Long Term Measurements of NOx and O3 in the Remote Tropical North Atlantic Marine Boundary Layer

Trace quantities of nitrogen oxides (NO and NO₂, collectively termed NO_x) affect the concentration of the hydroxyl radical, the main oxidising agent in the atmosphere, and often control the photochemical formation of ozone (O₃). Remote tropical areas are of special interest because the mixing ratios of NO_x are close to the critical mixing ratio where the prevailing O₃ destruction regime changes to one of photochemical O₃ production and the presence of high levels of hydroxyl radicals is critical in controlling the global lifetime and hence future concentrations of gases such as methane. The Cape Verde Atmospheric Observatory is situated in the tropical Atlantic ocean (16.85 ^oN, 24.87 ^oW) facing the North Easterly trade winds and receives air from this sector almost exclusively. It has a near 5 year continuous data series of reactive gases (including NO, NO₂ and O₃) measured in air with a variety of air mass origins, including the North Atlantic, African continental and Europe. Ozone depletion is observed almost every day, and this is due to very low NO mixing ratios (typical 11:00 - 15:00 average < 10 pptv) and a low but significant presence of halogen oxides such as BrO and IO (< 2.5 and 1.5 pptv respectively). Atlantic air originating over Africa contains the highest levels of NO_x (~35 pptv) and also reduced daily O₃ destruction, with O₃ production observed on a few days. Atlantic air originating from over North America or Europe shows lower NO_x levels (~25 pptv), with greater observed O₃ destruction. Annual NOx levels were higher in 2008 and 2009 than in 2006-2007. Box model simulations (including halogen oxide chemistry) predict that the transition from O₃ destruction to production may occur at NO mixing ratios as low as 17 pptv at certain times of year. Comparison of measurements with the output of the GEOS-CHEM global chemistry transport model show good agreement for O₃, but the model significantly underestimates NO and NO₂ mixing ratios. Reasons for this discrepancy including possible additional NO_x sources are discussed.