Aerosol particle evolution in an aircraft wake: Implications for the high-speed civil transport fleet impact on ozone

Previous calculations of the ozone impact from a fleet of high‐speed civil transports (HSCTs) have been carried out by global two‐dimensional (2‐D) models [Bekki and PyIe, 1993; Pitari et al., 1993] which have not included explicit wake processing of sulfur species. This processing could be important for the global sulfate aerosol and ozone perturbations [Weisenstein et al., 1996]. For an HSCT scenario with emission indices of NOx and sulfur equal to 5 and 0.4, respectively, and a cruise speed of Mach 2.4 [Stolarski and Wesoky, 1993b], the Atmospheric and Environmental Research (AER) 2‐D model gives 0.50–1.1% as the range of the annually averaged O3 column depletion at 40°–50°N. This range is determined by the extreme assumption that emitted SO2 is diluted into the global model grid box either as gas or as 10 nm sulfate particles. A hierarchy of models is used here to investigate the impact of processes in the wake on the calculated global ozone response to sulfur emissions by a proposed HSCT fleet. We follow the evolution of aircraft emissions from the nozzle plane using three numerical models: the Standard Plume Flowfield‐II/Plume Nucleation and Condensation model (SPF‐II/PNC), an AER far wake model incorporating microphysics of aerosol particles, and the AER global 2‐D chemistry‐transport model. Particle measurements in the wake of the Concorde [Fahey et al., 1995a] are used to place constraints on sulfur oxidation processes in the engine and the near field. To explain the Concorde measurements, we consider cases with different fractions of SO3 (2%, 20%, and 40%) in the sulfur emissions at the nozzle plane and also the possibility of other unknown heterogeneous or homogeneous oxidation processes for SO2 in the wake. Assuming similar characteristics for the proposed HSCT fleet, the global ozone response is then calculated by the 2‐D model. Using the model‐calculated wake processing of sulfur emissions under the above assumptions and constrained by the Concorde particle measurements, the range of annually averaged O3 column depletion at 40°–50°N is reduced from 0.5–1.1% to 0.75–1.0%. Our analysis shows that the global ozone response is more sensitive to the assumed partitioning of sulfur emissions between SO2 and SO3 at the nozzle plane than to the wake dilution rate. Outstanding uncertainties and recommendations for further wake‐sampling experiments are also discussed.