Green Sahara

The African humid period (AHP) (also known by other names) is a climate period in Africa during the late Pleistocene and Holocene geologic epochs, when northern Africa was wetter than today. The covering of much of the Sahara desert by grasses, trees and lakes was caused by changes in Earth's orbit around the Sun; changes in vegetation and dust in the Sahara which strengthened the African monsoon; and increased greenhouse gases.

During the preceding Last Glacial Maximum, the Sahara contained extensive dune fields and was mostly uninhabited. It was much larger than today, but its lakes and rivers such as Lake Victoria and the White Nile were either dry or at low levels. The humid period began about 14,600–14,500 years ago at the end of Heinrich event 1, simultaneously to the Bølling–Allerød warming. Rivers and lakes such as Lake Chad formed or expanded, glaciers grew on Mount Kilimanjaro and the Sahara retreated. Two major dry fluctuations occurred; during the Younger Dryas and the short 8.2 kiloyear event. The African humid period ended 6,000–5,000 years ago during the Piora Oscillation cold period. While some evidence points to an end 5,500 years ago, in the Sahel, Arabia and East Africa, the period appears to have taken place in several steps such as the 4.2-kiloyear event.

The AHP led to a widespread settlement of the Sahara and the Arabian Deserts, and had a profound effect on African cultures, such as the birth of the Ancient Egyptian civilization. People in the Sahara lived as hunter-gatherers and domesticated cattle, goats and sheep. They left archaeological sites and artifacts such as one of the oldest ships in the world, and rock paintings such as those in the Cave of Swimmers and in the Acacus Mountains. Earlier humid periods in Africa were postulated after the discovery of these rock paintings in now-inhospitable parts of the Sahara. When the period ended, humans gradually abandoned the desert in favour of regions with more secure water supplies, such as the Nile Valley and Mesopotamia, where they gave rise to early complex societies.

Research history

Herodotus in 440 BC and Strabo in 23 AD discussed the existence of a greener Sahara, although their reports were at first questioned owing to their anecdotal nature. In 1850 the researcher Heinrich Barth discussed the possibility of past climate change leading to increased wetness in the Sahara after discovering petroglyphs in the Murzuq Desert, and further discoveries of petroglyphs led desert explorer László Almásy to coin the concept of a '' Green Sahara'' in the 1930s. Later in the 20th century, conclusive evidence of a past greener Sahara, the existence of lakes and higher Nile flow levels was increasingly reported and it was recognized that the Holocene featured a humid period in the Sahara.

The idea that changes in Earth's orbit around the Sun influence the strength of the monsoons was already advanced in 1921, and while the original description was partly inaccurate, later widespread evidence for such orbital controls on climate was found. At first it was believed that humid periods in Africa correlate with glacial stages (" pluvial hypothesis") before radiocarbon dating became widespread.

The development and existence of the African humid period has been investigated with archaeology, climate modelling and paleoproxies, with archaeological sites, dunes and deposits left by lakes, aeolian deposits and leaf wax in the sea and wetlands playing an important role. Pollen, lake deposits and former levels of lakes have been used to study the ecosystems of the African humid period, and charcoal and leaf impressions have been used to identify vegetation changes. The time 6,000 years ago has received particular attention, especially since that period of the AHP has been used as an experiment in the Paleoclimate Modelling Intercomparison Project.

Research issues

While the precipitation changes since the last glacial cycle are well established, the magnitude and timing of the changes are unclear. Depending on how and where measurements and reconstructions are made, different beginning dates, ending dates, durations and precipitation levels have been determined for the African humid period. The amounts of precipitation reconstructed from paleoclimate records and simulated by climate modelling are often inconsistent with each other; in general, the simulation of the Green Sahara is considered a problem for earth system models. Erosion of lake sediments and carbon reservoir effects make it difficult to date when they dried up. Vegetation changes by themselves do not necessarily indicate precipitation changes, as changes in seasonality, plant species composition and changes in land use also play a role in vegetation changes. Isotope ratios such as the hydrogen/deuterium ratio that have been used to reconstruct past precipitation values likewise are under the influence of various physical effects, which complicates their interpretation.

