The Green Sahara Of The Past

Evidence from a whole range of sources shows that only a few thousand years ago, the climate of the whole Sahara region was very different from now. Animal bones in the desert sands show that giraffes and elephants once walked where there is now no vegetation and no water. The people who lived in the central Sahara at that time even recorded the animals they saw in rock paintings and engravings, vividly illustrating just how completely this place has changed in a few thousand years. A more detailed picture of the landscape at that time comes from pollen which has ended up preserved in the dried muds of old lake beds and empty river channels. It reveals a mosaic of scrublands, open woodlands and grasslands, consisting of plant species that now only grow hundreds of kilometers farther south. Even the extremely arid core of the Sahara, which nowadays gets less than 25 mm of rainfall a year, had a dense vegetation cover capable of sustaining cattle-herding and localized wheat-growing. All the evidence shows that the moistness of the Saharan climate at that time far exceeded the alternative "green Sahara" state of the present-day world. Perhaps we should honor the memory of this remarkable phase in climate history with the upper case "Green" Sahara, to distinguish it from the merely "green" Sahara.

The picture assembled from pollen, animal fossils and ancient lake sediments shows that the ancient Green Sahara phase lasted from about 9,000 to 6.000 years ago (Figure 5.6a, b*), followed by a halting decline in rainfall to reach essentially the present state of aridity by 4,000 years ago. So, when the first organized societies began in Egypt some 6,000 years ago the landscape beyond the edges of the Nile Valley may have been entirely different from now. When the great pyramids were being constructed 4,500 years ago, the landscape they were built in was probably not yet the bare sand that exists at present. Instead, there would have been small bushes and clumps of grass dotting the landscape, spaced perhaps a few feet apart from one another. So, the landscape the pyramid builders saw around them was rather different, perhaps offering more of a contrast with the yellow rock of the pyramids themselves. Some archaeologists have speculated that the trigger that first started the phase of monument-building by the ancient Egyptian civilization was the initial

* See also color section.

Mediterranean forest Mediterranean scrub

Montane forest

Montane forest

Sahara Holoc

Extreme desert Somi- desert Grasslands

Savanna (a few trees)


Woodland (open canopy)

Tropical rainforest

Figure 5.6. The distribution of vegetation /ones of (a) the present-day and (b) the Holocene "Green Sahara" (8,000-7,000 ,4C years ago). Grasslands (mixed with scrub) seem to have covered the whole Sahara desert at that time. Source: Author.

Mediterranean forest Mediterranean scrub

Tropical Sahara

Figure 5.6. The distribution of vegetation /ones of (a) the present-day and (b) the Holocene "Green Sahara" (8,000-7,000 ,4C years ago). Grasslands (mixed with scrub) seem to have covered the whole Sahara desert at that time. Source: Author.

Extreme desert Semi-desert


Woodland (open canopy

I Tropical rainforest

Extreme desert Semi-desert


Savanna (a few trees)


Woodland (open canopy

I Tropical rainforest

Mediterranean forest Mediterranean scrub drying out of climate that concentrated diverse peoples and talents into a narrow strip along the Nile Valley.

Why was the Sahara once so moist, and why did it dry out? Bearing in mind the modeling evidence showing that the present-day desert climate is very sensitive to changes in conditions, climate-vcgctation scientists took on the challenge of modeling the past of the Sahara. From the models, it seems that there were several factors at work in producing this moist phase. The major one. giving most of the difference in rainfall, is the amount of sunlight the region gets in summer and does not itself depend on vegetation cover. Because of an asymmetry in the earth's orbit, there is a regular cycle of about 21.000 years in the amount of solar energy the northern hemisphere gets in summer and in winter. So, every 21,000 years there is a peak of summer input of radiation over North Africa; the sun is 7% stronger at this time than during summers at the opposite point in the cycle. Climate models show that this increased summer radiation is enough to alter the monsoon How in the northern hemisphere. With more heating at the surface, air rises farther up into the atmosphere. and this strong convection pulls in air from the south that has picked up moisture over the tropical ocean. The moist air moving in northwards hits the ascending air from the land surface then rises and cools, generating rain. In the world as it was between 9.000 and 6.000 years ago, the monsoon rain came much farther north than it does now. because of this difference in summer sunlight.

