This study investigates several factors that could influence ozone chemistr
y occurring in subsonic aircraft plumes in the upper troposphere. The study
focuses on uncertainties in gas-phase rate parameters, but also examines t
he influence of selected heterogeneous reactions, the rate of expansion of
the plume, ambient and initial plume concentrations, and the time of emissi
ons. Monte Carlo analysis with Latin hypercube sampling was applied to an e
xpanding box model of an aircraft plume, in order to estimate the sensitivi
ties of O-3, perturbations (Delta O-3,) to uncertainties in rate constants
in the RADM2 chemical mechanism. The resulting coefficient of variation in
Delta O-3, at the end of a 36 h simulation was about 50%. Influential uncer
tainties in gas-phase rate parameters include those for photolysis of NO2 a
nd HCHO, O-3 + NO, HO2 + NO, and formation of PAN and HNO3. With high backg
round concentrations of non-methane hydrocarbons, uncertainties in rate par
ameters of reactions involving peroxy radicals from ethene and propene oxid
ation were also influential. The coefficient of variation for Delta O-3, du
e to uncertainties in emission indices of NOx, CO, and organic compounds wa
s less than 15%. The effects of the heterogeneous reaction of N2O5 leading
to HNO3 formation, and hypothesized reactions of HNO3 and NO2 on soot, were
also investigated. The results suggest that the latter two reactions could
be influential for Delta O-3, if published estimates of reaction probabili
ties and high estimates of soot concentrations in plumes are realistic. (C)
1999 Elsevier Science Ltd. All rights reserved.