Terminology

Earlier humid periods are sometimes known as "African humid periods" and a number of dry/wet periods have been defined for the Central Africa region. In general, these types of climate fluctuations between wetter and drier periods are known as " pluvials" and " interpluvials", respectively. Because the AHP did not affect all of Africa, Williams ''et al.'' 2019 recommended that the term be dropped.

Other terms that have been applied to the Holocene AHP or correlative climate phases are "Holocene humid period", which also covers an analogous episode in Arabia and Asia; "early to mid-Holocene humid episode"; "Holocene Pluvial"; "Holocene Wet Phase"; "" in Central Africa; "Makalian" for the Neolithic period of northern Sudan; "Nabtian Wet Phase" or "Nabtian period" for the 14,000–6,000 humid period over the Eastern Mediterranean and Levant; "Neolithic pluvial"; "Neolithic Subpluvial"; "Neolithic wet phase"; "" of the Western Sahara 6,500 – 4,000 years before present; "Subpluvial II" and "" in the Central Sahara 14,000 – 7,500 years before present. The terms "" and have been applied to the dry period in the last glacial maximum, the latter is equivalent to the "Kanemian"; "Kanemian dry period" refers to a dry period between 20,000 and 13,000 years before present in the Lake Chad area.

Background and beginning

The African humid period took place in the late Pleistocene and early-middle Holocene, and saw increased precipitation in Northern and Western Africa due to a northward migration of the tropical rainbelt. The AHP is the most profound climate change of the low latitudes during the last 100,000 years and stands out within the otherwise relatively climatically stable Holocene. It is part of the so-called Holocene climatic optimum, during which summers in the Northern Hemisphere were warmer than today. Liu ''et al.'' 2017 subdivided the humid period into an "AHP I" which lasted until 8,000 years ago, and an "AHP II" from 8,000 years onward, with the former being wetter than the latter.

The African humid period was not the first such phase; evidence for about 230 older such "green Sahara"/wet periods exist going back perhaps to the first appearance of the Sahara 7–8 million years ago, for example during Marine Isotope Stage 5 a and c. Earlier humid periods appear to have been more intense than the AHP of the Holocene, including the exceptionally intense Eemian humid period which provided the pathways for early humans to cross Arabia and Northern Africa and which, together with later moist periods, has been linked to expansions of the Aterian populations. Such humid periods are usually associated with interglacials, while glacial stages correlate to dry periods.

The Bølling-Allerød warming appears to be synchronous with the onset of the African humid period as well as to increased humidity in Arabia. Later, in the Blytt–Sernander sequence the humid period coincides with the Atlantic period.

Conditions before the African humid period

During the Last Glacial Maximum, the Sahara and Sahel had been extremely dry with less precipitation than today as reflected by the extent of dune sheets and water levels in closed lakes. The Sahara was much larger, extending farther south, a difference of 5° latitude. Dunes were active much closer to the equator, and rainforests had retreated in favour of afromontane and savannah landscapes as temperatures, rainfall, and humidity decreased.

There is little and often equivocal evidence of human activity in the Sahara or Arabia at that time, reflecting its drier nature. The aridity during the Last Glacial Maximum appears to have been the consequence of the colder climate and larger polar ice sheets, which squeezed the monsoon belt to the equator and weakened the West African Monsoon. The atmospheric water cycle and the Walker and Hadley circulations were weaker as well. Exceptional dry phases are linked to Heinrich events when there are a large number of icebergs in the North Atlantic; the discharge of large amounts of such icebergs between 11,500 and 21,000 years before present coincided with droughts in the subtropics.

Before the onset of the AHP, it is thought that the Lake Victoria, Albert, Edward, Turkana and the Sudd swamps had dried out. The White Nile had become a seasonal river whose course along with that of the main Nile may have been dammed by dunes. The Nile Delta was partially dry, with sandy plains extending between ephemeral channels and exposed seafloor, and it became a source of sand for ergs farther east. Other lakes across Africa, such as Lake Chad and Lake Tanganyika, also had shrunk during this time, and both the Niger River and Senegal River were stunted.

