Ozone production and transports in the tropical Atlantic region during the biomass burning season

Tropospheric O 3 distributions over the Atlantic Ocean have been calculated with a coupled chemistry‐general circulation model. Photochemically produced O 3 , mainly from biomass burning emissions, dominates O 3 abundances over the tropical Atlantic Ocean in the southern hemisphere (SH) during September/October. On the other hand, O 3 of stratospheric origin strongly contributes to the tropospheric O 3 column at latitudes poleward of 30°S. Simulated tropospheric flow patterns are in good agreement with European Centre for Medium‐Range Weather Forecasts analyses and trajectory studies. Tracer transports over the tropical South Atlantic Ocean are strongly influenced by wind shear between the boundary layer and the free troposphere, leading to a stratification of O 3 . At the surface relatively O 3 ‐poor air is transported from SH middle latitudes. At higher altitudes, relatively O 3 ‐rich air is transported from the African biomass burning regions between 2 and 5 km and from South American biomass burning regions in the mid‐to‐upper troposphere. The model simulates O 3 production rates of 10–50 parts per billion by volume (ppbv) O 3 d −1 in the lower troposphere over the biomass burning regions and 2–6 ppbv O 3 d −1 in the middle and upper troposphere. Photochemical destruction of O 3 prevails in the lower troposphere over the ocean, maximizing in the African and South American outflow regions. In the northern hemisphere, in situ photochemically produced O 3 dominates throughout most of the troposphere. Calculated O 3 volume mixing ratios are compared with a latitude‐altitude O 3 distribution measured during an Atlantic ship cruise along 30°W, between 55°N and 30°S, in September/October 1988, and with ozone sonde measurements from Southern African Fire Atmospheric Research Initiative/Transport and Atmospheric Chemistry Near the Equator—Atlantic (SAFARI/TRACE A) in September/October 1992. Calculated O 3 levels agree reasonably well with the ship cruise data, except for the tropical SH lower troposphere where the model significantly underestimates O 3 . However, modeled vertical O 3 distributions are underpredicted compared to TRACE A sonde measurements. Simulated O 3 columns over the ocean are somewhat lower compared to values retrieved from satellite data, in particular over the tropical Atlantic. The underestimation is probably due to the neglect of higher hydrocarbon chemistry in the model.