Sensitivity of the African and Asian Monsoons to Mid-Holocene Insolation and Data-Inferred Surface Changes

Abstract

Orbital forcing alone is not sufficient to explain the massive northward penetration of monsoon rains in Africa shown by data during the mid-Holocene (6000 yr ago). Feedbacks associated with changes in SSTs and land surface cover may be necessary to produce a sufficient increase in the monsoon. A step toward a better understanding of the respective role of oceans and land surfaces is to design sensitivity studies with prescribed forcings, inferred from observations. In the first study, SSTs are lowered in the upwelling regions offshore of West Africa and Somalia, and increased in the Bay of Bengal and South China Sea. In the second simulation, the modern Sahara desert is replaced by a combination of xerophytic woods/scrub and grassland. In both cases the amount of water vapor advected from oceanic sources is increased north of 10°N in Africa in response to the increased land?sea temperature contrast, thereby enhancing rainfall. But the magnitude of the simulated changes is much larger when land surface is modified. The lower albedo (compared to desert) increases the amount of radiation absorbed by the surface in northern Africa and warms it up, and the larger roughness length increases both the sensible and latent heat fluxes. Moreover, vegetation is more efficient in recycling water than a bare soil, and the release of latent heat in the atmosphere increases convection, which in turn helps maintain the onshore oceanic advection. The monsoon season is then lengthened by 1?2 months compared to all other simulations reported in the paper. The intensity of monsoon rains is also modified in Asia in both sensitivity experiments. Warmer SSTs in the Bay of Bengal and South China Sea reduce the land?sea contrast and therefore the inland penetration of monsoon rains. Changes in the position of the main large-scale convergence area in the case of a green Sahara enhances the precipitation in India. Changes are also discussed in terms of atmospheric circulation. For example, the tropical easterly jet at 200 hPa is increased in all 6-kyr-BP simulations, but only over Africa in the case of a prescribed green Sahara. The African easterly jet has been pushed at higher altitude in response to all prescribed forcings; wind speed is then reduced at 700 hPa but increased at higher altitude.