Simulation of the termination of the African Humid Period

Hans Renssen
Faculty of Earth and Life Sciences, Vrije Universiteit Amsterdam, Netherlands

Victor Brovkin
Potsdam Institut für Klimafolgenforschung, Potsdam, Germany

Thierry Fichefet and Hugues Goosse
Institut d´Astronomie et de Géophysique Georges Lemaître, Université catholique de Louvain, Louvain-la-Neuve, Belgium


Numerous paleoclimatic data show that the early Holocene (~9-7 kyr BP) was a relatively humid period in Northern Africa. During this phase, the African Humid Period (AHP), grasslands covered the Sahara/Sahel region, and many lakes and wetlands existed here1-2. The humid conditions at this time were associated with a strengthening of the summer monsoon circulation due to an increase in the land-sea thermal contrast under influence of the relatively high summer insolation in the early Holocene3. Some paleoevidence from ocean cores indicates that the AHP came to an abrupt end around 6 kyr BP4. This abrupt termination is linked to a positive biogeophysical feedback between vegetation and precipitation in the Sahara region: when precipitation is reduced, the vegetation cover decreases, thereby increasing the surface albedo, which in turn leads to a decrease in precipitation5. Previous climate model experiments have reproduced the abrupt termination of the AHP, thereby showing the effectiveness of this biogeophysical feedback6. Thus, in short, around 6 kyr BP, a shift took place from a "green" Sahara state to a "desert" state. Indeed, climate model studies have suggested that these two basic states could be stable in the Sahara/Sahel region7-8. In addition, some proxy data give evidence of rapid fluctuations between these green and desert states during the termination phase of the AHP9-10, indicating that the dynamics of the coupled climate system are more complex than previously thought.

Our numerical experiments

We have investigated the AHP termination in experiments performed with the ECBilt-CLIO-VECODE coupled atmosphere-ocean-vegetation model11-13. The model consists of three components:

(1) ECBilt, an atmospheric model (T21, three layers) based on quasi-geostrophic equations11,
(2) CLIO, an oceanic general circulation model coupled to a comprehensive dynamic-thermodynamic sea-ice model12, and
(3) VECODE, a model that describes the dynamics of grassland and forest, and desert as a third dummy type13.

In our main experiment, we have forced the model with insolation and atmospheric concentrations of CO2 and CH4 for the last 9,000 years. All other boundary conditions were fixed at their 1750 AD values. The initial conditions were taken from an experiment in equilibrium with boundary conditions for 9 kyr BP. In the latter experiment, grasslands covered the Sahara in agreement with palaeobotanical data.


Phase 1: 9-7.5 kyr BP, 'green' state
In the Western Sahara region, the model simulates from 9 to 7.5 kyr BP a 'green' equilibrium characterized by a mean annual precipitation of 290 mm/yr and a vegetation fraction of 70%. This green state is associated with a relatively strong land-sea thermal gradient, which strengthens the summer monsoons, leading to an increased transport of humid air towards the continent and enhanced convective precipitation over land.

Phase 2: 7.5-5.5 kyr BP, intermediate unstable state
After 7.5 kyr BP, precipitation and vegetation concentration decrease to values of 210 mm/yr and 50%, respectively. In addition, in response to the biogeophysical feedback, the variability in vegetation fraction increases significantly (standard deviation is 9.2% for 9-7.5 kyr BP and 12.2% for 7.5-5.5 kyr BP). The time period separating the "green" spikes ranges from 110 to 370 years, which is similar to the lake-level fluctuations observed in high-resolution palaeodata from the Western Sahara.

Phase 3: 5.5 kyr BP to present, desert state
After 5.5 kyr BP, the variability decreases substantially and the system moves towards a desert state. At 1 kyr BP, annual precipitation is as low as 60 mm/yr and vegetation fraction is only 10%.

In addition to the main 9 kyr-long simulation, we have performed four additional sensitivity experiments of 200-yr duration to study the stability of the desert and green states through time. In these sensitivity experiments, we fixed in the first 100 years the vegetation in the Sahara region to either desert or grassland, after which the model was allowed to evolve freely during the remaining 100 years. All other forcings were kept constant. The four experiments can be characterized as follows: 9k-desert, 6k-desert, 6k-green, 0k-green. In the 0k-green experiment, the model quickly returned to the desert state after 100 years, showing that at under present-day conditions, only the desert state is stable in our model. In the 9k-desert experiment, on the other hand, the model returned to the green state, suggesting that this state is favoured under 9 kyr BP conditions. In the 6k-desert and 6k-green experiments, the model evolved towards an intermediate state with a vegetation cover between 45 and 60%. Thus, under 6 kyr BP conditions, the model has no clear preference for either the green or desert state, explaining why, between 7.5 and 5.5 kyr BP, the stochastic variations in precipitation are able to induce transitions between the the two states.

