Evaluation of hydrocarbon reactivity in urban air

Three a priori methods for estimating the potential for O₃ formation of hydrocarbon (HC) mixtures were applied to a large air pollution data set collected aloft over Tokyo during 16-17 July 1981 and 6-7 August 1980. Individual HC samples were compared using the concentration weighted OH-HC reaction rate constant, -k_OH the effective O₃ formation rate constant, k_e; and the maximum O₃, O₃(max), formed during 12 h of irradation with a NO₂ photolysis rate constant, k₁, of 0.4 min^-1. Values of k_e and O₃(max) were estimated using a carbon-bond photochemical smog model (CBM-III). The maximum incremental ozone (O₃(fp)) above observed levels (O₃^*) was also determined from O₃(max)-O₃^*. The HC data set consisted of 192 samples containing 18 components from 1981 and 66 samples containing 47 components from 1980. Each sample was accompanied by measurements of O₃, NMHC, NO, NO₂, temperature and relative humidity (r.h.). Sampling was mostly at altitudes between 350 and 600 m. Six flights, usually covering the same flight pattern, were spaced at approximately 3 h intervals throughout each day starting at sunrise. In essence, this provided six chemical 'snapshots' of the air over Tokyo for each day of sampling. All of the reactivity parameters were found to be reasonably consistent with each other when compared by individual samples for a given year. In addition, when comparisons were carried out on a run averaged basis. O₃ (max) appeared to be linearly related to the actual O₃ concentration 3 h later; and k_e and k_OH to the change in observed O₃ with time 3-4.5 h later. Trajectory analysis demonstrated that air moved slowly through the sampling region. This was physically consistent with the 3-h lag interpretation. From this evaluation it appears that any of the reactivity parameters are reasonable ways of comparing HC mixtures, with more detailed input information supplying more detailed results.