Although the increased summer sunlight alone can explain a large part of the increase in rainfall during the Green Sahara phase, it cannot explain all of it. The combined picture from the flora and fauna of this time is that the Saharan climate was even wetter than the models can account for by the increased summer sunlight alone. Apparently then, something is missing from the calculations. Faced with the discrepancy between models and reality, the modelers added in another component. This is the vegetation itself, and the feedbacks it exerts on the monsoon rainfall that it also depends upon. The model builders took a vegetation distribution corresponding to what the fossil record suggests prevailed during the Green Sahara, and to their model they added the lower albedo, the greater roughness and the transpiration of water from the leaves of all this vegetation, flic result was an even moistcr climate: in the model the vegetation helps to set off convection in the atmosphere, pulling in the monsoon more strongly. Moreover, the vegetation across the Sahara recycles the rain that falls, allowing the monsoon to keep going strong as it travels farther up through North Africa. The model forecasts enough rainfall to sustain the abundant vegetation that we know prevailed at that time; so, the loose ends are tied up to make a loop. The vegetation made the climate moister; and bccause the climate was moister, that specific type of vegetation could live there. What the vegetation did was intensify a moist climate that would have existed to some extent anyway, because of the basic underpinning of increased summer sunlight.

So, the feedback loop that made the climate "optimum" in the Sahara went something like this:

Greater summer sunlight -> moister climate -> more vegetation

=> moister climate more vegetation

But. with this powerful feedback loop maintaining the moist climate, why did the Green Sahara end? The modern, weaker "green" (note the lower case) Sahara would be unstable, liable to end at any time given a slight push from the weather. But the models show that the ancient "Green" Sahara was far more stable, held in place by the stronger summer sunlight of the time. The only reason it ended was that the summer sunlight over the Sahara declined, to the point where the monsoon rains flickered and then died.

The final phase of drying of the ancient Sahara around 4.000 years ago apparently took only a few hundred years, much less than the sorts of timescales (thousands and millions of years) that geologists have got used to thinking of for climate changes. This rapid loss of rainfall is indicated by various forms of evidence that can be precisely dated from lake muds and other sediments. Despite any temporary reversals that may have happened, the overall shift from a lush green landscape to bare sand and rock was completed in at most a few centuries. Yet, during this time there was a slow gradual decline in summer sunlight, taking several thousand years. Since summer sunlight is really the underpinning cause of the Green Sahara, one might cxpect a similar gentle change in the climate of the Sahara during this period, and yet in fact it flipped relatively suddenly. Why?

The rapid end to the Green Sahara can only be explained by the way in which the vegetation system responds with its positive feedbacks. Sometimes positive feedbacks can help to stabilize a certain state, and this is what they did during a couple of thousand years (between about 7,000 and 5,000 years ago) when the summer sunlight was stronger, even though the sunlight was declining. But positive feedbacks also tend to reach a sudden breaking point, beyond which they push things in completely the opposite direction. Instead of slamming rainfall up against the top of the scale, they slammed it down against the bottom of the scale. The Sahara reached a point where the vegetation cover could no longer maintain the monsoon rains, even with its darker surface, its roughness and its abundant evaporative leaf area. The sunlight intensity could not quite ensure enough atmospheric upwelling, or enough evaporation, and the rains began to fail. Once they began to fail, the vegetation suffered and died back. And the more it died back, the more the rains failed until within a fewr centuries there was almost no vegetation and almost no rain (Figure 5.7).