According to our experiments, the Holocene evolution of the atmosphere-ocean-vegetation system in the Western Sahara region may be summarised by this simplified diagram. Hypothetically, the climate-vegetation system possesses multiple steady states, desert and 'green'. Potential minima, marked by black balls, correspond to equilibria that are stable in the absence of perturbations. Precipitation fluctuations induced by large-scale atmospheric and oceanic variability perturb the stable state, and a positive feedback between vegetation and atmosphere amplifies external variability. Grey balls and arrows indicate the maximum range of system variations.

A. 9 kyr BP. The dynamical system has two steady states with a preference for the green state (deeper potential minimum). The system fluctuates in vicinity of the green state.

B. 6 kyr BP. The potential became equal for both states. The system fluctuates between desert and green states with a stronger variability than at 9 kyr BP. Consequently, 6 kyr BP is not a good period to test models in equilibrium experiments.

C. 0 kyr BP. The system underwent bifurcation as the green state lost stability and disappeared. Desert is the only steady state. Precipitation fluctuations are reduced in comparison with the two-well system (A and B). Similar transitions between multiple equilibria may have occurred during other orbitally forced transitions in the geological past.

The IPCC expects a significant increase in precipitation in the Western Sahara/ Sahel region in the 21st century. This is likely to considerably increase the variability as it pushes the system in the direction of the 6 kyr BP state, with both the green and desert states being potentially stable.

For further information, please consult the following paper:

Renssen, H., V. Brovkin, T. Fichefet and H. Goosse, Holocene climate instability during the termination of the African Humid Period, Geophysical Research Letters 30, 1184, doi: 10.1029/2002GL011636, 2003.


(1) Prentice, I.C., D. Jolly and BIOME 6000 participants, Mid-Holocene and glacial-maximum vegetation geography of the northern continents and Africa, J. Biogeogr., 27, 507-519, 2000.

(2) Gasse, F., Hydrological changes in the African tropics since the last glacial maximum, Quat. Sci. Rev., 19, 189-211, 2000.

(3) Kutzbach, J.E., and F.A. Street-Perrott, Milankovitch forcing of fluctuations in the level of tropical lakes from 18 to 0 kyr BP, Nature, 317, 130-134, 1985.

(4) deMenocal, P., J. Ortiz, T. Guilderson, J. Adkins, M. Sarnthein, L. Baker and M. Yarusinsky, Abrupt onset and termination of the African humid period: rapid climate responses to gradual insolation forcing, Quat. Sci. Rev., 19, 347-361, 2000.

(5) Charney J., P.H. Stone and W.J. Quirk, Drought in Sahara - biogeophysical feedback mechanism, Science, 187, 434-435, 1975.

(6) Claussen, M., C. Kubatzki, V. Brovkin, A. Ganopolski, P. Hoelzmann and H.J. Pachur, Simulation of an abrupt change in Saharan vegetation in the mid-Holocene, Geophys. Res. Lett., 26, 2037-2040, 1999.

(7) Claussen, M., On multiple solutions of the atmosphere-vegetation system in present-day climate, Global Change Biol., 4, 549-559, 1998.

(8) Brovkin, V., M. Claussen, V. Petoukhov and A. Ganopolski, On the stability of the atmosphere-vegetation system in the Sahara/Sahel region, J. Geophys. Res., 103, 31,613-31,624, 1998.

(9) Fabre, J., and N. Petit-Maire, Holocene climatic evolution at 22-23°N from two palaeolakes in the Taoudenni area (Northern Mali), Palaeogeogr. Palaeoclim. Palaeoecol., 65, 133-148, 1988.

(10) Lézine, A.M., J. Casanova and C. Hillaire-Marcel, Across an early Holocene humid phase in Western Sahara: pollen and isotope stratigraphy, Geology, 18, 264-267, 1990.

(11) Opsteegh, J.D., R.J. Haarsma, F.M. Selten and A. Kattenberg, ECBILT: a dynamic alternative to mixed boundary conditions in ocean models, Tellus, 50A, 348-367 1998.

(12) Goosse, H., and T. Fichefet, Importance of ice-ocean interactions for the global ocean circulation: a model study, J. Geophys. Res., 104, 23,337-23,355, 1999.

(13) Brovkin, V., J. Bendtsen, M. Claussen, A. Ganopolski, C. Kubatzki, V. Petoukhov and A. Andreev, Carbon cycle, vegetation and climate dynamics in the Holocene: Experiments with the CLIMBER-2 Model, Global Biogeochem. Cycl., 16(4), 1139, Doi:10.1029/2001GB001662, 2002


To publication list Hans Renssen

Go to modelling of 8,200 yr BP event

Go to study on atmospheric variability during stadial-interstadial transitions

Go to simulation of the impact of global deforestation