Early humidity increases

Whether some parts of the desert such as highlands like the Red Sea Hills were reached by the westerlies or weather systems associated with the subtropical jet stream—and thus received precipitation—is contentious. It is only clearly supported for the Maghreb in northwestern Africa, though river flow/terrace formation and lake development in the Tibesti and Jebel Marra mountains and residual Nile flow may be explained in this way. The highlands of Africa appear to have been less affected by drought during the last glacial maximum.

The end of the glacial drought occurred between 17,000 and 11,000 years ago, with an earlier beginning noted in the Saharan mountains (possibly) 18,500 years ago. In southern and central Africa earlier starts 17,000 and 17,500 years ago, respectively, may be linked to Antarctic warming, while Lake Malawi appears to have been low until about 10,000 years ago.

High lake levels occurred in the Jebel Marra and Tibesti Mountains between 15,000 and 14,000 years ago and the youngest stage of glaciation in the High Atlas mountains took place at the same time as the early African humid period. Around 14,500 years ago, lakes started to appear in the arid areas.

Onset

The humid period began about 15,000-14,500 years ago. The onset of the humid period took place almost simultaneously over all of Northern and Tropical Africa, with impacts as far as Santo Antão on Cape Verde. In Arabia, wet conditions apparently took about two millennia to advance northward, a gradual advance is supported by tephrochronological data.

Lake Victoria reappeared and overflowed; Lake Albert also overflowed into the White Nile 15,000–14,500 years ago and so did Lake Tana, into the Blue Nile. The White Nile flooded part of its valley and reconnected to the main Nile. In Egypt widespread flooding by the "Wild Nile" took place; this "Wild Nile" period led to the largest recorded floods on this river, sedimentation in floodplains, and probably also impacted human populations along the river. Even earlier, 17,000–16,800 years ago, meltwater from glaciers in Ethiopia – which were retreating at that time – may have begun to increase the flow of water and sediment in the Nile. In the East African Rift water levels in lakes began to rise by about 15,500/15,000-12,000 years ago; Lake Kivu began overflowing into Lake Tanganyika by about 10,500 years ago.

About the same time that the AHP started, the cold glacial climate in Europe associated with Heinrich event 1 ended with climate changing as far as Australasia. A warming and retreat of sea ice around Antarctica coincides with the start of the African humid period, although the Antarctic Cold Reversal also falls into this time and may relate to a drought interval recorded in the Gulf of Guinea.

Causes

The African humid period was caused by a stronger West African Monsoon directed by changes in solar irradiance and in albedo feedbacks. These led to increased moisture import from both the equatorial Atlantic into West Africa, as well as from the North Atlantic and the Mediterranean Sea towards the Mediterranean coasts of Africa. There were complex interactions with the atmospheric circulation of the extratropics and between moisture coming from the Atlantic Ocean and the Indian Ocean, and an increased overlap between the areas wetted by the monsoon and those wetted by extratropical cyclones.

Climate models indicate that changes from a dry to a green Sahara and back have threshold behaviour, with the change occurring once a certain level of insolation is exceeded; likewise, a gradual drop of insolation often leads to a sudden transition back to a dry Sahara. This is due to various feedback processes which are at work, and in climate models there is often more than one stable climate-vegetation state. Sea surface temperature and greenhouse gas changes synchronized the beginning of the AHP across Africa.

Orbital changes

The African humid period has been explained by increased insolation during Northern Hemisphere summer. Due to precession, the season at which Earth passes closest to the Sun on its elliptical orbit – the perihelion – changes, with maximum summer insolation occurring when this happens during Northern Hemisphere summer. Between 11,000 and 10,000 years ago, Earth passed through the perihelion at the time of summer solstice, increasing the amount of solar radiation by about 8%, resulting in the African monsoon becoming both stronger and reaching farther north. Between 15,000 and 5,000 years ago, summer insolation was at least 4% higher than today. The obliquity also decreased during the Holocene but the effect of obliquity changes on the climate is focused on the high latitudes and its influence on the monsoon is unclear.

During summer, solar heating is stronger over the North African land than over the ocean, forming a low pressure area that draws moist air and precipitation in from the Atlantic Ocean. This effect was strengthened by the increased summer insolation, leading to a stronger monsoon that also reached farther north. The effects of these circulation changes reached as far as the subtropics.

