ENERGETICS OF THE 6000-YR BP ATMOSPHERIC CIRCULATION IN BOREAL SUMMER, FROM LARGE-SCALE TO MONSOON AREAS - A STUDY WITH 2 VERSIONS OF THE LMD AGCM

The mechanisms of northern summer monsoon changes at 6 kyr BP and thei r dependency to model parameterizations is investigated using two vers ions of the atmospheric general circulation model developed by the Lab oratoire de Meteorologie Dynamique, CNRS, Paris (LMD), within the Pale oclimate Modeling Intercomparison Project (PMIP). These two versions d iffer in their horizontal resolution and in their treatment of some su rface processes: one model has a low horizontal resolution and rather simple surface treatments; the other model has a better resolution and uses more complex surface parameterizations. The authors focus on the energy budgets, which are useful tools to understand the impact of th e model parameterizations. As a response to mid-Holocene insolation ch ange, the June-September monsoon is enhanced over northern Africa, nor thern India, and the western Pacific, both in terms of precipitation a nd low-level convergence changes; convection decreases over Central Am erica. These changes, which are simulated with different amplitudes an d locations by the two LMD versions, are associated with large-scale c hanges in the Hadley-Walker circulation-export of energy released by t he precipitation in the monsoon regions to the northern extratropics a nd to the Southern Hemisphere. Due to its simple treatment of the evap oration, the first model simulates large changes in the hydrological c ycle, but negative feedbacks compensate For the gain of energy associa ted with the latent heat release: because of a strong evaporation incr ease, the sensible heat flux decreases, and the high-albedo clouds for med in low levels have a negative impact an atmospheric heat gain. Wit h a more complex surface treatment, the other model simulates less eva poration changes and thus a weaker hydrological response, but similar energy source changes, because of increased sensible hear flux and the formation of more middle-level clouds that have a positive effect on the radiative budget. The differences between both versions are partic ularly large for the African monsoon, which is mainly fed by local rec ycling. It is shown that the mechanism of past monsoon intensification is sensitive to model surface parameterization, mainly because of the strong coupling between the surface evaporation and the cloud formati on and optical properties. This methodology will help to compare the d ifferent models that participate to PMIP.