the desert may have brought the beginnings of agriculture and of organized, hierarchical society - in short, of a post-Stone Age way of life - to the Nile Valley. At Farafra, Barich has found flint knives and other tools carved in a distinctive style that shows up later along the Nile, implying a transmission of know-how. "We think the western areas were the place from which the idea of cultivation came to the Nile," she says. "It was something that was completely different from the tradition there." The first pharaohs, meanwhile, came to power only a few centuries after people had to leave Nabta. Their pyramids can be seen as more elaborate expressions of an idea already contained in the Nabta megaliths, namely, as Close puts it, "that some people are more equal than others."Today people have returned in large numbers to Farafra and to other oases in Egypt and Libya. They are living there off the climatic past, with wells that descend over a thousand feet and tap groundwater that was last replenished when the Sahara was green. Tripoli and other cities on Libya's Mediterranean coast get water from oases that lie as much as 250 miles south in the desert. One of the giant pipelines tracks the path of one of Pachur's paleo-rivers. At some oases the fossil groundwater is used to irrigate vast circular fields. In the area around Farafra they grow wheat now instead of sorghum.

Those fields are not enough vegetation to call in the monsoon, in spite of the feedback that Claussen has identified - but that is not to say that such a project is beyond human powers. If we followed the Milankovitch clock, we would need to pass through another ice age before the Sahara turned green again, but we're not sticking to that timetable. Global warming caused by carbon dioxide and other greenhouse gases could, a century or two from now, provide the nudge to the monsoon that the tilt and wobble of Earth's axis did in the middle Holocene. Claussen's model suggests it's possible, though he doesn't much like to talk about it: He's afraid people will get the wrong idea. The real message of his simulations, he says, is that feedbacks in Earth's climate have in the past produced abrupt climate changes that disrupted human societies tremendously - and we're not yet capable of predicting which changes lie in our future.

Still, the notion that at least one of the unintended consequences of our fossil-fuel habit might be the greening of the Sahara is cheerful to contemplate. Pachur takes the possibility seriously. Though he likes the desert well enough to struggle through it every year, one gets the sense he would quite like to see it become a savanna once again, like the ancient one whose traces he uncovers. "There would be one essential difference, though," he says. "Those vast plains would no longer be populated by animals."

In the past, when the desert came, the giraffes, elephants, lions, and other large animals retreated toward the Mediterranean and into mountains like the Tibesti or the Atlas ranges; when the Sahara went green again, the animals could spread down onto the plains following the rivers and lakes. Those relict populations were still around in historical times - the Greeks and Romans met them - but in the past couple of millennia, human hunters have eradicated them all. If the Sahara ever becomes Eden again, it won't be the Eden of the Holocene Optimum. It will be Eden after the fall.