Obliquity and precession are responsible for two of the foremost Milankovich cycles and are responsible not only for the onset and cessation of ice ages but also for monsoon strength variations. Southern Hemisphere monsoons are expected to have the opposite response of Northern Hemisphere monsoons to precession, as the insolation changes are reversed; this observation is borne out by data from South America. The precession change increased seasonality in the Northern Hemisphere while decreasing it in the Southern Hemisphere.

Albedo feedbacks

According to climate modelling, orbital changes by themselves cannot increase precipitation over Africa enough to explain the formation of the large desert lakes such as Lake Megachad or the northward expansion of vegetation unless ocean and land surface changes are factored in.

Decreasing albedo resulting from vegetation changes is an important factor in the precipitation increase. Specifically, increased precipitation increases the amount of vegetation; vegetation absorbs more sunlight and thus more energy is available for the monsoon. In addition, evapotranspiration from vegetation adds more moisture, although this effect is less pronounced than the albedo effect. Heat fluxes in the soil and evaporation are also altered by the vegetation.

Reduced dust generation from a wetter Sahara influences the climate by reducing the amount of light absorbed by dust and also modifying cloud properties, making them less reflective and more efficient at inducing precipitation. In climate models, reduced amounts of dust in the troposphere together with vegetation changes can often but not always explain the northward expansion of the monsoon. There is not universal agreement on the effects of dust on precipitation in the Sahel, however, in part because the effects of dust on precipitation may be dependent on its size.

In addition to raw precipitation changes, changes in precipitation seasonality such as the length of dry seasons need to be considered when assessing the effects of climate change on vegetation, as well as the fertilizing effects of increased carbon dioxide concentrations in the atmosphere.

Other sources of albedo changes:
* Changes in soil properties result in changes in the monsoon; replacing desert soils with loamy ones results in increased precipitation, and soils that are wet or contain organic matter reflect less sunlight and accelerate the moistening process. Desert sand changes also modify the albedo.
* Albedo changes caused by lakes and wetlands can alter precipitation in climate models.

Intertropical Convergence Zone changes

Warmer extratropics during summer may have drawn the Intertropical Convergence Zone (ITCZ) northward, resulting in precipitation changes. Sea surface temperatures off North Africa warmed under orbital effects and through weaker trade winds, leading to a northward movement of the ITCZ and increasing moisture gradients between land and sea. Two temperature gradients, one between a cooler Atlantic during spring and an already warming African continent, the other between warmer temperatures north of 10° latitude and cooler south, may have assisted in this change. In Eastern Africa, ITCZ changes had relatively little effect on precipitation changes. The past position of the ITCZ in Arabia is also contentious.

Precipitation changes in East Africa

The African humid period that took place in East Africa appears to have been caused by different mechanisms. Among the proposed mechanisms are decreased seasonality of precipitation due to increased dry season precipitation, shortening of the dry season, increased precipitation and increased inflow of moisture from the Atlantic and Indian Oceans. The Atlantic moisture inflow was in part triggered by a stronger West African and Indian monsoon, perhaps explaining why the effects of the AHP extended into the Southern Hemisphere. The behaviour of the easterly trade winds is unclear; increased moisture transport by easterly trade winds may have aided in the development of the AHP but alternatively a stronger Indian Monsoon that draws easterly winds away from East Africa may have occurred.

Changes in the Congo Air Boundary or increased convergence along this boundary may have contributed; the Congo Air Boundary would have been shifted east by the stronger westerly winds directed by lower atmospheric pressure over Northern Africa, allowing additional moisture from the Atlantic to reach East Africa. The parts of East Africa that were isolated from Atlantic moisture did not become significantly wetter during the AHP although at one site in Somalia the seasonality of precipitation may or may not have decreased.

Various contributing factors may have led to the increased humidity in East Africa, not all of which were necessarily operating simultaneously during the AHP. That the "African humid period" reached this part of Africa has been doubted. Finally, increased greenhouse gas concentrations may have been involved in directing the onset of the AHP in tropical southeastern Africa; there, orbital changes would be expected to lead to climate variations opposite to those in the Northern Hemisphere. The pattern of humidity changes in south-eastern Africa are complex.

