A DRY DEPOSITION PARAMETERIZATION FOR SULFUR-OXIDES IN A CHEMISTRY AND GENERAL-CIRCULATION MODEL

A dry deposition scheme, originally developed to calculate the deposit ion velocities for the trace gages O-3, NO2, NO, and HNO3 in the chemi stry and general circulation European Centre Hamburg Model (ECHAM), is extended to sulfur dioxide (SO2) and sulfate (SO42-). In order to red uce some of the shortcomings of the previous model version a local sur face roughness and a more realistic leaf area index (LAI), derived fro m a high-resolution ecosystem database are introduced. The current mod el calculates the deposition velocities from the aerodynamic resistanc e, a quasi-laminary boundary layer resistance and a surface resistance of the surface cover, e.g., snow/ice, bare soil, vegetation, wetted s urfaces, and ocean. The SO2 deposition velocity over vegetated surface s is calculated as a function of the vegetation activity, the canopy w etness, turbulent transport through the canopy to the soil, and uptake by the soil. The soil resistance is explicitly calculated from the re lative humidity and the soil pH, derived from a high-resolution global soil pH database. The snow/ice resistance of SO2 is a function of tem perature. The SO2 deposition velocity over the oceans is controlled by turbulence. The sulfate deposition velocity is calculated considering diffusion, impaction, and sedimentation. Over sea surfaces the effect of bubble bursting, causing the breakdown of the quasi-laminary bound ary layer, scavenging of the sulfate aerosol by sea spray, and aerosol growth due to high local relative humidities are considered. An integ rated sulfate deposition velocity is calculated, applying a unimodal m ass size distribution over land and a bimodal mass size distribution o ver sea. The calculated sulfate deposition velocity is about an order of magnitude larger than that based on a monodisperse aerosol, which i s often applied in chemistry-transport models. Incorporation of the ne w dry deposition scheme in the ECHAM model yields significant relative differences (up to similar to 50%) in mass flux densities and surface layer concentrations compared to those calculated with a simple, cons tant dry deposition scheme.