A more detailed look at the environmental record of this critical time shows that actually the change from the Green Sahara to the brown Sahara was not a simple one-time flip. Some indicators from inland lakes and from the amount of river water coming down the Nile suggest that changes occurred in an even more sudden, chaotic manner with sudden flips and then reversals in climate each taking only a few dccadcs. It looks like the monsoon rains flickered on and off like a failing striplight; they turned off for a few decades or a century, then back on for a few decades more, and so on, before they eventually failed completely. Some of the vegetation climate models also seem to support this detail of the picture; they predict various metastable states during the transition period that would have flipped amongst themselves rapidly, before eventually settling down into one stable barren and dry state when the summer sunlight had declined sufficiently.





- CO

3 O



c -< «-* _





to <



-9000 -8000 -7000 -6000 -5000 -4000 -300-0 -2000 -1000 0


Figure 5.7. In the Sahara, during the last 9,000 years, the summer solar energy input changes slowly (a) hut because of vegetation feedback the rainfall (c) and especially the vegetation cover (d) changes much faster, flipping from quite dense vegetation to virtually no vegetation. After Bonan.



So far, relatively little modeling has been done on other grassland and desert regions of the world, but the suspicion must be that some of these also show instability in climate that is amplified by vegetation. Apart from the Sahara and Arabia, there are certainly some regions that have a history of large, repeated changes in climate over the past 10,000 years or so. One example is semi-arid northwestern China, which shows great instability in climate on the timescale of millennia. At various times, the climate on the fringes of the Mongolian desert became much moister, moist enough to grow an abundance of crops in areas that arc now too dry to cultivate. Farming communities thrived in areas now barely inhabited and mostly devoid of vegetation. Fossil pollen and wood from wild plants such as trees, and changes in the iron oxide chemistry of the soils, confirms that the climate was much wetter at these times. Using a combination of different indicators, geologists have put together a general history of climate change in the region:

Table 5.2. Climate history of northwestern China over the last 10,000 years.

ti y j

From: Petit-Maire and Guo.

Drier than present 9,900 9,400 yr

Moist 9,400-7,900 yr

Drier than present 7,900 6,500yr

Moist 6,500 4,900 yr

Drier than present 4,900 yr

Moist 3,200 yr

Relatively dry phase (but still moister than present) 3,200-2,800yr

About the same as at present (fairly arid) Since 2,800 yr

So, it seems that there were several separate moist phases when trees and crop-growing spread out across northwest China, the main times being between about 9,400 and 7,900 years ago. and between 6.500 and 4.900 years ago. Although these broadly fall within the same phase as the moist Sahara, when summer sunlight was at its greatest, in China there are some striking fluctuations in the climate that do not occur in the Green Sahara. At these times, the climate in western China switched from much moister than present to drier than present, before later switching back again.

From what we know from attempts at modeling the Green Sahara, it seems reasonable that the moist phases in northwestern China might have been accentuated by vegetation feedbacks. When the climate models arc applied to this region they do indeed show that the greener landscapes would have helped to pull in more rainfall. However, while these same models produce stable moist conditions during the phase between 9,000 and 6,000 years ago in the Sahara (and instability only later, as the summer sunlight decreased), they do not predict the instability of rainfall in northwestern China. The climate instability seen in the environmental record of north western China is suggestive of a system where climate-vegetation feedbacks are playing an important role, but the current models at least do not show it. It may be that a future generation of more sophisticated models will show up hidden feedbacks that arc occurring with vegetation.

So far then, modeling attempts in western China have not revealed any sign of multiple steady states, but the closely alternating history of arid and moist phases suggests that such instability must also exist there. Whatever the underlying cause, it is likely that part of this variability is caused by amplification of background climate changes by vegetation.

Continue reading here: Dust

Was this article helpful?

+4 0


  • tytti laukka
    What was the Egyptian climate 4,000 years ago ?
    4 months ago
  • martha
    When Sahara had tropical forests?
    2 years ago
  • ronja
    What animal where in egypt when it was a rainforest?
    3 years ago
  • Jan-Erik
    When was Sahara a rainforest?
    3 years ago