More than half of the annual human-induced biomass burning which occurs on Earth takes place on the African continent. Such practises, which are necessary for regenerating arable land during subsidence farming, release vast quantities of trace gas pollutants and aerosol particles into the atmosphere. The peak in the burning season displays a staggered seasonal maximum in intensity in northern and southern Africa. Pyrogenic mixing and convection can rapidly loft such emissions and aerosols out of the boundary layer, allowing long-range transport away from the main source regions. For instance, during the West African Monsoon (WAM) emissions released in southern Africa are typically blown westwards out over the tropical Atlantic Ocean. Satellite measurements have shown that significant increases in both CO and ozone also occur around Equatorial Africa as a result.

One focus of the African Monsoon Multi-disciplinary Analysis experiment (AMMA) was to determine whether state-of-the-art Chemistry Transport Models have to ability to capture the correct distribution of such species during July and August 2006. For this purpose a recently compiled emission database was applied (L3JRCv2) in four different models (namely TM4, p-TOMCAT, LMDZ-INCA and MOCAGE) in order to allow an inter-comparison of the resulting distributions of gases and assess short-comings in the current generation of offline CTM's. Due to the different implementation of parameterizations for the description of convection, advective mixing and vertical diffusion of surface emissions, a number of different passive tracer species were also included across different latitudinal bands throughout Africa, where they were prescribed a lifetime of 20 days.

Due to (e.g.) ozone being formed via the chemical processing of nitrogen oxides and hydrocarbons, the intercomparison also tested the performance of the chemical mechanisms employed in each of the different models. Although the relative magnitude of the maximum concentrations were different, it was found that the latitudinal position at which such concentrations occurred was typically between 0-5S. Thus, the northerly transport across the Equator was not captured by any of the large-scale global CTM's.

Moreover, using co-located ozonesonde profiles taken at Cotonou, Benin, comparisons revealed that none of the models were able to capture the correct vertical profile of tropospheric ozone observed in Western Africa during the WAM, where complex meteorology and active photo-chemistry are dominant. Figure 1 below shows the monthly mean distribution of the southern African passive tracer for each of the participating models averaged between 2W-6E. This provides an impression differences in westerly transport out over the Atlantic Ocean.

For more details the reader is referred to the early online release of the resulting article (pdf: 2.1 Mbyte) in the Bulletin of the American Meterological Society