Additional factors

* Climate change in the far northern latitudes may have contributed to the onset of the AHP. The shrinkage of the Scandinavian and the Laurentide Ice Sheets occurred at its beginning, and in climate models, a retreat of the ice sheets is often required to simulate the humid period. Their existence might also explain why the AHP did not start immediately with the early insolation peak, as still existing ice sheets would have cooled the climate.
* Sea surface temperature changes in the Atlantic influence the African monsoon and may have influenced the onset of the AHP. Weaker trade winds and higher insolation would lead to warmer sea surface temperatures, increasing precipitation by increasing moisture gradients between land and sea. Changes in North Atlantic temperature gradients were also involved.
* Warming of the Mediterranean Sea increases the amount of Sahel precipitation; this effect is responsible for the recent anthropogenic global warming mediated increase in Sahel precipitation. Warmer sea surface temperatures there might also explain the increased precipitation recorded in the Mediterranean and increased intensity of precipitation reconstructed from former rivers in the Sahara during the AHP.
* Increased precipitation during winter is correlated with a larger spatial extent of Mediterranean precipitation and might have aided in the establishment of the AHP, especially in North Africa including Morocco and Northern Egypt, around the northern Red Sea, in the Tibesti and in northern Arabia and generally at higher latitudes where the monsoon did not arrive. This precipitation may have extended to other parts of the Sahara; such would have led to the areas of summer and winter precipitation overlapping and the dry area between the monsoonal and westerlies-influenced climate zones becoming wetter or disappearing altogether. Such changes in Mediterranean-derived precipitation may correlate with changes in the North Atlantic and Arctic Oscillations.
* Trough-mediated northward transport of moisture during autumn and spring has also been proposed to explain the increased precipitation and its underestimation by climate models. In one climate model, increased northward moisture transport by such troughs increases autumn rainfall in the Sahara, especially in the mid-Holocene and when the climate is already moister than usual there.
* Weaker subtropical anticyclones were proposed as an explanation during the 1970s–1980s.
* In montane regions such as the Meidob volcanic field cold temperatures after the last glacial maximum may have reduced evaporation and thus allowed an early onset of humidity.
* Changes in the Earth's geomagnetic field may be linked to the humidity changes.
* Increased moisture supply from larger lakes like Lake Megachad may have increased the precipitation, although this effect is probably not adequate to explain the entire AHP. A similar role has been attributed to the extensive wetlands, drainages and lakes in the Eastern Sahara and to the ecosystem in general.
* Two high elevation winds, the African Easterly Jet and the Tropical Easterly Jet modulate atmospheric air flows over Africa and thus also the amount of precipitation; the Tropical Easterly Jet comes from India and is powered by temperature gradients between the tropics and the subtropics while the African Easterly Jet is powered by temperature gradients in the Sahel. A stronger West African Monsoon resulted in a weaker African Easterly Jet and thus decreased transport of moisture out of Africa.
* Increased atmospheric carbon dioxide concentrations may have played a role in triggering the AHP, especially its extension across the equator, as well as its resumption after the Younger Dryas and Heinrich event 1 through increased sea surface temperatures.
* In some parts of the Sahara increased water supply from montane regions may have assisted in the development of moist conditions.
* Larger forests in Eurasia may have led to a northward shift of the ITCZ.
* Other proposed mechanisms involve convection occurring above the atmospheric boundary layer, increased latent heat fluxes, low pressure in northwestern Africa drawing moisture into the Sahara, changes in the solar cycles and complex atmospheric flow phenomena.

Effects

The African humid period extended over the Sahara as well as eastern, southeastern and equatorial Africa. In general, forests and woodlands expanded through the continent. A similar wet episode took place in the tropical Americas, China, Asia, India, the Makran region, the Middle East and the Arabian Peninsula and appears to relate to the same orbital forcing as the AHP. An early Holocene monsoonal episode extended as far as the Mojave Desert in North America. In contrast, a drier episode is recorded from much of South America where Lake Titicaca, Lake Junin, the discharge of the Amazon River and water availability in the Atacama were lower.

The discharge of the Congo, Niger