Tropospheric ozone simulation with a chemistry-general circulation model: Influence of higher hydrocarbon chemistry

We present an improved version of the global chemistry-general circulation model of Roelofs and Lelieveld [1997]. The major model improvement is the representation of higher hydrocarbon chemistry, implemented by means of the Carbon Bond Mechanism 4 (CBM-4). Simulated tropospheric ozone concentrations at remote locations, which agreed well with observations in the previous model version, are not affected much by the chemistry of higher hydrocarbons. However, ozone formation in the polluted boundary layer is significantly enhanced, resulting in a more realistic simulation of surface ozone in regions such as North America, Europe, and Southeast Asia. Our model simulates a net global tropospheric ozone production of 73 Tg yr(-1) when higher hydrocarbon chemistry is considered, and -36 Tg yr(-1) without higher hydrocarbon chemistry. The simulated seasonality of surface CO agrees well with observations. However, the southern hemispheric maximum for O-3 and CO associated with biomass burning emissions is delayed by 1 month compared to the observations, which demonstrates the need for a better representation of biomass burning emissions. Simulated peroxyacetyl nitrate (PAN) concentrations agree well with observed values, although the variability is underestimated. OH decreases strongly in the continental boundary layer due to its reaction with higher hydrocarbons. However, this is almost compensated by an increase of OH over oceans in the lower half of the troposphere. Consideration of higher hydrocarbon chemistry decreases the global annual tropospheric OH concentration by about 8% compared to a background tropospheric chemistry scheme. Further, the radiative forcing by anthropogenically increased tropospheric ozone on the northern hemisphere increases, especially in July. The forcing also increases on the southern hemisphere where biomass burning emissions produce tropospheric ozone, except between December and June, that is, outside the biomass burning season, when ozone formation is suppressed due to formation of PAN and MPAN from isoprene oxidation. Globally and annually averaged, the forcing increases only by a few percent due to higher hydrocarbon chemistry.